Online Tutoring Platforms: How to Choose One

What are the best Zoom alternatives for online tutoring?

The best Zoom alternatives for online tutoring in 2026 are purpose-built virtual classrooms that bake teaching tools into the platform itself, rather than asking you to bolt them on after a generic video call. Six worth shortlisting: Koala Go (best for independent K-12 tutors who teach younger kids), Lessonspace (best for STEM tutors and tutoring organizations), LearnCube (best for ESL and language teaching with a built-in curriculum), Vedamo (best for group classes and traditional classroom-style live tutoring), Whereby (best for clean 1-on-1 video calls with adult learners), and Google Meet (best as a free fallback when you don't need teaching-specific tooling). Every option runs in the browser, so students don't install anything. Paid tiers start at roughly $10-$30/month.

The 6 best Zoom alternatives for online tutors

None of these tools is right for every tutor. Each replaces a specific weak spot of Zoom — the missing whiteboard, the painful screen share on tablets, the lack of built-in scheduling or rewards, the self-serve support that goes silent when something breaks mid-lesson. Pick the one whose strengths match how you actually teach.

1. Koala Go — best for independent K-12 tutors and engagement-driven lessons

What it is: A browser-based virtual classroom built specifically for online tutors. Includes an interactive whiteboard that takes PDF, PowerPoint, and image uploads; a cobrowser (you and the student interact with the same live webpage together); a 3D Playground used as a brain break or full lesson environment with student avatars; built-in scheduling and worldwide invoicing (parents pay in local currency, including via WeChat in China); 24/7 in-app support; and lesson recording on the Pro plan.

Strengths: Strong engagement for younger learners — kids actively want to come to class because they want to be in the Playground. Cobrowser is genuinely unusual; nothing on Zoom or most alternatives lets you share a live website where both of you can click and type. Everything in one tool, so you stop juggling video + whiteboard + scheduler + invoicer + rewards system.

Tradeoffs: Koala Free caps group sessions at 4 students. Koala Pro is $25.99/month ($21.99/month billed annually); some tutors find that higher than a generic video tool — though it tends to replace several other paid tools. The Playground is actively developed and still adding requested features like timers and group-select of Playground objects.

Best for: Independent tutors teaching K-12, especially under-10s; ESL teachers of kids; reading and Orton-Gillingham practitioners; anyone whose retention problem is "the student is bored on Zoom."

2. Lessonspace — best for STEM and subject-specific tutoring

What it is: A virtual classroom built around a powerful collaborative whiteboard with subject-specific tools (equation editor, graph editor, code editor), document upload and annotation, session recording, and a resource library. Students join via a Space link with no signup.

Strengths: Whiteboard is widely considered one of the strongest in the category, especially for math, physics, and coding. Tutor-side controls let you lock what students can edit and unlock when it's their turn. Available in 10+ languages. APIs for tutoring companies to embed.

Tradeoffs: More whiteboard-first than engagement-first — less optimized for very young learners who need movement and rewards. Some reviewers report audio/video hiccups on weaker connections.

Best for: Math, science, and coding tutors; tutoring companies that need a polished whiteboard and developer APIs.

3. LearnCube — best for ESL and language teaching with a ready curriculum

What it is: A virtual classroom focused on language teaching. Includes an online whiteboard with multiple tabs, an integrated ESL content library (CEFR A1-C2 and Business English), an AI teacher assistant in beta, white-label branding, lesson recording, and reload-past-class features.

Strengths: Subject-specific for language teachers — if you teach standard CEFR content, the built-in curriculum cuts prep time meaningfully. White-label option for tutors building a branded school. Used by some of the larger language tutoring companies.

Tradeoffs: Optimized for language teaching; if you teach math, piano, or non-language subjects, the content library doesn't help and you're paying for a feature you won't use. Pricing splits between the standalone Virtual Classroom and a higher tier that includes the ESL content library.

Best for: ESL or other language teachers, especially those who'd rather use a ready-made CEFR curriculum than build their own; small language schools.

4. Vedamo — best for group classes and classroom-style live tutoring

What it is: A browser-based virtual classroom with interactive whiteboard, breakout rooms, screen sharing, reusable lesson templates, gamification options, and support for up to 50 active participants per session. Around $25/month for the Virtual Classroom plan, with a 30-day free trial.

Strengths: Genuine support for small group classes (up to 50 participants) and breakout rooms make this stronger than most for language schools and group-tutoring organizations. Reusable lesson templates save prep time across recurring lessons. Integrates with Moodle, Canvas, and other LMSes via LTI.

Tradeoffs: Interface feels more traditional and meeting-like than engagement-driven. Younger students don't necessarily love it the way they love an avatar environment.

Best for: Group classes (3-50 students), tutoring organizations, or solo tutors who want a classroom-style experience at a low monthly price.

5. Whereby — best for clean 1-on-1 video calls with adult learners

What it is: A browser-based video conferencing tool from Norway, designed around a simple proposition: persistent custom room URLs, no downloads, no logins for guests. Includes screen sharing, recording (on Pro), live captions, and integrations with Miro and Google Docs. Pro starts at $10.99/month; the free plan caps every meeting at 30 minutes.

Strengths: The smoothest "click a link, you're in the call" experience on this list. Permanent custom room URL parents and students can memorize. Reliable on mobile browsers. Strong on privacy — stated policy is not to store or train on call content.

Tradeoffs: It is a video tool, not a virtual classroom. No whiteboard built for teaching, no rewards system, no scheduling or invoicing for tutors. You'll bring your own teaching content (slides, docs) and share via screen share or by linking to a separate tool.

Best for: Adult 1-on-1 conversation practice, business English, IELTS/TOEFL prep with mature students, or any context where you mostly need a clean video call.

6. Google Meet — best free fallback when you don't need teaching tools

What it is: Google's video conferencing product. Free with a Google account, browser-based, ubiquitous. Parents and kids already know how to use it.

Strengths: Free, reliable, no learning curve for parents on any device. Trusted brand reduces the "is this a real thing?" conversation with new families.

Tradeoffs: A meeting tool, not a teaching tool. Google retired Jamboard (the whiteboard companion most tutors used with Meet) at the end of 2024, so the standard Meet + Jamboard tutoring setup no longer exists out of the box. You'll bolt on a separate whiteboard, a separate scheduler, a separate invoicing tool, a separate rewards system — exactly the stack-juggling that drove tutors off Zoom in the first place.

Best for: Tutors where cost is a hard constraint and the lesson format genuinely doesn't need more than video + screen share. For the full read on where Meet works for paid tutoring and where it stops, see can I use Google Meet for online tutoring?

Feature comparison at a glance

Approximate as of mid-2026; vendor pricing and features drift, so check the source pages before committing.

Feature	Koala Go	Lessonspace	LearnCube	Vedamo	Whereby	Google Meet
Built specifically for online tutoring	Yes	Yes	Yes	Yes	No	No
Interactive whiteboard with PDF/PowerPoint upload	Yes	Yes	Yes	Yes	No (Miro embed only)	No (since Jamboard retired)
Live shared web browsing (cobrowser)	Yes	No	No	No	No	No
3D playground / avatars for younger students	Yes	No	No	No	No	No
Built-in scheduling for tutors	Yes	Partial (via APIs/integrations)	Yes (in Online School tier)	Partial	No	Via Google Calendar
Built-in invoicing / payment collection	Yes (worldwide, including WeChat)	No	Yes (in Online School tier)	No	No	No
Lesson recording	Yes (Pro)	Yes	Yes	Yes	Yes (Pro)	Yes (paid Workspace)
Built-in rewards / gamification	Yes	Limited	Limited	Some	No	No
Group session limit (one host)	4 on Free, more on Pro	~10 in group classes	Up to ~25	Up to 50	Up to 100-200 (by plan)	100+
Starting paid price (per month, approx)	$25.99 (Pro)	From ~$9	From ~$9	~$25	From $10.99	Free / part of Workspace
Has a free tier	Yes	Trial	14-day trial	30-day trial	Yes (45 min cap on groups)	Yes

How to choose between Zoom alternatives

Four questions, in order, narrow the field fast:

How old are your students? Under 10, engagement is your real problem — favor platforms with rewards, avatars, and interaction baked in (Koala Go). Teenagers and adults, a clean whiteboard and reliable video is usually enough (Lessonspace, LearnCube, Vedamo, Whereby).
What subject? Language teaching with a ready curriculum: LearnCube. Math/science with strong subject tools: Lessonspace. General K-12 across subjects: Koala Go. Adult conversation: Whereby.
Solo tutor or organization? Solo tutors care about all-in-one (video + scheduling + invoicing) so they don't run five subscriptions: Koala Go. Tutoring organizations care about white-labeling, APIs, and team management: LearnCube's Online School or Lessonspace.
Group size? 1-on-1 and small groups: any of the above. Classes of 10+: Vedamo or LearnCube; Whereby for adult workshops up to ~100.

A practical shortcut: run a real lesson on each shortlist option during its free trial. The platform that feels right after a single full lesson is almost always the right answer; the one you have to talk yourself into is almost always wrong.

Free Zoom alternatives for online teaching

If "free" is the constraint — you're piloting an idea, you only run a handful of paid lessons a week, or your students' families are paying you in a market where every dollar of overhead matters — three of the platforms above have free tiers genuinely usable for teaching, not just trials that expire in a couple of weeks: Koala Go, Whereby, and Google Meet. Each has a real cap somewhere — that's what makes it free. The right pick depends on which cap you can live with: Koala Free caps group size at 4 students, Whereby Free caps every meeting at 30 minutes, and Google Meet's free tier caps group meeting length. Koala Free is the only one of the three that doesn't put a clock on a 1-on-1 lesson.

Koala Go (Free) — best for free 1-on-1 and small-group K-12 tutoring

The Free plan is a real virtual classroom, not a stripped-down teaser: you get the interactive whiteboard with PDF and PowerPoint upload, the 3D Playground with student avatars, the cobrowser for shared live web browsing, drawing and sticky notes, and built-in scheduling. A few things are capped to keep it free — group size is the main one (Koala Free is capped at 4 students per session, enough for 1-on-1 and most small-group tutoring but not a 20-student class), PDF and PowerPoint uploads are limited on Free and unlimited on Pro, and the cobrowser is time-limited per session on the free tier. What it doesn't cap is lesson length — there's no time limit on lessons the way there is on free Whereby or Meet. Use it free as long as you want; upgrade to Koala Pro ($25.99/month, or $21.99/month billed annually) when you need lesson recording, unlimited PPT and PDF uploads, full-length cobrowser sessions, or larger group sizes.

Best when: you teach K-12 1-on-1 or small groups, want a real teaching platform rather than a generic meeting tool, and can live with the 4-student cap.

Whereby (Free) — best for short adult 1-on-1 calls with no install

The free plan gives you a persistent custom room URL with no downloads and no logins required for guests, fully browser-based, with screen sharing built in. The main cap is time — free meetings are limited to 30 minutes per session, and that applies to 1-on-1 lessons as much as to group calls, so a standard hour-long lesson doesn't fit on the free plan at all. Whereby is a video tool rather than a virtual classroom: it lists a virtual whiteboard for sketching mid-call, but there's no lesson canvas you can load a worksheet into, so you'd add a separate tool (Miro, a shared Google Doc, or similar) for shared work. There's a fuller breakdown on can I use Whereby for online tutoring?

Best when: you teach adults 1-on-1, value zero-friction join above teaching tools, and don't need a shared canvas inside the call itself.

Google Meet (Free) — best free fallback with maximum parent familiarity

Free with any Google account, browser-based, instantly recognizable to almost every parent. The free tier caps group meeting length, but 1-on-1 calls aren't affected by the same cap. The catch is the missing teaching layer: since Google retired Jamboard at the end of 2024 there's no built-in whiteboard for shared work, no scheduling for paying clients, no invoicing, no rewards system. To run a real tutoring practice on free Meet you'll stack a separate whiteboard, a separate scheduler, a separate invoicing tool, and probably a separate rewards or engagement system — exactly the stack-juggling that pushed many tutors off Zoom in the first place.

Best when: cost is a hard constraint, parents already use Meet for work or school, and your lesson genuinely doesn't need more than video plus a screen share.

What about Lessonspace, LearnCube, and Vedamo?

These three are paid platforms with free trials, not free tiers — Lessonspace varies by plan, LearnCube is a 14-day trial, Vedamo is a 30-day trial. They're worth running a real lesson on during the trial window (especially Lessonspace if you teach STEM, or LearnCube if you teach standard CEFR language curriculum), but they aren't a long-term free option once the trial ends. The honest move with any of these is to set a calendar reminder for the last day of the trial, decide whether the platform earned the subscription, and either pay or fall back to one of the genuinely free tools above.

A note on free open-source video tools

There's a separate category of free open-source community video projects (Jitsi Meet and BigBlueButton are the two most commonly cited) that some teachers use as free Zoom alternatives, either via public instances or by self-hosting. We don't try to cover them in depth here because they aren't purpose-built for tutoring — they're general-purpose conferencing tools that you'd configure or set up yourself. If you have the technical comfort to run your own instance or audit a public one, they're worth knowing about; if you don't, one of the three free options above will get you teaching faster.

Tradeoffs of going free for online teaching

Free tiers exist for a reason — they let you try a platform without commitment. They are not always the right long-term choice for a tutoring practice that's earning real money.

Time and group caps interrupt teaching. A typical free-tier meeting cap (Whereby's is 30 minutes, on every meeting) cuts a standard hour lesson in half, and a 4-student group cap is fine for tutoring but not for whole-class delivery.
Lesson recording is usually a paid feature. If parents want to review what was covered, or you want to review your own teaching, recording typically requires upgrading.
Support is slower on free plans. When something breaks mid-lesson and you need a human inside five minutes, paid plans get you there faster than free ones do.
You'll end up stacking other paid tools anyway. A "free" video tool plus a paid whiteboard plus a paid scheduler plus a paid invoicer often costs more in total — and more juggling time per week — than a single paid all-in-one virtual classroom.

If your tutoring practice is generating consistent income, a paid all-in-one virtual classroom usually pays for itself by replacing two or three separate subscriptions and reclaiming the half-hour-plus per week you'd otherwise spend gluing tools together.

When sticking with Zoom is still the right call

We won't pretend leaving Zoom is always the right move. Stay on Zoom if:

Your students and their parents already use Zoom for work and school, and the "familiar tool" win matters more than the missing teaching features.
You teach almost entirely through one-way explanation (lectures, exam prep walkthroughs) where a webcam + screen share is genuinely enough.
You teach in a region where the alternatives have weaker connectivity than Zoom and your students are sensitive to that. (Notably, mainland China connectivity varies platform to platform — always test before committing.)
You're billing through an agency or marketplace that requires Zoom as the delivery tool.

If none of those apply, the question stops being "should I leave Zoom" and starts being "which of the alternatives above fits how I actually teach."

Questions to ask any platform before you commit

Does it work in the browser on iPad, Android tablets, and Chromebooks — without an app install? Most of your students are not on MacBooks.
Can you upload PDFs, PowerPoint slides, and worksheets and annotate them with the student in real time?
What happens when something breaks mid-lesson — is there a human you can reach in under 5 minutes, or a help center and a ticket queue?
Is there a free tier or trial long enough to run at least 3 real lessons?
Where are the servers, and how does it perform in your students' regions? (China, Southeast Asia, and parts of Africa are common pain points.)
Does the platform record lessons, and where are the recordings stored?
Can you collect payment through the platform, or will you still need a separate invoicing tool?
If you ever scale to a small team of tutors, does the platform support multiple teachers under one brand?

If you'd like to talk through which of these matters most for your practice — or just want to ask whether Koala Go is a fit for the kind of students you teach — write to us at koala@teachwithkoala.com. We're a small team and we read everything.

View full answer page →

What's the best online whiteboard for online tutoring?

The best online whiteboard for tutoring depends on what you teach and whether you want a whiteboard-only tool or a whiteboard inside a full virtual classroom. Six worth shortlisting in 2026: BitPaper (best for general 1-on-1 tutoring with handwriting), Ziteboard (best for math tutors who want a vector canvas), Lessonspace (best for STEM tutors who need equation, graph, and code editors), Koala Go (best when you want one tool that bundles a whiteboard with a cobrowser, scheduling, and a 3D playground for younger students), Miro (best for adult learners and design or business teaching), and Microsoft Whiteboard (best as a free option for tutors already on Teams or Microsoft 365). All six run in the browser, so students don't install anything. Paid tiers cost roughly $9–$25 per month, with free tiers available on most.

The 6 best online whiteboards for online tutors

None of these tools is right for every tutor. Each replaces a different weak spot of a shared screen or a generic video tool — the missing pen pressure for handwriting, the cramped canvas for long algebra problems, the awkward PDF workflow, the lack of multi-page lesson saving. Pick the one whose strengths match the subject you actually teach and the age of your students.

1. BitPaper — best for general 1-on-1 tutoring with handwriting

What it is: A browser-based whiteboard built specifically for tutors, originally by working tutors in the UK. Auto-saved multi-page "papers" you share with each student via a permanent link, with a built-in audio/video call, document upload (PDF, images), and a pen-friendly drawing canvas that's smooth on a stylus or trackpad.

Strengths: Tutor-specific from day one. Each student gets their own paper that persists across lessons, so you and the student can scroll back through last week's work without re-uploading anything. The handwriting feel is one of the best in the category — important if you teach maths, handwriting, or a script-heavy language. Reasonable pricing for solo tutors.

Tradeoffs: It is a whiteboard plus a basic call — not a full virtual classroom. No scheduling, no invoicing, no rewards system, no built-in marketplace of activities. If you want one tool to run your whole tutoring practice, BitPaper isn't it.

Best for: Solo tutors doing 1-on-1 lessons in any subject who want a clean shared workspace and don't need scheduling or invoicing baked in.

2. Ziteboard — best for math tutors on a vector canvas

What it is: A vector-based infinite canvas whiteboard with shape recognition, equation-friendly drawing tools, PDF import, and real-time collaboration. Stylus-friendly on iPad and graphics tablets.

Strengths: The infinite vector canvas matters more than it sounds — you can zoom in to neaten a fraction or pan across to a different worked example without the lines pixelating. Strong with a stylus, which is the realistic way most math tutors actually write equations on screen. Free tier is genuinely usable for occasional lessons.

Tradeoffs: Interface is utilitarian rather than warm — fine for adult learners or older teens, less engaging for young kids. No real "lesson" or "student" model; it's a canvas, not a classroom. No audio/video call built in (you pair it with Zoom, Meet, or similar).

Best for: Math, physics, and engineering tutors working with older students or adults; anyone who teaches primarily with a stylus.

3. Lessonspace — best for STEM tutors who need equation, graph, and code editors

What it is: A full virtual classroom built around a powerful collaborative whiteboard with subject-specific tools — an equation editor, a graph editor, a code editor with syntax highlighting, document annotation, and session recording. Students join via a "Space" link with no signup.

Strengths: The whiteboard is widely considered one of the strongest in the category, especially for technical subjects. Tutor-side controls let you lock what students can edit and unlock when it's their turn. Available in 10+ languages and used by larger tutoring organizations via embeddable APIs.

Tradeoffs: More whiteboard-first than engagement-first — less optimized for very young learners who need movement and rewards. Some reviewers report audio/video hiccups on weaker connections. Premium tiers can get expensive for solo tutors compared to whiteboard-only tools.

Best for: Math, science, and coding tutors with older students; tutoring companies that need a polished whiteboard and APIs.

4. Koala Go — best whiteboard inside a full virtual classroom for kid-focused tutors

What it is: A browser-based virtual classroom for online tutors with an interactive whiteboard at the centre of every lesson. The whiteboard takes PDF and PowerPoint uploads (unlimited on the Pro plan), runs as a Fabric.js canvas with drawing, sticky notes, text, images, stamps, and embedded video, and sits alongside two features no standalone whiteboard offers: a cobrowser (you and the student interact with the same live webpage together), and a 3D Playground used as a brain break or full lesson environment with student avatars. Built-in scheduling, worldwide invoicing (including WeChat in China), and 24/7 in-app support round out the package.

Strengths: One tool replaces the stack-juggling that most tutors fall into — video + whiteboard + scheduler + invoicer + rewards system collapsed into one browser tab. Strong engagement for younger learners because the whiteboard sits next to the Playground rather than in a separate window. PDFs and PowerPoints come straight in and become annotatable slides, so existing materials work without re-creation.

Tradeoffs: Koala Go's whiteboard now ships math tools on its paid tiers (a symbol palette covering operations, fractions, powers, and algebra), but it's still not the most flexible canvas on this list — Lessonspace's subject-specific editors and BitPaper's handwriting feel each win on a narrow axis. If your tutoring is whiteboard-only and you have no use for the cobrowser or Playground, you're paying for capabilities you won't use. Koala Pro is $25.99/month ($21.99/month billed annually); free tier exists with a cap on group size and on PPT/PDF uploads. The Playground is actively developed and still adding requested features like timers and group-select of Playground objects.

Best for: Independent tutors teaching K–12, especially under-10s; ESL teachers of kids; reading and Orton-Gillingham practitioners; any tutor whose retention problem is "the student is bored on Zoom."

5. Miro — best for adult learners and design or business teaching

What it is: A general-purpose collaborative whiteboard from Miro Inc., used by product, design, and consulting teams. Infinite canvas, templates, sticky notes, drawing, comments, frames, plus a large library of integrations (Google Docs, Figma, Jira, and many more).

Strengths: Extremely capable canvas with the deepest template library on this list — anything from mind maps and Kanban boards to flowcharts and SWOT frames. Strong on mobile and tablet. Great fit for adult learners who are already comfortable in business tools.

Tradeoffs: Not designed for tutoring. No "lesson" concept, no student-management view, no scheduling, no audio/video call built in. Free tier limits you to three editable boards, which is fine for one or two students but tight for a full practice. Interface assumes desktop-class comfort — wrong for young kids on tablets.

Best for: Adult-learner tutoring (business English, IELTS prep with mature students, executive coaching), design teaching, or any context where the template library and integrations matter more than tutor-specific features.

6. Microsoft Whiteboard — best free option for tutors already on Teams or Microsoft 365

What it is: Microsoft's collaborative whiteboard, free with any Microsoft account and integrated into Microsoft Teams. Infinite canvas, ink and shape tools, sticky notes, image and document insertion, and live collaboration.

Strengths: Genuinely free for personal Microsoft accounts. If you (and your students) already live in Teams or Microsoft 365 for school, the whiteboard is one click away and works without onboarding. Reliable, well-maintained, with strong stylus support on Surface devices and iPad.

Tradeoffs: A general-purpose whiteboard, not a tutoring tool — no lesson structure, no student dashboard, no scheduling or invoicing. The strongest "Teams + whiteboard" workflow assumes both you and the student are on Microsoft accounts, which is true in school settings but rarely true for independent K–12 tutoring with families.

Best for: Tutors whose students are already on Microsoft 365 (common in schools and some districts); cost-constrained tutors who want a real whiteboard without a subscription.

Feature comparison at a glance

Approximate as of mid-2026. Vendor pricing and features drift, so check the source pages before committing.

Feature	BitPaper	Ziteboard	Lessonspace	Koala Go	Miro	MS Whiteboard
Built specifically for tutoring	Yes	Partial (popular with tutors)	Yes	Yes	No	No
PDF / PowerPoint upload onto the canvas	Yes (PDF, images)	Yes (PDF)	Yes	Yes (Pro: unlimited)	Yes (PDF, images)	Yes (PDF, images)
Subject-specific editors (equation, graph, code)	No	Equation-friendly drawing	Yes	No	No (templates only)	No
Infinite / multi-page canvas	Multi-page papers	Infinite	Multi-page	Multi-slide	Infinite	Infinite
Built-in audio / video call	Yes	No	Yes	Yes	No	Via Teams
Live shared web browsing (cobrowser)	No	No	No	Yes	No	No
3D playground / avatars for younger students	No	No	No	Yes	No	No
Built-in scheduling and invoicing	No	No	Partial (via integrations)	Yes (worldwide, incl. WeChat)	No	No
Lesson recording	Yes (paid)	Limited	Yes	Yes (Pro)	Limited	Yes (via Teams)
Has a free tier	Yes (limited)	Yes	Trial	Yes (group + upload caps)	Yes (3 editable boards)	Yes
Starting paid price (per month, approx)	From ~$10	From ~$9	From ~$9	$25.99 (Pro)	From ~$8 / editor	Free / part of M365

How to choose between online whiteboards for tutoring

Four questions, in order, narrow the field fast:

What subject and what age? Math, physics, and coding tutors with older students get the most out of Lessonspace and Ziteboard. ESL, reading, and any K–4 tutoring is better served by a tool with built-in engagement (Koala Go). General handwriting-led tutoring works well on BitPaper. Business English and adult learners pair naturally with Miro or Microsoft Whiteboard.
Whiteboard-only or full classroom? If you already have a scheduling tool, an invoicing tool, and a video tool you like, a whiteboard-only product (BitPaper, Ziteboard, Miro, Microsoft Whiteboard) lets you keep the rest of your stack. If you want one platform to handle scheduling, payments, and the lesson itself, a full virtual classroom (Koala Go, Lessonspace) replaces several subscriptions at once.
What devices do your students actually use? Many young students join from iPads or Chromebooks borrowed from school. Pen pressure, pinch-to-zoom behaviour, and touch responsiveness vary widely between whiteboards on those devices. Run a trial with a student on the device they'll actually use before committing.
How important is support? Whiteboard-only tools are largely self-serve. Full classrooms vary — Koala Go offers in-app support that replies in under two minutes around the clock; some competitors are email-only. If a parent can't connect mid-lesson, the answer to "how do I reach a human in 30 seconds?" matters more than the third tier of whiteboard polish.

The honest version of the decision: for pure whiteboard quality, the standalone whiteboards on this list — BitPaper for handwriting, Ziteboard for vector math, Lessonspace for subject-specific editors — each beat a whiteboard that's bundled into a larger product. For everything around the whiteboard (engagement, scheduling, invoicing, support, parent experience), a full virtual classroom like Koala Go pulls ahead. Most independent tutors care more about the second list than the first; some specialist tutors care more about the first. Pick accordingly.

A note on Google Jamboard

If you used to teach with Google Jamboard, that product was retired at the end of 2024 and the companion Jamboard devices reached end-of-life at the same time. Google's recommended successor is FigJam (via a partnership with Figma), though it isn't tutor-specific. Most former Jamboard users we've spoken to have moved to one of the tools above — usually Microsoft Whiteboard if they wanted the closest free equivalent, or a tutor-built tool (BitPaper, Lessonspace, or Koala Go) if they wanted features Jamboard never had.

What makes a whiteboard good for tutoring, specifically?

Generic collaboration whiteboards (Miro, Microsoft Whiteboard, Figma's FigJam) are powerful tools designed for adult professionals running design or strategy sessions. Tutoring puts different demands on a whiteboard:

Handwriting smoothness. Most tutoring still involves writing — letters, numbers, equations. Whiteboards optimised for clean ink with a stylus (BitPaper, Ziteboard, Lessonspace) feel meaningfully better than office-collaboration tools for this.
Live document annotation. Tutors don't usually draw from scratch — they mark up a worksheet, a textbook page, or a screenshot. Drop-in PDF / PowerPoint upload with annotation on top is the single most-used whiteboard workflow in tutoring.
Persistence across lessons. A student who returns next week should be able to see what you did last week. Whiteboards that save per-student "papers" or per-class history beat one-shot canvases.
Control over student edits. Tutors need to lock and unlock what students can change at different points in a lesson. Office whiteboards assume peers; tutoring assumes one driver and one passenger most of the time.
Engagement scaffolding. For young students, stickers, stars, drag-and-drop activities, and a sense of progress on the same canvas you're teaching from outperforms a sterile drawing surface, even if the drawing surface is more capable.

The whiteboards that win in tutoring tend to score highly on the last three of these; the office whiteboards that get used by tutors tend to score highly on the first two but miss the others entirely.

What a cobrowser actually is, under the hood

A cobrowser is not a screenshot, a video, or a recording. It is a real browser that runs in the cloud and streams its rendered output to both participants, with input (clicks, typing, scrolling) flowing back to the same shared instance. From the teacher's side and the student's side, it looks and behaves like a browser tab — because it is one. Both participants share the same live page, each with their own cursor: there's a single underlying webpage, so at any given instant one person is acting on a given element, but control flows naturally — one person finishes dragging something into place, the other clicks the next button — instead of everyone wrestling over a single shared cursor.

Three things follow from that design:

The session is shared, not your password. When you open a workbook, a curriculum site, or a worksheet platform inside the cobrowser, the student can work in that same session with you — without you ever handing over credentials or screensharing a login screen. Use it with sites your own plan or teacher license covers; the point is that the student gets to interact with the material in real time, not that you skip a seat you're meant to buy.
The page is real, not a picture. Forms work. Drag-and-drop works. Audio in the page works. Videos play. Sites that require interaction (Wordwall, Boom Learning, Quizlet Live, Khan Academy, YouTube embeds, Google Docs) all behave like they would in a normal tab.
Devices barely matter. Because the heavy lifting happens in the cloud, what each participant needs locally is a modern browser and a network connection. iPads, Android tablets, Chromebooks, and older Windows laptops all participate equally — the student is not running the site on their device, they're streaming it.

How screen sharing fails for tutoring (the do-the-thing-yourself problem)

Screen share is fine for showing slides to a board of adults. It struggles in tutoring for a specific, repeated reason: the student can't act on what they're seeing. A few concrete failure modes we hear from working tutors every week:

Interactive lesson sites become "watch me click." You open Wordwall, drag the correct answer to the box, and the student watches your cursor. The whole point of the activity was for the student to drag. The lesson reverts to passive consumption — exactly what Wordwall (or Boom Learning, or Quizlet) was built to avoid.
iPads break screen share comprehension. Zoom and Google Meet's screen share on a 9-inch iPad shrinks every text size down to unreadable. Pinch-to-zoom fights the host's scrolling. Most parents put their 7-year-old on the iPad they already own. The screen share that works on your 27-inch monitor doesn't survive that journey.
Remote control is fiddly and tab-specific. Zoom offers "Request Remote Control," which can hand a single application's cursor to a guest. It requires the guest to install Zoom (not the browser version), takes 4-5 clicks to set up, hands control of your whole desktop (not just one tab), and breaks every time you switch applications. A handful of tutors get it working; most give up and just narrate.
The student can't type. Asking "type your answer in the box" is the most natural ESL prompt in the world. On a screen share, the student types into chat and you copy it over. The friction is enough that most tutors stop bothering.

None of these are unfixable; tutors who use screen share solve them with workarounds (sending the student a direct Wordwall link to a parallel tab, narrating instead of asking, switching to verbal-only activities). But the workarounds compound: a 30-minute lesson with a young learner loses 3-5 minutes to "let me share my screen again" / "can you see this?" / "type it in chat" — minutes that aren't going to vocabulary.

What changes when you teach on a cobrowser instead

The student stops watching and starts acting. Practically, that means:

Interactive lesson sites become interactive again. You open the Wordwall set, the student drags the answer themselves. Boom Learning cards: the student taps the picture. Khan Academy practice problems: the student types into the answer box. Whatever the page was designed to do, the student does it.
Real curriculum websites become teachable. ESL tutors on Koala Go teach directly inside the published curricula they already use — Abridge Academy, Crystal Clear ESL, Flip the Classroom, Reading A-Z, Oxford Owl — with the student doing the activities in real time instead of watching the teacher click. Tutors of older students do the same with Khan Academy, Brilliant, Duolingo, or a standard homework portal. (Teach from sites your own plan or teacher license covers, and within each site's terms — the cobrowser changes who can interact with the page, not what you're licensed to use.)
Worksheets, PDFs, and live websites stop being three different workflows. Anything that lives on a web page — a YouTube clip, a Google Doc, an embedded form, a Padlet, a Jamboard replacement, a Wordwall, a parent-facing report you're filling in together — goes through the same surface.
Tablets become first-class. The student is not screen-sharing your desktop into their iPad; they're touching their own browser. Pinch-to-zoom does what pinch-to-zoom should. Tapping a button taps the button. Reading is readable.

The honest summary, from tutors who switched from Zoom: a cobrowser doesn't make a new kind of lesson possible — it makes the kind of lesson everyone already wanted to teach actually happen. Less narrating what you're doing, more "you try it."

Where screen sharing is still the right tool

We aren't going to pretend a cobrowser replaces screen sharing for everything. Stay on screen share if:

You're sharing a desktop app, not a website. A cobrowser only shares a browser. If you're teaching with a desktop installation of GarageBand, Final Cut, AutoCAD, or any locally-installed software, a regular screen share is the only option.
You only need a one-way display. A lecture-style explanation where the student is genuinely meant to watch, then ask questions — a screen share with a webcam overlay is the simpler tool. Cobrowsing is for activities, not lectures.
You're teaching adult learners who already have their own logins. If your IELTS student has their own paid Brilliant account, a parallel-tabs workflow plus screen share for shared moments is often enough.
You're walking through your own dashboard. Showing a parent their child's progress in your tutoring CRM is a one-way display problem. Screen share is fine.

Which online tutoring platforms have a cobrowser, and how they differ

"Cobrowser" here means: a shared browser where both participants can act on the same running page — not a screenshot, not a one-way remote control, not a static image embedded in a whiteboard. A handful of tutoring platforms ship one; they differ mostly in how control of the shared page is handed around. Descriptions below are written to what each vendor's own marketing page says at the time of writing — vendors do ship and remove features, so re-check their pages if a specific behavior matters to your evaluation.

Koala Go. Cobrowser is a built-in feature, available on every paid lesson and (with a 10-minute per-session limit) on the free tier. Every participant has their own cursor on the shared page — unlike a Zoom screen-share, where a single remote-control cursor gets passed around and people end up fighting over it. The teacher can give control to one student or to several, or take it back at any point. The session routes through regional infrastructure (North America / Europe / Asia, with separate routing for mainland China) for latency, and the cobrowser lives inside the same classroom as the whiteboard, 3D Playground, scheduling, and worldwide invoicing. There's a section below with more of the Koala-specific behavior.
Lessonspace. Ships a feature called Cobrowser (in beta as of mid-2026, per Lessonspace's own marketing). Multi-tab support, the session leader can lock the browser to keep students view-only, and otherwise everyone in the space can click, scroll, and open tabs in the shared window.
Vedamo. Ships Co-browser (Vedamo calls it "collaborative browsing") — an embedded browser that hosts and participants navigate together in real time. Vedamo's tutorial states the cobrowser can be used by one user at a time; the host can hand the controller role to a chosen participant rather than having multiple participants act simultaneously.
LearnCube, BitPaper, Ziteboard, Whereby. Purpose-built classrooms / whiteboards for tutoring, all with a whiteboard or "board sharing" feature, but no shared-browser feature we could find on their docs at the time of writing. Teachers on these platforms typically share a browser the same way they would on Zoom: a screen share, or a document-upload workflow that flattens the page into the whiteboard.
Zoom, Google Meet, Microsoft Teams. General videoconferencing. Screen share with optional remote control on Zoom; no shared-browser primitive on any of the three.

If you're trying to tell whether a "cobrowser" feature is a real shared browser or a fancier screen-share, two checks usually settle it: (1) can the student type into a form on the page and submit it, not just point a cursor; (2) is the student genuinely inside your live session — acting on the same running page — rather than watching a relayed picture of your screen. If both answers are yes, it's a real cobrowser. If either is no, it's a remote-control or relay setup.

What a cobrowser is not

Not a clone of your local browser. The cobrowser is a separate, cloud-hosted browser. Your bookmarks, your password manager, and your local extensions don't come with it. (For most tutoring workflows this is a feature, not a bug — you don't accidentally share personal logins.)
Not unlimited compute. Heavy 3D web games or sites that pin a CPU to 100% will be sluggish in any cobrowser, the same way they're sluggish in a low-end local browser.
Not a substitute for the whiteboard. For drawing, annotating worksheets, modeling letter formation, or working through an algebra problem, the whiteboard is the right tool. Cobrowsing is for sites; the whiteboard is for ink. Most full-classroom platforms (Koala Go included) put them side by side and you flip between them inside the same lesson.
Not a security or licensing shortcut. If you log in to an account inside a cobrowser, the student is in that session — so use a dedicated tutoring login rather than your personal one, and only on sites whose terms let you teach from your account this way. The cobrowser is for interacting with content together, not for stretching one seat across people a license doesn't cover.

Real classroom workflows tutors use a cobrowser for

From working tutors on Koala Go, the most common day-to-day uses:

Teaching inside a published curriculum. Open Abridge Academy, Crystal Clear ESL, or Flip the Classroom; the student does the activities directly. No "share your screen, can you see my cursor" loop.
Wordwall, Boom Learning, Quizlet, Blooket, Gimkit. Drag-and-drop and tap-to-answer activities that depend on the student physically acting on the page.
Reading practice on leveled reader sites. Oxford Owl, Reading A-Z, Epic. The teacher opens the book; the student reads aloud and uses the audio controls.
Math practice with worked examples. Khan Academy, IXL, ALEKS. The student attempts the problem; the teacher sees the same screen and can intervene on the same input boxes.
YouTube and video clips. Open a short Super Simple Songs clip or a TED-Ed; the student gets clean audio and can pause and rewind themselves.
Filling in shared documents together. A Google Doc, a worksheet, a parent intake form. Both people type into the same fields in real time.
Real-world tasks for older students. A business-English student practicing a job application; a teenager being walked through enrolling in a university course site; an adult learner navigating a real government form in English. The cobrowser turns "in real life" tasks into things you can rehearse inside a lesson.

Cobrowser support in Koala Go specifically

A few facts about how Koala Go's cobrowser behaves in practice, in case you're evaluating it:

Browser-only. Nothing to install for you or your student. The teacher opens the cobrowser from inside the classroom; the student joins by clicking the same room link they use for the rest of the lesson.
Region routing. The cobrowser session is routed through North America, Europe, or Asia, with separate handling for mainland China. As a tutor, you set your default region in settings once; for one-off sessions in another region, you can switch it in the lesson.
Per-student control, individual cursors. Each participant has their own cursor on the page, rather than fighting over one shared remote-control cursor the way a Zoom screen-share works. The teacher chooses whether each student starts with control or has to be handed it (a default you set in settings, overridable per student), and can grant or revoke it during the lesson — to more than one student at once. There's one underlying webpage, so at any given moment a single person is acting on a given element; in practice it flows turn by turn (one student finishes dragging a card into place, another clicks the next button), which feels far more natural than wrestling for one cursor.
Audio is piped through the lesson's audio device. A YouTube video playing in the cobrowser comes out of the same speaker setup you and the student are already using, with no separate routing.
The whiteboard sits next to it, not on top of it. When you want to annotate (circle a paragraph, draw an arrow, model a letter), the lesson canvas opens over the cobrowser. When you're done, both views go back to normal. You're not flipping between two windows.
Free vs. Pro. Koala Free includes cobrowser sessions with a 10-minute cap per session so you can try it; Koala Pro ($25.99/month monthly, $21.99/month billed annually) lifts the cap and includes the cobrowser as a standard feature with the rest of the classroom.

How to decide whether you need a cobrowser

A 30-second decision rule that holds up in practice:

Do you teach activities, or do you lecture? If your lessons are mostly you talking and the student listening, you don't need a cobrowser; a clean screen share is enough. If the student is supposed to do things during the lesson — drag, type, click, drop, drill — and they currently can't because the activity lives on a website only you can interact with, a cobrowser is the missing piece.
Do your students learn on tablets? If most of them are on iPads, Android tablets, or Chromebooks, screen share is the wrong primitive and you'll feel the friction every lesson. Cobrowsers are device-agnostic in a way screen shares aren't.
Do you depend on published or interactive web content? Wordwall, Boom Learning, Khan Academy, leveled reader sites, ESL curriculum portals — if any of those are core to your teaching, you'll get value out of a cobrowser within the first lesson.

If you answered "yes" to any of those, the right move is to run a real lesson on a cobrowser-equipped classroom and see what changes. You can open a free Koala Go room at classroom.teachwithkoala.com — no install, no card — and try it with a real student on the device they actually use. Questions about how it'd work for your specific subject, age range, or region: write to koala@teachwithkoala.com with a sentence about the kind of tutoring you do and we'll give you our honest read.

What Google Meet actually is

Google Meet is Google's video conferencing product, part of Google Workspace. It's free with any Google account, runs entirely in the browser, and is a familiar sight to almost every parent and student because it's what their school, their office, and their family group calls run on. The core shape is a video call: a grid of participant tiles, one shared screen at a time via screen share, a chat pane, live captions, and a hand-raise button.

Meet's job is to connect people who need to see and hear each other. It does that job well. What it wasn't built for is teaching a live lesson to a student who needs something to do rather than just watch. Google's built-in whiteboard (Jamboard) was retired at the end of 2024; Google's recommended successor for the shared-canvas use case is FigJam, added via a partnership with Figma — a separate tool with its own account, login, and interface. So the "Meet with a Jamboard tab open" tutoring setup a lot of teachers ran during 2020-2024 no longer exists in its old shape. If you want a canvas alongside Meet today, you'll add one from outside the Google stack (a separate whiteboard app, a Google Slides file that both people edit, or a Miro board in a second tab).

What's missing for online tutoring

Six gaps show up regularly, in roughly the order tutors notice them:

No shared interactive canvas built for teaching. The screen-share model — one person shares their screen, everyone else watches — collapses the moment you want a student to drag a letter into a box, type into a form, circle a paragraph, or place a piece on a number line. On a video call, only the host's mouse and keyboard touch the shared content; the student sees a picture of it and comments. In-person you'd hand the student the pen. On Meet you can't.
No activities the student can act on. Related but distinct: even if you bolt on a whiteboard, a video-call-plus-whiteboard workflow gives you a canvas, not a lesson library. Phonics matching, dragging fraction bars, sorting words into categories, snapping shapes together — the interactive activities working tutors of young kids build a lesson around are not in Meet, and Meet doesn't have a home for them.
No scheduling for paying clients. Google Calendar exists and is fine at what it does, but it isn't set up for parents to book paid tutoring slots against your open availability without a scheduling tool layered on top (Calendly, YouCanBook.me, or a purpose-built tutoring scheduler). A parent who wants to reschedule a Tuesday lesson to Wednesday has no self-serve path through Meet itself.
No invoicing or payment collection. Meet doesn't bill parents. You'll add Stripe, PayPal, Wise, or manual invoicing outside the platform — and reconcile which student paid for which package inside a spreadsheet somewhere. For tutors with international families this is the piece that costs the most time: currencies, WeChat Pay for China, refunds when a student cancels a package mid-way.
No engagement layer for young learners. A face-on-webcam interface is the wrong shape for a six-year-old on lesson eight of the same week. There are no stars, no gems, no avatars, no shared space for the student to be in as a character rather than as a face on a grid. Some tutors patch this with a rewards spreadsheet the parent can see, sticker printouts sent by mail, or a Class Dojo integration. All of those are workflow to run in parallel with Meet, not part of Meet.
Free tier caps that bite paid tutoring. The free Google Meet tier is generous for one-on-one calls but caps the length of group meetings (three or more participants), which is exactly the small-group workflow that tutors trying to move beyond one-on-one economics run into. Paid Google Workspace tiers lift the caps and add recording, breakout rooms, and longer meetings — but at that point you're paying for a Workspace subscription that is really designed around email, drive, and business collaboration, and the tutoring layer is still bolted on.

The typical "Google Meet + bolt-ons" tutoring stack

Working independent tutors who run their practice on Meet usually end up with something like this — and the total cost, both in dollars and in minutes per week of stack management, is often higher than a single purpose-built classroom:

Google Meet — the live video call
FigJam, Miro, Google Slides, or Whiteboard.fi — a shared canvas (Google Jamboard used to be the obvious choice; Google retired it end of 2024)
Calendly, YouCanBook.me, or Google Calendar — booking and rescheduling
Stripe, PayPal, or Wise — invoicing and getting paid across currencies
A separate rewards tracker — Class Dojo, a shared Google Sheet, or a physical sticker chart the family updates between lessons
A Drive folder — where the lesson materials, PDFs, worksheets, and past recordings live
A spreadsheet — to track which student paid for which package, how many lessons are left, and who owes a make-up

None of those are bad tools on their own. The problem is the seams between them: every time a tool updates, the integration breaks; every time you switch students mid-lesson, you need to switch four tabs; every time a parent asks "what did we cover last week and what's next?" you're pulling the answer from three places. That friction is tolerable at two or three students and becomes the shape of the workday somewhere around ten to fifteen.

When Google Meet is a reasonable fit for tutoring

You teach one-on-one adults or teenagers where the lesson is mostly talking (conversation practice, IELTS speaking prep, business English, university-level tutoring by discussion). A clean video call plus a shared document is often enough.
Cost is a hard constraint and you're testing the idea of taking on paid students before committing to a paid stack. Free Meet plus free Google Calendar plus a manual invoice will get you to your first three lessons.
Your students' families are already deep in the Google ecosystem — Gmail, Drive, Classroom, Meet — and asking them to sign up for another tool is a real friction point. Meet's zero-install familiarity is worth something you can feel in trial-to-paid conversion.
You are, or were, a K-12 classroom teacher using Google Workspace for Education professionally, and picking up a few tutoring students on the side. Reusing the tools you already know is the correct call at that stage.
You need a fallback for when your main teaching tool has a bad day. Every tutor should have one; free Meet plus a Google Calendar invite is the fallback most tutors we know keep in their back pocket.

When it isn't

You teach young learners (roughly four to ten) and the retention story your parents write in reviews is about engagement — "she never wants to log off" — rather than curriculum. Video calls give a young learner one thing to do (talk when unmuted) and one thing to look at (a face on a grid). It falls short of what an engagement-designed classroom gives them.
You teach reading, phonics, math manipulatives, or any subject where the student needs to act on the material — dragging letters, sorting shapes, placing counters on a number line. Screen share plus voice narration is not a substitute for the student doing the thing themselves.
You teach students on tablets or Chromebooks — which is most K-12 students. Meet's screen share on a nine-inch iPad shrinks every text size below readable and turns pinch-to-zoom into a wrestling match with the host's scrolling. See what is a cobrowser? for why the shared-browser primitive fixes this and screen share doesn't.
You teach students in mainland China. Google Meet — like Google Search, Gmail, and the rest of the Google Workspace stack — is not reachable from mainland China without a VPN, which many families would rather not install on a child's device. See teaching English to students in mainland China for the connectivity workaround pattern.
You want a single login that handles video, the lesson canvas, activities, scheduling, and invoicing — and you're past the "am I really doing this?" stage of building a tutoring business. The stack collapse is what a purpose-built virtual classroom is for.
You run recurring small-group classes (three to eight students) and the free-tier group meeting length cap is starting to interrupt lessons. Paid Google Workspace lifts the cap, but the tutoring layer is still absent — a purpose-built virtual classroom gives you the meeting and the classroom together.

What about recording on Meet?

Google Meet does record — but recording is a paid Google Workspace feature, not a consumer-tier one. Recorded meetings save into the host's Google Drive, and the sharing and retention model is Drive's rather than a purpose-built tutoring one. If you rely on lesson recordings for parent updates, make-up sessions for kids who missed a class, or your own self-review, that's real friction: recordings live in a folder alongside your work docs, filenames default to timestamps, and the family-sharing story is whatever Drive's default sharing does. Purpose-built tutoring classrooms tend to ship recording as a first-class feature with a per-lesson home. See how do I record online tutoring lessons? for the broader workflow, including the consent step you should be walking parents through regardless of which tool you record in — the safe move is to get explicit parental consent in writing, check what your local rules require (recording law varies by jurisdiction), and only save recordings for as long as you actually need them.

Purpose-built alternatives — what tutors typically switch to

Tutors who outgrow the Meet-plus-bolt-ons stack usually move to a virtual classroom designed for live tutoring from the start. The category is described in depth on what is a virtual classroom?; the head-to-head shortlist (with tradeoffs, subject fits, and free-tier caps) lives on the best Zoom alternatives for online tutoring, which covers Meet as one option among six and is worth reading if you're actively picking a replacement.

Where Koala Go (the classroom we build) fits the picture, specifically:

An interactive whiteboard with PDF, PowerPoint, and image upload. The file becomes the lesson surface — the student annotates the same page you're teaching from. This is the piece a Meet screen share can't do because a screen share is a picture the student can't act on.
A cobrowser (a live shared browser window). You and the student open the same real webpage together, both with your own cursor, both able to click and type. Wordwall, Boom Learning, Khan Academy, IXL, a leveled-reader site — the student does the activity instead of watching you click. Koala Free ships a limited cobrowser (session-length capped); Koala Pro removes the cap.
A 3D Playground for younger students. A shared 3D world where each student has a customizable avatar. Used as a brain break, a themed lesson surface, or the reason a seven-year-old wants to come back next week. This is the piece a face-on-webcam interface can't do because a webcam grid gives the child nothing to be in.
Worldwide invoicing built in. Parents pay in their local currency, including via WeChat in China. There's a 4.5% Koala fee on Free-tier invoicing that drops to 3.5% on Pro. No separate Stripe reconciliation.
Integrated scheduling on Pro. A teacher-and-parent scheduling portal where you set weekly availability once and parents book their own slots via a link. No Calendly bolt-on.
Lesson recording on Pro and Platinum. Koala Pro records the lesson locally to your own machine; Koala Platinum adds cloud recordings that save inside the classroom, per-lesson, so a family whose kid missed a session can catch up without derailing next week.
24/7 live support on Pro with a real person. When something breaks mid-lesson, there's a human to reach rather than a help-centre article.

Koala Free is a genuine free tier (up to 4 students per session, worldwide invoicing, limited cobrowser, up to 3 activities), not a trial that expires — so you can run a real trial lesson end-to-end with a real student on the device they actually use, and decide from there. Koala Pro is $25.99/month billed monthly or $21.99/month billed annually. The difference isn't really the price; it's what you get for it — a tool built for teaching, not a tool built for meetings with teaching bolted on.

Honest limits worth naming so you can weigh them against the fit: iPad works but is rougher than desktop; mainland-China connectivity is served by our Hong Kong proxy but is a moving target; the whiteboard ships math tools on the paid tiers — the symbol palette's operations, comparisons, and fraction slash on Pro, with the powers, the advanced symbols, and the drag-and-drop fraction bars on Platinum — but it stops short of a full graphing or equation editor, so heavy STEM tutors should test the workflow on a real lesson first; and the Playground is actively developed with requested features (timers, group-select of Playground objects) still on the roadmap. Roughly 81% of the tutors on our 2026 product/market fit survey said they'd be very disappointed if they could no longer use Koala Go — the "very disappointed" threshold conventionally taken to indicate strong product/market fit is 40%, so that number is well past the bar. That 81% figure is the same one we publish on the main landing page.

Bottom line

Google Meet is a competent video conferencing tool and a workable free fallback for one-on-one tutoring with older students or adults. It's the wrong shape for teaching young learners a live, interactive lesson — and the right answer there is not "Meet plus five other apps" but a purpose-built tutoring classroom that collapses the stack into one login. If your students are under ten, if they're on tablets, if they need to do things during the lesson rather than watch them, or if you're spending more than a couple of hours a week glueing your tool stack together, the switch usually pays for itself inside the first month. If you'd like our honest read on whether it fits your specific practice — subject, age range, region, current tool stack — write to koala@teachwithkoala.com with a sentence about the kind of tutoring you do. We're a small team and we read everything.

What actually happened to Skype

Microsoft announced Skype's retirement on February 28, 2025 and shut down the consumer service on May 5, 2025. The retirement affects both free and paid consumer Skype accounts — Skype Credit, Skype Numbers, all of it — and does not affect Skype for Business (a separate enterprise product on its own end-of-life track). From May 5 onward the Skype app itself has stopped working; opening it just shows an error or nothing.

Microsoft's official migration path is Microsoft Teams Free. If you were an active Skype user, you can sign into Teams with your existing Skype credentials — your chats and contacts move across automatically. The other route — exporting your history through the Skype Export Portal — has closed: Microsoft began deleting Skype data on April 1, 2026, and the portal stopped accepting export requests on June 15, 2026. There is no revival plan; consumer Skype is gone.

Why Skype worked for tutors (and what it always left you doing manually)

A generation of online tutors ran their whole practice on Skype. It was one of the first tools with cheap, decent-quality voice and video calls that worked across borders — long before Zoom made the same thing normal. Language teachers, ESL specialists, and independent 1-on-1 tutors on every continent had a Skype workflow of some kind, and many still had a "skype:username" line on their booking page when the retirement announcement landed.

What Skype was good at: a stable voice/video call, a text chat alongside it for typing corrections or pasting links, and a file drop for sending a PDF worksheet mid-lesson. What it never did: a shared canvas where the student could draw, a lesson library, scheduling for paying clients, invoicing that handled currencies and refunds, or any concept of "engagement" for younger learners. Tutors filled those gaps outside the app — a whiteboard tool, a scheduler, a payment tool, a rewards spreadsheet — and moved between four or five tabs during every lesson. That friction was familiar; it wasn't fine, but it was the ambient cost of running an online tutoring business in 2015, 2018, or 2022.

What Microsoft Teams Free actually gives you

Teams Free is genuinely fuller-featured than Skype was. On top of the video calls, chat, and file sharing you're used to, Microsoft's retirement post highlights meeting hosting, calendar integrations, and communities for keeping in touch with a group. If you migrate straight across from Skype, that's a net upgrade for the pieces of your workflow that were Skype-shaped: chats and contacts come with you, and you get a slightly richer communications hub sitting on top.

What Teams Free doesn't add is anything specific to teaching. Microsoft's own framing is that Teams is a "communications and collaboration hub" — for work, for family, for community groups. That's a legitimate positioning; it's just not a virtual classroom. The gaps that made the Skype-plus-bolt-ons tutoring stack painful are all still gaps here:

No lesson-surface canvas. A meeting whiteboard is a rectangle everyone can draw on. That's a useful primitive for brainstorming; it's not the same thing as a lesson surface where a student drags fraction bars onto a number line, sorts letters into phonics categories, or works through the specific worksheet you uploaded for today's session.
No activities the student can act on. Related but distinct: even if you bolt on a whiteboard, a video-call-plus-whiteboard workflow gives you a canvas, not a lesson library. Phonics matching, dragging fraction bars, sorting words into categories, snapping shapes together — the interactive activities working tutors of young kids build a lesson around are not in Teams, and Teams doesn't have a home for them.
No parent-facing scheduling for paid lessons. Teams can schedule meetings on a linked calendar, but not the parent-facing "book this Tuesday at 4 PM against my open availability" flow a tutoring business runs on. You'll add Calendly, a dedicated tutoring scheduler, or a shared Google Calendar with a manual back-and-forth.
No invoicing or payments. Parents pay outside the app. If your students are in different currencies — or in mainland China — you're bolting on Stripe, PayPal, Wise, WeChat Pay, or a manual invoice, and reconciling which student paid for which package inside a spreadsheet somewhere.
No engagement layer for young learners. Face-on-webcam is still face-on-webcam. There are no stars, no gems, no avatars, no shared space for the student to be in as a character rather than as a face on a grid. Some tutors patch this with a rewards spreadsheet the parent can see, sticker printouts sent by mail, or a Class Dojo integration. All of those are workflow to run in parallel with Teams, not part of Teams.
Optimised for the Microsoft ecosystem. Teams is at its best when everyone in the call has a Microsoft or Outlook account and works inside the broader Microsoft 365 stack. Some tutoring families do; many don't — and where they don't, a parent may need to set up a Microsoft account for their child, which is a step a Skype ID didn't require.

The typical "Teams Free + bolt-ons" tutoring stack

If you ran a tutoring practice on Skype, the migration path Microsoft is steering you to is Teams Free plus roughly the same bolt-ons Skype needed — same shape, different app:

Microsoft Teams Free — the live video call, chat, and file sharing
Miro, Whiteboard.fi, or a Google Slides file — a shared canvas
Calendly, Google Calendar, or a dedicated tutoring scheduler — booking and rescheduling
Stripe, PayPal, Wise, or WeChat Pay — invoicing and getting paid across currencies
Class Dojo, a Google Sheet, or a printed sticker chart — rewards for younger learners
A OneDrive or Drive folder — where lesson materials and any recordings live
A spreadsheet — to track which student paid for which package, how many lessons are left, and who owes a make-up

None of that is broken. It's the same friction you were living with before the retirement, in a different app. For tutors with a small steady book of older students, that friction is tolerable. For tutors teaching young learners on shared screens — where every "can you see this?" tax compounds — it adds up fast.

When Teams Free is a reasonable fit for tutoring

You teach one-on-one adults or teenagers where the lesson is mostly talking — conversation practice, business English, IELTS speaking prep, university-level tutoring by discussion. A clean video call plus a shared document is often enough.
You already had your Skype practice running smoothly and you want to move the least possible amount of your workflow; the automatic import of chats and contacts is worth something concrete.
Your students' families are already deep in the Microsoft ecosystem — a school with Teams for Education, a workplace on Microsoft 365, or a parent whose whole calendar lives in Outlook. Meet them where they already are.
Cost is a hard constraint and you're testing whether you want to keep tutoring at all before committing to a paid stack.
You need a fallback for when your main teaching tool has a bad day. Every tutor should have one; free Teams plus a Google Calendar invite is a reasonable one to keep in your back pocket.

When it isn't

You teach young learners (roughly four to ten) and the retention story your parents write in reviews is about engagement — "she never wants to log off" — rather than curriculum. Video calls give a young learner one thing to do (talk when unmuted) and one thing to look at (a face on a grid). It falls short of what an engagement-designed classroom gives them.
You teach reading, phonics, math manipulatives, or any subject where the student needs to act on the material — dragging letters, sorting shapes, placing counters on a number line. Screen share plus voice narration is not a substitute for the student doing the thing themselves.
You teach students on tablets or Chromebooks. Screen share on a small tablet screen shrinks text below readable size and turns pinch-to-zoom into a wrestling match with the host's scrolling. See what is a cobrowser? for why the shared-browser primitive fixes this and screen share doesn't.
You teach students in mainland China. Cross-border connectivity through Western consumer video tools is a moving target inside the Great Firewall, and even when a tool nominally works, tutors report drops and audio hiccups on longer sessions. See teaching English to students in mainland China for the connectivity workaround pattern.
You relied on Skype's recording for parent updates, make-up sessions for kids who missed a class, or your own self-review. Whatever you were doing on Skype for recording, you'll need to re-set-up on the new stack — see how do I record online tutoring lessons? for the workflow, including the consent step you should be walking parents through regardless of which tool you record in.
You want a single login that handles video, the lesson canvas, activities, scheduling, and invoicing — and you're past the "am I really doing this?" stage of building a tutoring business. The stack collapse is what a purpose-built virtual classroom is for.

Purpose-built alternatives — what tutors typically switch to

The bigger reason a lot of long-time Skype tutors won't be happy on Teams Free is that the "video call plus a stack of other apps" shape has been getting old for a decade. The category that grew up around it — purpose-built virtual classrooms for online tutoring — is described in depth on what is a virtual classroom?, with a head-to-head shortlist of options on the best Zoom alternatives for online tutoring. Both are worth reading if you're actively picking a Skype replacement rather than just accepting the default Teams path.

Where Koala Go (the classroom we build) fits, specifically:

An interactive whiteboard with PDF, PowerPoint, and image upload. The file becomes the lesson surface — the student annotates the same page you're teaching from. This is the piece a Teams screen share can't do because a screen share is a picture the student can't act on.
A cobrowser (a live shared browser window). You and the student open the same real webpage together, both with your own cursor, both able to click and type. Wordwall, Boom Learning, Khan Academy, IXL, a leveled-reader site, a YouTube video for the ones that aren't blocked in your student's region — the student does the activity instead of watching you click. Koala Free ships a limited cobrowser (session-length capped); Koala Pro removes the cap. The cobrowser is region-aware, with separate routing for mainland China so students there aren't fighting the Great Firewall inside a Western tool.
A 3D Playground for younger students. A shared 3D world where each student has a customisable avatar. Used as a brain break, a themed lesson surface, or the reason a seven-year-old wants to come back next week. Face-on-webcam gives a kid nothing to be in; the Playground gives them a place.
Worldwide invoicing built in. Parents pay in their local currency, including via WeChat in China. There's a 4.5% Koala fee on Free-tier invoicing that drops to 3.5% on Pro and 0% on Platinum. No separate Stripe reconciliation.
Integrated scheduling on Pro. A teacher-and-parent scheduling portal where you set weekly availability once and parents book their own slots via a link. No Calendly bolt-on. Platinum adds a Scheduling Plus layer with 30/45/60-minute class lengths, booking-policy controls, and recurring classes.
Cloud lesson recording on Platinum. Recordings save inside the classroom, per-lesson, so a family whose kid missed a session can catch up without derailing next week.
24/7 live support on Pro with a real person. When something breaks mid-lesson, there's a human to reach rather than a help-centre article.

Koala Free is a genuine free tier (up to 4 students per session, worldwide invoicing, limited cobrowser, up to 3 activities), not a trial that expires — so you can run a real trial lesson end-to-end with a real student on the device they actually use, and decide from there. Koala Pro is $25.99/month billed monthly or $21.99/month billed annually; Koala Platinum, which adds cloud recording and richer scheduling, is $49.99/month billed monthly or $39.99/month billed annually. The difference from Skype-shaped tools isn't really the price; it's what you get for it — a tool built for teaching, not a communication tool with teaching bolted on.

Honest limits worth naming so you can weigh them against the fit: iPad works but is rougher than desktop; mainland-China connectivity is served by our Hong Kong proxy but is a moving target; the whiteboard's math tools ship in Pro (the symbol palette's operations, comparisons, and fraction slash) and Platinum (the powers, the advanced symbols, and the drag-and-drop fraction bars) but stop short of a full graphing or equation editor, so heavy STEM tutors should test the workflow on a real lesson first; and the Playground is actively developed with requested features (group-select of Playground objects) still on the roadmap. Roughly 81% of the tutors on our 2026 product/market fit survey said they'd be very disappointed if they could no longer use Koala Go — the "very disappointed" threshold conventionally taken to indicate strong product/market fit is 40%, so that number is well past the bar. That 81% figure is the same one we publish on the main landing page.

Bottom line

Skype is gone. Teams Free is the direct successor Microsoft is steering you toward, and it works fine if what you needed was Skype-with-more-features. If what you actually need is a virtual classroom — with a shared canvas, activities the student can act on, scheduling and invoicing for the business side, and an engagement layer for young learners — the Skype retirement is a good moment to stop stacking apps and pick a purpose-built tool. Skype's own export window closed on June 15, 2026, so if you kept any record of who you taught and when inside Skype itself, rebuild it in whatever you pick next rather than assuming you can go back for it. If you'd like our honest read on whether Koala Go fits your specific practice — subject, age range, region, current stack — write to koala@teachwithkoala.com with a sentence about the kind of tutoring you do. We're a small team and we read everything.

Where WhatsApp actually fits in most tutors' workflows

Across the tutors we hear from — a lot of them ESL, a lot of them working with families in India, Latin America, the Middle East, and Southeast Asia — WhatsApp is almost always in the stack somewhere. Parents open it before they open email; it's where the "can we push tomorrow's lesson by 15 minutes?" message lands; it's what they forward the meeting link into so their kid can tap it on their tablet. If your students' families use WhatsApp, you use WhatsApp. Trying to drag every family onto email or a proprietary parent portal is a losing fight.

The distinction this page is about is a different one: WhatsApp as your family-communication channel (great, keep it) versus WhatsApp as your actual teaching platform (the video call is where the lesson happens, the chat is where you paste homework, the media tab is where worksheets live). The second one is the setup a lot of new tutors start with, especially in markets where families don't already pay for a purpose-built classroom. It works — until the practice grows enough that its limits start costing you time and enrollments.

What WhatsApp itself gives you

Meta describes WhatsApp as a personal messaging app for chat, voice calls, video calls, and file sharing, end-to-end encrypted by default. The WhatsApp Business app is a separate download aimed at small businesses — it adds a business profile, an away-message auto-reply, quick canned replies, and a chat-label system, but the underlying calling and messaging is the same product. Both apps use your phone number as your identity. For a solo tutor, the pieces you'd actually use in a lesson are:

Voice and video calls. One-to-one calls are the workhorse; small-group calls exist but the participant cap is meant for informal group chats, not classroom-shaped groups.
Text chat alongside the call. Type a correction, paste a link, drop a definition.
File and media sharing. Send a PDF worksheet, a photo of a page, a voice note homework prompt. The file lands in the chat and both sides can download it.
Screen sharing during a call. Available on WhatsApp Desktop and on mobile; you share what's on your screen, the student watches.
End-to-end encryption on calls and chats. A real, marketed feature of the app.
WhatsApp Business extras (in the Business app). Business profile, greeting/away messages, quick replies, labels for organising chats.

That is a legitimately useful communications toolkit. It just isn't a teaching toolkit. The rest of this page is where that difference bites for tutoring specifically.

Where a WhatsApp-as-classroom stack breaks for tutoring

No lesson surface the student can act on. Screen share and video are things the student watches. There's no shared canvas where the student drags a fraction bar onto a number line, sorts letters into phonics categories, annotates the worksheet you uploaded, or works a problem while you watch the pen. For younger learners and for any subject where doing beats watching, this is the biggest gap.
No activity library. Phonics matching, dragging fraction bars, sorting words into categories, snapping shapes together — the interactive activities working tutors of young kids build a lesson around are not something WhatsApp has a home for. You end up screen-sharing another tool over the call, which introduces the next problem down the list.
Screen share isn't interactive. If you share the browser tab where a student practice site lives, the student sees your cursor moving; they can't click, drag, or type into the page themselves. On a phone or a small tablet the readable-text problem compounds — you're squeezing a laptop-sized site into a five-inch screen and the student has to fight your scrolling. See what is a cobrowser? for the specific mechanic that fixes this.
Phone-number-as-identity mixes your personal and business life. Your students see the number that's also on your family group chats. Every "quick question about tomorrow's lesson" lands in the same app as everything else you get pinged about at 10 PM. WhatsApp Business's separate download and dedicated business number soften this, but you're still routing your practice through a phone identity — not an account you can hand off or share visibility on.
Only the parent should be the account holder for a minor. WhatsApp's own account policy expects an adult account holder; a young student should never be on your practice as their own contact. In practice that means every lesson runs through the parent's phone, which is fine for booking and messaging but awkward when the parent isn't the one meant to be looking at the screen for 45 minutes. If you're moving past casual tutoring, this is a signal to keep the messaging on WhatsApp but move the lesson itself onto a tool the student opens in their browser.
No booking flow parents can use themselves. Every reschedule, every "can we move to Thursday?", every new package is a manual back-and-forth in the chat. It works while you have five students; it does not scale to twenty. You'll bolt on Calendly, a Google Calendar link, or a dedicated tutoring scheduler.
No invoicing or payments. Parents pay outside the app. Cross-border tutors end up on Wise, PayPal, Payoneer, Stripe (where available), a local bank transfer, or a mix — and reconciling which student paid for which package inside a spreadsheet is your job.
No lesson recording. WhatsApp calls are end-to-end encrypted and the app doesn't ship a "Record call" button. If you want a lesson recording — for a family whose kid missed a session, or for your own review — you're running a separate screen recorder (OBS, QuickTime, a browser extension) alongside the call. See how do I record online tutoring lessons? for the workflow, including the consent step you should be walking parents through regardless of which tool you record in.
No engagement layer for young learners. Face-on-webcam is face-on-webcam. There are no stars, no gems, no avatars, no shared space for the student to be in as a character rather than as a thumbnail on a parent's phone. Some tutors patch this with a sticker chart the parent tracks or a rewards spreadsheet — that's workflow to run in parallel with WhatsApp, not part of it.
WhatsApp is blocked in mainland China. If any of your students are inside mainland China, WhatsApp is not reachable there without a VPN, and using a VPN as a student is increasingly unreliable — it slows the connection, drops mid-call, and is something many families would rather not install on a child's device. See teaching English to students in mainland China for the connectivity workaround pattern.

The typical "WhatsApp + bolt-ons" tutoring stack

Tutors who run their whole practice on WhatsApp usually end up assembling a stack that looks roughly like this:

WhatsApp (personal or Business) — video call, chat, file drop, screen share
Miro, Whiteboard.fi, Jamboard alternative, or a Google Slides file — a shared canvas the student can (in some cases) draw on
A shared Google Drive or OneDrive folder — where worksheets and any lesson recordings live
Calendly, a Google Calendar link, or a dedicated tutoring scheduler — booking and rescheduling that doesn't cost you a chat every time
Wise, PayPal, Payoneer, Stripe, or a local bank transfer — getting paid across currencies
OBS, QuickTime, or a browser extension — if you record lessons at all
A sticker chart, a Google Sheet, or Class Dojo — rewards for younger learners
A spreadsheet — to track which student paid for which package, how many lessons are left, and who owes a make-up

None of that is broken. It's just a lot of moving parts held together in your head. For tutors with a small steady book of older students, that friction is tolerable — many practices run this way for years. For tutors teaching young learners on shared screens, or for anyone whose practice is growing past ten or fifteen weekly students, the tab-switching tax compounds every lesson.

When WhatsApp-as-classroom is a reasonable fit

You teach one-on-one adults or teenagers and the lesson is mostly talking — conversation practice, business English, IELTS or Cambridge speaking prep, university-level tutoring by discussion. A clean video call plus a shared PDF over chat is often enough.
You're running a first paid lesson with a family that already uses WhatsApp for everything, and asking them to install a new tool before they've decided to keep working with you is the wrong ask. Meeting a family where they already are is a real, honest starting move.
You're testing whether you want to keep tutoring at all before committing to a paid tools stack.
Your students' families are on unreliable connections and the tool that survives the worst 10% of network conditions matters more than the feature list of the tool that shines on the best 90%. WhatsApp calls handle bad networks unusually well.
You need a fallback for when your main teaching tool has a bad day. Every tutor should have one — a WhatsApp video call plus a PDF the parent already has is a reasonable one to keep in your back pocket.

When it isn't

You teach young learners (roughly four to ten) and the retention story your parents write in reviews is about engagement — "she never wants to log off" — rather than curriculum. Video calls give a young learner one thing to do (talk when unmuted) and one thing to look at (a face on a screen). It falls short of what an engagement-designed classroom gives them.
You teach reading, phonics, math manipulatives, or any subject where the student needs to act on the material — dragging letters, sorting shapes, placing counters on a number line. Screen share plus voice narration is not a substitute for the student doing the thing themselves.
You teach students on tablets or Chromebooks. Screen share on a small tablet shrinks text below readable size and turns pinch-to-zoom into a wrestling match with the host's scrolling. A cobrowser fixes this; a WhatsApp screen share doesn't.
You teach students inside mainland China. WhatsApp isn't reachable there without a VPN.
You're growing past the number of students you can hold in your head. Ten students booking and rescheduling in your inbox is fine; twenty-five isn't, and the fix isn't a bigger inbox — it's a scheduler with a policy attached and an invoicing tool that reconciles itself.
You want a single login that handles video, the lesson canvas, activities the student can act on, scheduling, and invoicing — and you're past the "am I really doing this?" stage of building a tutoring business. Collapsing the stack is what a purpose-built virtual classroom is for.

Purpose-built alternatives — what tutors typically switch to

The bigger reason a lot of long-time WhatsApp tutors eventually stop teaching there is that the "video call plus a stack of other apps" shape stops paying its rent. The category that grew up around it — purpose-built virtual classrooms for online tutoring — is described in depth on what is a virtual classroom?, with a head-to-head shortlist of options on the best Zoom alternatives for online tutoring. Both are worth reading if you're actively picking a WhatsApp replacement rather than just accepting the "well, this is what the family already uses" default.

Where Koala Go (the classroom we build) fits, specifically:

An interactive whiteboard with PDF, PowerPoint, and image upload. The file becomes the lesson surface — the student annotates the same page you're teaching from. This is the piece a WhatsApp screen share can't do because a screen share is a picture the student can't act on.
A cobrowser (a live shared browser window). You and the student open the same real webpage together, both with your own cursor, both able to click and type. Wordwall, Boom Learning, Khan Academy, IXL, a leveled-reader site, a YouTube video for the ones that aren't blocked in your student's region — the student does the activity instead of watching you click. Koala Free ships a limited cobrowser (session-length capped); Koala Pro removes the cap. The cobrowser is region-aware, with separate routing for mainland China so students there aren't fighting the Great Firewall inside a Western tool.
A 3D Playground for younger students. A shared 3D world where each student has a customisable avatar. Used as a brain break, a themed lesson surface, or the reason a seven-year-old wants to come back next week. A phone-screen video call gives a kid nothing to be in; the Playground gives them a place.
Worldwide invoicing built in. Parents pay in their local currency, including via WeChat Pay in mainland China. There's a 4.5% Koala fee on Free-tier invoicing that drops to 3.5% on Pro and 0% on Platinum. No separate Wise or Stripe reconciliation.
Integrated scheduling on Pro. A teacher-and-parent scheduling portal where you set weekly availability once and parents book their own slots via a link. No Calendly bolt-on. Platinum adds a Scheduling Plus layer with 30/45/60-minute class lengths, booking-policy controls, and recurring classes — the "book all four Wednesdays for the month at once" flow that a WhatsApp chat can't hold.
Lesson recording on Pro and Platinum. Koala Pro records the lesson locally to your own machine; Koala Platinum adds cloud recordings that save inside the classroom, per-lesson, so a family whose kid missed a session can catch up without derailing next week.
24/7 live support on Pro with a real person. When something breaks mid-lesson, there's a human to reach rather than a help-centre article.
Runs in the browser. The student and parent open a link — nothing to install, no separate app account, no phone-number-linked identity. The link works on the device the family already uses.

Koala Free is a genuine free tier (up to 4 students per session, worldwide invoicing, limited cobrowser, up to 3 activities), not a trial that expires — so you can run a real trial lesson end-to-end with a real student, on the device they actually use, and decide from there. Koala Pro is $25.99/month billed monthly or $21.99/month billed annually; Koala Platinum, which adds cloud recording and richer scheduling, is $49.99/month billed monthly or $39.99/month billed annually. The difference from WhatsApp isn't really the price; it's what you get for it — a tool built for teaching, not a communication tool with teaching bolted on.

Honest limits worth naming so you can weigh them against the fit: iPad works but is rougher than desktop; mainland-China connectivity is served by our Hong Kong proxy but is a moving target; the whiteboard's math tools ship in Pro (the symbol palette's operations, comparisons, and fraction slash) and Platinum (the powers, the advanced symbols, and the drag-and-drop fraction bars) but stop short of a full graphing or equation editor, so heavy STEM tutors should test the workflow on a real lesson first; and the Playground is actively developed with requested features (group-select of Playground objects) still on the roadmap. Roughly 81% of the tutors on our 2026 product/market fit survey said they'd be very disappointed if they could no longer use Koala Go — the "very disappointed" threshold conventionally taken to indicate strong product/market fit is 40%, so that number is well past the bar. That 81% figure is the same one we publish on the main landing page.

Keep WhatsApp for what it's good at

None of the above is an argument to remove WhatsApp from your practice. It's an argument to use it where it earns its keep: as the single channel per family for booking messages, quick reminders, homework-check-ins, and the "she's running late" message on the day. Send one link a parent can bookmark for every lesson (your classroom room link), and let WhatsApp be the place you notify them when something changes. That is the shape that both survives and scales — one link for the lesson, one channel for the messaging, both pointed at people who need to actually reach you. See how do I onboard a new online tutoring student? for the "one channel per family" onboarding discipline this rests on.

Bottom line

WhatsApp is where a lot of tutoring families already live, and the discipline of meeting them there for messaging is worth keeping. It works as a lesson-delivery tool for a specific slice — one-on-one adults or teenagers, mostly-talking lessons, first-lesson stopgaps, and reliable-fallback duty. For anyone teaching young learners, teaching subjects the student needs to act on, growing past a hand-held book of students, or serving families in mainland China, the honest answer is that WhatsApp is where you talk to the parent — not where the lesson itself should live. If you'd like our read on whether Koala Go fits your specific practice — subject, age range, region, current stack — write to koala@teachwithkoala.com with a sentence about the kind of tutoring you do. We're a small team and we read everything.

Where Discord actually fits in tutoring today

Discord is genuinely popular with a specific slice of tutors, and if you're one of them we're not here to tell you to switch. The clearest fits:

Coding, CS, and software tutoring for teens and adults. Your students already have Discord open — most bootcamps, open-source projects, and university CS departments run servers. The workflow of "hop into a voice channel, share your screen, look at each other's code together" is Discord-native, and text channels are a natural home for code snippets, error messages, and follow-up questions between sessions.
Music theory, production, and audio work. Discord's voice channels handle audio well enough for the "listen to this progression, tell me what you hear" workflow, and screen share on a DAW works. Producer and music-teacher communities are among the largest categories on Discord.
TTRPG game-mastering and narrative-heavy adjacent work. A Dungeons & Dragons group runs beautifully on Discord (voice channel plus a bot for dice), and some tutors run narrative history / creative-writing / roleplay-driven lessons in the same shape.
Adult subject cohorts — languages, exam prep, university-level tutoring. If your students are old enough to have a Discord account of their own, comfortable in the interface, and mainly needing conversation, discussion, or code/text work, a Discord server can be a legitimate primary classroom.
Community layer around another primary teaching tool. A lot of tutors keep Discord as their "between-lessons" channel — homework check-ins, study-group voice hangouts, a shared library of past resources — while running the live lesson somewhere else. That combined stack is a real, working setup.

What Discord actually is (as a category)

Discord describes itself as "your place to talk" — a communication platform organised around servers (a community you host or join), channels inside those servers (text, voice, forum, or stage), and direct messages between users. Voice channels are the piece most tutors build lessons on: you drop into a channel, other members follow, and you get shared audio, per-member video toggles, and screen share while you're there. Text channels sit alongside for typed conversation and file drops.

Discord is free with an email or phone-number sign-up. A paid tier (Discord Nitro) raises some quality caps and adds cosmetic and community-management extras; Discord's monetisation for creators lives at the server level (subscriptions and shops), not at the per-user Nitro tier. Discord's Terms of Service require account holders to be at least 13 years old, and higher in some regions — this matters for tutoring specifically because it means a K-8 student should not have their own Discord account. The workaround for that age group is either "a parent is on the call with them" or "don't use Discord for that student", not a workaround inside Discord.

Where a Discord-as-classroom stack breaks for tutoring

Discord is a legitimately useful communications toolkit. It just isn't a teaching toolkit, and the gaps show up in the same places every time:

No lesson surface the student can act on. Screen share and video are things the student watches. There is no shared canvas where the student drags a fraction bar onto a number line, sorts letters into phonics categories, annotates the worksheet you uploaded, or works a problem while you watch the pen. Voice channels are for talking; a "screen share plus a mic" workflow gives you a video call in a Discord skin.
No activity library. Related but distinct: even if you screen-share another tool over Discord, the interactive activities working tutors of young kids build a lesson around are not in Discord, and Discord doesn't have a home for them.
Screen share isn't interactive. The student sees your screen; they can't click, drag, or type into the page you're showing. On a tablet the readable-text problem compounds. See what is a cobrowser? for the specific mechanic that fixes this and screen share doesn't.
Servers and channels are the wrong primitive for a 1-on-1 paid lesson. Discord's architecture is optimised for community — a room that stays open, that other members can drop into, that runs asynchronous conversation alongside synchronous voice. A paid 1-on-1 tutoring lesson is a private, scheduled, bounded event with two people in it. You can force this shape onto Discord (a private server per family, a DM voice call) but you're working against the app's grain the whole way.
Account age policy means young students can't have their own account. Discord's own terms require users to be at least 13 (and older in some regions). If your students are K-8, the parent's account has to be the one on the call, which is fine for a check-in and awkward for a 45-minute lesson.
No parent-facing scheduling for paid lessons. Discord has server events, which is fine for community-wide "we're all meeting at 7 PM Thursday" cadence. It isn't a parent-facing "book my Tuesday 4 PM slot against your open availability" flow that a growing tutoring business runs on. You'll bolt on Calendly, a dedicated tutoring scheduler, or a shared calendar with a manual back-and-forth in DMs.
No invoicing or payments for individual tutoring lessons. Discord does have server-level monetisation (Server Subscriptions, Shops) aimed at creators charging for community access, but that's a subscription-to-a-community shape, not a per-lesson-paid-invoice shape. Cross-border tutors handling different currencies still end up on Stripe, PayPal, Wise, or a local bank transfer alongside Discord, and reconciling which student paid for which package in a spreadsheet somewhere.
No first-party lesson recording. Discord's voice channels do not have a "Record call" button. The community workaround is a third-party bot (the most common is Craig); some tutors run a local screen recorder alongside instead. Either way, the recording workflow is bolt-on, not built-in — see how do I record online tutoring lessons? for the broader flow, including the consent step you should be walking parents through regardless of which tool you record in.
No engagement layer built for K-12 learners. Face-on-webcam-plus-voice is the wrong shape for a seven-year-old on lesson eight of the same week. Discord's engagement primitives — roles, XP bots, emoji, Nitro cosmetics — are calibrated for teenage and adult gaming and creator communities, not for a young learner in the middle of a phonics session with a parent watching from the sofa. Some tutors patch this with a rewards spreadsheet or a sticker chart the parent tracks — that's workflow to run in parallel with Discord, not part of it.
Discord's community vibe is a mixed feature in a parent-facing setting. The chat culture and default UI of Discord signals gaming and community more than a professional tutoring service. Adult students who already live there won't notice; parents evaluating a tutor for their eight-year-old often will. There's no fix inside the app for that framing.

The typical "Discord + bolt-ons" tutoring stack

Tutors who run the actual teaching layer on Discord usually assemble a stack that looks roughly like this:

Discord (free or a per-server Nitro Boost) — voice channels, video, screen share, text chat, file drops
Miro, Excalidraw, Whiteboard.fi, or a Google Slides file — a shared canvas the student can (in some cases) draw on
A shared Google Drive, GitHub repo, or Notion workspace — where lesson materials and any submitted work live
Calendly, Google Calendar, or a dedicated tutoring scheduler — booking and rescheduling that isn't manual DM negotiation
Stripe, PayPal, Wise, Payoneer, or a local bank transfer — invoicing and getting paid across currencies
Craig, another recording bot, OBS, or a screen recorder — for lessons you need to record
A rewards tracker (only if you have young learners at all) — a Google Sheet, Class Dojo, a printed sticker chart
A spreadsheet — to track which student paid for which package, how many lessons are left, and who owes a make-up

None of that is broken. It's just a lot of moving parts held together in your head. For a coding tutor with a steady book of five to ten adult students who all live on Discord already, that friction is tolerable — many practices run this way for years. For a tutor whose students are on tablets, or on the parent's phone, or under thirteen, or for anyone whose practice is growing past ten to fifteen weekly students, the tab-switching tax compounds every lesson.

When Discord is a reasonable fit for tutoring

You teach coding, CS, or software topics to teenagers and adults who already live on Discord. Meeting them where they already are, with the screen-share + voice-channel workflow they already use inside their own study groups, is the right call.
You teach music theory, music production, or audio work with adult students. Discord's voice-channel audio handles the "listen to this and let me hear you play it back" workflow, and screen share on a DAW is a legitimate teaching surface.
You run a cohort-style community program — a study group, a language club, a writing circle — where the community feel is the product, not an inconvenience. Discord's server + channels architecture was built for exactly this.
You want a between-lessons channel with your students that isn't your personal WhatsApp or SMS number. A private server per family, or a class-wide server your students opt into, gives you a professional-ish home for follow-ups without spilling into your personal apps.
You need a fallback for when your main teaching tool has a bad day. Every tutor should have one; a Discord voice channel plus a Google Doc is a reasonable one to keep in your back pocket for older students.

When it isn't

You teach young learners (roughly four to twelve) and the retention story your parents write in reviews is about engagement — "she never wants to log off" — rather than curriculum. Voice channels give a young learner one thing to do (talk when unmuted) and one thing to look at (a face on a screen). Discord's account-age policy also means the student's account is really the parent's account, which is workable for a check-in and awkward for a 45-minute lesson.
You teach reading, phonics, math manipulatives, or any subject where the student needs to act on the material — dragging letters, sorting shapes, placing counters on a number line. Screen share plus voice narration is not a substitute for the student doing the thing themselves.
You teach students on tablets or Chromebooks. Screen share on a small tablet shrinks text below readable size and turns pinch-to-zoom into a wrestling match with the host's scrolling. A cobrowser fixes this; a Discord screen share doesn't.
Your students' families are not already on Discord and would experience a Discord invite as unusual. For a lot of parents outside the coding / gaming / creator overlap, "join our Discord server" reads as unprofessional in a paid-tutoring context — no matter how good your setup is.
You teach students inside mainland China. Cross-border connectivity to Western community-messaging platforms is a moving target through the Great Firewall — see teaching English to students in mainland China for the connectivity workaround pattern rather than assuming any specific Western tool will hold up.
You want a single login that handles video, the lesson canvas, activities the student can act on, scheduling, and invoicing — and you're past the "am I really doing this?" stage of building a tutoring business. The stack collapse is what a purpose-built virtual classroom is for.

Purpose-built alternatives — what tutors typically switch to

The reason a lot of tutors eventually leave the "Discord + a stack of other apps" shape is the same reason tutors leave Zoom / Meet / WhatsApp / Skype-successor-Teams for a purpose-built tool: past a certain number of students, the tab-switching tax gets larger than the "meet them where they are" benefit. The category is described in depth on what is a virtual classroom?, with a head-to-head shortlist on the best Zoom alternatives for online tutoring. Both are worth reading if you're actively picking a Discord replacement rather than just accepting the default.

Where Koala Go (the classroom we build) fits, specifically:

An interactive whiteboard with PDF, PowerPoint, and image upload. The file becomes the lesson surface — the student annotates the same page you're teaching from. This is the piece a Discord screen share can't do because a screen share is a picture the student can't act on.
A cobrowser (a live shared browser window). You and the student open the same real webpage together, both with your own cursor, both able to click and type. Wordwall, Boom Learning, Khan Academy, IXL, a leveled-reader site, an online IDE — the student does the activity instead of watching you click. Koala Free ships a limited cobrowser (session-length capped); Koala Pro removes the cap. The cobrowser is region-aware, with separate routing for mainland China so students there aren't fighting the Great Firewall inside a Western tool.
A 3D Playground for younger students. A shared 3D world where each student has a customisable avatar. Used as a brain break, a themed lesson surface, or the reason a seven-year-old wants to come back next week. This is the piece a voice-channel-and-video-tile interface can't do because it gives the child nothing to be in.
Worldwide invoicing built in. Parents pay in their local currency, including via WeChat in China. There's a 4.5% Koala fee on Free-tier invoicing that drops to 3.5% on Pro and 0% on Platinum. No separate Stripe reconciliation.
Integrated scheduling on Pro. A teacher-and-parent scheduling portal where you set weekly availability once and parents book their own slots via a link. No Calendly bolt-on. Platinum adds a Scheduling Plus layer with 30/45/60-minute class lengths, booking-policy controls, and recurring classes.
Cloud lesson recording on Platinum. Recordings save inside the classroom, per-lesson, so a family whose kid missed a session can catch up without derailing next week. Local recording is available on Pro. Either way, the consent step is your job — the workflow is on the record-online-tutoring-lessons page.
24/7 live support on Pro with a real person. When something breaks mid-lesson, there's a human to reach rather than a help-centre article or a community thread.

Koala Free is a genuine free tier (up to 4 students per session, worldwide invoicing, limited cobrowser, up to 3 activities), not a trial that expires — so you can run a real trial lesson end-to-end with a real student on the device they actually use, and decide from there. Koala Pro is $25.99/month billed monthly or $21.99/month billed annually; Koala Platinum, which adds cloud recording and richer scheduling, is $49.99/month billed monthly or $39.99/month billed annually. The difference from a Discord-shaped stack isn't really the price; it's what you get for it — a tool built for teaching, not a tool built for community with teaching bolted on.

Honest limits worth naming so you can weigh them against the fit: iPad works but is rougher than desktop; mainland-China connectivity is served by our Hong Kong proxy but is a moving target; the whiteboard's math tools ship in Pro (the symbol palette's operations, comparisons, and fraction slash) and Platinum (the powers, the advanced symbols, and the drag-and-drop fraction bars) but stop short of a full graphing or equation editor, so heavy STEM tutors should test the workflow on a real lesson first; and the Playground is actively developed with requested features (group-select of Playground objects) still on the roadmap. Roughly 81% of the tutors on our 2026 product/market fit survey said they'd be very disappointed if they could no longer use Koala Go — the "very disappointed" threshold conventionally taken to indicate strong product/market fit is 40%, so that number is well past the bar. That 81% figure is the same one we publish on the main landing page.

Bottom line

Discord is the right tool for a specific slice of tutoring — coding and CS, music and audio, TTRPG-adjacent narrative work, adult subject cohorts, and any community-shaped program where the students already live there. It's the wrong tool for anyone teaching young learners, subject-do work, tablet-first students, parent-facing paid lessons, or a practice that's outgrowing a hand-held stack. If you started on Discord because your first students were already there, you don't need to leave on our account — but the day the tab-switching starts to feel like a job on top of the teaching, that's the signal to move the lesson surface onto a tool built for teaching and keep Discord for what it's actually good at. If you'd like our honest read on whether Koala Go fits your specific practice — subject, age range, region, current stack — write to koala@teachwithkoala.com with a sentence about the kind of tutoring you do. We're a small team and we read everything.

What FaceTime actually is

FaceTime is Apple's video and audio calling app, pre-installed on iPhone, iPad, and Mac, free with any Apple ID. Historically it was a walled garden — you needed an Apple device to make or receive a call. Since iOS 15 and macOS Monterey (autumn 2021), an Apple user can create a FaceTime call link and share it with anyone; Windows and Android users open the link in Chrome or Edge, enter a name, and join from the browser. Only an Apple user can initiate the call; Windows/Android participants can only join a link that was created for them.

In-call, on the Apple side, the current app supports the usual video-call basics — a grid of tiles, chat, mute, camera flip — plus a handful of Apple extras: Live Captions, Live Translation, Screen Sharing, and SharePlay (for watching a video together or sharing a compatible app's content). On the browser side (Windows/Android joiners), the picture is slimmer: Apple's own support page explicitly notes that SharePlay and screen sharing aren't available in FaceTime on the web. So the tool works as a lowest-common-denominator video call for a mixed-device family, and works as a slightly-fuller video call if everyone's on Apple.

One documented gap worth knowing before you plan around FaceTime: Apple states that FaceTime links are not supported in mainland China, and creating a FaceTime link is not supported on iPhone models purchased in mainland China. If a family is inside the Great Firewall, or bought their phone there, the link-based join workflow won't work for them.

Why parents keep suggesting it

It isn't random. In parts of the world with high iPhone/iPad penetration — particularly the US, and increasingly urban families in the UK, Australia, Canada, and parts of Asia — FaceTime is the default app for a "let's video-call each other" moment the same way iMessage is the default for a text. Parents use it to call grandparents, arrange playdates, and reach the babysitter. When a parent books their first paid tutoring lesson and the tutor says "we'll do it over video," the parent's mental model is often "great, we'll FaceTime you." It's what "video call" means to them.

For a first trial lesson with an Apple family that's still deciding whether they want to work with you at all, going along with the FaceTime request costs you nothing: the child is on a familiar app, the parent isn't fighting new software during a first impression, and you can focus on whether the tutoring itself is a fit rather than on the setup. The problem is what happens after the trial — when the lesson goes recurring, when the child needs to do more than talk, when a sibling on a Chromebook wants to try a session, or when the family with a WeChat-first workflow lands next.

What's missing for online tutoring

Six gaps show up regularly, in roughly the order tutors notice them:

No shared interactive canvas built for teaching. On the Apple side you can screen-share; on the browser (Windows/Android) side you can't screen-share at all. Neither is the same thing as a canvas the student can draw on. Screen share is a picture: only the host's mouse and keyboard touch the shared content, and the student comments. On a real lesson surface the student drags a letter into a box, types into a form, circles a paragraph, or places a piece on a number line. In-person you'd hand the student the pen. On FaceTime you can't.
No activities the student can act on. Related but distinct: even if you point the child at a shared browser tab somewhere, a video-call-plus-webpage workflow gives you a webpage — it's not a lesson library. Phonics matching, dragging fraction bars, sorting words into categories, snapping shapes together — the interactive activities working tutors of young kids build a lesson around are not in FaceTime, and FaceTime doesn't have a home for them.
Cross-platform gap. Only Apple users can start a call; only Chrome or Edge (not Safari, since it isn't on Android/Windows) can join from the other side; and SharePlay and screen sharing aren't available in the web version at all. If your student's family is fully Apple you'll only feel this occasionally. If half your students are on Chromebooks, or the child uses a shared family Windows PC while you're on your iPad, the mixed-device seam bites every session.
No scheduling for paying clients. Apple's Calendar app is fine at what it does, but there's no "parent books this Tuesday at 4 PM against my open availability" flow inside FaceTime. You'll add Calendly, YouCanBook.me, a shared Google Calendar, or a dedicated tutoring scheduler; the parent's rescheduling requests come to you in a text they expect a reply to.
No invoicing or payments. FaceTime doesn't bill parents. You'll add Stripe, PayPal, Wise, Venmo, or manual invoicing outside the app — and reconcile which student paid for which package inside a spreadsheet somewhere. If your students are in different currencies — or in mainland China, where the FaceTime link workflow doesn't work in the first place — the friction is bigger than one bolt-on.
No engagement layer for young learners. Face-on-a-tile is still face-on-a-tile. There are no stars, no gems, no avatars, no shared space for the student to be in as a character rather than as a face on a grid. Some tutors patch this with a rewards spreadsheet the parent can see, sticker printouts sent by mail, or a Class Dojo integration. All of those are workflow to run in parallel with FaceTime, not part of FaceTime.

The typical "FaceTime + bolt-ons" tutoring stack

Working tutors who run their practice on FaceTime usually end up with something like this — and the total cost, in dollars and in minutes per week of stack management, is often higher than a single purpose-built classroom:

FaceTime — the live video call (Apple side), a browser link for anyone else
Miro, Whiteboard.fi, or a shared Google Slides file — a shared canvas
Calendly, YouCanBook.me, or a shared calendar — booking and rescheduling
Stripe, PayPal, Venmo, or Wise — invoicing and getting paid across currencies
Class Dojo, a Google Sheet, or a printed sticker chart — rewards for younger learners
An iCloud, Drive, or Dropbox folder — where lesson materials live
A spreadsheet — to track which student paid for which package, how many lessons are left, and who owes a make-up

None of that is broken. It's the same friction Zoom-shaped and Skype-shaped tutoring stacks live with, in a different app. For tutors with a small steady book of adult 1-on-1 clients, that friction is tolerable. For tutors teaching young learners on shared screens, or for practices that mix Apple and non-Apple families, it adds up fast.

When FaceTime is a reasonable fit for tutoring

You're running a first, free trial call with an Apple family and want the setup to feel familiar so the child and parent can focus on whether the tutoring itself is a fit. FaceTime removes one variable from the first impression.
You teach one-on-one adults or teenagers where the lesson is mostly talking — conversation practice, business English, IELTS speaking prep, university-level tutoring by discussion. A clean video call plus a shared document is often enough, and if the client's on an iPhone every session, FaceTime is one less password to hand out.
Cost is a hard constraint and you're testing whether you want to keep tutoring at all before committing to a paid stack. Free FaceTime plus a shared iCloud note will get you to your first three lessons if everyone's already inside Apple's world.
You need a fallback for when your main teaching tool has a bad day. Every tutor should have one; a FaceTime link is a reasonable one to keep in your back pocket for Apple families.

When it isn't

You teach young learners (roughly four to ten) and the retention story your parents write in reviews is about engagement — "she never wants to log off" — rather than curriculum. Video calls give a young learner one thing to do (talk when unmuted) and one thing to look at (a face on a grid). It falls short of what an engagement-designed classroom gives them.
You teach reading, phonics, math manipulatives, or any subject where the student needs to act on the material — dragging letters, sorting shapes, placing counters on a number line. Screen share (Apple-side only) plus voice narration is not a substitute for the student doing the thing themselves. See what is a cobrowser? for the primitive that fixes this.
Your students' families are on mixed devices. If half your students are on Chromebooks, if the child sometimes joins from a parent's Windows laptop, or if a sibling wants to try a lesson on an Android tablet, the FaceTime workflow degrades — no SharePlay or screen sharing on the browser side, and no way for a non-Apple parent to initiate a fallback call if you're the one with the flaky connection today.
You teach students in mainland China. Apple's own documentation says FaceTime links aren't supported in mainland China, and creating a link isn't supported on iPhone models purchased there. That is the exact workflow you'd need to reach a family inside the Great Firewall. See teaching English to students in mainland China for the connectivity workaround pattern.
You want a single login that handles video, the lesson canvas, activities, scheduling, and invoicing — and you're past the "am I really doing this?" stage of building a tutoring business. The stack collapse is what a purpose-built virtual classroom is for.
You need lesson recording as part of your workflow — for parent updates, for make-up sessions, or for your own self-review. FaceTime itself doesn't ship a "record this call" button; you'd fall back to your device's screen-recording tool (built into iOS and macOS) and manage the file yourself, which is a workflow, not a feature. See below.

What about recording on FaceTime?

FaceTime does not offer a built-in "record this call" button the way some meeting tools do. If you want a recording of a FaceTime lesson, the standard route is to use the operating system's own screen-recording tool — the Screen Recording toggle in iOS Control Center, or the built-in screen recorder on macOS (Shift-Command-5, or QuickTime). The file lands in your Photos or on your Desktop; sharing it with a parent, storing it somewhere they can find it later, and deleting it when you don't need it any more is manual. There is no per-lesson home for a recording, no automatic labelling by student and date, no built-in retention policy. Purpose-built tutoring classrooms tend to ship recording as a first-class feature. See how do I record online tutoring lessons? for the broader workflow — including the consent step you should be walking parents through regardless of which tool you record in; the practical rule of thumb is to get explicit parental consent in writing, check what your local rules require (recording law varies by jurisdiction), and only keep recordings as long as you actually need them.

Purpose-built alternatives — what tutors typically switch to

Tutors who outgrow the FaceTime-plus-bolt-ons stack usually move to a virtual classroom designed for live tutoring from the start. The category is described in depth on what is a virtual classroom?; the head-to-head shortlist (with tradeoffs, subject fits, and free-tier caps) lives on the best Zoom alternatives for online tutoring, which covers Google Meet, Whereby, LearnCube, Lessonspace, Vedamo, and Koala Go and is worth reading if you're actively picking a replacement.

Where Koala Go (the classroom we build) fits the picture, specifically:

An interactive whiteboard with PDF, PowerPoint, and image upload. The file becomes the lesson surface — the student annotates the same page you're teaching from. This is the piece a FaceTime screen share can't do (and that a FaceTime browser joiner can't do at all) because a screen share is a picture the student can't act on.
A cobrowser (a live shared browser window). You and the student open the same real webpage together, both with your own cursor, both able to click and type. Wordwall, Boom Learning, Khan Academy, IXL, a leveled-reader site — the student does the activity instead of watching you click. Koala Free ships a limited cobrowser (session-length capped); Koala Pro removes the cap. The cobrowser is region-aware, with separate routing for mainland China so students there aren't fighting the Great Firewall inside a Western tool.
Browser-based, so device-mix stops being a wall. A student on a Chromebook, a parent joining briefly from a Windows laptop, and you on an iPad all open the same URL — no app install, no Apple ID needed on the child's device, no separate join workflow for the non-Apple side.
A 3D Playground for younger students. A shared 3D world where each student has a customisable avatar. Used as a brain break, a themed lesson surface, or the reason a seven-year-old wants to come back next week. Face-on-a-tile gives a kid nothing to be in; the Playground gives them a place.
Worldwide invoicing built in. Parents pay in their local currency, including via WeChat in China. There's a 4.5% Koala fee on Free-tier invoicing that drops to 3.5% on Pro and 0% on Platinum. No separate Stripe or Venmo reconciliation.
Integrated scheduling on Pro. A teacher-and-parent scheduling portal where you set weekly availability once and parents book their own slots via a link. No Calendly bolt-on. Platinum adds a Scheduling Plus layer with 30/45/60-minute class lengths, booking-policy controls, and recurring classes.
Cloud lesson recording on Platinum. Recordings save inside the classroom, per-lesson, so a family whose kid missed a session can catch up without derailing next week.
24/7 live support on Pro with a real person. When something breaks mid-lesson, there's a human to reach rather than a help-centre article.

Koala Free is a genuine free tier (up to 4 students per session, worldwide invoicing, limited cobrowser, up to 3 activities), not a trial that expires — so you can run a real trial lesson end-to-end with a real student on the device they actually use, and decide from there. Koala Pro is $25.99/month billed monthly or $21.99/month billed annually; Koala Platinum, which adds cloud recording and Scheduling Plus, is $49.99/month billed monthly or $39.99/month billed annually. The difference from FaceTime-shaped tools isn't really the price; it's what you get for it — a tool built for teaching, not a family video-call app with teaching bolted on.

Honest limits worth naming so you can weigh them against the fit: iPad works but is rougher than desktop; mainland-China connectivity is served by our Hong Kong proxy but is a moving target; the whiteboard's math tools ship in Pro (the symbol palette's operations, comparisons, and fraction slash) and Platinum (the powers, the advanced symbols, and the drag-and-drop fraction bars) but stop short of a full graphing or equation editor, so heavy STEM tutors should test the workflow on a real lesson first; and the Playground is actively developed with requested features (group-select of Playground objects) still on the roadmap. Roughly 81% of the tutors on our 2026 product/market fit survey said they'd be very disappointed if they could no longer use Koala Go — the "very disappointed" threshold conventionally taken to indicate strong product/market fit is 40%, so that number is well past the bar. That 81% figure is the same one we publish on the main landing page.

Bottom line

FaceTime is a competent, familiar consumer video-call app, and going along with a parent's "let's just FaceTime" for a first trial is fine. It's the wrong shape for a paid, ongoing tutoring practice — especially with young learners, on mixed-device families, or if any of your families are in mainland China. The right answer there is not "FaceTime plus five other apps" but a purpose-built tutoring classroom that collapses the stack into one login and works the same way for the Chromebook kid as for the iPad kid. If you'd like our honest read on whether Koala Go fits your specific practice — subject, age range, region, current stack — write to koala@teachwithkoala.com with a sentence about the kind of tutoring you do. We're a small team and we read everything.

What Telegram actually is

Telegram describes itself, on its own homepage and FAQ, as a cloud-based messaging app focused on speed and security, with over a billion users and messages that sync across every device you sign in on. It is free — messaging and calls stay free forever per the FAQ, and Telegram Premium is an optional subscription for larger uploads and cosmetic extras rather than a paywall around the calling and chatting most tutors would use. Apps ship for iOS (13 and above), Android (5.0 and up), a native macOS app, a universal desktop app for Windows, macOS, and Linux, and Telegram Web. Cross-platform reach is one of the honest strengths.

The pieces you'd actually use in a lesson:

Voice and video calls. One-to-one calls and group calls of up to 200 participants, all end-to-end encrypted, per telegram.org/faq.
Text chat alongside the call. A rich message thread per contact or per group, permanently searchable across devices.
File and media sharing up to 2 GB per file (4 GB with a Telegram Premium subscription, per the FAQ). The lesson PDF, a photo of a page, a voice-note homework prompt, a short video clip — all drop into the same thread and both sides can download.
Screen sharing in video calls. Available on Telegram's mobile and desktop apps; you share what's on your screen and the other side watches.
Groups (up to 200,000 members) and Channels (unlimited subscribers). A group is a two-way space where members chat with each other; a channel is broadcast-only and useful for one-to-many announcements — bulletin-style updates to every family at once, for example.
Voice Chats and Live Streams inside groups and channels. Persistent audio rooms members join when they want, and channel Live Streams with an unlimited audience. Neither is designed as a paid-lesson primitive, but both exist.
Bots via the open Bot API. Third-party developers build small programs that live inside Telegram — quiz bots, reminder bots, payment bots ("accept payments from users around the world", per the FAQ). Bots are how anything commerce-shaped bolts on.
Two chat encryption modes. Regular Cloud Chats live in Telegram's cloud with server-client encryption (so they sync across devices); Secret Chats are end-to-end encrypted, device-specific, and can self-destruct. Voice and video calls are end-to-end encrypted per the FAQ regardless of which chat mode you were in.

That is a legitimately useful communication toolkit — and it is a communication toolkit, not a teaching toolkit.

Where parents (and some tutors) already are on Telegram

Telegram's market share isn't evenly distributed. In parts of the world — Russia, Ukraine, much of Central Asia, and various MENA and Eastern European markets — Telegram is the default messenger the way WhatsApp is in Latin America or India, and iMessage is in the US. If your students' families are anchored there, Telegram is where they already talk to their kids' school, their landlord, and their book club — and asking them to install a different messenger just to talk to their tutor is a friction that WhatsApp-anchored tutors don't have.

There is also a separate niche: adult tech-comfortable students, adult exam-prep students, adult subject-matter cohorts (crypto, coding, self-hosted-tools communities, and adjacent), and language teachers who run public channels as a lightweight audience-building tool. For that slice, Telegram is often the tool everyone already has, and asking a client to install something new for the lesson feels like the wrong ask before the client has committed to working with you.

The distinction this page is about is the same one that applies to WhatsApp: Telegram as your family-communication channel (fine, keep it) versus Telegram as your actual teaching platform (the video call is where the lesson happens, the chat is where you paste homework, the media tab is where worksheets live). The second one works up to a point; past that point its limits start to cost you time and enrollments.

Where a Telegram-as-classroom stack breaks for tutoring

No lesson surface the student can act on. Screen share and video are things the student watches. There's no shared canvas where the student drags a fraction bar onto a number line, sorts letters into phonics categories, annotates the worksheet you uploaded, or works a problem while you watch the pen. For younger learners and for any subject where doing beats watching, this is the biggest gap.
No activity library. Phonics matching, dragging fraction bars, sorting words into categories, snapping shapes together — the interactive activities working tutors of young kids build a lesson around are not something Telegram has a home for. You end up screen-sharing another tool over the call, which introduces the next problem down the list.
Screen share isn't interactive. If you share the browser tab where a student practice site lives, the student sees your cursor moving; they can't click, drag, or type into the page themselves. On a phone or a small tablet the readable-text problem compounds — you're squeezing a laptop-sized site into a five-inch screen and the student has to fight your scrolling. See what is a cobrowser? for the specific mechanic that fixes this and screen share doesn't.
Phone-number-as-identity mixes your personal and business life. Telegram accounts are pinned to a phone number by design (per telegram.org/faq). Your students see the number that's also on your family chats. A public username lets people find you without knowing your number, but you're still logging into your practice as the same account you use for everything else — and, unlike a business tool, there's no shared-access model for a partner or an assistant.
Group calls up to 200 participants isn't a classroom cap; it's a party cap. Being able to run a 200-person voice call is a nice technical property; a paid lesson is one, two, or a small group of students, and the actual gap in Telegram-as-classroom isn't the number — it's what's inside the call. No lesson surface, no shared canvas, no student-response dashboard, no per-student attendance record you can bill from. The cap tells you the tool is designed for community events, not tutoring lessons.
Only the parent should be the account holder for a minor. Any messaging app designed for adult phone-number accounts sits awkwardly around messaging children directly. Practical answer regardless of app: run everything — booking, messaging, calls — through the parent's account, not the child's; check what your local rules require around messaging minors and around consent to record; don't add student contacts under student numbers. This is the same practical rule the WhatsApp and Discord pages make; it's an app-shape issue, not something wrong with Telegram specifically.
No booking flow parents can use themselves. Every reschedule, every "can we move to Thursday?", every new package is a manual back-and-forth in the chat. It works while you have five students; it does not scale to twenty. You'll bolt on Calendly, a Google Calendar link, or a dedicated tutoring scheduler.
No first-party invoicing. Parents pay outside the app. The Telegram Bot API can accept payments through third-party bots ("accept payments from users around the world", per the FAQ), which some tutors wire up — but a payment bot is engineering work, not a first-party invoicing feature, and the reconciliation of "which student paid for which package" still lands on you.
No first-party lesson recording. Telegram's calls are end-to-end encrypted and the app doesn't ship a "Record call" button. If you want a lesson recording — for a family whose kid missed a session, or for your own review — you're running a separate screen recorder (OBS, QuickTime, a browser extension) alongside the call. See how do I record online tutoring lessons? for the workflow, including the consent step you should be walking parents through regardless of which tool you record in.
No engagement layer for young learners. Face-on-webcam is face-on-webcam. There are no stars, no gems, no avatars, no shared space for the student to be in as a character rather than as a thumbnail on a parent's phone. Some tutors patch this with a sticker chart the parent tracks or a rewards spreadsheet — that's workflow to run in parallel with Telegram, not part of it.
Reachability varies by region. Telegram is on the widely-published list of services blocked or restricted in mainland China, and access rules in some other countries change from time to time. If any of your students are inside mainland China or in a country whose access to Telegram is unreliable, the app can be intermittently unreachable for them — usually meaning a VPN, which is a friction and a network-quality hit. See teaching English to students in mainland China for the connectivity workaround pattern.

The typical "Telegram + bolt-ons" tutoring stack

Tutors who run a whole practice on Telegram usually end up assembling a stack that looks roughly like this — the specifics differ by market but the shape doesn't:

Telegram — video call, chat, big file drop, screen share, optional broadcast channel for the practice
Miro, Whiteboard.fi, or a Google Slides file — a shared canvas the student can (in some cases) draw on
A shared Google Drive, OneDrive, or the Telegram thread itself — where worksheets and any lesson recordings live
Calendly, a Google Calendar link, or a dedicated tutoring scheduler — booking and rescheduling that doesn't cost you a chat every time
Stripe, PayPal, Wise, a local bank transfer, or a Telegram payment bot — getting paid across currencies
OBS, QuickTime, or a browser extension — if you record lessons at all
A sticker chart, a Google Sheet, or Class Dojo — rewards for younger learners
A spreadsheet — to track which student paid for which package, how many lessons are left, and who owes a make-up

None of that is broken. It's a lot of moving parts held together in your head. For tutors with a small steady book of older students, that friction is tolerable — plenty of practices in Telegram-anchored markets run this way for years. For tutors teaching young learners on shared screens, or for practices growing past ten or fifteen weekly students, the tab-switching tax compounds every lesson.

When Telegram-as-classroom is a reasonable fit

You teach one-on-one adults or teenagers where the lesson is mostly talking — conversation practice, exam speaking prep, business English, university-level tutoring by discussion, coding review, professional coaching. A clean video call plus a shared PDF over chat is often enough.
Your students' families anchor on Telegram (Russia, ex-USSR states, MENA, adult tech niches) and asking them to install a different tool just for the tutoring is a friction that costs you the booking. Meet the family where they already are.
You want a broadcast channel — a Telegram channel is a decent, low-friction way to publish weekly tips, homework prompts, or "we have a slot open next Tuesday" announcements to every family at once, without email.
You're running a first paid lesson with a Telegram-first family and want to see whether the tutoring itself is a fit before either of you commit to a new tool.
You need a fallback for when your main teaching tool has a bad day. Every tutor should have one; a Telegram video call plus a PDF the parent already has in the thread is a reasonable one to keep in your back pocket for Telegram-anchored families.
You need to send a big file to a family — a 500 MB PDF workbook, a scanned textbook, a set of exam-prep videos. Telegram's 2 GB (or 4 GB Premium) per-file cap is honestly one of the more useful things about the app for tutoring.

When it isn't

You teach young learners (roughly four to ten) and the retention story your parents write in reviews is about engagement — "she never wants to log off" — rather than curriculum. Video calls give a young learner one thing to do (talk when unmuted) and one thing to look at (a face on a screen). It falls short of what an engagement-designed classroom gives them.
You teach reading, phonics, math manipulatives, or any subject where the student needs to act on the material — dragging letters, sorting shapes, placing counters on a number line. Screen share plus voice narration is not a substitute for the student doing the thing themselves.
You teach students on tablets or Chromebooks. Screen share on a small tablet shrinks text below readable size and turns pinch-to-zoom into a wrestling match with the host's scrolling. A cobrowser fixes this; a Telegram screen share doesn't.
You teach students inside mainland China. Telegram is on the widely-published blocked-in-China list, and a VPN-behind-a-family setup is an unreliable base to build a paid lesson on. See teaching English to students in mainland China for the connectivity workaround pattern.
You're growing past the number of students you can hold in your head. Ten students booking and rescheduling in your Telegram inbox is fine; twenty-five isn't, and the fix isn't a bigger inbox — it's a scheduler with a policy attached and an invoicing tool that reconciles itself.
You want a single login that handles video, the lesson canvas, activities the student can act on, scheduling, and invoicing — and you're past the "am I really doing this?" stage of building a tutoring business. Collapsing the stack is what a purpose-built virtual classroom is for.

What about recording on Telegram?

Telegram's voice and video calls are end-to-end encrypted, and the app doesn't ship a first-party "Record this call" button the way some meeting tools do. If you want a recording of a Telegram lesson, the standard route is to use the operating system's own screen-recording tool — the Screen Recording toggle in iOS Control Center, the built-in Windows Game Bar recorder, or the built-in macOS screen recorder (Shift-Command-5, or QuickTime). The file lands on your device; sharing it with a parent, storing it somewhere they can find it later, and deleting it when you don't need it any more is manual. There is no per-lesson home for a recording, no automatic labelling by student and date, no built-in retention policy. Purpose-built tutoring classrooms tend to ship recording as a first-class feature. See how do I record online tutoring lessons? for the broader workflow — including the consent step you should be walking parents through regardless of which tool you record in; the practical rule of thumb is to get explicit parental consent in writing, check what your local rules require (recording law varies by jurisdiction), and only keep recordings as long as you actually need them.

Purpose-built alternatives — what tutors typically switch to

The bigger reason a lot of long-time Telegram tutors eventually move the lesson surface elsewhere is that the "cloud messenger plus a stack of other apps" shape stops paying its rent past a certain practice size. The category that grew up around it — purpose-built virtual classrooms for online tutoring — is described in depth on what is a virtual classroom?, with a head-to-head shortlist of options on the best Zoom alternatives for online tutoring. Both are worth reading if you're actively picking a Telegram replacement rather than defaulting to "well, this is what the family already uses."

Where Koala Go (the classroom we build) fits, specifically:

An interactive whiteboard with PDF, PowerPoint, and image upload. The file becomes the lesson surface — the student annotates the same page you're teaching from. This is the piece a Telegram screen share can't do because a screen share is a picture the student can't act on.
A cobrowser (a live shared browser window). You and the student open the same real webpage together, both with your own cursor, both able to click and type. Wordwall, Boom Learning, Khan Academy, IXL, a leveled-reader site, a YouTube video for the ones that aren't blocked in your student's region — the student does the activity instead of watching you click. Koala Free ships a limited cobrowser (session-length capped); Koala Pro removes the cap. The cobrowser is region-aware, with separate routing for mainland China so students there aren't fighting the Great Firewall inside a Western tool.
A 3D Playground for younger students. A shared 3D world where each student has a customisable avatar. Used as a brain break, a themed lesson surface, or the reason a seven-year-old wants to come back next week. A phone-screen video call gives a kid nothing to be in; the Playground gives them a place.
Worldwide invoicing built in. Parents pay in their local currency, including via WeChat Pay in mainland China. There's a 4.5% Koala fee on Free-tier invoicing that drops to 3.5% on Pro and 0% on Platinum. No separate Stripe reconciliation, no payment-bot to build.
Integrated scheduling on Pro. A teacher-and-parent scheduling portal where you set weekly availability once and parents book their own slots via a link. No Calendly bolt-on. Platinum adds a Scheduling Plus layer with 30/45/60-minute class lengths, booking-policy controls, and recurring classes — the "book all four Wednesdays for the month at once" flow that a Telegram chat can't hold.
Lesson recording on Pro and Platinum. Koala Pro records the lesson locally to your own machine; Koala Platinum adds cloud recordings that save inside the classroom, per-lesson, so a family whose kid missed a session can catch up without derailing next week.
24/7 live support on Pro with a real person. When something breaks mid-lesson, there's a human to reach rather than a help-centre article.
Runs in the browser. The student and parent open a link — nothing to install, no separate app account, no phone-number-linked identity. The link works on the device the family already uses.

Koala Free is a genuine free tier (up to 4 students per session, worldwide invoicing, limited cobrowser, up to 3 activities), not a trial that expires — so you can run a real trial lesson end-to-end with a real student, on the device they actually use, and decide from there. Koala Pro is $25.99/month billed monthly or $21.99/month billed annually; Koala Platinum, which adds cloud recording and richer scheduling, is $49.99/month billed monthly or $39.99/month billed annually. The difference from Telegram isn't really the price; it's what you get for it — a tool built for teaching, not a cloud messenger with teaching bolted on.

Honest limits worth naming so you can weigh them against the fit: iPad works but is rougher than desktop; mainland-China connectivity is served by our Hong Kong proxy but is a moving target; the whiteboard's math tools ship in Pro (the symbol palette's operations, comparisons, and fraction slash) and Platinum (the powers, the advanced symbols, and the drag-and-drop fraction bars) but stop short of a full graphing or equation editor, so heavy STEM tutors should test the workflow on a real lesson first; and the Playground is actively developed with requested features (group-select of Playground objects) still on the roadmap. Roughly 81% of the tutors on our 2026 product/market fit survey said they'd be very disappointed if they could no longer use Koala Go — the "very disappointed" threshold conventionally taken to indicate strong product/market fit is 40%, so that number is well past the bar. That 81% figure is the same one we publish on the main landing page.

Keep Telegram for what it's good at

Nothing above is an argument to remove Telegram from your practice. It's an argument to use it where it earns its keep: as the single channel per family for booking messages, quick reminders, homework check-ins, and the "she's running late" message on the day. If your practice is Telegram-anchored, a small broadcast channel for your students' families is also a low-friction way to publish weekly announcements without email. Send one link a parent can bookmark for every lesson (your classroom room link), and let Telegram be the place you notify them when something changes. That is the shape that both survives and scales — one link for the lesson, one channel for the messaging, both pointed at people who need to actually reach you. See how do I onboard a new online tutoring student? for the "one channel per family" onboarding discipline this rests on.

Bottom line

Telegram is where a real slice of tutoring families already live, and the discipline of meeting them there for messaging is worth keeping — especially in Russia, the CIS, MENA, and various adult 1-on-1 niches. It works as a lesson-delivery tool for a specific slice: one-on-one adults or teenagers, mostly-talking lessons, first-lesson stopgaps, and reliable-fallback duty. For anyone teaching young learners, teaching subjects the student needs to act on, growing past a hand-held book of students, or serving families in mainland China, the honest answer is that Telegram is where you talk to the parent — not where the lesson itself should live. If you'd like our read on whether Koala Go fits your specific practice — subject, age range, region, current stack — write to koala@teachwithkoala.com with a sentence about the kind of tutoring you do. We're a small team and we read everything.

What Whereby actually is

Whereby describes itself as a browser-based video-conferencing product built in Europe, "GDPR and HIPAA compliant, and ISO 27001 certified" (whereby.com). It ships as two separate products, and it's worth being clear which one you're looking at:

Whereby Meetings — the standalone video-call product you sign up for as an individual. You get a persistent room URL (something like whereby.com/your-room-name) that stays yours; you share the link and anyone can join in their browser. This is what an indie tutor would use.
Whereby Embedded — an API and SDK for developers who want to bake Whereby's video call into someone else's product (a telehealth platform, an eLearning platform, a therapy scheduling app). This is not what you'd buy as a tutor; it's what a tutoring platform vendor might build with. Out of scope here.

Everything in the rest of this page is about Whereby Meetings. Its whole pitch, in Whereby's own words: "All calls run directly in your browser, making it easy to join and connect during a session. No additional logins or app downloads required. Works seamlessly for desktops and mobile" (whereby.com/information/meetings). That is honestly one of the nicer things about it — for a family that's spent the last five years installing new video apps every six months to talk to a doctor, a school, a therapist, and now a tutor, Whereby's "click the link, be in the room" experience is a small quality-of-life win.

The pieces you'd actually use in a lesson:

A persistent, memorable room URL. The parent bookmarks it once and clicks the same link every week. This is closer to how a real classroom works (walk into the same room every Tuesday) than the meeting-link-per-lesson dance Zoom / Meet still make you do.
Video and audio calls in the browser. The Free plan caps calls at 30 minutes and 4 attendees per meeting per Whereby's pricing page; Pro lifts both — 100 attendees per meeting and no time limit — for $10.99 per host per month (Whereby's own pricing page, USD, excluding VAT). Business is $13.99 per host per month with a 3-host minimum and is priced for teams rather than solo tutors.
Screen sharing. Available on all tiers per the pricing table. Same shape as Zoom or Meet: you share what's on your screen and the student watches.
Chat inside the call. Send messages, reactions, and files during a session (whereby.com/information/meetings).
A virtual whiteboard. Whereby's Meetings page lists "breakout groups, file sharing, and virtual whiteboards to support seamless collaboration" (whereby.com/information/meetings). It's a meeting whiteboard — a shared surface for sketching an idea mid-call — and it doesn't appear in the per-tier pricing comparison, so check which plan you land on before you count on it. What it isn't is a lesson canvas: there's nothing to upload a worksheet or a PDF into and teach from, and no library of activities sitting behind it.
File sharing. Listed as a Pro-tier feature in the pricing comparison — not available on Free.
Local recording (Pro and up). Whereby's pricing page lists "Local Recording" as a Pro feature and "Unlimited recording" in the Pro plan card. The file records to your machine; sharing, storing, and deleting it are your workflow.
Custom branding (Pro and up). Your logo and colours on the room; useful for looking like a real practice rather than a personal call. Also a custom subdomain on Business.
Breakout Groups and Hybrid meetings (Pro and up). Useful mostly if you run small-group sessions with mixed adults; a solo 1-on-1 tutor rarely needs either.
Locked rooms and a waiting-room experience. Control who joins your room; the Business tier adds custom waiting-room configuration.
Background effects and noise reduction. Standard camera / microphone quality-of-life features; both sit on Pro and up in the pricing comparison, not on Free.

That is a legitimately well-designed meeting tool — and it is a meeting tool, not a tutoring tool. A meeting whiteboard aside, the pricing page's per-tier feature comparison lists no lesson canvas you can load teaching material into, no activity library, no scheduling flow, no invoicing, no engagement layer for young learners. Every capability above is one you'd want in a therapy session, a coaching call, a consultant's client meeting, or a one-on-one adult lesson; none of them are ones a working tutor of young children would build a lesson around.

Why solo tutors keep considering Whereby

Three reasons come up honestly, and they're worth naming rather than dismissing:

The "no download" experience. The parent has already had to install Zoom for the pediatrician, Meet for the school Zoom-fatigue committee, Teams for the workplace, and a portfolio of "this specific specialist uses this specific app" installs. "Click this link and you're in the lesson" is a real relief to them, and if you've ever spent the first ten minutes of a first paid lesson helping a parent update Zoom, you know why it matters.
The persistent room URL. A single link a family bookmarks once matches the mental model of "we go to the same place at the same time every Tuesday." It removes the "which link was the one for tonight?" question a per-lesson-link tool creates.
The friendly, European, privacy-forward positioning. Whereby markets itself as European-owned and GDPR / ISO 27001 compliant; for a family that has actively opted out of American Big Tech video, that positioning is not marketing fluff — it is the reason they picked their doctor's telehealth vendor, and it will feel familiar. This matters more in some European markets than others; if your practice sits inside one where families ask about data protection, it's a genuine leg up over Zoom on the trust axis before you've even started the trial.

What none of these three do is fix the shape of the thing itself. A friendly video call plus a persistent URL is still a friendly video call; a paid tutoring lesson with a nine-year-old learning fractions is a different job.

Where a Whereby-as-classroom stack breaks for tutoring

The whiteboard is a meeting whiteboard, not a lesson surface. Whereby does list a virtual whiteboard, and for sketching a diagram mid-call it's fine. What it isn't is a place the lesson lives: there's nothing to upload the worksheet or the PDF into and teach from, no fraction bars or number lines to drag, no phonics tiles to sort, and no library of activities sitting behind it. Screen share fills that gap, and screen share is a thing the student watches. For younger learners and for any subject where doing beats watching, this is the biggest gap.
No activity library. Phonics matching, dragging fraction bars, sorting words into categories, snapping shapes together — the interactive activities working tutors of young kids build a lesson around are not something a video-call tool has a home for. You end up screen-sharing another tool over the call, which introduces the next problem down the list.
Screen share isn't interactive. If you share the browser tab where a student practice site lives, the student sees your cursor moving; they can't click, drag, or type into the page themselves. On a small tablet the readable-text problem compounds — you're squeezing a laptop-sized site into a tablet screen and the student has to fight your scrolling. See what is a cobrowser? for the specific mechanic that fixes this and screen share doesn't.
Free-tier caps hit a tutor's real day. Whereby's Free plan caps meetings at 30 minutes and 4 attendees. A one-hour paid lesson doesn't fit; a small-group session past 4 people doesn't fit. Real tutoring lands on the Pro plan ($10.99 per host per month, per whereby.com's pricing page) from lesson one, so the "free forever" framing doesn't apply to a paid practice — you're paying from day one, and you're paying for a meeting tool, not a teaching tool.
No scheduling flow parents can use. Whereby's Meetings page offers to "integrate your rooms with Google calendars" (whereby.com/information/meetings), but that's putting a room on a calendar, not a "parent books this Tuesday at 4 PM against my open availability" flow. Every reschedule is a manual back-and-forth in email or messages. You'll bolt on Calendly, YouCanBook.me, or a dedicated tutoring scheduler.
No first-party invoicing. Whereby doesn't bill parents. You'll add Stripe, PayPal, Wise, Venmo, a local bank transfer, or manual invoicing outside the app — and reconcile which student paid for which package inside a spreadsheet somewhere. If your students are in different currencies the friction is bigger than one bolt-on.
No engagement layer for young learners. Face-on-tile is still face-on-tile. There are no stars, no gems, no avatars, no shared space for the student to be in as a character rather than as a thumbnail on the parent's laptop. A background effect isn't an engagement mechanic; a rewards spreadsheet the parent tracks is workflow you run in parallel with Whereby, not part of it.
Recording is a file, not a workflow. Local Recording on Pro gives you an MP4 on your machine. There's no per-lesson home for it, no automatic labelling by student and date, no built-in retention policy or share-with-parent flow — that's yours to build with folders and a naming convention. See below.
Region-specific reachability is your problem to check. Whereby is European-hosted and doesn't publish the kind of prominent "we work behind these firewalls" statement that a China-first family would want to see. If any of your students are inside mainland China or a region whose access rules to European telecoms services change, you own the connectivity check. See teaching English to students in mainland China.
One room URL on Free, three on Pro. Free gives you 1 room URL, Pro gives you 3 (per whereby.com's pricing page). If you teach five different students at five different times and want a different room URL per family, you'd need Business ($13.99 per host per month, 3-host minimum). Most tutors work around this by using the same room URL and staggering lessons — which is fine, but it means the parent's mental model of "our room" is actually "the tutor's room, that we happen to share with four other families."

The typical "Whereby + bolt-ons" tutoring stack

Tutors who run a whole practice on Whereby usually end up assembling a stack that looks roughly like this — the specifics differ by market but the shape doesn't:

Whereby — video call, chat, screen share, local recording (Pro), file sharing (Pro), persistent room URL
Miro, Whiteboard.fi, a shared Google Slides file, or a paid whiteboard tool — a lesson canvas you can load material into, once Whereby's own whiteboard stops stretching
A shared Google Drive, OneDrive, or Dropbox folder — where worksheets and lesson recordings live
Calendly, a Google Calendar link, or a dedicated tutoring scheduler — booking and rescheduling that doesn't cost you a message every time
Stripe, PayPal, Wise, or a local bank transfer — getting paid across currencies
OBS, QuickTime, or a browser extension — if the local recording Pro provides isn't enough
A sticker chart, a Google Sheet, or Class Dojo — rewards for younger learners
A spreadsheet — to track which student paid for which package, how many lessons are left, and who owes a make-up

None of that is broken. It's the same friction Zoom-shaped and Meet-shaped tutoring stacks live with, in a friendlier browser video shell. For tutors with a small steady book of adult 1-on-1 clients that's tolerable — plenty of coaching and therapy practices run this way for years, and the Whereby experience genuinely is nicer than the alternatives on that specific job. For tutors teaching young learners on shared screens, or for practices growing past ten or fifteen weekly students, the tab-switching tax compounds every lesson.

When Whereby is a reasonable fit for tutoring

You teach one-on-one adults or teenagers where the lesson is mostly talking — conversation practice, business English, IELTS speaking prep, university-level tutoring by discussion, coding review, professional coaching. A clean video call in a persistent room plus a shared PDF over chat is often enough, and Whereby's UX is easier on both sides than Zoom's.
You value the "no download for the family" experience over any specific feature. If a big fraction of your no-shows or bad first impressions come from "the parent couldn't figure out how to install X," moving to Whereby will genuinely fix that, and that alone can be worth the switch.
Your practice sits inside a European market where the GDPR / European-hosted positioning is a real trust signal for families. Whereby's compliance framing — GDPR, ISO 27001, and HIPAA per whereby.com — will resonate with the kind of family that already opted for European-hosted email and a European-hosted note tool.
You're running a first paid trial lesson with a family that finds new software difficult, and you want the setup to disappear so both of you can focus on whether the tutoring itself is a fit. Whereby's browser-first experience removes one variable from the first impression.
You need a fallback for when your main teaching tool has a bad day. Every tutor should have one; a Whereby room the parent can bookmark is a reasonable one to keep in your back pocket.

When it isn't

You teach young learners (roughly four to ten) and the retention story your parents write in reviews is about engagement — "she never wants to log off" — rather than curriculum. Video calls give a young learner one thing to do (talk when unmuted) and one thing to look at (a face on a screen). It falls short of what an engagement-designed classroom gives them.
You teach reading, phonics, math manipulatives, or any subject where the student needs to act on the material — dragging letters, sorting shapes, placing counters on a number line. Screen share plus voice narration is not a substitute for the student doing the thing themselves.
You teach students on tablets or Chromebooks. Screen share on a small tablet shrinks text below readable size and turns pinch-to-zoom into a wrestling match with the host's scrolling. A cobrowser fixes this; a Whereby screen share doesn't.
You teach students inside mainland China or in another region whose access to European services you can't confirm. Whereby doesn't publish prominent "works behind X" reachability guidance, so the connectivity risk falls on you.
You're growing past the number of students you can hold in your head. Ten students booking and rescheduling in your Whereby room is fine; twenty-five isn't, and the fix isn't a bigger inbox — it's a scheduler with a policy attached and an invoicing tool that reconciles itself.
You want a single login that handles video, the lesson canvas, activities the student can act on, scheduling, and invoicing — and you're past the "am I really doing this?" stage of building a tutoring business. Collapsing the stack is what a purpose-built virtual classroom is for.

What about recording on Whereby?

Whereby's Pro tier (and above) includes Local Recording, per its pricing page. That means the recording file lands on your device — you press record, the app captures the call, and you get an MP4 when the lesson ends. Sharing it with a parent, storing it somewhere they can find it later, and deleting it when you don't need it any more is manual: there's no per-lesson home for a recording, no automatic labelling by student and date, no built-in retention policy, no "share recording with the parent for the lesson she missed" flow inside the tool. Purpose-built tutoring classrooms tend to ship recording as a first-class feature that ties the file back to a specific lesson. See how do I record online tutoring lessons? for the broader workflow — including the consent step you should be walking parents through regardless of which tool you record in; the practical rule of thumb is to get explicit parental consent in writing, check what your local rules require (recording law varies by jurisdiction), and only keep recordings as long as you actually need them.

Purpose-built alternatives — what tutors typically switch to

The bigger reason a lot of long-time Whereby tutors eventually move the lesson surface elsewhere is that the "friendly meeting tool plus a stack of other apps" shape stops paying its rent past a certain practice size. The category that grew up around it — purpose-built virtual classrooms for online tutoring — is described in depth on what is a virtual classroom?, with a head-to-head shortlist of options on the best Zoom alternatives for online tutoring. Both are worth reading if you're actively picking a Whereby replacement rather than defaulting to "well, the browser experience is nice."

Where Koala Go (the classroom we build) fits, specifically:

Also browser-based, no download, no separate app account for the family. The parent bookmarks a link the same way they'd bookmark a Whereby room. So the "no install" experience that draws tutors to Whereby is not something you give up moving to a real teaching tool.
An interactive whiteboard with PDF, PowerPoint, and image upload. The file becomes the lesson surface — the student annotates the same page you're teaching from. This is the difference between a meeting whiteboard and a teaching one: it isn't a blank sketch surface next to the call, it's where your actual material goes.
A cobrowser (a live shared browser window). You and the student open the same real webpage together, both with your own cursor, both able to click and type. Wordwall, Boom Learning, Khan Academy, IXL, a leveled-reader site — the student does the activity instead of watching you click. Koala Free ships a limited cobrowser (session-length capped); Koala Pro removes the cap. The cobrowser is region-aware, with separate routing for mainland China so students there aren't fighting the Great Firewall inside a Western tool.
A 3D Playground for younger students. A shared 3D world where each student has a customisable avatar. Used as a brain break, a themed lesson surface, or the reason a seven-year-old wants to come back next week. A browser video call gives a kid nothing to be in; the Playground gives them a place.
Worldwide invoicing built in. Parents pay in their local currency, including via WeChat Pay in mainland China. There's a 4.5% Koala fee on Free-tier invoicing that drops to 3.5% on Pro and 0% on Platinum. No separate Stripe reconciliation.
Integrated scheduling on Pro. A teacher-and-parent scheduling portal where you set weekly availability once and parents book their own slots via a link. No Calendly bolt-on. Platinum adds a Scheduling Plus layer with 30/45/60-minute class lengths, booking-policy controls, and recurring classes — the "book all four Wednesdays for the month at once" flow a persistent room URL and a shared email chain can't hold.
Lesson recording on Pro and Platinum. Koala Pro records the lesson locally to your own machine (the same shape Whereby offers); Koala Platinum adds cloud recordings that save inside the classroom, per-lesson, so a family whose kid missed a session can catch up without derailing next week.
24/7 live support on Pro with a real person. When something breaks mid-lesson, there's a human to reach rather than a help-centre article.

Koala Free is a genuine free tier (up to 4 students per session, worldwide invoicing, limited cobrowser, up to 3 activities), not a trial that expires — and it doesn't cap the lesson length the way Whereby Free's 30 minutes does — so you can run a real trial lesson end-to-end with a real student, on the device they actually use, and decide from there. Koala Pro is $25.99/month billed monthly or $21.99/month billed annually; Koala Platinum, which adds cloud recording and richer scheduling, is $49.99/month billed monthly or $39.99/month billed annually. That is more than Whereby Pro's $10.99 per host per month, and the honest comparison is that you're paying for a different thing: a tool built for teaching, not a friendly meeting tool with teaching bolted on.

Honest limits worth naming so you can weigh them against the fit: iPad works but is rougher than desktop; mainland-China connectivity is served by our Hong Kong proxy but is a moving target; the whiteboard's math tools ship in Pro (the symbol palette's operations, comparisons, and fraction slash) and Platinum (the powers, the advanced symbols, and the drag-and-drop fraction bars) but stop short of a full graphing or equation editor, so heavy STEM tutors should test the workflow on a real lesson first; and the Playground is actively developed with requested features (group-select of Playground objects) still on the roadmap. Roughly 81% of the tutors on our 2026 product/market fit survey said they'd be very disappointed if they could no longer use Koala Go — the "very disappointed" threshold conventionally taken to indicate strong product/market fit is 40%, so that number is well past the bar. That 81% figure is the same one we publish on the main landing page.

Keep Whereby for what it's good at

Nothing above is an argument to close your Whereby account. It is a genuinely good meeting tool, and for adult 1-on-1 talking lessons, therapy sessions, coaching calls, first-lesson stopgaps, and reliable-fallback duty, it earns its keep. Where it doesn't stretch is being your primary tutoring surface for young learners, for subjects the student needs to act on, or for a practice growing past the size where you can hold every booking in your head. Use it where it fits; use something built for teaching where it doesn't. Two tools, one for meetings and one for teaching, is not more complicated than one wrong tool held together with five bolt-ons — and past a certain practice size, it is materially simpler.

Bottom line

Whereby is one of the friendlier meeting tools around and a fine choice for one-on-one adult and teenage lessons that are mostly talking, for a first-lesson stopgap, and for reliable-fallback duty. For anyone teaching young learners, teaching subjects the student needs to act on, growing past a hand-held book of students, or serving families where a full stack of bolt-ons has stopped paying its own rent, the honest answer is that Whereby is where you take the meeting — not where the lesson itself should live. If you'd like our read on whether Koala Go fits your specific practice — subject, age range, region, current stack — write to koala@teachwithkoala.com with a sentence about the kind of tutoring you do. We're a small team and we read everything.

Should I record my online tutoring lessons at all?

Not every lesson. Recording is genuinely useful for four specific situations, and unhelpful or counterproductive in most others. Decide on a default first, then exception in either direction.

Record on purpose when:

The parent has asked to see how lessons go. Especially common with parents of young children who don't sit in on lessons. A short clip — not the whole hour — answers the parent's question without intruding on the lesson.
The student or family wants to review later. Adult learners and exam-prep students often value a recording of a complex explanation (a tricky grammar pattern, a worked math problem, a vocabulary set) so they can rewatch instead of re-asking.
You're reviewing your own teaching. Watching yourself on video is the single best free professional-development tool an indie tutor has. Most tutors never do it. It's worth recording one lesson a month and watching the first 10 minutes.
You expect you may need evidence of what happened. A disputed cancellation, a payment chargeback, or any safeguarding concern. Rare, but the recording is the only objective record.

Don't default to recording when:

The student is visibly less natural on camera once recording starts. Some kids (and some adults) freeze. If recording changes the lesson, the lesson is worth more than the recording.
The parent hasn't agreed. "I'll just record in case" without explicit consent is the single most common compliance mistake we see tutors make.
You don't have a plan for what to do with the file. A recording you'll never watch, never share, and never delete is a privacy liability sitting in a folder.

Get consent before you press Record

Recording a minor in an online lesson is governed by the same baseline as recording any other private conversation, plus an extra layer because the subject is a child. The specifics depend on your jurisdiction — and the parent's — but the working principles hold almost everywhere:

Ask the parent in writing before the first recording, not after. A two-sentence email is enough: what you plan to record, why, where it'll be stored, and that they can ask you to delete it at any time. Save the reply.
Tell the student that recording is starting. A short verbal "I'm going to record this lesson so your parent can see how we did with the vowel sounds today, okay?" is good practice and matches what a reasonable parent would expect to hear. Most teaching platforms (Koala Go included) also show the student a visible "recording" indicator on their screen the whole time, but the verbal heads-up matters more.
Know whether your jurisdiction is one-party or two-party consent. In most of the US, federal law and many states allow recording with only one party's consent (the tutor's); a minority of states ("two-party" or "all-party" consent states — California, Florida, Illinois, Maryland, Massachusetts, Pennsylvania, Washington, and a handful of others) require everyone in the call to agree. The US Department of Education's COPPA rules apply if you're collecting information about a child under 13. Most of Europe, the UK, and Australia operate under data-protection regimes (GDPR, UK GDPR, the Australian Privacy Principles) that treat lesson recordings of identifiable children as personal data, with parental consent and a clear purpose required. Check your own situation; this Q&A is not legal advice.
Put it in your terms. If you have a one-page lesson agreement parents sign before lesson 1, add a line: "I may record lessons for [these purposes]; recordings are stored at [where] and deleted at [when]; you can opt out or ask for deletion at any time by writing to [email]." If you don't yet have an agreement, this is a good reason to start one.

What to record (and what not to record)

The single safest default: record the lesson itself — the shared learning canvas, the teacher and student video tiles, and the audio — and don't record what happens before and after. A few specific things to keep out of the recording:

The pre-lesson chat with the parent. If the parent appears on camera before the lesson to hand the device over to the child, start recording after that conversation, not during it.
Anything inside a logged-in account that isn't the student's. If you've signed in to your own Google account, a curriculum portal, or a tutor CRM in another tab, don't record over to it. Stay inside the classroom you've set up for the lesson.
Other students. If you teach back-to-back students from the same room, stop the recording fully between lessons. Different student, different consent, different file.
Family members in the background. Especially with younger students at home, siblings or parents may wander into frame. Either ask them to stay out, frame the camera tighter, or stop recording for those minutes.
Sensitive personal information. If a parent shares an address, a school name, or a medical detail during the lesson, that's exactly the kind of thing a recording shouldn't capture.

Where the recording is actually stored

You broadly have three choices, with very different implications:

Cloud recording inside your teaching platform. The recording is captured in the browser and uploaded to the platform's cloud storage as the lesson runs. If your browser crashes mid-lesson, the chunks that have already uploaded are safe. This is what most purpose-built tutoring classrooms do (Koala Go included — see below). Pros: the file survives a crash, you don't manage local files, and the platform usually puts the recording behind your account login. Cons: the file lives on the platform's storage, so deleting it from your end depends on the platform's controls.
Local browser recording. The browser writes the recording to your machine while the lesson runs and you save it at the end. Pros: full control, no third party. Cons: a crashed browser or accidental tab close before you save can lose the whole recording (Koala Go warns about this for its local-recording mode — "Make sure to stop recording before refreshing"), and you're then responsible for storing the file safely yourself.
An external screen recorder (OBS, QuickTime, Loom, an OS-level recorder). Pros: works regardless of what platform you teach on. Cons: you're recording your whole screen, including any notifications, browser tabs, or chat windows that pop up; you're storing the file locally; and it's the easiest of the three setups to fumble the consent / retention side of.

For independent online tutors, cloud recording inside the teaching platform is almost always the right default. The one case the other options win is when the platform you're forced to use simply doesn't record lessons — in which case a local browser recorder or OBS is the fallback.

How to share a recording with a parent

Three rules that hold up across almost every setup:

Share by private link, not by email attachment. Email attachments get forwarded; private links can be revoked. If your platform exposes a per-recording sharing link tied to your account, use that. If you're sharing through your own cloud drive, use the per-file "anyone with the link" or, better, "only people I invite by email" setting.
Don't post recordings to a public URL. No public YouTube uploads, no open Google Drive folder. Both are easy mistakes — both are searchable for as long as they exist.
Share a clip when a clip is enough. If the parent asked "how is she doing with reading aloud?", a 90-second clip of the read-aloud is more useful than a 50-minute file no one has time to watch. Most platforms let you trim or share a segment; if not, a quick edit before sending is usually worth the five minutes.

How long should I keep recordings?

Default to short, not long. A reasonable working rule for an indie tutor: keep a recording as long as it's serving a specific purpose, and delete it once that purpose is done. Concrete versions of that rule:

Parent-comms recordings: delete within a few weeks of sharing. The parent saw it; you've made your point.
Student-review recordings: follow what the family asks for — some want each lesson for a month, some want only specific ones kept indefinitely.
Your own self-review recordings: watch within a month or delete unwatched.
Dispute/safeguarding recordings: keep until the matter is closed, then delete. If it goes to a serious safeguarding process, follow your jurisdiction's guidance on what to retain and for how long.

A standing "auto-delete after 90 days unless I mark this one to keep" mental policy is enough for most tutors. The point is that recordings don't accumulate without a reason.

How recording works in Koala Go specifically

If you're using Koala Go (the platform we build), recording comes in two tiers: Koala Pro records locally to your own machine, and Koala Platinum adds cloud recordings of scheduled lessons stored inside the platform. A few practical facts:

One button inside the lesson. The Record button is in the classroom toolbar. The first time you record, the browser will ask for screen-share permission so it can capture the lesson canvas alongside the camera and audio. After that, it's one click to start, one click to stop.
Recording is gated on student presence. Recording only starts (and stays running) while at least one student is in the room. If the student leaves and you're alone in the classroom, recording is disabled until they re-join. This makes "I accidentally recorded myself prepping for an hour" close to impossible.
Students see a visible recording indicator. Once you start recording, the student sees a notice on their side of the classroom that the lesson is being recorded — which we treat as a baseline, not a substitute for the verbal heads-up and the parent's written consent.
On Platinum, the file uploads as the lesson runs. Cloud recordings are captured in 3-second chunks in the browser and uploaded continuously to our cloud storage (AWS S3) — if your laptop closes the tab partway through, the chunks already uploaded are safe; you lose at most the last few seconds. Pro's local recording instead saves the file to your own machine when you stop, so keep the tab open until the download lands.
Recordings are private by default. Cloud recordings are stored behind your login, not in a public folder; local recordings live only on your machine until you choose to share them. Either way, sharing with a parent is a deliberate action you take per recording.
Local recording from Pro ($25.99/month monthly, $21.99/month annual); cloud recordings on Platinum ($49.99/month monthly, $39.99/month annual). You can run free lessons indefinitely on Koala Free and only upgrade when you actually need recording, larger group sizes, unlimited PDF/PPT uploads, or full-length cobrowser sessions. Platinum adds the cloud side — recordings of scheduled lessons stored inside the platform, plus auto-generated class notes drafted from each recorded lesson.

For white-label / B2B setups (a tutoring company on its own branded domain), recording is also available on Koala for Business and is billed as an optional usage-based add-on rather than a per-tutor feature — write to koala@teachwithkoala.com if that's the shape you're running.

A quick pre-record checklist

Written parental consent on file covering recording, storage, and retention.
A line in your lesson agreement describing what you record, where it's stored, and when it's deleted.
A verbal heads-up to the student at the start of the first recorded lesson.
Cloud storage tied to your account login (not a public folder, not your personal Google Drive root).
A retention default — even just "delete after 90 days unless I mark it to keep".
A sharing default — private link only, never an unlisted public URL.
A test recording with yourself before lesson 1, so you know exactly what the student sees and what ends up in the file.

What "affordable" actually means at institutional scale

"Affordable" looks very different on an institutional procurement spreadsheet than it does to a solo tutor comparing $9 and $25 a month. At the institutional level, the real cost of a whiteboarding tool has four components, only one of which is the sticker price:

License cost across your real teacher headcount. A $10/month tool for one tutor is $1,200/year for ten tutors and $24,000/year for two hundred. Per-seat math compounds fast at scale, and most education organizations have more teachers on the roster than they have teaching at any given hour.
Onboarding cost. If every new teacher needs an hour of training plus a week of "where's the button?" support tickets, the tool's free tier saved you nothing. The cheapest whiteboard for an institution is usually one your teachers already know — which is why "is it built into Microsoft Teams or Google Workspace?" matters more than a $5/month price difference.
Stack cost. A whiteboard-only tool doesn't replace your video platform, your scheduling, your invoicing, your parent comms, your recording pipeline, or your admin oversight. Adding a $15/month whiteboard on top of five other subscriptions for every teacher is rarely the cheapest path; sometimes one slightly-more-expensive classroom replaces three other line items.
Hours actually used. Most education organizations don't teach 40 hours a week, 52 weeks a year. They teach in school terms, around exams, around national holidays, across summer dips and back-to-school spikes. Per-seat licensing is paying full rate during the slow weeks; usage-based licensing isn't. The cheaper model depends on which way your hours actually run.

Hold every option below up against all four — not just the sticker.

The 6 most affordable institutional whiteboarding options in 2026

Listed roughly cheapest-first within each tier. All run in the browser, so students don't install anything. Pricing drifts — re-check each vendor's pricing page before procurement, especially for institutional-volume quotes.

1. Microsoft Whiteboard — free at institutional scale via Microsoft 365 for Education

What it is: Microsoft's collaborative whiteboard, included in Microsoft 365 (consumer and Education tiers) and integrated into Microsoft Teams. Infinite canvas, ink and shape tools, sticky notes, image and document insertion, and live collaboration. Reliable stylus support on Surface and iPad.

How it's affordable at scale: Microsoft 365 A1 is offered free to qualifying K-12 institutions and includes Whiteboard alongside the core Office apps. If your district is already on Microsoft 365 for Education — most US K-12 districts are on either Microsoft or Google — the whiteboard is a zero-marginal-cost rollout.

What it's not: a tutoring platform. There's no lesson structure, no student dashboard, no scheduling or invoicing, no engagement system for younger learners. It's a whiteboard inside the video/comms suite you already run, which is fine if a whiteboard is all you need.

Best for: K-12 schools and districts already standardized on Microsoft 365 / Teams; higher-ed institutions where Microsoft is the default identity provider.

2. Google Workspace for Education + FigJam — free Workspace tier, FigJam as the canvas successor to Jamboard

What it is: Google Workspace for Education Fundamentals is free to qualifying institutions and includes Google Meet, Classroom, Drive, and Calendar. Jamboard — Google's previous whiteboard — was retired at the end of 2024; Google's recommended successor for the canvas use case is FigJam, via a partnership with Figma.

How it's affordable at scale: The Workspace tier itself is free for qualifying schools, which makes the Meet / Classroom / Drive backbone a zero-cost rollout. FigJam has a free starter tier and education pricing; for institutions, that turns into a per-editor license rather than free-at-scale.

What it's not: a single integrated whiteboard the way Jamboard was. If you used to teach with Jamboard at the institution level, FigJam isn't a 1:1 replacement and isn't designed for K-12 tutoring specifically. Many former Jamboard users have moved to Microsoft Whiteboard if they wanted the closest free equivalent or to a tutor-built tool if they wanted features Jamboard never had.

Best for: K-12 schools and districts already on Google Workspace for Education whose primary use is general classroom collaboration rather than 1-on-1 tutoring.

3. BitPaper — from ~$10/mo per tutor, multi-page papers with built-in audio/video

What it is: A browser-based whiteboard built specifically for tutors, originally by working tutors in the UK. Auto-saved multi-page "papers" you share with each student via a permanent link, with a built-in audio/video call, document upload (PDF, images), and a pen-friendly drawing canvas.

How it's affordable at scale: Solo-tutor licensing starts from around $10/month per tutor. For a tutoring company with twenty tutors that's $200/month — workable, but per-seat math means it climbs linearly with headcount whether or not those tutors are actively teaching.

What it's not: a full institutional classroom. There's no admin dashboard for oversight across a tutor team, no shared activity library teachers can copy between accounts, and no white-label option to put your tutoring company's brand on the experience. Each tutor has their own BitPaper account.

Best for: Small tutoring practices (1-10 tutors) where each tutor manages their own students directly and the org doesn't need centralized oversight.

4. Vedamo — ~$25/mo per teacher, with LMS integrations and up to 50 students per session

What it is: A browser-based virtual classroom with interactive whiteboard, breakout rooms, screen sharing, reusable lesson templates, and support for up to 50 active participants per session. Integrates with Moodle, Canvas, and other LMSes via LTI. About $25/month for the Virtual Classroom plan, with a 30-day free trial.

How it's affordable at scale: Reasonable per-teacher pricing combined with genuine support for group classes (up to 50 students), so a language school or group-tutoring organization can run small classes without paying for an enterprise tier. The LTI integrations mean institutions already running Moodle or Canvas can drop Vedamo into their existing LMS rather than running two parallel systems.

What it's not: the most engagement-driven option for very young students — the interface is more meeting-like than play-like. Per-seat math still applies; a hundred-teacher organization is looking at roughly $2,500/month base before any LMS- or volume-tier negotiation.

Best for: Group-class organizations (3-50 students per session), language schools, and institutions already running Moodle or Canvas that want a whiteboard-equipped live classroom alongside their LMS.

5. Lessonspace — embeddable subject-specific whiteboards for organizations

What it is: A virtual classroom built around a powerful collaborative whiteboard with subject-specific tools — an equation editor, a graph editor, a code editor with syntax highlighting, document annotation, and session recording. Students join via a "Space" link with no signup. Available in 10+ languages. Embeddable APIs for tutoring organizations.

How it's affordable at scale: Lessonspace's institutional model is the API: rather than paying per teacher seat, larger tutoring organizations can embed the whiteboard inside their own product and pay for usage. For a tech-capable tutoring company that already runs its own front-end and wants the whiteboard as one component, this is typically the cheapest path because you stop paying for everything else in a virtual-classroom product you weren't going to use.

What it's not: a turnkey institutional product. The API path assumes you have engineering capacity to integrate it. Pricing for the embedded-API option isn't published; speak with their sales team for a quote at your usage profile.

Best for: STEM-heavy tutoring companies, math/science tutoring organizations, and EdTech businesses that want a whiteboard as an embedded component inside their own classroom product.

6. Koala for Business — usage-based pricing, white-labeled classroom with the whiteboard included

What it is: A white-labeled version of Koala Go for tutoring organizations, schools, and language schools. Includes the interactive whiteboard with PDF and PowerPoint upload, a cobrowser (you and the student interact with the same live webpage together), a 3D Playground for younger learners, scheduling, worldwide invoicing (including WeChat Pay for families in mainland China), session recording, and 24/7 live support for every tutor and every family. Hosted on your own domain or on a Koala subdomain you choose — your logo, your colors, your name on the login page.

How it's affordable at scale: Pricing is $1.10 per hour per student for the classroom and all features, plus $0.10 per hour per student if you want session recordings, with a $120 monthly minimum covering your first 100 active hours each month. Only active hours are billed — hours where a student is actually in the classroom with a tutor. A tutor in the classroom alone (prepping, testing, or waiting for a no-show) is not billed.

What that means in practice for a few institutional shapes:

A small tutoring practice running 200 active student-hours/month pays $220/month, regardless of how many tutors are on the roster.
A growing organization running 1,000 active hours/month pays $1,100/month — no extra seat fees as tutors are added.
An organization that fluctuates seasonally — 1,500 hours/month in school terms, 400 hours/month in breaks — pays accordingly each month, rather than full per-seat rates while students are on holiday.

What it's not the cheapest at: very small operations or orgs already deeply embedded in Microsoft 365 / Google Workspace whose existing whiteboard is "good enough." The $120/month minimum is a tax on a 50-hour/month operation; below ~110 active hours/month, the per-hour math is worse than the headline rate.

What's bundled that other options usually aren't: the white-label experience (your brand on the login page, your domain, your customer-facing emails), an admin dashboard with filterable, searchable, CSV-exportable lesson and session data, a 30-day contractual go-live from signature, a Customer Success Manager available via Slack, and a Copy Activities tool in the business portal that lets an admin clone activities from one teacher's library into others' — useful for onboarding new tutors with a vetted starter set or distributing a curriculum update across the team. Used in production by Tutor Doctor franchises, Nao Now, The Literacy Corner, Nihaoma Mandarin, Ladder Learning Services, Learning Success Academy, and Imerica Education.

Best for: Tutoring companies with 5+ tutors who want one platform replacing video + whiteboard + scheduling + invoicing + recordings + parent comms; reading-intervention and learning-difference practices; language schools teaching young learners; cross-border tutoring organizations (especially mainland China — the Hong Kong proxy infrastructure is documented at /answers/teach-english-online/teach-students-in-mainland-china).

Per-seat vs. usage-based pricing — which is cheaper for your institution

This is the single biggest cost decision and it doesn't have a universal answer — it depends on the ratio of teacher headcount to actually-taught hours. A simplified rule of thumb:

Per-seat is cheaper when teaching hours are high and steady. Twenty tutors who each teach 25 hours a week, 48 weeks a year, on a $25/month classroom = $6,000/year for 24,000 student-hours = $0.25/hr. That's hard to beat on usage rates.
Usage-based is cheaper when teaching hours fluctuate, when teachers are part-time, or when the roster has high turnover. The same twenty tutors averaging 8 hours/week each across the year (160 hours/week × 48 weeks = 7,680 student-hours) at $1.10/hr/student = $8,448/year — but compare that to per-seat licensing across a more realistic roster (where roughly half of "active" tutors stop teaching mid-year), and the picture changes again. Per-seat licensing pays full rate for a tutor whether they taught 30 hours or 3 hours in a given month; usage-based pays nothing for the 27-hour gap.
Free institutional tiers win for general-classroom use. A K-12 school whose students are using a whiteboard for in-class collaboration (not 1-on-1 tutoring) is almost always best off on the whiteboard bundled with their existing Microsoft 365 or Google Workspace for Education tier — the marginal cost is zero and the workflow doesn't need anything more than that.

Build a small spreadsheet of your actual taught hours over the last 12 months (or your projection) before talking to a vendor. The conversation goes much better when you can say "we taught 9,400 student-hours last year, here's the seasonal distribution" than when you're guessing.

What institutional buyers should evaluate beyond price

Once a tool clears the price bar, the boring operational questions become the deciding ones. In rough order of how often they trip up institutional rollouts:

How does a parent or a student reach a human when something breaks mid-lesson? If the answer is "email, two business days", your operations team becomes tier-1 support by default — and that operational cost dwarfs any whiteboard license fee. Koala for Business includes 24/7 live support for tutors and families; some standalone whiteboards are email-only.
What does your operations team see? An admin dashboard that shows every lesson, every tutor, every session — filterable, CSV-exportable, with per-session revenue and device compatibility flagged — is operational infrastructure, not a nice-to-have. Without it, your finance and ops teams are reconciling spreadsheets manually.
How do new tutors get up to speed? A whiteboard that's part of Microsoft Teams trains itself; a whiteboard with custom tutoring tooling needs onboarding. Tools with a shared activity/curriculum library (so you can clone a starter set into a new tutor's account) cut onboarding time materially compared to "every tutor builds their own materials from scratch."
Where are your students physically? If you teach a meaningful number of students in mainland China — common for ESL and Mandarin schools — most general-purpose tools struggle with the Great Firewall. Confirm the routing situation before you commit a cohort. Koala for Business runs its mainland-China classroom traffic through a Hong Kong proxy and has serving B2B customers (Nihaoma Mandarin among them) with that setup as standard.
Is the experience on your brand or theirs? Parents recommend brands they remember. A whiteboard at vendor.com/your-org teaches families that the vendor is the school. A whiteboard at classroom.your-org.com teaches families that your school is the school. White-label availability — and whether it's a logo swap or genuine domain-level branding — is a real business consideration, not vanity.
What's the exit clause? Multi-year lock-ins should make you nervous. A 90-day either-side termination clause (which is standard in Koala for Business agreements) lets you re-evaluate annually without renewal pressure. Ask for it explicitly.

A note on Google Jamboard for institutions that used to standardize on it

If your district or organization used to teach with Google Jamboard, that product was retired at the end of 2024 and the companion Jamboard devices reached end-of-life at the same time. Most institutions we've spoken to moved one of three ways: to Microsoft Whiteboard if they were dual-running Workspace and Microsoft 365 and wanted the closest free equivalent; to FigJam if they wanted a richer canvas and could absorb the per-editor cost; or to a tutor-built tool (BitPaper, Lessonspace, Koala for Business) if their primary use case was 1-on-1 / small-group tutoring rather than whole-class collaboration. There isn't a like-for-like Jamboard successor inside Google Workspace; pick the option that maps to how your teachers actually used it.

Where Koala for Business fits — and where it doesn't

Koala for Business is built for tutoring organizations, language schools, reading-intervention practices, and small online schools that want one branded platform handling the live lesson, the whiteboard, the scheduling, the invoicing, the recordings, and the parent support. The usage-based pricing is designed to make that work for organizations whose teaching hours vary — by season, by student cohort, by tutor turnover — and to remove the per-seat-license trap where you're paying full rate for tutors who only teach part-time.

It is not the cheapest option for every institutional shape. If you're a K-12 school district whose whiteboarding need is general in-class collaboration, Microsoft Whiteboard (free with Microsoft 365 A1) or FigJam (with Google Workspace) will be cheaper and adequate. If you're a small practice with two tutors and 80 student-hours per month, the $120/month Koala for Business minimum is a tax — the indie tier of Koala Go (Koala Pro at $25.99/month per tutor) is the right price-point at that scale. If your tutoring is whiteboard-only with no need for video, cobrowser, scheduling, or invoicing, a standalone whiteboard saves you what you'd pay for capabilities you won't use.

Where it does fit, the math typically works out cheaper than per-seat alternatives once you cross ~5 active tutors with variable teaching loads. If you'd like a numbers comparison against your actual taught hours, write to koala@teachwithkoala.com with a rough hours-per-month figure and a sketch of your team, and we'll work through it with you. Applications for Koala for Business go to business.teachwithkoala.com/apply; we read every application and respond within one business day. The standard go-live is 30 days from signature, written into the agreement.

The short list — what you actually need before lesson one

A computer that runs a modern browser smoothly (any laptop from the last 5 years).
A USB microphone or USB headset — skip the built-in laptop mic.
A webcam (built-in is fine to start).
A reasonably quiet room, with light hitting your face from the front rather than behind you.
Stable internet — wired Ethernet if you can; 10+ Mbps up and down is comfortable.
A virtual classroom for the lesson itself.
A way to schedule lessons (calendar tool or scheduler).
A way to take payment (Stripe, PayPal, Wise, or a classroom with built-in invoicing).

Total starter spend: $0-$200 if you already own a laptop. Tutors with full calendars run for years on this setup.

Hardware — your kit, in order of impact

Spend in this order, not the other way around.

1. Audio (this is the lever)

A USB microphone or USB headset. A $40-$80 USB condenser mic — the Blue Yeti and Samson Q2U are common picks at that price band; any reputable USB mic of similar specs works — lifts every lesson immediately. Parents and students will forgive a slightly soft camera; they will not forgive audio that's painful to listen to. If you only upgrade one thing, upgrade this.
Closed-back headphones with a built-in mic are a good alternative — fewer cables, no echo from your speakers, fine for indie tutors. Avoid open-ear earbuds (AirPods and similar) for back-to-back lessons; the mic quality is workable but they leak the student's audio out into the room.
Test your audio with someone who isn't on your network before lesson one. The room sounds different to your laptop than it does to you.

2. Computer

Any laptop or desktop from the last five years that runs Chrome, Firefox, or Edge cleanly. Modern browsers handle WebRTC video plus a whiteboard plus a shared browser at the same time.
If you're choosing new: 16 GB of RAM is the comfortable floor when you'll have the classroom, the lesson plan, and a few reference tabs open at once.
An iPad as your primary teaching device is doable but rougher than a laptop — most virtual classrooms run on iPad and feel "mobile" in places, and some controls are reduced. iPad works better as a secondary annotation tablet alongside your laptop than as the only device.

3. Camera

Start with your laptop's built-in camera. It's usually good enough.
If you upgrade later, an external 1080p USB webcam at eye level with the student is the right next step. The framing — eyes-to-eyes, not eyes-to-chin — matters more than the megapixel count.
A document camera is a real upgrade for tutors who teach handwriting, math working, or paper worksheets. Inexpensive USB doc cameras work fine.

4. Lighting

A single soft light in front of you, slightly off to one side, is the difference between "professional" and "cave dweller". A $20-$40 USB ring light or small LED panel is enough.
Avoid back-lighting (light behind you puts you in silhouette) and direct overhead lighting (raccoon-eye shadows). A window in front of you, during daylight, beats most artificial setups for free.

5. Internet

10 Mbps up and 10 Mbps down is comfortable for one video lesson plus a whiteboard. 25+ Mbps is comfortable with margin. Test at the time of day you actually teach, not at midnight when the network is empty.
Wired Ethernet to your laptop beats Wi-Fi for reliability. If your laptop has no Ethernet port, a USB-C to Ethernet adapter costs about $15 and is the cheapest upgrade you'll ever make to lesson stability.
If you teach across continents — especially to students in mainland China — your platform's routing matters more than your home speed. See how to teach English online to students in mainland China for the routing detail.

6. Optional but worth it

A second monitor. Lesson plan on one screen, classroom on the other. Doubles your prep efficiency once you teach more than a few lessons a week.
A stylus or pen tablet. Useful for phonics, letter formation, math working, or anything where dragging a mouse is too clumsy. Apple Pencil with an iPad as a secondary device, or a low-cost Wacom Intuos for laptop users.
A door that closes on the room you teach in. Pets and family members are the most common lesson interruption for indie tutors working from home.

Software — the seven jobs your stack has to cover

Whichever path you take, an independent online tutor needs to handle these seven jobs:

Live video — the call itself, with a stable connection on whatever device the family already owns.
A shared canvas / whiteboard — for drawing, annotating, working through problems with the student rather than at them.
Lesson materials — slides, PDFs, flashcards, worksheets, sometimes a published curriculum (see should I use a published curriculum or build my own?).
Scheduling — booking, rescheduling, automated reminders so parents don't forget the lesson.
Invoicing and payments — billing parents and getting paid in their currency without three days of bank-transfer waiting.
Recording (optional) — lesson capture for parent review or self-review, with consent and a storage plan. See how do I record online tutoring lessons?
Parent communication — quick updates after each lesson, monthly progress notes, and a way for the parent to reach you between lessons.

You can cover those seven jobs two ways.

Path A — Stitch a stack together

A typical first stack for an indie tutor:

Zoom or Google Meet for video ($0-$15/month).
Miro, FigJam, or Whiteboard.fi for a shared whiteboard ($0-$20/month).
Google Slides + Drive for lesson materials (free).
Calendly or Google Calendar for booking (free-$10/month).
Stripe, PayPal, or Wise for invoicing (transaction fees only).
A spreadsheet to track who paid for what package and how many lessons remain.

Cost: roughly $0-$45/month, depending on which free tiers you stay inside.

Where it works: tutors teaching adult learners on clean 1-on-1 calls, where the lesson is mostly conversation or worked examples and you don't need rewards, avatars, or a touch-friendly canvas.

Where it shows its seams: with young learners (free-tier video calls have 40- or 60-minute caps that bite mid-lesson, no rewards system, the whiteboard is painful on a tablet); with parent communication (lessons live in five tools that don't cross-reference); and with the bookkeeping at scale (you'll reconcile Stripe payments against a calendar against a spreadsheet by hand at month end).

Path B — One purpose-built tutoring classroom

A purpose-built virtual classroom bundles most of the seven jobs into a single login. The category includes Koala Go (the platform we build), Lessonspace, BitPaper, Vedamo, LearnCube, and Bramble. They differ in detail; the common shape is "one login covers the lesson, the materials, the schedule, and the bill". For a working comparison, see the best Zoom alternatives for online tutoring and the best online whiteboards for tutoring.

Cost: typically $10-$30/month per tutor on the paid tier, with most platforms offering a free tier you can start on.

Where it works: tutors of younger learners (the engagement features pay for themselves in retention), tutors who teach 10+ paid hours a week (the time saved from not juggling tools shows up fast), and tutors who want to look professional to parents from lesson one.

Where it doesn't: if you only teach adults, only on 1-on-1 video calls, and need nothing more, you're paying for capabilities you won't use.

Stitched stack vs. all-in-one — the honest comparison

The stitched stack is cheaper at face value — most pieces have free tiers. The bill is in your time: stack-juggling, "where's the link?", payments that don't reconcile, parents who can't find last month's lessons.
The all-in-one classroom trades $20-$30/month for the consolidated workflow, plus engagement features built specifically for tutoring (rewards, avatars in some platforms, drawing tools designed for the lesson rather than for office meetings). For tutors of younger learners, those engagement features are usually the actual reason to switch.
Either path can work. The honest test: run both for two weeks, count the times you switched tools mid-lesson, and notice whether your students stay in the lesson longer.

How Koala Go fits — and how it doesn't

Koala Go is the all-in-one classroom we build. It runs in the browser (so the student installs nothing) and includes built-in video, an interactive whiteboard that takes PDF, PowerPoint, and image uploads, a cobrowser (you and the student interact with the same live webpage together, not a screen share), a 3D Playground used as a brain break or a full lesson environment with student avatars, built-in scheduling, worldwide invoicing (parents pay in their local currency, including via WeChat Pay in mainland China), and 24/7 in-app support. Free to start. Koala Pro unlocks lesson recording, group sessions, unlimited PDF/PPT uploads, and full-length cobrowser sessions; Koala Platinum adds Scheduling Plus (30/45/60-minute class lengths, recurring classes, booking policy controls), cloud recordings, and auto-generated class notes after each lesson.

Where Koala Go fits well: independent K-12 tutors, ESL tutors of younger kids, reading and Orton-Gillingham practitioners, and anyone whose retention problem is "the student is bored on Zoom."

Where Koala Go isn't the right pick: tutoring exclusively to adult learners on clean 1-on-1 video calls (a video tool plus a Stripe invoice is usually enough); STEM tutoring whose biggest need is a full equation/graph-editor whiteboard (Koala Go's whiteboard now includes a math symbol palette on its paid tiers, but dedicated STEM canvases like Lessonspace still go deeper); whole-class delivery to a 25-student K-12 classroom (Nearpod and Pear Deck are built for that shape).

Three budget tiers — what each gets you

Approximate dollar figures; spend more or less depending on what you already own.

Starter (~$0-$200 setup, $0-$30/month)

The laptop you already own.
A $40-$80 USB mic or a decent USB headset.
Your laptop's built-in webcam.
A free or low-tier virtual classroom (Koala Free, Whereby Free, or a free Zoom/Google Meet account if your lessons fit inside their group-call time caps).
A Stripe or PayPal account for invoicing, or a classroom plan with invoicing included.

Enough to run paid lessons from day one. Most tutors get started exactly here.

Polished (~$500-$1,000 setup, $25-$50/month)

The starter list, plus an external 1080p webcam, a small LED panel, closed-back headphones, and a USB doc camera if you teach handwriting or paper-worksheet subjects.
A paid virtual classroom plan with recording, scheduling, and parent-friendly invoicing built in.
A subscription to whichever published curriculum you teach from, if your subject has one (Cambridge "Super Minds", Oxford "Family and Friends", and similar are common ESL picks — see published curriculum vs DIY).

This is the kit most full-time indie tutors converge on within their first year.

Established (~$1,500-$3,000 setup, $50-$150/month)

The polished list, plus a higher-end USB condenser mic on a boom arm, a 24-27" external monitor, an iPad or Wacom tablet as a secondary annotation device, and a small library of physical realia (puppets, manipulatives, flashcards) for younger learners.
A paid tier of your virtual classroom that includes recording and any group-size capacity you actually need.
If you've grown past a solo practice into a small team, consider a white-label classroom rather than per-tutor licensing — see affordable online whiteboarding for educational institutions for the math.

Spending more than this rarely makes you a better tutor. The ceiling on what gear adds tops out fast; the next compounding investment is curriculum and lesson craft.

Common setup mistakes new tutors make

Buying a green screen before fixing the mic. Audio is the lever; visuals are downstream.
Choosing a platform that locks you into one student device. iPad-only or Windows-only platforms shrink your addressable family base. Pick a browser-based classroom that works on whatever device the family already owns.
Recording every lesson "just in case" without a consent and storage plan. See how do I record online tutoring lessons? for the consent, storage, sharing, and retention defaults.
Running on a free-tier video plan with a 40- or 60-minute group-call cap. The mid-lesson disconnect is one of the most common reasons a trial doesn't convert to a paying student.
Skipping the cancellation policy. Tools matter less than rules. Set yours before lesson one — see how to handle student no-shows and last-minute cancellations.
Optimising for "free" over "consolidated". A stack of six free tools you switch between mid-lesson can cost more in time than a single $25/month classroom that does the same jobs in one tab.

A pre-lesson checklist — run it before your first paid lesson

Test the mic. Record yourself for 30 seconds and play it back. If it sounds muffled, hissy, or echoey, fix it now.
Test the camera. Frame eyes-to-eyes with where the student's face will be. Light on your face, not the wall behind you.
Test the lesson link on a second device — a phone, a friend's laptop, a family member's tablet — the way the parent will see it. Anything that asks for an install or a sign-up is a friction point you'll repeat with every new family.
Test the recording, if you'll record. Confirm it captures the lesson canvas, the audio, and your camera — not just one of the three.
Test the invoicing flow with a $1 test invoice to yourself. Knowing exactly what the parent's email looks like before they receive one is worth five minutes.
Open the materials for the lesson in the same browser window before the parent joins. Switching apps with a six-year-old already on screen is how lessons fall apart in the first three minutes.

Questions about Koala Go specifically — what's bundled, whether it fits your subject and student ages, pricing for a particular shape of practice — go to koala@teachwithkoala.com. You can also try it without installing anything at classroom.teachwithkoala.com.

The four inputs that set your rate

Don't pick a rate from a single reference point. Triangulate from these four:

Your floor. What price makes the hour worth it after platform fees, payment processor fees, taxes, and the unpaid time around the lesson — prep, parent messages, scheduling. New tutors spend 0.5x-1x the teaching time on prep; experienced tutors with reusable materials drop close to zero. If your published rate divided by your real total-time-per-lesson lands under your local minimum wage, the floor is too low.
The market your target families anchor on. Parents anchor on local tutoring prices, not on a global average. A working US suburban family will calibrate against $40-$80 local in-person tutoring; a working family in many Asian or European markets calibrates against a different band. Spend an hour reading the rates of independent tutors on Wyzant, Preply, italki, your local Facebook groups, and a couple of competitor tutoring sites in your target region. Don't price against the agency floor on Preply or Cambly — those agencies subsidise a low rate by taking a cut and bringing the student. Price against other indie tutors of your subject and age group.
Your differentiator. Credentials (a relevant degree, a specialist certification like Orton-Gillingham, TEFL, IELTS examiner training), years of experience, an unusual subject (less competition), a track record with a specific student population (test scores, university acceptances, parents who'll say so), and the language(s) you can communicate with parents in. Every credible differentiator is worth a step up the band.
Your retention model. A tutor who runs monthly packages with a clear cancellation policy can hold a slightly higher per-lesson rate than a pay-per-lesson tutor at the same level of experience, because the package is structurally less risky for the parent. See how to structure and price lesson packages for the packaging mechanics — this page is about the per-lesson number underneath them.

Triangulating from all four beats picking a number that "feels right". The rate that feels right to a new tutor is almost always too low — the floor and the market both point higher than imposter syndrome wants to let you charge.

Rough rate bands by experience tier

The numbers below are directional ranges, not gospel; they drift with the market and shift with the subject (the next section). Treat them as starting brackets to triangulate inside, not as a price list to copy.

Agency / marketplace platforms (Preply, italki, Cambly, Outschool, Wyzant before fees): roughly $10-$22/hr after the platform's cut, depending on credentials and the platform's rate floor. Useful for getting first reps and reviews; weak as a long-term income base because the platform owns the relationship and takes 20-40%.
New independent tutor (0-2 years, no specialist credential): $20-$35/hr. Strong enough to be taken seriously, low enough to convert your first paying families before you have testimonials. Don't price below this — discount-heavy entry pricing attracts rate-sensitive families who churn at the first full-price renewal.
Established independent tutor (2+ years, full or near-full calendar, repeat-renewing families): $35-$60/hr for generalist tutoring (general ESL, primary-school subjects, conversational lessons).
Specialist tutor (exam prep, special-needs, technical subjects, university admissions, high-end test prep): $60-$120/hr and up. Specialists who routinely produce specific outcomes — a particular IELTS band, a SAT delta, a place at a specific university — can sustain rates well above this band.

If your calendar is full and your trial-to-paid conversion is staying above 50%, you're under-priced. The signal of a correctly priced rate is that you turn away the occasional family — usually about 1 in 10.

How subject and specialty change the rate

Same level of experience, very different rates. Some directional differences that hold up across markets:

Conversational / generalist ESL, primary-school subjects, homework help: the base band above. Lots of supply, lots of competition. Where rate compression is most visible.
Math and science tutoring: typically a 15-30% premium over the generalist band, because demand is high and effective tutors are visibly fewer. The premium scales with student level — primary math sits near the base; AP/A-level/IB math and physics sit well above it.
Exam preparation (SAT, ACT, IELTS, TOEFL, GMAT, GRE, Cambridge YLE/CAE/CPE, AP, A-level, IB, university admissions): commonly a 50-100% premium over the generalist band — exam prep is outcome-driven and parents will pay for a tutor who reliably hits targets. The premium grows with the stakes of the exam (a $200/hr tutor for SAT prep at a competitive US high school is normal; a $40/hr tutor for the same is the suspicious one).
Special-needs tutoring (Orton-Gillingham for dyslexia, structured literacy intervention, math intervention, ADHD coaching, autism-specific support): usually a 25-75% premium. Families who need this tutoring know what they're looking for, the qualified tutors are visibly scarce, and the work is harder. Pricing too low here can signal under-qualification, not value.
Music tutoring (piano, violin, guitar, voice): roughly the generalist band for beginners and the specialist band for advanced students preparing for ABRSM, RCM, or conservatory entry. Local in-person rates are a strong anchor for music — many families compare directly to their child's last in-person teacher.
Coding / programming for kids: roughly the generalist-to-specialist band. The market is younger and rates are still calibrating; expect more spread than a mature subject.
Business English, professional communication, accent coaching for adults: the specialist band, billed to the learner (or their employer). Hourly rates of $80-$150 are normal for credentialed coaches working with mid-career professionals.
Foreign-language tutoring (non-English target): wide spread depending on the language. Common Western European languages sit near the generalist band; less-commonly-taught languages can sit higher because supply is thin.

If your subject sits in two of these (e.g. you tutor exam prep and the test-takers have learning differences), the premium compounds — you're now in the smaller pool of tutors who can do both, and rates reflect that.

How geography and market positioning change the rate

Online tutoring is global, but pricing isn't.

Charge in the currency the parent thinks in. A US family expects USD; a UK family expects GBP; a European family expects EUR. Charging in USD to a parent in Vietnam loses some of the value to the parent's currency conversion and bank fees. Pick the currency the parent will compare against local tutors in — most invoicing tools (Koala Go's included) let you bill in different currencies per client.
The market your target families anchor on, not the market you live in. A US-based tutor of US students charges US rates. A US-based tutor of mainland-China-based students charges what working Chinese urban families pay independent foreign tutors — which is its own band, usually lower than US in-person and higher than a Preply listing. Decide who your students are before you set the price.
Premium vs accessible positioning. A tutor at the top of the local band tends to sell on outcome (visible progress, specific results) to fewer, longer-staying families. A tutor at the middle of the band tends to sell on engagement and consistency to more families. Both work; pick a position and price to it. The unsustainable shape is "middle band but treats the practice like premium" — too many slots to spend the time premium service requires, but priced as though you can.

Hourly vs package — what should be in the price

For ongoing students after a trial, default to packages — usually a monthly subscription of 4 or 8 lessons. Packages don't change your per-hour rate; they change the parent's commitment unit. Two specifics this page won't repeat (the packages answer covers them in full):

How to size and discount packages — see how to structure and price lesson packages for the size / cadence / discount-tier shape.
Trial-to-paid pricing. Don't deep-discount the trial. A free or token-priced trial slot followed by a "starter 4-pack at your standard rate" converts more durably than a 50%-off intro promo — the discount-priced intro attracts families who churn at the first full-price renewal.

For the per-lesson rate you set on top of those packages, the inputs are the same as the hourly question: floor, market, differentiator, retention model.

When (and how) to raise rates

Most indie tutors raise rates too late, not too soon. The signal-to-raise is visible:

Your calendar is 70%+ full at your current rate for two months running.
Your trial-to-paid conversion stays above 50%.
Renewals on your current families are running 90%+ month-over-month.
You have testimonials, before/after evidence, and a clear differentiator to point at.

If three of those four are true, raise. A working shape:

Raise by 10-20% at a time. Less than 10% isn't worth the awkwardness; more than 20% provokes churn even from happy families.
Grandfather existing families at the current rate for 6-12 months when announcing the increase. The new rate applies to new students immediately. Tell existing families when they'll move to the new rate, well before they get there.
Announce in a single message per family, not a generic broadcast. Two sentences: "starting [month], my rate goes from X to Y for new students; you're locked in at X until [date]". Don't apologise for raising rates — it primes families to negotiate.
Some families will leave. Expect 5-15% churn on a rate raise. The tutors who never lose anyone never raised enough.

How to actually bill the rate

The tooling matters less than the discipline. Whatever stack you use, an independent online tutor's billing needs to handle five jobs:

Generate an invoice the parent can pay. Multi-line-item invoices ("June 1st class, 20 minutes"; "June 3rd class, 20 minutes"; etc.) read more professionally than a single "tutoring services" line and travel better through a parent's bookkeeping.
Accept payment in the parent's local currency on the rail they actually use. Cards via Stripe or PayPal in most of the world; WeChat Pay or Alipay for families in mainland China; occasional Wise transfers for parents your Stripe doesn't reach.
Send automated reminders for unpaid invoices. Late renewals are the biggest source of unintentional churn — a parent meant to renew and just didn't get to it. One reminder a few days after an invoice goes unpaid recovers a meaningful share.
Track what's paid against what's taught. Who's on a current package, how many lessons remain, what's outstanding. A spreadsheet is enough; an invoicing tool that surfaces a package view is better.
Hold a clear cancellation and no-show policy. Tools matter less than rules. See how to handle student no-shows and last-minute cancellations for the operational defaults.

You can stitch this together — Stripe or PayPal for invoicing, Calendly or Google Calendar for scheduling, a spreadsheet for the package balance — or use a virtual classroom that bundles invoicing and scheduling into the same product. Both work. The stitched path is cheaper at face value; the bundled path saves real time at month-end and removes the "where did that payment go?" reconciliation step.

How Koala Go handles the billing side

Koala Go is the virtual classroom we build; we ship the invoicing and scheduling pieces inside the same product so the per-lesson rate you set on this page actually gets billed and paid without a second tool. Honest about what's included on which plan:

Built-in worldwide invoicing. You generate an invoice from inside the classroom with as many line items as you want (the input prompts for things like "June 1st class, 20 minutes" — the shape a package invoice needs for the parent's records and your own tax return). The parent receives a link and pays in their own currency. Payouts go to your Stripe, PayPal, or Wise account depending on which payment method you have set up.
WeChat Pay end-to-end for parents in mainland China. If you teach families in China, the invoice link works on WeChat without the parent installing anything else. There's a small QR gotcha — the same WeChat app can't both display and scan the QR — and the parent-facing workaround is at /wechat-pay-instructions. Same payment plumbing as the existing answer on teaching students in mainland China.
One-click payment reminders. If an invoice goes unpaid, you send a reminder from inside the classroom rather than drafting a follow-up email.
Scheduling on the Pro plan; Scheduling Plus on Platinum. Koala Pro ($25.99/month, or $21.99/month billed annually) includes the scheduling tool. Koala Platinum ($49.99/month, or $39.99/month billed annually) adds recurring classes (so a monthly package books all four Wednesdays at once), booking policy controls (so the 24-hour cancellation rule is enforced by the platform rather than your inbox), and 30/45/60-minute class lengths.
Koala invoicing fee tiered to your plan. Koala Free carries a 4.5% Koala fee on each invoice; Koala Pro is 3.5%; Koala Platinum is 0%. Card-network fees from Stripe, PayPal, or Wise are separate and unavoidable on any platform. If you're invoicing a few thousand a month, the difference between the Pro and Platinum fees usually covers the Platinum plan on its own.
Auto-generated class notes on Platinum. If part of what justifies your rate is the progress report you send parents after each lesson (see how to keep parents renewing month after month), Platinum drafts that summary automatically — you edit and send in a minute or two instead of skipping it on a busy week.

Two things Koala Go doesn't do that some tutors assume any billing tool does: it doesn't auto-charge a parent's saved card on a monthly cycle (a monthly package on Koala Go means you invoice the parent each month — one click, they pay one link — and the recurring scheduling books the lessons; this is the same honest framing used in the packages page), and it isn't a tax tool (your accountant or your local tax filing service still handles the year-end side). For unusual billing shapes — corporate language tutoring billed to an employer, a tutoring company on a white-label Koala for Business setup, a sibling-discount multi-month package — write to koala@teachwithkoala.com before you build the rate around it.

Common pricing mistakes new tutors make

Pricing on imposter syndrome instead of the floor. "I haven't been doing this long enough to charge X" is rarely an accurate read of what your hour is worth — and a too-low rate signals "low value" to the families you most want.
Heavy intro discounts. 50%-off first month, "free first three lessons" stacked with discounts. These attract families who will churn at the first full-price renewal, and they train every future negotiation.
Quoting in USD to families who think in another currency. Sticker-price friction at a point of sale costs you conversions you'll never see.
Forgetting to count the unpaid time. Prep, parent messages, scheduling friction, and admin take real hours. If your real total-time-per-lesson is 1.5x the teaching hour, your rate has to reflect that.
Refusing to raise rates. Holding the same rate for three years while the calendar fills is the single most common indie-tutor income mistake. Raise on schedule; grandfather existing families.
Negotiating the rate after publishing it. Publishing a rate then quietly cutting it for an asking family teaches every future parent that the rate is a starting position. Hold the published price; offer a structural alternative if a family genuinely can't reach (a shorter lesson, a smaller package, a less-frequent cadence).

A short worked example

An indie tutor with two years of experience, a TEFL certificate, no specialist exam-prep credential, teaching primary-school ESL to families in a high-income market. They run 22 paid hours a week, all on monthly 8-lesson packages.

Floor. Local effective minimum wage post-tax plus 30 minutes of unpaid time per lesson ≈ $20/hr published rate. Don't go below this.
Market. Local indie tutors of the same subject and age group publish $30-$45/hr.
Differentiator. Two years of repeat-renewing families, a small library of testimonials, TEFL but no specialty. Modest differentiator — one step up the local band, not two.
Retention model. Monthly packages with a clear cancellation policy. A modest premium over a comparable pay-per-lesson tutor is defensible.

Landing rate: $38-$42/hr. Published as $40/hr standard; $38/hr inside the 8-lesson monthly package (a small package discount, per the packages page). Six months later, calendar 75% full and renewals at 90%+ — they raise to $45 for new students and grandfather existing families at the old rate until the new calendar year.

The numbers are illustrative; the framework — floor, market, differentiator, retention — is the part to copy.

What "gamified" means in an online classroom (and what it doesn't)

"Gamified" in tutoring usually means one or more of these mechanics:

A reward currency the teacher controls. Stars, points, coins, gems — something the student can earn during the lesson, see accumulate on screen, and (in many platforms) spend on cosmetics or in-classroom unlocks. Crucially, the teacher decides when and why, not an algorithm. Reward-on-correct-answer is the most common pattern; reward-for-effort and reward-for-on-task-behavior are usually the more powerful ones with younger kids.
An avatar the student owns. A character the student customizes — clothes, hair, eyes, accessories — and uses as their on-screen presence. Owning the character creates buy-in: when a child has spent five minutes choosing their avatar's hair, they are already invested in the lesson the avatar is about to enter.
A shared 3D or 2D world. A space — sometimes a playground, sometimes a themed environment, sometimes a simple 2D scene — where the teacher and student's avatars can move around together. The space is not the lesson; the lesson runs alongside it. The space is the part the student wants to come back to.
Progress visualization. A streak counter, a level bar, a sticker book, a "lesson chain" — anything that makes a continuous practice habit visible to the student and the parent.
Brain-break activities. Quick mini-games or interactive moments between lesson phases — not the lesson, but the breath between teaching segments that resets the student's attention.

What "gamified" does not mean: replacing teaching with games. The mechanics earn their cost when they multiply your teaching minutes; they don't replace them. A lesson that's 90% playground and 10% teaching is not a gamified lesson — it's an unsupervised play date. A lesson that's 80% teaching with a five-minute brain-break in the middle and a points-on-screen system that rewards on-task behavior is the shape that holds up over months of weekly sessions.

Why young learners respond to it (in plain terms)

Three forces are doing the work, and they're worth understanding so you can use them deliberately rather than throw mechanics at a lesson and hope:

Variable rewards hold attention longer than predictable ones. A child who earns one gem for every correct answer disengages faster than a child who earns one to five gems based on effort, occasionally a "bonus" of ten, and the teacher's hand on the slider. This is well-documented in operant-conditioning research and equally well-documented in slot machines — the same mechanism, used in service of literacy rather than against a player's wallet. Predictable, fixed rewards become wallpaper within a few lessons; the rhythm of variable rewards keeps the student leaning in.
Ownership creates investment. A student who has chosen their avatar's hair color, jacket, and shoes treats the avatar — and the classroom the avatar lives in — as theirs. They don't want to skip lessons because that's where their character lives. Compare with a Zoom call: there is nothing on the screen that belongs to the student.
Autonomy, competence, and relatedness. Self-Determination Theory describes three needs that underpin motivation: feeling that you chose what you're doing (autonomy), feeling that you're getting better at it (competence), and feeling connected to another person doing it with you (relatedness). A good gamified classroom hits all three: the student picks their avatar (autonomy), watches their gem count grow (competence), and shares a space with the teacher's avatar in real time (relatedness). The mechanics aren't decoration; they're delivery mechanisms for those three.

None of this is unique to children — adult learners respond to the same forces — but children's attention windows are shorter, so the mechanics show up faster. If your students are seven, you will see the difference within a single lesson. If your students are forty, you may not see it at all, which is fine; gamification is a tool not a value.

Where a gamified classroom helps in tutoring

From working tutors on Koala Go, the situations where gamification reliably moves the lesson:

Young learners (ages 4-10). The single clearest fit. The 30- to 45-minute lesson with a four- to seven-year-old is a different shape from the same lesson with a teenager; the mechanics are usually what keep the student in the chair for the back half.
Students who have visibly disengaged on Zoom or Google Meet. If a parent has told you "she won't sit still on Zoom but she'll play games on the iPad for an hour," the constraint isn't attention span — it's that meetings-software gives a child nothing to do. A classroom that gives them something to do recovers the lesson.
ESL with very young or absolute-beginner learners. Vocabulary is harder when you don't share a language; a shared space where you can both point, drag, and act on the same objects collapses a lot of the "translate, gesture, repeat" friction.
Students with ADHD, autism, dyslexia, or other attention/learning differences. Tutors specifically working in this space — Orton-Gillingham practitioners, dyslexia and dyscalculia specialists, reading-intervention tutors — repeatedly mention that platforms with engagement mechanics let them run a full session where Zoom would have lost the student in ten minutes. The mechanics aren't magic; the multi-modal interaction (visual + spatial + auditory + reward) is what helps.
Building habit and retention. Streak counters and avatar progression give the student a reason to show up beyond "my parent signed me up for this." Across months of weekly lessons, the difference between "I want to go to class" and "I have to go to class" is most of the retention math for an indie tutor's business.

Where a gamified classroom is the wrong tool

Don't use it everywhere. A few cases where the mechanics get in the way:

Adult learners on focused work. A 40-year-old IELTS student or a business-English client does not need an avatar. The cosmetic layer feels infantilizing, the reward currency feels patronizing, and the time spent choosing hair is time off the lesson.
Teenagers who would be embarrassed by it. Roughly between ages 11 and 16 is the awkward zone — old enough to find a cartoon avatar uncool, young enough to need engagement mechanics. Read the room. Some teenagers love it; many will visibly disengage from it; a clean shared whiteboard and a competent teacher is often the better tool.
Pure lecture-style content delivery. If your lesson is mostly you explaining something complex and the student listening, the mechanics are noise. A clean whiteboard, a slide deck, and good audio are enough.
Test prep with a deadline. Two weeks before a board exam, you and the student both want efficient drill-and-review. The avatar is in the way.
Subjects with their own native-interaction surface. A music-theory tutor needs a notation editor more than a 3D playground; a programming tutor needs a real code editor; a physics tutor needs a graphing whiteboard. Use a classroom that puts the right primitive at the center of the lesson.

If you're not sure whether a gamified classroom fits your practice, the cleanest test is to run a real lesson with a real student on the device they actually use, and watch where their attention goes. If they hold attention longer than they did on your current tool, the mechanics are earning their keep. If they spend the lesson distracted by the avatar's wardrobe, they're not.

The engagement-vs-distraction tradeoff (be honest about it)

The same mechanics that hold attention can also pull it the wrong direction. We've seen these failure modes often enough to call them out:

The student wants to play, not learn. A first lesson on a new platform sometimes goes 80% playground exploration and 20% teaching. That's fine for a trial — it's the parent's "see, she likes it" moment — but if it's still happening in lesson five, you've ceded the structure. The fix is not to remove the gamification; it's to gate it. "We do the lesson first, then five minutes of playground at the end if we finish the activity" is a reliable structure most kids accept once you set it.
Choosing the avatar takes longer than the lesson. Many platforms include avatar customization (Koala Go included — body, eyes, hair, shirt, pants, shoes are all editable). For a first lesson with a new student, set aside five minutes for this on purpose; for every lesson after, treat changing the avatar as a between-lessons activity, not a lesson activity.
Rewards stop working when they're predictable. One gem per correct answer is engagement wallpaper after lesson three. The mechanic earns its cost when the student doesn't know exactly what they'll get — a small reward for an okay answer, a bigger one for unprompted effort, an occasional surprise burst. The teacher's judgment, not a fixed rule, is what makes it work.
Younger siblings, pets, or parents wander into the playground. If the student's space at home is shared, the on-screen world can pull attention from other people in the room more than the lesson does. A door that closes on the lesson room is the cheapest fix.
Performance on older devices. A 3D environment in the browser asks more of the student's device than a 2D whiteboard does. On older Chromebooks, low-end Windows laptops, and older iPads, you'll need to drop the rendering quality (see why are the avatars walking in slow motion in the Koala Go Playground? for the Quality setting fix). If your students are mostly on hardware that can't render a 3D scene smoothly, a 2D classroom with reward mechanics is the better fit.

How Koala Go's Playground works specifically

If you're evaluating Koala Go (the platform we build) as a gamified classroom, here's what you actually get:

A 3D browser-based environment. The Playground runs in the browser using WebGL, rendered with Three.js, with physics from Rapier3D. There's no app to install for you or the student — they click the same lesson link they'd use for a normal class and land in the same shared space. The teacher and student each have their own avatar in the world; you can both move around, interact with placed objects, and share the scene.
Customizable avatars. Each student (and teacher) can edit their avatar across six categories: body, eyes, hair, shirt, pants, and shoes. The customization screen is one click from the Playground. Choices persist across lessons, so the student picks their look once and it's still there next week. This is the "owning the character" mechanic in concrete form.
Gems as the reward currency. Gems are the in-classroom currency. The teacher has a gem balance shown in the room UI; so does each student. During a lesson, the teacher opens the student panel and gives gems — one click for one gem, or up to 100 at once with a typed amount — to any student in the room. The student sees their gem count tick up in real time. There's also a visible gem icon next to each student's tile, so younger learners can watch the number grow without you narrating it. (We cap a single transaction at 100 gems on purpose — the goal is a rhythm of small frequent rewards, not a one-shot burst that defeats the variable-reward effect.)
How teachers get gems to give. A teacher's gem balance is funded two ways: by activity inside Koala Go (you earn gems for actions like inviting other teachers — currently 310 gems — and for sending invoices through the built-in invoicing — currently 45 gems, both visible in the Gems Store inside the home menu) and by buying gem packs from the Gems Store when you want a deeper pool. There's no monthly cap on giving gems to students; you can keep awarding as long as your balance lasts.
The Playground is one tab of the classroom, not the classroom. Day-to-day teaching happens on the lesson canvas — whiteboard, slides, PDF uploads, the cobrowser. The Playground is a separate space the teacher can drop into for brain breaks, vocabulary games in a 3D context, end-of-lesson rewards, or themed lessons. It's not where the lesson lives unless you decide it should be.
Quality settings for any device. The Playground has four rendering quality levels — Low, Medium, High, Ultra — set per teacher account. On a recent Mac or a dedicated-GPU desktop, Ultra looks the best. On a five-year-old Windows laptop or a Chromebook, Low gets you smooth motion. See the Quality settings guide for the exact UI path and what each level changes.
Free to try, Pro to scale. The Playground, the avatars, and the gem-giving mechanics are part of the Koala Free tier — you can run lessons indefinitely without upgrading. Koala Pro ($25.99/month monthly, $21.99/month billed annually) adds lesson recording, larger group sessions, unlimited PDF/PPT uploads, and full-length cobrowser sessions; Koala Platinum ($49.99/month monthly, $39.99/month billed annually) adds Scheduling Plus, cloud recordings, and auto-generated class notes. The gamification features themselves aren't gated behind Pro.

The single quote we hear back the most often, from a private tutor: "My students look forward to the playground each day as the highlight of our classes." That's both the case for the mechanics and the warning to use them deliberately — when the playground is the highlight, the lesson before it is what the student is willing to do in order to get there. Most working tutors structure around that, not against it.

How to decide whether you need a gamified virtual classroom

A short decision rule that holds up in practice:

How old are your students? Under 10, lean toward gamified — the engagement mechanics multiply your teaching time. 11-15, depends on the student; preview before committing. 16+, a clean classroom (whiteboard, video, shared canvas) is usually better.
What's currently breaking? If your problem is "the student stops paying attention 12 minutes in," gamification is a direct fix. If your problem is "I can't find a stable platform with worldwide invoicing," gamification isn't the lever — see the best Zoom alternatives for online tutoring for the broader stack question.
What devices do your students use? 3D environments ask more of the student's device than 2D whiteboards do. If most of your students are on modern laptops or recent tablets, a 3D classroom works comfortably; if many are on aging Chromebooks or older iPads, look for a 2D classroom with strong reward mechanics or a 3D one that ships an explicit Low-quality mode.
Are you the parent's "the only platform she'll sit through" tutor? If your retention story already depends on "kids love it" — and a lot of indie tutors discover this is what their reviews actually say — gamification is what's making your retention math work. Picking a platform that supports that story is not optional; it's the whole product.

If you're considering Koala Go specifically: the Playground, avatars, and Gems are all part of the free tier — you don't have to commit to a subscription to see whether they fit your practice. Open Koala Go at classroom.teachwithkoala.com, run a single real lesson with a real student on the device they actually use, and watch their attention. If they stay engaged longer than they did on your current tool, that's the mechanics working. Questions specific to your subject, age range, or device mix — write to koala@teachwithkoala.com with a sentence about the kind of tutoring you do and we'll give you our honest read on whether a gamified classroom is the right fit.

The four things a virtual classroom has that a video call does not

The category is defined by four capabilities, in rough order of how often they show up:

An interactive whiteboard, not a shared screen. A virtual classroom's whiteboard is a shared canvas both the teacher and the student can draw on, type on, and move objects around on. You can upload a PDF or a slide deck and turn it into the lesson surface — the student annotates the same page you're teaching from. A video call, by contrast, gives you screen-share: one person's screen is displayed to everyone else, and only that person can act on it. The gap between the two shows up the first time you try to teach a five-year-old to write a letter on a shared Zoom screen and realize they have no way to hold the pen.
Activities the student can act on. A virtual classroom lets a student drag, drop, sort, match, click, and type on the material. It might be a phonics activity where the student drags letters into words, a math activity where they position blocks on a number line, or a live web activity you're doing together. A video conferencing tool gives the student one thing to do: talk when unmuted. Everything else has to happen through the teacher's screen, which limits pacing and puts the student in the audience.
An engagement layer built for younger students. Rewards, points or gems, avatars, a shared space they can be present in as a character rather than a face on a webcam. A video call is a face-on-webcam interface — reasonable for adults on focused work, but the wrong shape for a seven-year-old who's on lesson eight of the same week. A virtual classroom gives the young learner a reason to want to come back, which is where most of the retention math lives for an indie tutor.
A workflow layer around the lesson itself. Scheduling parents can book against; invoicing that generates the receipt; a room link that persists between lessons; recording (usually on the paid tier); a place to store activities and lesson history. A video conferencing tool gives you a meeting URL. Everything else — the calendar, the invoice, the recording, the activity storage — you assemble from other apps. A real virtual classroom assumes this workflow is the job, not an add-on.

Not every virtual classroom ships all four in the same depth. A whiteboard-first classroom (strong shared canvas, weaker engagement layer) is a good fit for teenagers and STEM subjects; an engagement-first classroom (strong shared space and rewards, whiteboard as one tab among several) is a better fit for K-8 and language learning. When we say "virtual classroom" as a category we mean the software has all four capabilities to some meaningful depth. If it only has one — a nice whiteboard, say, or nice avatars — it's a lesson-layer tool riding on top of a video call, not a virtual classroom.

What "virtual classroom" is not

The phrase gets used loosely, so it's worth naming three things that are called virtual classrooms and actually aren't:

A video conferencing tool with a whiteboard feature toggled on. Zoom's built-in whiteboard, Google Meet's Jamboard integration, Microsoft Teams's whiteboard — these are useful, but the shape is still "meeting first, teaching bolted on." The meeting screen is what dominates the interface; the whiteboard is a tab. For lecture-heavy lessons with adults this is fine. For interactive lessons with kids it usually falls short of what an actual virtual classroom does, because the primary surface is a video grid.
A learning-management system (LMS). Google Classroom, Canvas, Blackboard, Moodle, Schoology — these are async-assignment-focused platforms. The teacher posts work, the student submits work, the teacher grades work. That's a valuable workflow, but it isn't a live classroom. If you want to teach a live lesson through an LMS you have to add either a video call (Google Meet) or a virtual classroom to it. LMSes handle the between-lessons layer; they don't handle the lesson itself. (See can I use Google Classroom for private online tutoring? for the specific case of using an LMS as a tutoring stack.)
A slide-based lesson delivery tool. Nearpod, Pear Deck, Kahoot — these are excellent for adding student-response interactivity to a slide deck. But they don't ship a live video room, a persistent whiteboard, or a scheduling and invoicing layer. They're a lesson layer, not a classroom. Many teachers pair one of these with a video call plus a scheduler plus an invoicer — which is often exactly the stack a real virtual classroom replaces.

When you actually need a virtual classroom (and when a video call is fine)

Not every teacher needs a full virtual classroom. A short decision rule that holds up in practice:

Ages under 12: usually virtual classroom. The engagement layer matters. If your students are 4-10, especially if their parents are paying you specifically because "school is on Zoom and it isn't working," a virtual classroom is doing real teaching work you can feel in retention.
Ages 13-17: it depends on the subject. Language and reading benefit from the shared surface and the engagement layer. Test prep, physics, and math can go either way — some tutors do fine with a clean whiteboard + video call, others get value from the persistent room and workflow layer even without heavy engagement mechanics.
Adult 1-on-1 conversation or focused-work clients: often a video call is enough. A business-English learner or an adult IELTS student doesn't need an avatar. If your lesson is mostly talking and a shared document, Zoom or Meet plus a shared Google Doc covers it. You lose the workflow layer (scheduling, invoicing, activity storage) which is real friction over time — many tutors add a virtual classroom for that reason alone once they hit 5-10 recurring students.
You're running a business that will keep teaching the same students for months: virtual classroom, almost always. The workflow-layer math tips the decision even if your subject and student age would let you use a video call. A tutor with 12 recurring students juggling Zoom + Google Calendar + Google Sheets + Stripe + an activity folder in Dropbox will save more time collapsing that stack into one classroom than they'll spend on the classroom's subscription.
You're running a school-sized class (15+ students): probably neither. Look at classroom platforms designed for whole-class delivery — Nearpod, Pear Deck for the lesson layer alongside a video conferencing solution built for that group size, or LMS-integrated tools like Vedamo for cohort-sized synchronous teaching. Virtual classrooms in the sense this page describes are calibrated for 1-on-1 and small-group tutoring, not for a full school class.

Why video conferencing was never quite right for teaching kids

Zoom, Google Meet, and Microsoft Teams shipped originally as tools for grown-ups meeting each other for work. The interface reflects that: a grid of faces, one screen shared at a time, a chat pane, and not much else. The muscle memory is meeting muscle memory — mute yourself, unmute yourself, share your screen when it's your turn.

None of that was designed against the reality of teaching a seven-year-old for the fifth time this week. Seven-year-olds don't need a mute button; they need something to do. A video call gives them one thing to do — talk when unmuted — and one thing to look at — a face on a grid. The lesson has to do all the work of holding attention with no help from the tool. Most tutors who moved from in-person to Zoom during the 2020 remote-teaching push discovered this within a month, and either patched it (an iPad app for interactivity, a whiteboard tab, a rewards spreadsheet the parent could see) or lost the students.

Virtual classrooms exist because that patching didn't scale. Once you're running 20 lessons a week, the assembled stack of "video call + whiteboard app + activity app + scheduler + invoicer + reward spreadsheet" costs more time to run than the teaching itself, and each seam between tools is a place where something breaks — the wrong parent can't find the link, the whiteboard didn't save, the invoice went out in the wrong currency. A tool built for teaching from the start collapses those seams into one workflow. That is the whole product: not any single feature, but the fact that the room, the materials, the scheduling, the billing, and the engagement layer all live in the same place.

How much does a virtual classroom typically cost?

Pricing splits roughly into three bands. The exact numbers drift, so treat these as shapes rather than quotes:

A meaningful free tier exists on some virtual classrooms (Koala Go's included) — enough to run real lessons at small group sizes and volumes without paying. Free tiers usually come with a cap somewhere: student count per session, activity storage, cobrowser session length, or the presence of an ad on the free-tier interface. The point of a real free tier is that you can teach paid lessons on it while you decide whether the paid tier earns its price.
Independent-tutor paid tiers tend to sit in the $20-$50 per month range and are calibrated for one working tutor. They unlock the workflow layer (scheduling, unlimited activity storage, recording, larger sessions) and remove the caps on the free tier. Koala Go's Pro tier is $25.99/month (or $21.99/month billed annually), which sits inside this band.
School / tutoring-company tiers come next and typically ship per-seat pricing, admin tooling, single-sign-on, and multi-tutor management. If you're running a tutoring company or a small school, this is the right seat to look at (see how do I run an online tutoring business with multiple tutors? for the multi-tutor case, and affordable online whiteboarding for educational institutions for the institutional buying question).

Two costs that don't show up on the pricing page but do show up in real budgets: the time saved by having scheduling and invoicing in the same tool (measurable in tutor-hours per week) and the retention effect from a stronger student experience (measurable in months of average student lifetime). Both usually swamp the subscription price by a wide margin once you're at more than a handful of students — which is the point.

What Koala Go is, specifically

Koala Go (the classroom we build) is a browser-based virtual classroom purpose-built for independent tutors and small teaching businesses. Everything runs in a normal Chrome / Safari / Edge browser — there's nothing to install for you or for the student, they click the same lesson link you'd send for a Zoom call and land in the classroom. What you actually get:

An interactive whiteboard. A shared canvas both you and every student in the room can draw on, type on, and place objects on. On Koala Pro the whiteboard takes PDF, PowerPoint, and image uploads — the file becomes the lesson surface, and the student annotates the same page you're teaching from.
A cobrowser. A live shared browser window inside the classroom — you and the student open the same webpage together and both interact with it (not screen share, actual shared control). Useful with reading platforms, worksheet sites, and any web activity you'd want the student to actually do rather than watch. Koala Free ships a limited cobrowser; Koala Pro removes the session-length cap. See what is a cobrowser? for the mechanic in depth.
A 3D Playground. A shared 3D browser environment where the teacher's avatar and the student's avatar are in the same space — used as a brain break, a themed lesson surface, or the shared-space engagement layer that keeps younger students showing up. Everything renders in the browser via WebGL. See what is a gamified virtual classroom? for how the Playground, avatars, and Gems reward mechanic fit together, and the Quality settings guide for tuning it on older devices.
Worldwide invoicing built in. Parents pay in their local currency, including via WeChat in China. Koala Free carries a 4.5% invoicing fee; Koala Pro drops it to 3.5%. This is the workflow-layer piece most independent tutors underestimate before they hit their first international student paying in the wrong currency through PayPal.
Integrated scheduling on Pro. A teacher-and-parent scheduling portal (at classroom.teachwithkoala.com/portal) where you set weekly availability once and parents book against your open slots via an invite link — no signup dance for the parent. See how do I schedule online tutoring lessons? for the workflow.
Recording on Pro and Platinum. Koala Pro records the lesson locally to your own machine; Koala Platinum adds cloud recordings that store inside the classroom, so a family whose kid missed a session can catch up without derailing next week's lesson. See how do I record online tutoring lessons? for the details.
Real-person support. Koala Pro includes 24/7 live support with a real person on the other end. This is a workflow-layer feature that also matters when it's your live class time and something isn't working.
Small-group friendly by default. Koala Free supports up to 4 students per session (this is exactly the small-group tutoring band — see how do I teach small group classes online?). Koala Go is not the right pick for whole-class 25-student delivery — for that, see the platform categories on best Zoom alternatives for online tutoring.

Honest limits worth naming, because they show up in real teaching: iPad works but is rougher than desktop; students in mainland China sometimes hit connection issues (we route through Hong Kong proxies but the Great Firewall is a moving target — see teaching English to students in mainland China); the whiteboard now ships math tools on its paid tiers (a symbol palette covering operations, fractions, powers, and algebra) but stops short of a full graphing or equation editor, so heavy STEM tutors should test their workflow on a real lesson first; and the Playground is actively developed, with requested features (timers, group-select of Playground objects, more customizable environments) still on the roadmap rather than shipped today.

Roughly 81% of the tutors on our 2026 product/market fit survey said they would be very disappointed if they could no longer use Koala Go — the "very disappointed" threshold that's conventionally taken to indicate strong product/market fit is 40%, so that number is well past the bar. That 81% figure is the same one we publish on the main landing page; we mention it here because the frame this page opens with — "does a virtual classroom actually earn its place over a video call?" — is precisely the question those tutors have answered for themselves. One reply from that survey, from a private tutor: "The use of Playground and Cobrowser has helped me set myself apart from other educators. It's a game changer."

How to try one without committing

Two ways to test whether you need a virtual classroom at all, in order of what we'd suggest:

Run one real lesson on a real free tier. Not a scripted demo — a real lesson with a real student on the device they actually use. If the student's attention holds longer than it did on your current tool, the virtual classroom is doing real work. If it's the same, your current tool is enough. Koala Go's Free tier is designed for exactly this test: up to 4 students per session, worldwide invoicing, limited cobrowser, up to 3 activities — enough to run a real trial lesson end-to-end. Open classroom.teachwithkoala.com and try it.
List the tools you'd collapse into one. If your current stack is Zoom + Google Calendar + a whiteboard app + Google Docs + Stripe + a rewards spreadsheet, count the monthly cost and the weekly hours of workflow that comes from moving between them. If both numbers are meaningful, a virtual classroom is a stack collapse, not just a lesson tool.

If you're weighing Koala Go for a specific subject, student age, or workflow — or you're not sure whether a virtual classroom is even the right shape for what you're teaching — write to koala@teachwithkoala.com with a sentence about what you're trying to do and we'll give you our honest read.

What "small group" means in online tutoring (and what it doesn't)

The line between formats matters because each one is a different business, not just a different seat count:

1-on-1 tutoring. One student, one tutor, one lesson. The highest per-hour rate you can charge and the highest per-hour intensity you'll teach — a full session of turn-taking with a single learner is tiring in a different way than group teaching is.
Pair tutoring (2 students). Two students, usually a sibling pair or two families who know each other. Practically this is a 1-on-1 lesson with extra scheduling friction and a small discount per seat — the pedagogy doesn't change much and neither does your fatigue. Useful as a friction-reducer for parents (they were going to enroll both kids anyway); not really a group-teaching workflow yet.
Small-group tutoring (3-6 students). The band this page is about. Different lesson design from 1-on-1 — you need activities where students take turns, work in shared documents, or engage in the same 3D or 2D activity space together. The economics start to compound (more revenue per hour to you, lower per-seat price to families) and the peer effect starts to help (students hear each other, correct each other, get pulled forward by a slightly stronger group member).
Class-sized tutoring (7-15 students). Past 6 or so, the "shared attention" model breaks and the format becomes a class. You need breakout rooms, mute/unmute discipline, an LMS-shaped homework loop, and a working classroom-management style. Doable, but it's a different job — closer to running a school class than to tutoring — and you should pick tooling designed for it.
Whole-class / cohort (15+). Not tutoring in the sense this page uses. If your business shape is 20-student cohorts, you're a small teaching school; look at LMS-plus-classroom platforms that ship breakout rooms and formal seat licensing (Vedamo runs group classes up to 50 participants and integrates with Moodle/Canvas via LTI — described in more detail in the best Zoom alternatives page).

Why the small-group model works (the economics and the pedagogy)

Two forces make small-group tutoring worth setting up, and neither is a marketing story:

Per-hour revenue goes up without your hourly effort scaling linearly. The classic move: a 1-on-1 tutor charging $40/hr moves 3 committed families into a 4-student group at $18-$22 per seat per hour. Each family's monthly bill drops noticeably, your per-teaching-hour revenue moves from $40 to $54-$66, and you gain back the hours you'd have booked as three separate lessons. The family gets a lower sticker price; you get more revenue and a more predictable calendar. (See how much should I charge for online tutoring? for the per-lesson-rate side, and how to structure lesson packages for how the group seat sits inside a monthly package.)
The peer effect does real teaching work. Students learn faster in a small group than 1-on-1 in some subjects — especially language (they hear other learners of their own level speak), reading (they hear other students decode), and any subject with a discussion component (they hear each other's reasoning). The evidence base behind "high-dosage tutoring" — the tutoring model schools have been scaling since the 2020s — specifically points at group sizes of 3-4 as the size where per-student outcomes stay comparable to 1-on-1 while the format is cheap enough to sustain. Above the group of 4, the per-student attention drops and the pedagogy usually needs to shift.

The catch on both: small groups only work if the students are close enough in level and pace that the group is teaching itself and you're not running three parallel lessons in your head. Grouping mismatched levels is where indie tutors most often burn out on the model.

Where small-group tutoring is the right fit

From working tutors we've spoken to, the situations where the group format reliably outperforms 1-on-1:

ESL and language teaching with young learners. Kids learn a language faster with peers to talk to — they hear another child's mistake and correct their own, they play the same game together in the target language, and the room stops being "a stranger's face on a screen." Groups of 3-4 same-level students are the sweet spot.
Reading and phonics with beginner readers. Same mechanism — the student who's stuck on a sound hears the student one week ahead say it correctly. Orton-Gillingham and structured-literacy tutors sometimes run small groups deliberately for this reason, though many keep it 1-on-1 for the older intervention population.
Exam-prep cohorts with a fixed target. IELTS, TOEFL, SAT, ACT, Cambridge, AP — students preparing for the same test at the same date benefit from group problem-solving, and the price per seat lets families who couldn't afford 1-on-1 test prep get access to it. The tutor's per-hour goes up materially at group rates.
Sibling / friend pairs and trios. Two or three kids in the same family or the same friend group are usually the easiest first group to fill — the parents already coordinate schedules, and the students already know each other. This is often how a solo tutor's first small-group class starts, and it's a legitimate way to test the format without a marketing push.
Any subject where discussion is the point. Debate, essay writing, literature analysis, business English, ethics, university-admissions interview prep. In these, 1-on-1 constrains the pedagogy — you want more voices in the room by design.

Where small-group tutoring is the wrong tool

Don't force the format. Cases where 1-on-1 is better:

Students with specific learning differences that require intensive individual attention. Serious dyslexia intervention, autism spectrum work, ADHD coaching — most specialist tutors keep this 1-on-1, and families paying specialist rates expect it. A group can dilute the attention the student is actually paying for.
Test prep close to the deadline. Two weeks before the exam, the student needs targeted drilling of their weak spots, not group work on the median student's weak spots. Move test-prep students to 1-on-1 for the final push even if they were in a group cohort until then.
Levels too far apart. If your two candidates are an absolute beginner and an intermediate learner, don't group them — one will disengage and the other will be lost. Grouping is only cheaper when the pedagogy scales, and it doesn't scale across levels.
Very young learners with short attention spans and no peer familiarity. A 4- or 5-year-old often takes weeks to acclimate to a 1-on-1 tutor; adding two more kids they don't know can make the acclimation collapse. Start these students 1-on-1 and only move them into groups once they're comfortable with the classroom.
Adult 1-on-1 conversation clients. A business-English learner paying a specialist rate wants a specialist hour of their own — putting them in a group is a downgrade even if the rate makes sense to you.

How many students should be in a small group?

Assume 3-4 as the default and stretch to 5-6 only for specific formats. The tradeoffs:

3 students. The safest first group. Enough voices to make the group teach itself, few enough that if one is absent you still have a proper group (not a 1-on-1 in disguise). Turn-taking cycles cleanly in a 30-minute lesson — each student gets ~7-8 minutes of direct attention.
4 students. Highest revenue per hour that still holds the small-group pedagogy. Turn-taking is tighter and you need slightly more disciplined activity design; each student gets ~5-6 minutes of direct attention in a 30-minute lesson. This is the cap on Koala Free and it's not an accident — 4 is where most independent tutors settle.
5-6 students. Workable for discussion-heavy formats (literature, debate, exam prep with strong readers) and for language classes where students spend most of the time speaking to each other rather than to you. Below-average students get less attention here; above-average students carry the room. Not the shape for young learners.
7+ students. Stop calling it small-group. Start using a platform designed for classes.

Whichever seat count you pick, hold it. A "sometimes 3, sometimes 5" group ruins your lesson planning and confuses the families. Set the size in your policy, waitlist over-flow, and downgrade to 1-on-1 slots if the group falls below your minimum for two lessons running.

How to structure a small-group lesson

The 30- to 60-minute small-group lesson has a different shape than 1-on-1. A working template for a 4-student 45-minute lesson:

Warm-up together (5 min). Something the whole group does at once — a short game, a quick round-robin check-in, a group brainstorm. Sets the "we're in a room together" tone before the teaching starts.
Whole-group teach (8-10 min). New content introduced once for everyone. Use a shared whiteboard, slide, or activity page — one input surface the whole group is looking at. This is the part where a real virtual classroom (shared canvas, PDF upload, cobrowser for a web activity) does much better than a video-call-plus-screen-share.
Round-robin production (12-15 min). Every student produces the new content in turn while the others listen or respond. Use the students' names — "Anya, your turn — try that with 'went'" — because the student who's not sure it's their turn will silently disengage. This is where 3-4 students holds up and 6+ falls apart.
Pair or small-team activity (10-12 min). Split the group into two pairs (or two mini-teams if 5-6 students) working on the same task in parallel. In a virtual classroom that supports it, this can be actual pairs working on separate parts of the whiteboard or moving around the Playground; in a plainer setup, it's you giving each pair a task and rotating between them. The peer effect does the teaching work here.
Whole-group wrap and reward (5 min). Recap what everyone learned; award something visible to each student for a specific thing they did (see the reward-currency section below); preview next lesson. The visible reward is what turns "this was a nice lesson" into "I want to come back next week" — even more than in 1-on-1, because now each student wants their peers to see them earn.

Two structural choices worth making up front. First, if you teach primary-age students in groups, plan a movement or brain break every 15 minutes rather than every 25 — the group amplifies any single student's restlessness. Second, decide before lesson 1 whether every student attends every session or whether the group has flexible attendance (a "drop-in cohort"). Fixed attendance groups teach better; flexible attendance is easier to sell. Pick one and hold it.

What platform capabilities you actually need

A shortlist for judging whether a virtual classroom can carry a small-group format:

A group room that supports at least the seat count you plan to run. Sounds obvious; often forgotten. If your intended group size is 4 students and the tool's free tier caps at 3, that free tier isn't your free tier.
A shared visual surface everyone in the room can see — and, ideally, act on. A shared whiteboard where every student can annotate; a PDF or slide upload the group is looking at together; a shared 3D or 2D space if young learners need "somewhere to be." A video call plus screen-share fails this test — one screen shared by the teacher, everyone else watching, is not a shared surface, it's a webinar.
Reliable turn-taking mechanics. Named students in the participant list, per-student mute/unmute the teacher can control, a way to know quickly who's spoken and who hasn't. Rooms where students appear as anonymous tiles or where you can't quickly cue "Anya, your turn" cost you attention every lesson.
A per-student reward or attention mechanic. Something you can give one student in the room without affecting the others — a point, a gem, a star, a visible acknowledgment. This is what keeps a group format from flattening every student's individual engagement into "the lesson" rather than "my part of the lesson."
Recurring scheduling and roster management. Groups only work if the group meets on the same day at the same time every week. You want the classroom to book that cohort's schedule once (not four times, one per family) and to send each family the same lesson link.
Recording — for the family whose kid is off sick. Missed lessons happen more in group formats (any of 4 families having a bad week hits your attendance more than any of 1 family does). A recording the group can review means one absence doesn't wreck their arc.
Predictable performance on the students' actual devices. Group formats amplify performance issues — one student on an underpowered Chromebook lagging the shared activity affects the other three. If your students are mostly on older hardware, factor that into which classroom you choose (see the Quality-settings guide for the specific case of 3D environments on older devices).

How Koala Go handles small groups specifically

If you're evaluating Koala Go (the classroom we build) for small-group tutoring, here's what you actually get and where the current limits sit:

Up to 4 students per session on Koala Free. This is the shipping cap — the tier's "Up to 4 students per session" line is exactly the small-group band this page is about, not an accident of pricing. If your plan is 3-4-student groups, Koala Free is a real classroom to teach them in, not a trial. Group size above 4 needs Koala Pro; the Pro plan expands beyond the free-tier cap for larger group sessions.
A shared visual surface the whole group can look at and act on. The whiteboard takes PDF, PowerPoint, and image uploads; every student in the room can annotate the same canvas; drawings and sticky notes are visible to everyone. When the lesson involves a live web activity — a Wordwall game, a leveled reader, a browser-based exercise — the cobrowser gives every student in the group control of the same live webpage, not a passive view of your screen (see what is a cobrowser? for the mechanic in detail).
Per-student reward currency inside the group. Gems are awarded one student at a time. The teacher opens the student panel, picks a student, and gives 1-100 gems in a single click — the amount input is capped between 1 and 100 on purpose (small frequent rewards work better than one big burst). Each student's balance updates live, visible to the whole group, which is what turns individual attention into a small friendly competition rather than a private teacher-to-student channel. Grounded in more detail in gamified virtual classroom.
Shared 3D Playground with multi-player avatars. The Playground is a shared 3D space — the teacher's avatar and every student's avatar are in the same environment, moving around, seeing each other, sharing the same objects. For young-learner groups this is often what recovers the format — 4 seven-year-olds who'd have squabbled on Zoom cooperate in the Playground because there's a shared thing to be in. The Playground is a browser environment; students click the same lesson link and land in the same space (no install).
Recording lesson content for the absent student. Lesson recording is part of Koala Pro. If one of your 4 group students is out sick, the recording lets that family catch up without derailing next week's group lesson (see how to record online tutoring lessons for the recording specifics, including consent and retention).
Recurring cohort scheduling on Platinum. The Scheduling tool comes with Koala Pro; recurring classes — the mechanic that lets a group of 3-4 students book every Wednesday at 4pm without you rebooking each week — sit inside Scheduling Plus on Koala Platinum ($49.99/month monthly, $39.99/month billed annually), together with 30/45/60-minute class lengths and booking-policy controls. If your business is going to run more than one recurring group cohort, Platinum is the plan where the schedule side stops being manual.
Group invoicing that bills each family separately. A small group of 4 kids from 4 different families means 4 monthly invoices, not one — and each family should pay in their own currency. The built-in invoicing generates one invoice per family per package, and the parent pays a link in their local currency (see how much should I charge for online tutoring? for the per-seat pricing side).

Honest limits: Koala Go is built around independent-tutor and small-group workflow. If your target class size is 20+ students, it's the wrong pick — Koala Free's 4-student cap is a real cap, Koala Pro is calibrated for larger small groups rather than school-sized cohorts, and the Playground's shared-space model doesn't map to a 20-student lecture. For that shape, look at the whole-class-oriented platforms on the best Zoom alternatives page. Also worth flagging: features tutors have specifically asked for in group settings — lesson timers and more granular breakout mechanics — are still on the roadmap rather than shipped today (group-select of whiteboard objects, once on this list, has shipped: drag a box around several objects and move, resize, or rotate them together); if any of these is a hard requirement for you, ask before you commit (write to koala@teachwithkoala.com).

Pricing the seat and filling the group

Two questions every tutor moving into small groups has to answer:

How much do I charge per seat? A common shape: the per-seat rate is 40-55% of your 1-on-1 rate. A tutor charging $40/hr 1-on-1 typically prices a 4-student group at $18-$22 per seat per hour. Do the math on your side: $40 × 1 vs. $20 × 4 = $80, so the group hour is worth roughly twice the 1-on-1 hour to you. Don't price so low that the group hour makes less than 1.5× the 1-on-1 hour — you're taking on more prep, more parent messages, and more attendance risk for the same money. See how much should I charge for online tutoring? for the underlying rate framework, and how to structure lesson packages for how the group seat sits inside a monthly package.
How do I fill the group? Two paths that reliably work:
- Start with a sibling or friend group. Ask a 1-on-1 family whether their child would learn better with a friend or sibling in the room, and price the pair at a modest discount. Two families becomes three organically as one recommends a friend.
- Convert your own 1-on-1 waitlist into a cohort. If you have three families waiting for a 1-on-1 slot at your current price, offer them a group class at your group per-seat rate starting on a specific date. Families who wanted 1-on-1 will sometimes prefer the group price; families who need 1-on-1 will stay on the list, which is fine.
Cold-marketing a small-group cohort ("open enrollment, group class every Wednesday, join here") is harder than it looks — the group needs 3 committed families before the first lesson to work as a group. Start from your existing student base.

Common failure modes to avoid

Grouping mismatched levels. The single most common mistake. If two of your four students are visibly ahead and two are visibly behind, you're running two lessons in one hour and both halves are getting a worse experience than 1-on-1. Split the group.
Turning the group into a lecture. If one student is talking 60% of the time and the other three sit silent, it isn't small-group tutoring anymore. Round-robin production (from the lesson-structure section above) is what prevents this — force yourself to name each student's turn.
Flexible attendance. "Come when you can" cohorts feel accommodating but destroy the group dynamic — no student ever meets the same peers twice in a row, the peer effect collapses, and the group becomes a series of low-quality 1-on-1s. Hold fixed attendance from lesson 1 unless the format is explicitly a drop-in workshop.
Under-pricing the seat. A group hour at $10 per seat feels attractive to families and looks like decent revenue to you until you count prep, four separate parent-message threads, and four separate scheduling flows. Group teaching takes more admin, not less. Price to the value of the group hour, not to the parent's price sensitivity.
No visible per-student reward. Without something that acknowledges individual students inside the group, quieter students disappear. Whatever your mechanic — Koala Go's Gems, a running scoreboard, a shared sticker chart — use it deliberately on the students who spoke least, not the students who spoke most.
Not planning for absences. Any given lesson, at least one of your 4 students has a chance of missing. Have a rule up front: recorded catch-up (Pro-plan recording covers this), a make-up 1-on-1 on the tutor's schedule (not a group-lesson slot), or a "you miss it, you miss it" package policy. Decide before lesson 1 (the operational side of no-shows and cancellations is covered in how do I handle student no-shows and last-minute cancellations?).

How to decide whether to run small groups at all

A short decision rule:

Do you have 3+ same-level students who could form a group? If yes, small-group is worth piloting. If no, keep teaching 1-on-1 until your student base is deep enough to source a cohort — a group of 2 mismatched-level students isn't a group.
Is your subject one where peer voices help (language, discussion, exam prep)? If yes, groups will teach better than 1-on-1 for most students. If your subject is intensive individual intervention (dyslexia therapy, ADHD coaching, advanced test prep close to the exam), keep it 1-on-1.
Does your platform support a shared surface, per-student rewards, and recurring cohort scheduling? If yes, run the group. If your current tool is Zoom or a plain video call, either fix the tool first or expect the group format to be harder than it should be.
Are you willing to hold a fixed seat count and fixed attendance policy? Half-committing to a group format is what causes it to fail. Pick 3 or 4 students, pick the day and time, hold both.

If you're considering Koala Go specifically for the small-group format, the group room, the shared whiteboard, the Playground, and the per-student Gems mechanic are on the Free tier — you can run a real 3- or 4-student group lesson without upgrading first. Open Koala Go at classroom.teachwithkoala.com, run a single real group lesson with your candidate cohort, and see whether the pedagogy works before committing to the model. Questions about your specific subject, group size, or workflow — write to koala@teachwithkoala.com and we'll give you an honest read.

What a trial lesson is really doing

Trials work as three jobs stacked into one 20-30 minute session — miss any of the three and the trial fails silently:

Qualify the family. Is this student's level, age, subject, and schedule something you can teach well? Is the parent someone you want as a client (responsive, respectful of your time, prompt about payment)? The trial is your last chance to walk away before you take a package. Bad-fit families cost more than the revenue they bring — they eat message time, dispute cancellations, and generate the churn you'll spend the next quarter trying to fix.
Deliver a real teaching sample. Not a demo of "what you're like as a person," not a Q&A about your credentials — a small self-contained piece of teaching the student can walk away with. The parent watching decides based on what they saw the student learn, not on what you said you could do. This is the part that most under-prepared trials get wrong: they turn into a chat.
Get commitment to a next step. By the end of the trial, either the family books a recurring package on the spot, or they book a specific follow-up date to decide, or you learn it's a no. Vague endings ("let me know what you think") are the single largest source of trial-to-paid drop-off. See how to structure lesson packages for the package shape you're offering at the close.

If your trial does only one of the three, it becomes something else. Sample-only with no qualification is a workshop. Qualification-only with no sample is a sales call. Sample-plus-qualification with no ask is a friendly hour that ends in silence. All three at once is a trial.

Free trial or paid trial?

Both work; each carries a specific tradeoff, and the right answer depends on where you sit in your practice.

Fully free trial. Removes the price barrier — the parent's decision to book is one click, no card, no payment friction. Highest volume of trial bookings. Also the highest no-show rate: a parent who hasn't put anything on the line drops out at 20-40% (some tutors report worse). Free trials are the right pick when you're building your first 5-10 students, when you're launching a new subject or new age range, or when the market you're in expects a free trial as the norm (most ESL markets, most young-learner markets).
Paid trial at a small discount off your standard rate. Say $15-$25 for a 20-minute session against a $40/hr standard rate. Same session length as a free trial, but the parent is now a paying customer from minute one, which anchors the transaction and cuts no-shows sharply. Right pick when your calendar is already 50%+ full and every trial slot has a real opportunity cost, or when you charge specialist rates ($60+/hr) where a free trial signals a different market position than you want.
Small refundable hold on a free trial. A hybrid: the parent puts $5-$25 on the trial slot as a hold, and you refund it (or roll it into the first package) if they show up. This is already the pattern we describe on how to structure lesson packages — it cuts trial no-shows dramatically without discouraging the initial booking, because the parent knows they're getting the money back. Setup is a one-line-item invoice sent before the trial and marked refunded (or applied against the first package) after.
Fully paid trial at full rate. Not really a trial — it's just a first paid lesson. Some experienced tutors run first lessons at their standard rate and don't call it a trial at all. Works when you have a strong referral base or reviews doing the qualification for you, and doesn't when a family needs to see you teach before they'll pay full rate. Safe to graduate into once your calendar is consistently full at full price and you're actively raising rates.

The two decisions you should NOT make: do not stack discounts on top of a free trial (a free trial and "50% off your first month" tells the parent you're rate-sensitive and attracts families who churn at the first full-price renewal — this is why the lesson-packages page flags heavily-discounted intro offers as a trap), and do not stretch the trial to 45 or 60 minutes to "give real value" — a trial that's the same length as a paid lesson is a paid lesson given away. Keep the trial to 20-30 minutes, do it well, and price the actual lessons.

What to do before the trial

The trial has to be teachable-out-of-the-box the moment the family logs in. Everything you can do before the trial reduces the amount of the trial you spend on intake instead of on teaching. A working pre-trial checklist:

A short intake form the parent completes when they book. Five questions, maximum: student's name and age, subject and current level (or the specific problem — "she's struggling with reading fluency"), what they've tried before, what a successful trial looks like to them, and the student's timezone. Anything longer than five questions gets partly filled or skipped; anything you skip is time you spend asking during the trial instead of teaching.
An expectations email 24 hours before the trial. Confirms the time in both timezones, restates what will happen ("we'll spend the first 3 minutes saying hi and doing a tech check, then about 15 minutes on a short reading passage, then 2 minutes to talk about next steps"), and includes the lesson link. This one email prevents the "wait, was it 4pm my time or yours?" no-show and sets the parent up to see the structure of the session rather than wonder what it is. Same trick that halves the miss rate on regular lessons (see how to schedule online tutoring lessons for the two-timezone pattern).
A tech check on the student's actual device. If the family is joining from an iPad or an older Chromebook, five minutes of "does the classroom load and can we see each other" before the trial itself is cheaper than losing the trial to a lag. This is more true for tools that ship a shared whiteboard or a 3D environment, less true for plain-video-call setups (where the tradeoff is a less-engaging trial — see the section below on how the classroom shapes what you can show them).
A specific teaching plan for the trial. A single learning objective the student will visibly meet in 15 minutes: read three CVC words, blend two consonant clusters, solve a simple two-step equation, use a phrase correctly in three sentences. Not a generic "let's see where you are" — a concrete win they can point to at the end. Parents remember the win, not the assessment.
Your own opening 60 seconds, written down. The first minute of a trial is 30% of the parent's impression. Rehearse it: greeting, one sentence about who you are, one sentence about what the student is about to do, then straight into teaching. If you're improvising the opening every trial, you'll waste the opening every trial.

How to structure the trial itself (a working 25-minute template)

The trial has to fit a very tight time envelope — parents are watching the clock more than they'll admit, and a trial that runs 40 minutes because "we were having fun" makes the parent worry every future lesson will run long. A working 25-minute shape:

Warm greeting and tech check (2-3 min). Say hello to the student by name, say hi to the parent if they're on camera, confirm audio both ways, and set expectations: "We'll spend about 15 minutes on some reading today, then at the end we'll talk about what next steps could look like." Parents relax the moment they hear that structure. Do not spend 5 minutes on chitchat here — you're setting a lesson, not a coffee.
Micro-assessment (2-3 min). One targeted question or task that tells you where the student is. Not a formal placement test — a real teaching signal. For ESL: read three sentences and answer one question. For math: solve one problem two levels below where the parent said the student is, and read the result. For phonics: three CVC words and one CVCC. This is enough to place the teaching sample.
The teaching sample (15 min). The core of the trial. Pick one narrow, self-contained thing the student will learn, and teach it. Not "an overview of everything I could teach you." One thing, taught well, with clear before/after. Use the shared surface — a whiteboard, a slide, a cobrowsed web activity, an interactive game — not just talking. This is where a real virtual classroom outperforms a video call: a student who spent 15 minutes actively doing something on a shared canvas remembers a different lesson than a student who spent 15 minutes watching a screen-share.
Explicit close and the ask (3-5 min). Recap what the student learned in one sentence the student agrees with ("So today you learned to blend the -st sound in five new words — right?"). Then to the parent, direct: "If you'd like to continue, I have a Tuesday-4pm and a Thursday-4pm slot open — which works better?" Not "let me know what you think." Not "here's my email if you want to book more." A specific concrete offer with two choices. This is where trials convert or don't.

Two structural rules that keep the format honest. First, stop teaching at minute 22 no matter where you are — a trial that runs 35 minutes is not a favor, it's a signal. Second, if the student is visibly not enjoying it by minute 10, do not push through — pivot to something they can succeed at, salvage the last 10 minutes, and note the mismatch for yourself. Forcing a trial through a bad fit produces the worst outcome: no sale AND a parent who left annoyed.

Making the ask (the moment the trial actually converts)

The close is where trials convert. Almost all of the "the trial went great and they never wrote back" outcomes are actually "the trial went fine and I never made the ask." Three patterns that work:

Two-choice offer with a specific slot. "I have Tuesday 4pm or Thursday 4pm open starting next week — which works?" Named times force a yes/no on scheduling rather than a yes/no on principle. Parents who like the trial almost always pick a slot when offered specific ones; parents who need to think about it will say so, and then you're in a specific-follow-up conversation instead of an open-ended one.
Package-shaped, not lesson-shaped. The ask is a monthly package or a 4-lesson pack, not "let's book next Tuesday." Recurring commitment is where the retention lives (see how to structure lesson packages) — the trial-to-paid transition is your one clean moment to establish it. A family that says yes to "$X for 4 lessons" is a family; a family that says yes to "one more lesson next Tuesday" is a family you'll be re-selling every week until they drift.
Follow-up date if the answer isn't yes now. If the parent needs to check with a spouse, check a schedule, or think about it — good, but pin a specific date. "Would Friday be a good day for me to send a short follow-up with those two slots?" A dated follow-up converts at maybe a third of the rate of an on-the-spot yes, but converts far better than an undated "get back to me."

What to avoid at the close: multiple package options presented like a menu (families paralyzed by choice do nothing), a rate quote without a slot ("my rate is $X" without "and I have Tuesday-4pm open" hands the parent a homework assignment), and any language that lets the parent leave without a decision. The ask is a favor to the parent as much as to you — most people would rather be told the next step than have to invent it.

Common trial-lesson failure modes

Talking at the parent instead of teaching the student. The parent is evaluating you on what their kid learned, not on what you said you could do. Keep credentials-and-approach-explanation under 60 seconds and get to teaching.
Turning the trial into a placement test. Assessment-heavy trials feel like a school evaluation, not a lesson. Parents want to see teaching. Do 2-3 minutes of micro-assessment and then teach.
Overrunning the trial by 15 minutes. Signals every future lesson will overrun. Also devalues your paid lessons — if a trial can go 45 minutes, why should the paid one stop at 30?
Winging the technology. Discovering during the trial that the parent's iPad can't do X, or that the student can't share the whiteboard on Chromebook, means you spent the trial troubleshooting. Do the tech check before.
No teaching win the student can name. If you asked the student at the end "what did we do today?" and they'd say "we talked" — the trial won't convert. The student needs a specific one-sentence answer: "I read three big words. I did fractions. I said a sentence in Spanish." Design the trial so that answer exists.
Free-form close. "Let me know" ends trials. Two-choice specific ask converts them.
Discount-stacking. Free trial plus "50% off your first month" plus "a free lesson if you refer a friend" attracts churn-prone families and undervalues your calendar. Pick one lever and use it deliberately.
Running a trial without a working paid offer. If your package pricing isn't set before the trial, the close falls apart — you can't make a specific offer for a shape you haven't decided on. See how much should I charge for online tutoring? and how to structure lesson packages — decide these before your first trial.

How Koala Go fits the trial-lesson workflow specifically

If you're evaluating Koala Go (the classroom we build) for running trials, the shape:

Koala Free is a real classroom you can run trials in at $0 platform cost. The Free tier ships an actual working classroom — shared whiteboard, capped PPT and PDF uploads (Pro removes the cap for unlimited PPT and PDF uploads), the Playground with multi-avatar 3D space, per-student Gems, up to 3 saved activities, and up to 4 students per session. Not a demo. Not a 14-day countdown. The one Free-tier constraint that matters most for trials is the "Limited Co-browser" — Free sessions get a 10-minute cobrowser window with a 30-minute cooldown between sessions (Pro upgrades this to unlimited). For a 20-25 minute trial the 10-minute cobrowser window is usually enough to include one cobrowsed activity as part of the teaching sample. That's the honest tradeoff: Free is a real trial classroom, and if the cobrowser is central to the pedagogy you'd rather show, Pro (starting at $21.99/month billed annually, or $25.99/month monthly) removes the session limit.
The Playground makes young-learner trials demonstrably different from a video-call trial. A 5-8-year-old visiting a plain video-call trial spends the first five minutes staring at your face and the next twenty getting distracted; a Playground trial gives them a shared browser-based 3D space with their own avatar to be present in, which changes the whole trajectory of the session for that age group. This is what the gamified virtual classroom mechanic buys you at the trial — an engaged student rather than a bored one, in the exact 20 minutes where you need to prove you can teach them.
The cobrowser lets you demonstrate a real interactive activity. Instead of screen-sharing a Wordwall game and asking the student what they'd do, you can hand the student control of the same live webpage — they click, drag, and type on it while you watch. Same web activity, but the student is actively doing rather than passively watching. For trials in subjects that lean on interactive web content (phonics games, Wordwall, Boom Learning, leveled readers, geography puzzles), this is the one demo that converts parents who've already sat through Zoom trials elsewhere. Grounded in more detail on what is a cobrowser?. On Free the cobrowser is time-limited per session (10 min + 30 min cooldown); on Pro it's unlimited.
Gems give the trial student a visible win they take home. During the trial, awarding the student 20-40 Gems for a specific thing they did ("that was great sounding-out — 30 Gems") gives them a tangible record of what they achieved. Kids talk to their parent about the Gems on the way out of the trial; the trial stops being an abstract "was it good?" conversation and becomes a specific "look what I earned." Small mechanic, disproportionate effect on trial conversion for the K-4 band.
Portal booking removes trial no-show friction (Pro). On Koala Pro the built-in Scheduling gives you a public booking URL — you drop it in your Instagram bio, your affiliate page, or a Reddit comment, and parents pick a trial slot on your calendar. On booking the parent gets an invite link that creates their profile automatically — no separate portal signup, no password reset, no "how do I log in" email the day of the trial. This is the same invite-code flow described on how to schedule online tutoring lessons. Honest limit: the Portal itself sits behind Koala Pro; a tutor on Koala Free still runs trials by sending the parent the direct classroom link.
Invoicing supports the refundable-hold pattern for paid or hold-secured trials. The built-in Invoicing sends a one-line-item invoice ($5-$25 hold for the trial slot) and the parent pays it in their local currency — Stripe card, PayPal, Wise, or WeChat Pay if the family is in China. On Koala Free the platform fee is 4.5%; Pro drops it to 3.5%; Platinum ($39.99/month annual, $49.99/month monthly) drops it to 0%. For a $10 trial hold on Free that's $0.45 in fees — not a decision-changer at trial-lesson volume, but worth knowing before you scale a trial-heavy funnel. Refunding after the trial is a one-click action inside the same invoice, or you can leave the invoice open and apply it against the first package.
Recording a trial (Pro, with consent). Some tutors record trials so they can review the close afterwards and improve their own conversion — a legitimate reason to want the recording feature during the trial. Recording is Pro-tier and worth a note to the parent at the start. The consent, storage, retention, and mechanics are covered on how do I record online tutoring lessons?.

The honest limits: Koala Go is calibrated for indie and small-group tutoring workflows, not for high-volume trial funnels of the "1,000 trials a month" agency shape — the free 4-student cap and the Pro-gated Portal booking are both real. If your practice is 5-20 trials a month against a solo indie business, the Free tier is a real trial classroom and Pro is the plan you'd upgrade to when the booking friction starts to cost you more than the subscription. If you're not sure which side of that line you're on, open classroom.teachwithkoala.com, run one real trial with your next prospective family on the Free tier, and let the trial tell you whether the pedagogy converts before you decide on the plan. Questions specific to your subject, student age, or trial funnel — write to koala@teachwithkoala.com with a sentence about the shape of your practice and we'll give you an honest read.

What "structured-literacy tutoring online" is (and what it isn't)

The reader we're writing for is a tutor already trained in a structured-literacy approach. If you're an Orton-Gillingham practitioner, a Wilson-certified tutor, working through Barton or IMSE, using Take Flight, LiPS, Lindamood-Bell, or any of the systematic-phonics programs the International Dyslexia Association groups under the umbrella term "structured literacy," this page is for you. We're not teaching you the method — you already know it. We're covering the parts that change when the delivery moves from an in-person table to a browser tab.

Structured-literacy tutoring online is not:

Reading practice with an app. Apps that drill sight words or phonics can be useful supplementary tools, but a structured-literacy lesson is a live, guided session with explicit teacher-controlled sequencing — not a game the student plays alone.
A generic ESL reading lesson. ESL reading pedagogy overlaps with structured literacy but doesn't replace it. Dyslexic learners (in any language) need the systematic, cumulative, multisensory sequence — see how do I teach reading to ESL students online? for the generic ESL side.
A video call with a worksheet on screen-share. Screen-share turns the student into a passive watcher; a structured-literacy lesson turns on the student doing — tapping tokens, writing sounds, blending onto a card, tracking a finger under a word. The delivery has to preserve that.

Three constraints that change when the lesson moves online

Every SpLD tutor we've spoken to who's moved a practice online reports the same three constraints in the first month:

Multisensory delivery through a screen. In-person, you use physical letter tiles, sand trays, tapping fingers, arm sweeps, coloured cards. Online, the tokens are digital — the student needs to be able to manipulate them, not just see them. That means your platform's whiteboard has to support drag-and-drop objects that both you and the student can move, with real-time updates on both sides. A whiteboard that only the teacher can annotate isn't enough.
Attention stamina for a full session. Dyslexic students often have overlapping attention and working-memory profiles (ADHD comorbidity is common); a 45-minute session at a screen is harder than the same session at a table. You'll structure differently — shorter drill segments, more movement/brain breaks, more frequent reinforcement — and your platform needs to support the breaks without kicking the student out of the room.
Parent involvement without you in the house. The parent isn't in the room, but the intervention only works if home reinforcement continues between sessions. That means the parent needs a clear window into what happened in each lesson — a written recap, a recording of the decoding drill, a specific "practice this word list this week" — delivered on a channel the parent actually opens.

What the shared canvas has to do (multisensory in the browser)

Preserving the multisensory piece is the single hardest thing to get right online. The specific whiteboard capabilities that make it work:

Drag-and-drop tokens both sides can move. Letter tiles for building words. Coloured tokens for Elkonin-box style phoneme boxes. Syllable-type cards. Morpheme chunks. The student has to be able to grab a token and move it, not just watch. On Koala Go's whiteboard, Sticky-note objects and Stamp images both drag on the shared canvas and update live for every participant in the room — you place a row of empty boxes on the canvas, the student drags a phoneme token into each box as they segment the word, and both of you see the same state.
Fine drawing for letter formation. If the tutor's method includes handwriting or sky-writing, the whiteboard needs a smooth pen — a stylus on iPad, a graphics tablet, or even a good trackpad. Precision matters for letter-form modelling, especially for the reversal-prone letters (b/d/p/q) where the tutor is drawing the correct form and the student is copying it beneath.
Text you can highlight, colour, and mark. Decoding practice on a controlled text: mark the vowel red, underline the consonant blend, dot every syllable break, sweep an arrow under the words the student reads. The annotations need to persist across the session so you (and the parent, later, on a recording) can trace the exact sequence the student worked through.
Slide / PDF upload for your existing materials. Structured-literacy tutors typically have a large personal library — decodable readers, word lists, syllable-division worksheets, spelling dictation sheets — accumulated over years. A whiteboard that takes PDF and PowerPoint upload with annotation on top preserves that library instead of asking you to rebuild it in a new tool.
A student view that shows what you show. Simple, but not universal — the student sees the same canvas you see, not a mirrored / laterally-flipped view. Reversal-prone learners don't need any additional orientation friction.

Structuring the online session for shorter attention windows

A typical 45-minute in-person Orton-Gillingham lesson has a sequence something like: phoneme review → new phonogram introduction → controlled reading → dictation → decodable-text reading → spelling → summary. Moved online, most tutors we've heard from compress the drill segments (they run faster on-screen with digital tokens than in-person with physical ones) and add explicit micro-breaks. A working template for a 45-minute online lesson:

Warm-up review (5 min). Rapid drill of previously-taught phonograms — flash the sound cards, student says the sound, you tap the next one. Use a slide deck of cards; cycle through faster than you would in-person. This is your temperature check on last week's material.
New content or focused review (10 min). New phonogram, new syllable type, new morpheme, new spelling rule — whatever the method's next step is. Teacher-led, canvas-heavy. This is the segment where the shared canvas earns its cost — you're writing the new grapheme on the whiteboard, the student is copying it beneath, you're both marking the same example word.
Movement or attention break (2 min). A stretch, a short physical activity, sky-writing the new letter with a big arm, walking to a wall and back. Don't skip this; the student's attention needs to reset before the harder segments. Some tutors use the Playground here — a short "walk to the sign, come back" beat with the student's avatar — because it's low-stakes movement inside the same tab.
Word-building / Elkonin boxes (10 min). Drag-and-drop segment. Place empty phoneme boxes on the canvas, dictate a word, student drags tokens into the boxes as they segment. Reverse the exercise: place tokens, student says the whole word. This is the segment where the online delivery is often better than the in-person one — you have unlimited tokens, no lost letter tiles, and the sequence auto-records if you're recording the session.
Decodable text reading (8 min). Open a PDF of the decodable reader on the whiteboard, or open a leveled-reader site through the cobrowser (see the next section). Student reads aloud, you mark difficulty on the canvas, and you follow the student's finger — or their mouse — under the text.
Dictation / spelling (5 min). You dictate, student types or writes on the canvas. If handwriting is part of the method, use the whiteboard's pen tool; if the student is typing, use the Text tool. Both keep the record on the shared canvas for review.
Wrap-up and reward (5 min). Recap the day's target, award effort-based reinforcement (see below), preview the week's home practice for the parent. The last five minutes is where the student decides whether they want to come back next week — make it warm and specific.

The template is a starting point, not a script. Younger learners (age 6-8) usually need shorter drill segments, more breaks, and a hard 30-minute cap; older intervention students (age 12+) can hold longer segments and often prefer them.

Effort-focused reinforcement (why a rewards mechanic matters here specifically)

Structured-literacy work is hard for the student. A dyslexic 8-year-old is putting more cognitive effort into a decoding drill than a typically-reading peer puts into their whole reading period. If your reinforcement pattern rewards only "correct answers," you're rewarding the student's ease, not their effort — and for a dyslexic learner, ease is the least available signal. Every experienced SpLD teacher's manual describes some version of this: reinforce effort, reinforce approximation, reinforce persistence through a hard drill, not just accuracy.

Translated online, the mechanic that makes this work is a small, frequent, teacher-controlled reward currency. On Koala Go, that's Gems: the teacher opens the student panel, picks a student, and gives 1-100 gems in a single click (the amount input is capped between 1 and 100 on purpose — small frequent rewards work better than a single burst). Each gem drops with a sound and a visual, and the student sees their balance update live. Why this matters for a structured-literacy lesson:

You control the timing. You can give a gem for a good sky-write attempt, for tapping through a hard word even if the blend didn't work, for trying a phonogram the student got wrong last week — reinforcing the process at the moment the process is happening.
The size of the gem matches the size of the effort. A quick correct answer gets 1 gem; a hard attempt on a novel word gets 5; a full week of shown-up-and-tried gets a bigger reward at the end. This is the "variable rewards" pattern established learning-design research supports — variable magnitude holds attention longer than a fixed one-per-correct schedule.
The student sees the balance grow. A dyslexic student often can't see their own progress — reading feels no easier at the end of a month of intervention, even though the decoding data says otherwise. A visible reward balance gives them one clear signal that they did do something today. That's more of a retention lever than most tutors expect it to be.

None of this replaces the tutor's real reinforcement — the direct, warm, specific "that was a good try" that the student came for. It's a lightweight second channel that carries some of the reinforcement load between your verbal moments. If gems (or any similar mechanic) feel infantilising for an older student, drop them — for a 14-year-old working on advanced morphology, the on-screen currency is often the wrong tool. For a 7-year-old two months into intervention, it's often what makes the child agree to come back on Wednesday.

Using external structured-literacy tools via the cobrowser

Most SpLD tutors have a set of structured-literacy web resources they already use — decodable text libraries, syllable-division practice sites, phoneme-manipulation web tools, leveled reader libraries. Traditionally on Zoom you'd screen-share and drive the mouse yourself; the student watches you click. A cobrowser (the same feature explained in what is a cobrowser?) changes that — you and the student are on the same live webpage, and either of you can move the cursor, click, type. For structured-literacy work, that unlocks:

Leveled reader sites. Reading A-Z, Oxford Owl, Epic, and similar libraries — open the book, the student reads aloud, controls the audio, turns the page. You can highlight or annotate on top of the shared page. Teach from sites your own plan or teacher license covers, and within each site's terms.
Web-based structured-literacy activity tools. Whatever your program uses for its digital practice pieces — the student does the activity in real time, you're on the same page, you can pause and take over if the student stalls.
Interactive phoneme / syllable practice. Any browser tool that supports drag-and-drop letter or sound tokens becomes usable inside the lesson — the student manipulates the tokens on the same page you're seeing, no "can you see my screen?" loop.
Reference sites and word lists. Pull up a phonics reference, an etymology database, or a decodable-word generator during the lesson; the student sees exactly what you're referencing.

One honest caveat on Koala Free: cobrowser sessions on the free tier are time-limited (you get 10 minutes at a time, with a 20-minute cooldown between sessions), which is fine for a single web-activity segment in a 45-minute lesson but not for a lesson where the cobrowser is the main teaching surface. Full-length cobrowser sessions come with Koala Pro.

Recording the lesson for the parent (and for your own records)

Recording is optional but for SpLD work it often pays for itself in retention. A parent who watches a 5-minute clip of their child working through a decoding drill sees the intervention working in a way that a written recap can't convey — the parent can see the hesitation, the correction, the tutor's cue, and the student's second attempt getting to the word. That's the difference between "I hope this is helping" and "I can see this is helping." Practically:

Get consent up front. Recording a minor requires parent permission and, in some regions, additional consent — see how do I record online tutoring lessons? for the consent, storage, and sharing side, including a link to a written policy you can adapt.
Share short clips, not full recordings. A full 45-minute recording is more content than any parent will watch. Clip the 3-5 minutes that show the target progress — the drill on the new phonogram, the moment the student blended the difficult word, the correction pattern — and share those. Some tutors send one clip a week alongside the written recap.
Store for the intervention arc, not forever. An intervention runs on the order of months to years; recordings that long deep-file quickly. Set a retention policy (delete after 90 days, or after the intervention completes) so you're not carrying an unbounded personal archive.

On Koala Go, lesson recording starts on Koala Pro (saved locally to your machine) and cloud recordings of scheduled lessons come with Koala Platinum; either way the recording captures the shared canvas alongside the video, so the parent sees the drill sequence, not just the tutor's face.

1-on-1 vs small group for intervention

The short answer: most SpLD tutors keep serious intervention 1-on-1. Small groups (2-3 same-level students) sometimes work for the phonemic-awareness or spelling components, or for a well-matched pair of siblings, but the individual pacing that structured literacy needs usually breaks in a group of mismatched-need learners. If you're considering the group model for the economics, see how do I teach small group classes online? for the general small-group tradeoffs — and note the "students with specific learning differences that require intensive individual attention" caveat in that page. The families paying specialist rates for SpLD tutoring generally expect 1-on-1 and are right to.

Practical setup on Koala Go for an SpLD tutor

If you're evaluating Koala Go (the platform we build) for a structured-literacy online practice, here's what you actually get, and where the current limits sit:

The whiteboard is the workspace. Fabric.js canvas with pen (Path), Sticky-note tokens, Stamp images, Text, and image upload — everything drag-and-drop-able and visible live on both sides of the room. Upload your existing decodable readers, dictation sheets, and phonics worksheets as PDF or PowerPoint (Free-tier caps on uploads; unlimited on Pro), annotate on top of them, and each session's canvas is saved to reopen next week.
Per-student Gems for effort-based reinforcement. Each student in the room has their own gem balance. You award 1-100 gems in a single click from the student panel; the transaction shows visibly to the student. Design the gem pattern the way your program prescribes reinforcement — small for participation, larger for a hard attempt, largest for persistence through a difficult drill.
Cobrowser for structured-literacy web tools. Open Reading A-Z, Oxford Owl, Epic, ImseLab-style interactive tools, or any browser resource your program uses; the student has real control of the page, not a passive view of your screen. 10 minutes at a time on Free (with cooldown), full-length sessions on Pro.
Playground for micro-breaks and stamina. The 3D Playground is a separate tab of the classroom — teacher and student avatars in a shared 3D space, useful for the 2-minute movement break between drill segments or as end-of-lesson reward. Not the teaching surface. On older student devices, drop the Playground Quality setting to Low (see the Quality settings guide) so the break doesn't stutter.
Lesson recording on Pro; cloud recordings on Platinum. Records the shared canvas alongside video — the parent sees the drill sequence, not just the tutor's face. Consent, storage, and sharing rules covered in how do I record online tutoring lessons?.
Browser-only, no install for the family. The student clicks your lesson link and lands in the classroom. No download, no account setup for the child. Works on the family's existing laptop / iPad / Chromebook. This matters because SpLD families are usually already managing more than the average — one less tool to install is one less friction to lose the family over.
Auto-generated class notes (Platinum). If the parent recap is the part you skip on a busy week, Platinum drafts a summary after each lesson which you edit and send in a minute or two. Consistent recaps are one of the biggest retention levers for SpLD families — see how do I keep parents renewing month after month?.

One private tutor described the outcome this way: "It has allowed me to set up a thriving online SpLD practice, now with a waiting list." The point is not that the platform is doing the teaching — it isn't; the tutor is. The point is that the delivery finally holds up online well enough for a specialist practice to run at scale.

Honest limits — what Koala Go doesn't do (and what else you'll want in your stack)

Frank about the gaps, because SpLD work runs long-term and small mismatches compound:

iPad experience is rougher than desktop. Koala Go runs in an iPad browser, and many students use it that way, but the pen precision and drag-and-drop feel are best on a Mac / Windows laptop with a mouse or a graphics tablet. If your student is doing letter formation on a first-generation iPad without a stylus, expect it to be workable but not great.
Stability on unstable connections. Video and shared-canvas sync do occasionally hiccup — a real-time collaborative environment is more sensitive to a bad Wi-Fi moment than a plain video call is. Families with underpowered internet will notice this; a wired connection where possible on the tutor side (Ethernet beats Wi-Fi for reliability) reduces it.
Diagnostic and progress-tracking tools are not built in. Koala Go is a lesson-delivery classroom, not an intervention-tracking system. If your program requires you to log detailed decoding-accuracy data per session or generate structured-literacy progress reports, you'll continue to use whatever tracking system you use today alongside Koala. We ship class notes, not clinical intervention records.
No included curriculum. We do not ship an Orton-Gillingham / Wilson / Barton / IMSE curriculum. You bring your own program; Koala Go is the delivery surface.
Playground gaps. Requested features that specifically help SpLD workflow — customisable timers on drill segments, more finely-grained syllable/phoneme token libraries — are still on the roadmap rather than shipped. (Group-select on whiteboard objects has shipped: drag a box around several tiles to move or resize them as one.) If any of these is a hard requirement, write to koala@teachwithkoala.com before you commit.

A short decision rule for whether to move an SpLD practice online

Are the families you'd serve reachable geographically only online? If yes — you're the only trained OG / Wilson / Barton tutor within a two-hour drive of a family that needs you — moving online is the intervention, not an alternative to it. Set up the tooling and ship. The families gain access they couldn't otherwise have.
Can your current or intended platform reliably do drag-and-drop tokens both sides can manipulate, and bring your existing structured-literacy web resources into the room? If yes, the pedagogy translates. If no — if the tool is a video call with a whiteboard on top — the multisensory delivery breaks.
Is your target student population one where engagement mechanics help? Under-12 dyslexic learners almost always; older intervention students usually less so, and often outright want no cosmetic layer. Pick a platform whose mechanics you can turn up for the young and down for the older.
Are you willing to hold the parent-communication cadence that makes the intervention retain? Short recap + occasional clipped recording, every session, for months. If you're not, the retention lever isn't the platform — it's the routine you build around the platform.

If you'd like to try Koala Go specifically for a structured-literacy lesson, the whiteboard, Gems, Playground, and cobrowser mechanics are on the Free tier — you can run a real 45-minute session with your candidate student on their real device before deciding. Open Koala Go at classroom.teachwithkoala.com, and if you want to talk through your specific program or student population before you commit, write to koala@teachwithkoala.com with a sentence about the method you use and we'll give you our honest read.

What scheduling actually has to do (six moving parts)

"Scheduling" sounds like one thing; it's actually six workflows that have to hang together over months of weekly lessons:

Availability. You publish which weekly slots you'll teach — Tuesday 4-6pm, Thursday 4-6pm, Saturday 9-11am — in your own timezone. Availability is the source of truth; every other step reads from it.
Bookings. Parents (or older students) pick a slot from your published availability and book it for a specific date. The system has to check that the slot is still open, that the student isn't already booked somewhere else at that time, and that the requested date sits inside whatever booking window you're accepting.
Timezones. The tutor and the student are usually in different ones. Availability lives in the tutor's timezone (which does not shift with the student's daylight saving); every other view — parent booking screen, calendar invite, reminder email — displays in the viewer's timezone. Getting this half-wrong is the single most common cause of missed lessons.
Recurrence. A working tutoring business runs on recurring weekly slots — the same 4pm Tuesday for as many weeks as the family stays. One-off bookings are the exception (trial lessons, make-ups, occasional exam prep), not the shape.
Cancellation policy. Minimum booking notice ("must book at least 12 hours ahead"), minimum cancellation notice ("cancel at least 24 hours before or the lesson is billed"), and maximum booking advance ("open the next 8 weeks; not the next 8 months"). Policy without enforcement is a suggestion; the platform has to enforce it or you'll be arguing about it in DMs.
Notifications. Booking confirmations, 24-hour reminders, 1-hour reminders, cancellation notices, and — if the family is on a monthly package — a warning when the package is running out. Missed reminders are the second-most common source of no-shows after timezone confusion.

Any tool you use has to cover all six. The tools that don't cover all six create the exact "stack juggling" they promise to remove — you end up with a scheduler that doesn't send reminders, a calendar tool that doesn't handle timezones, and a lesson link that lives in a different tab than either. The scheduling half of your tutoring business is either a solved problem or it's constantly on fire.

The three real options past 5 students

Solo tutors converge on one of three shapes:

Manual (email/DM/spreadsheet). Free, familiar, works fine for 1-5 recurring students. Fails at ~10 students because the spreadsheet + inbox stops matching reality after the first parent asks to reschedule and the second parent asks about a make-up in the same week. Every hour past that point is a hidden tax on your teaching time.
Generic bolt-on scheduler. A standalone scheduling tool — the category is well-populated (Calendly, Acuity Scheduling, SetMore, TidyCal, YouCanBook.me, and Google Calendar's booking pages, among others). You publish availability, parents book, calendar invites go out, reminders go out. Handles the six-workflow list above competently for one-off bookings. Tradeoffs: it does not know anything about your virtual classroom — the lesson link is a static URL you paste into the confirmation email, so if your room URL changes or you want per-student rooms, that's another manual step. Recurring bookings are supported on most tools but often via a workaround. And you'll pay $10-$25/month for the tier that has recurring + reminders + calendar sync.
Scheduling built into the virtual classroom. One login handles availability, bookings, cancellations, and the lesson link. When a parent books, they receive a link that opens the actual classroom room the tutor uses — not a separate scheduler URL that then hands off to a video URL. Cancellation policy is enforced by the same tool that runs the lesson. Tradeoffs: you're tied to that classroom's scheduling shape (if the classroom's scheduling doesn't do exactly what you want, you can't swap in a better scheduler and keep the classroom), and the scheduling feature usually sits behind the paid tier of the classroom itself.
Full LMS (Google Classroom, Moodle, Canvas, learning-management systems). Includes scheduling as a small feature of a much larger product. Overkill for solo tutoring — as discussed on whether Google Classroom fits private tutoring, an LMS is built around graded assignments and 25-student classes, not around 1-on-1 recurring lesson bookings. Standard for schools and tutoring organizations with formal cohorts.

The choice is usually between the middle two (bolt-on scheduler vs classroom-integrated scheduling). Manual is the pre-decision state; a full LMS is a different business.

Where each option is the right pick

A short decision rule that holds up in practice:

Manual is fine when you have ≤5 recurring students, all in the same timezone, all on the same weekly slot for months at a time. If you've never had to reschedule two families in the same week, don't over-build.
A bolt-on scheduler is right when your classroom tool doesn't ship a scheduler you're happy with, when you want a public-facing booking page (a URL you drop into your Instagram bio or a Reddit comment), or when you have a mix of student types and want to run different booking flows per type. It's also the safe default if you plan to switch classroom platforms in the next year — a scheduler you own is portable.
Classroom-integrated scheduling is right when you're settled on your virtual classroom and you want to stop maintaining two calendars. The single-login, single-link workflow is the win — every "which link is today's lesson?" question your parents ask is a cost you can eliminate. It's also the right pick for tutors whose parents are on WhatsApp/WeChat rather than email — the scheduler URL that parents actually book on becomes their whole relationship with your tutoring business, not a technicality behind an email.

Timezones — the hidden hard part

More lessons are missed to timezone confusion than to anything else in indie tutoring. The failure mode is always the same: the tutor sets Tuesday 4pm in their own timezone, the parent hears "Tuesday 4pm" in their timezone, both sides write it down, and one week later — usually the week daylight saving flips somewhere — the lesson gets missed. The fix is a small set of rules the scheduling tool has to follow whether or not the tutor thinks about them:

The tutor's timezone is the fixed point. Your Tuesday-4pm slot is 4pm where you live, not 4pm in some abstract UTC space. When you're in Los Angeles and daylight saving ends, your Tuesday-4pm is still your Tuesday-4pm — a student in Beijing sees the exact time on their side shift by an hour, and that's correct. If the scheduling tool is doing the opposite (storing slots in the student's timezone and shifting yours), you'll teach at 3am one week per year.
Store times as wall-clock time in the tutor's zone, not as a fixed UTC instant. Wall-clock storage is what makes DST behave sanely — "every Tuesday at 4pm my time" survives DST correctly; "every Tuesday at 23:00 UTC" doesn't. Any decent scheduler handles this; a naive spreadsheet does not.
Display in the viewer's timezone, always with the zone name visible. A parent booking should see "Tuesday, 8:00 AM (Asia/Shanghai)" — not just "8:00 AM". The IANA zone name is what disambiguates the +12 different offsets that all say "8am".
Send reminders that state both timezones. The lesson reminder to the parent should show the parent's time AND the tutor's time, so the parent can sanity-check across the trip they took last weekend. Two-timezone reminders halve the miss rate at almost no cost.
Be explicit with the parent at first booking. "You booked Tuesday 8am in Shanghai time, which is Monday 5pm in my time — please confirm." One extra confirmation email at the moment of the first booking prevents the ambiguity from surfacing three weeks later.

If any of this sounds obvious, it is. It is also, in our experience, exactly what indie tutors get wrong when they hand-roll scheduling on top of Google Calendar and a spreadsheet. The tools that get it right are boring about it; the setups that get it wrong lose one lesson per student per year to it.

Recurring vs one-off bookings

Recurring is the shape that makes the business work; one-off bookings are useful only around the edges.

Recurring bookings — the same slot every week, every fortnight, or every month, held open for as long as the family stays — are the calendar equivalent of a monthly package (see how to structure lesson packages). They lock the calendar (the parent doesn't re-decide every week whether to book), smooth cash flow, and cut the scheduling admin per student to near zero. Every mature indie tutor's calendar is 80%+ recurring slots.
One-off bookings are for trial lessons (before the family commits to a slot), make-ups (a family missed their recurring slot and wants an alternate this week), occasional exam-prep pushes (three extra sessions in the two weeks before the test), and drop-in students who don't fit any weekly pattern. Keep them as the exception, not the default.
Bookings that convert from one-off to recurring are the trial-to-paid workflow. A parent books a single trial lesson; after the trial, if it went well, you offer them a recurring slot at your normal package rate. The scheduling tool should make it one click to promote the trial slot into a weekly repeat, not force the parent to re-enter everything.

Whatever tool you use, check that recurring bookings survive edits — if you shift your Tuesday-4pm slot to Tuesday-5pm mid-term, the recurring bookings for the next month should update in the same operation, not require you to cancel and re-book each week manually.

Cancellation and no-show policy — enforcement, not intention

The policy itself (typical shapes: 24 hours' cancellation notice, no refund for no-shows, one make-up allowance per month) is covered in more detail on how to handle student no-shows and last-minute cancellations. The scheduling side of it is enforcement: is the policy something the parent has to click past when they try to cancel late, or is it a paragraph in your onboarding email that you have to argue about in DMs?

The mechanical controls that make policy enforceable — and that the scheduling tool has to expose — are usually three toggles:

Minimum booking notice. "Parents can't book a slot less than X hours ahead." Prevents 30-minutes-before-the-lesson bookings that force you to prep in a rush. A typical setting is 12 or 24 hours for regular slots; 0 hours (i.e., "instant lessons") if you're using drop-in office hours for existing students.
Minimum cancellation notice. "Parents can't cancel a slot less than Y hours ahead without it being billed." This is the mechanic that enforces your 24-hour cancellation rule. If the tool doesn't have this, your rule is a paragraph in an email; if it does, the tool refuses the late cancellation and the conversation ends there.
Maximum booking advance. "Parents can only book within the next Z days." Prevents a family from booking every Tuesday for the next 12 months, which would lock your calendar in ways you can't recover from cleanly. A typical setting is 60-90 days; families on a monthly package usually re-book each month around the invoice, which is the pattern the max-advance window supports.

Two more that are situational: an instant-lessons toggle for tutors who want to accept last-minute drop-ins (useful for existing-student office hours, generally bad for new-student trials), and a booking-window blackout for holidays and travel weeks (per-date overrides on top of the weekly recurring availability).

How Koala Go's Scheduling works specifically

If you're evaluating Koala Go (the virtual classroom we build) for the scheduling half of your tutoring workflow, here's the shape:

Where it lives. Scheduling is a separate SvelteKit app at classroom.teachwithkoala.com/portal that shares authentication with the main classroom — one login covers both. You open it from the classroom menu (the "Scheduling" item in the top-right menu). The whole teacher scheduling section is gated behind Koala Pro; teachers on the Free tier see the menu item as locked and can either upgrade or run manual scheduling until they're ready.
Availability calendar. You set weekly recurring availability slots — Tuesday 4-6pm, Thursday 4-6pm, Saturday 9-11am — in your own timezone. Times are stored as wall-clock time in that timezone (so daylight saving handles cleanly without you touching anything). You can also add per-date overrides for holidays, travel days, or special one-off hours ("closed on July 4; open Sunday 10-12 that week only"). Each override has a start time, end time, and an optional reason.
Invite-code onboarding for parents. Parents don't create their own account first — you generate an invite link from your teacher dashboard, share it with the family (email, WhatsApp, WeChat, whatever they use), and when the parent opens it they land on a unified page that accepts the invite, auto-creates their profile if they don't have one, connects them to your teacher account, and lets them immediately book a slot. It's one link to send and one link for the parent to open — the invite-and-book flow is one page, not a signup form followed by a scheduling flow.
Timezones handled correctly. The parent's booking view displays your availability in the parent's timezone (using Luxon); their booking confirmation renders in their timezone too, while your own dashboard renders in yours. If a parent moves — say, from a US-based family to a summer in Spain — the availability re-projects into the new timezone automatically. You never have to translate 4pm to another zone in your head or in a spreadsheet.
Recurring bookings up to 52 weeks. When a parent books, they pick one-off or a recurring weekly commitment — the recurrence options are 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 16, 26, or 52 weeks. A parent who wants "every Tuesday at 4pm for the next 8 weeks" makes one booking and the system creates all 8 lessons. If you set a maximum-advance policy (see the next bullet), the recurrence options are trimmed accordingly.
Booking policy controls (Platinum). Minimum booking notice, minimum cancellation notice, maximum booking advance days, and an instant-lessons toggle are all in the teacher profile, gated behind Koala Platinum ($49.99/mo monthly, $39.99/mo annual). On Koala Pro these fields exist in the data model but the UI shows an upgrade prompt — teachers can see them but not edit them. If enforced policy matters to your business (and past ~10 students it almost always does), that's the reason Platinum is priced where it is.
Custom lesson lengths (Platinum). On Koala Pro, lessons are fixed at 30 minutes. On Koala Platinum you can choose 15, 20, 25, 30, 40, 45, 50, 55, 60, 75, 90, or 120 minutes, plus an optional break between back-to-back lessons (0, 5, 10, or 15 minutes). The break setting is what prevents a full afternoon of back-to-back 60-minute slots turning into a bathroom-and-water crisis on your side.
Cross-teacher double-booking prevention. If a student is already booked with a different teacher on Koala Go at a given time, the booking is refused. Rare for solo tutors; matters for families whose kids are on multiple tutors' schedules and for tutoring companies running teams (Koala for Business).
Cancellation with an audit trail. When a teacher or parent cancels a lesson, the lesson row is marked cancelled with a timestamp and the id of whoever cancelled — useful if a family's story about the cancellation and yours diverge a month later.
Parent UI in 8 languages. The parent booking flow supports English, Spanish, French, Japanese, Korean, Russian, Vietnamese, and Chinese (browser-detected, no configuration needed). If your students' families are primarily non-English speakers, they book in their own language without you translating anything.
The lesson link opens the actual classroom room. The booking's lesson link opens directly into your Koala Go classroom — not a separate scheduler URL that then hands off to a video URL. One click to join, same room the teacher's already in, no bounce between tools.

Honest gaps worth flagging up front. We don't send automated lesson reminders — the booking and the recurring schedule live in your dashboard and the parent's portal view, but the day-before "don't forget tomorrow's lesson" message is one you send yourself on the family's channel (it takes ten seconds and doubles as the place to mention anything the family should have ready). We don't (yet) ship native 2-way Google Calendar sync — lesson bookings live in the portal dashboard, but they don't push into your Google Calendar as native events with 2-way updates; if your other clients live in Google Calendar and you need one unified view, most tutors bridge it with a calendar-forwarding tool for now, and native sync is on our list. And, as flagged on structure lesson packages, we don't auto-charge a parent's saved card on a monthly cycle — you invoice each month (one click; parent pays one link) and the recurring schedule books the lessons on top of the invoicing, but the parent is not silently billed by a background process. Same rules apply for all three: we'd rather be honest about a shipped-vs-not gap than sell you on a feature you'd find out about the wrong way.

Common failure modes to avoid

Not setting a policy at all. "I'll be reasonable about cancellations" is a policy that costs you money on the third late cancellation. Set the numbers before your first student and let the tool enforce them.
Manual scheduling at 15+ students. The tax on your teaching time from "let me check my spreadsheet" is real. Move to a scheduling tool before you resent your own students — usually somewhere between 5 and 15 students, depending on how many timezones and how much rescheduling.
Not confirming the timezone in writing. First-booking confirmations should say "Tuesday 4:00 PM Los Angeles time, which is Wednesday 8:00 AM Beijing time — reply YES to confirm." One extra step at booking prevents the missed lesson later.
Sending calendar links from multiple tools. If your scheduler emails a Google Calendar invite AND your classroom emails a lesson link AND you paste the lesson link into WhatsApp for the parent, one of the three will drift out of sync when a room URL changes. Send one link — the one the parent should click on lesson day — and treat everything else as a reminder that points back to it.
Booking too far in advance. Some families will happily book the next 12 months on a Sunday afternoon. That locks your calendar in ways you can't unwind cleanly when you decide to change your rates, take a term off, or move a slot. Set a maximum advance of 60-90 days and re-open the next window each month.
Ignoring the data-protection side for EU/UK families. If any of your students are in the EU or the UK, you are storing personal data of both the parent and the child (names, timezones, lesson history), which brings the standard consumer-privacy considerations along with it. This is a "check your local rules and, if you're growing past a few families in that region, talk to a lawyer" situation — the specifics vary by country and are not something a page like this can conclude for you.
Rebuilding scheduling from scratch every year. Whatever tool you land on, plan to stay there for at least 12 months. Every switch loses history, retrains parents on new links, and eats a week of admin. Pick once, deliberately, and stay.

How to decide, in one paragraph

If you have ≤5 recurring students all in one timezone, stay manual. If your virtual classroom already ships an integrated scheduler you're comfortable with, use it — the single-login, single-lesson-link win is the biggest scheduling upgrade solo tutors ever make, and it removes the constant "which link is today's?" question. If it doesn't, and you're already committed to that classroom, add a generic bolt-on scheduler (any of Calendly, Acuity, SetMore, TidyCal, or YouCanBook.me will do the job — the pricing and feature differences between them are small at solo-tutor scale). Whichever you pick, set your minimum-notice and cancellation-window numbers before your first booking, make the tool enforce them, and put the tutor's timezone in every confirmation email.

If you're considering Koala Go specifically for the scheduling half of your workflow, the built-in Scheduling is included with Koala Pro ($25.99/mo monthly, $21.99/mo annual) — you can turn it on, generate one invite link, share it with a real family, and see the full teacher-plus-parent flow inside an hour. If you need enforced booking policy or lesson lengths other than 30 minutes, that lives on Koala Platinum ($49.99/mo monthly, $39.99/mo annual). Questions about your specific timezone mix, subject, or business shape — write to koala@teachwithkoala.com and we'll give you an honest read.

What "multi-tutor tutoring business" actually means (three shapes)

The line between the shapes matters because each one is a different business, not just a different tutor count:

Solo tutor with student overflow. One tutor, more students than hours. The obvious move is to either raise prices or refer overflow to a colleague — not to hire. Very few solo tutors actually need to become a company; most just need a tighter waitlist policy.
Solo tutor + 1-3 subcontracted tutors. You still teach yourself, but you route overflow to trusted colleagues at a per-lesson revenue split. Your "company" is really you plus a Rolodex; each tutor uses their own account on whatever platform they already use; families pay you (or pay the tutor, and the tutor pays you a referral fee). Cheapest shape to try; the highest per-hour margin stays with each individual tutor.
Small tutoring team (3-15 tutors) under one brand. You're now running a business, not a practice. Families see one brand name, one booking link, one invoicing address; internally you have multiple tutors delivering lessons with an expected consistent shape. This is where the operational overhead starts to matter — curriculum consistency, tutor onboarding, quality review, tutor support all become jobs somebody has to do. Some tutors are employees, some are contractors; the classroom platform is usually shared and branded.
Established tutoring school (15+ tutors). Full operations team, HR, curriculum lead, dedicated admin dashboard, formal quality-review process, LMS-adjacent tooling. Not a solo-tutor evolution any more — a different kind of business with different problems and different platform needs. Named customers on Koala for Business at this size include Tutor Doctor.
Enterprise / district / franchise tutoring network. Multi-country, multi-currency, multi-brand-line. Different again; not what this page is about.

This page mostly covers the transition from "solo tutor" to "small tutoring team (3-15 tutors)" — because that's the shape where the platform decision matters most, and where the wrong platform decision hurts the business fastest.

The three operational problems multi-tutor businesses actually hit

Every tutoring-company operator we've spoken with runs into the same three problems roughly in the same order:

Lesson quality drifts across tutors. The reason your first ten families signed up is a specific style of lesson — your style. Your second tutor teaches almost the same way; your fourth tutor doesn't. Six months in, one family loves you and another has quietly churned because their new tutor's lessons feel generic. The cause isn't usually that the new tutor is bad — it's that there's no shared source of truth for "how a lesson at our company runs," so every tutor rebuilds it from their own habits. The fix is a shared activity/lesson library the whole team teaches from, not a Notion doc titled "how we teach."
The tools stack fragments across tutors. You started on Zoom; Tutor 2 came from Google Meet; Tutor 3 uses Skype for older students. Each tutor sends parents a different lesson link, a different scheduling link, a different invoice format. Families experience three different tutoring companies wearing the same logo. This is the "professional-with-parents" pain point that indie tutors already flag as a differentiator; at a small tutoring company it compounds into a churn problem.
The operator becomes full-time tier-1 tech support. Every parent connection issue, every "which link do I click" question, every Zoom-crashed-mid-lesson escalation ends on your desk because you're the visible face of the brand. You can either staff around the clock (expensive), let tutors lose lessons (churny), or move to a platform whose support team handles it for the tutor and the family directly. This is exactly the case for a purpose-built tutoring-company platform — the support cost you offload matters more than the license cost you take on.

Two secondary problems tend to arrive later, once the first three are handled: (a) tutor onboarding — how quickly can a new tutor be productive on lesson one? — and (b) quality review — how does the operator watch a lesson without derailing it, and how are shared decisions about "we don't teach it that way" made? Both are downstream of the three above and are best handled by the same shared platform, not by yet another tool.

The three real shapes past 3 tutors

Small tutoring companies converge on one of three platform patterns:

Stitched-tools stack. Zoom (or Google Meet) for each tutor + a shared Google Drive for lesson materials + a spreadsheet for the team schedule + separate scheduling and invoicing tools per tutor. Cheap; every tool is free or nearly free. Familiar. Fails at ~3 tutors because the Google Drive folder becomes a compost heap that only the founder can navigate, the shared spreadsheet drifts out of sync with what each tutor's actual calendar says, and there is no way for parents to book across the team from one place. The hidden cost is the founder's time: at 3 tutors and 30-40 families the ops overhead usually hits 15-20 hours a week of unpaid coordination.
Shared per-seat solo platform. Every tutor buys their own paid seat on whatever solo-tutor classroom platform the company standardizes on (Koala Go's paid tiers are one option in this category). Each tutor gets a strong classroom, each family gets a consistent lesson experience within a single tutor, and the company's tech stack finally rhymes across tutors. Works to about ~10 tutors. Where it stops scaling: there is no shared admin over tutors (you can't see all lessons across the team in one dashboard), no branded classroom (families still see the platform's name and not yours), no shared lesson library (Tutor A's Reading Level 3 activity has to be rebuilt from scratch by Tutor B), and each tutor's invoicing is their own — families experience 5 different bills or a manual reconciliation you do every month.
Purpose-built multi-tutor platform. A branded classroom hosted on your domain (or a subdomain the platform provisions for you), an admin dashboard over your team, one shared lesson library the operator can push into any tutor's library, unified billing and scheduling, and one support team the tutors and families both talk to. This is the shape most working small-and-mid tutoring companies land on eventually. Tradeoffs: it's a real commitment (a contract, a minimum monthly floor of hours, and a launch project rather than a signup), and you're tied to the platform's shape rather than to individual per-tutor choices. For most tutoring companies past 5 tutors the operational math flips in favour of this shape — but not before.

The choice is usually between the middle two (per-seat solo vs purpose-built multi-tutor). The stitched-tools stack is the pre-decision state; anything with 15+ tutors or a formal HR structure is a different problem and usually needs a full LMS layered on top of whichever classroom you pick.

Sharing lesson content across the team (the hardest sub-problem)

Of the three operational problems above, curriculum consistency is the one operators most often underestimate. "We'll keep our activities in a shared Drive folder" is the plan; six months later, half the team teaches from files they've since re-edited on their own laptops, the shared folder has three subfolders named "READING v2" / "reading final" / "reading NEW", and new tutors get a Loom video that no longer matches the current lesson shape. The fix is not more Notion pages — it's a mechanic that takes one lesson from one tutor's library and copies it into the libraries of the specific tutors you want teaching from it.

Concretely, the pattern that works is:

One tutor (or the company's curriculum lead) authors the canonical version of an activity. Not "the master copy in Drive" — the actual working activity, inside the classroom, that has been taught in real lessons and refined based on what worked.
The operator selects one or more activities and one or more recipient tutors, and copies. Each recipient gets a private editable copy in their own activity library — not a link to someone else's file, not a template they have to open-then-save-as. The copy is theirs, they can adjust it for their students, and their edits don't propagate back to the original.
Recipients see the copies clearly labeled (usually inside a distinct group in their library) so they can tell what the operator sent them apart from what they authored themselves. This matters because tutors are protective of their own material; a shared library that overwrites their personal work is a shared library nobody uses.
Copies from the platform's public marketplace (activities purchased from third-party sellers) are not redistributable inside the team. If your curriculum lead bought a phonics pack on a marketplace, the licensing usually restricts redistribution — a good platform enforces that guardrail server-side so you don't accidentally break a licence.

Everything else — versioning, editing rights, "please update your copy," change-log — is downstream of this basic push-copy mechanic. Without the copy step working cleanly, curriculum drift is inevitable.

What a platform needs to ship for the multi-tutor case

A shortlist for judging whether a virtual classroom is set up for a tutoring company rather than a solo tutor:

Branded classroom on your own domain (or a company subdomain the platform provisions). Families should see your name, your logo, and your colors — not the platform vendor's. The classroom URL is the primary thing families type into their browsers weekly; if it doesn't say your brand, you're building someone else's brand every lesson.
Admin dashboard over your team. The operator needs a view of all tutors, all connected families, all lessons, and lesson-level session recordings — from one login, without asking each tutor to screenshot their calendar. This is the difference between "we have a company" and "we have a group of freelancers wearing our shirt".
Shared, copyable lesson library. Per the previous section — the operator (or a designated curriculum lead) can push activities into any tutor's personal library, cleanly grouped so recipients know what came from the company vs. what they authored.
Unified scheduling and booking. Families book across the team from one URL. Cross-tutor double-booking (a student booked with two of your tutors at the same time) is prevented at the platform layer, not by a spreadsheet check.
One invoicing surface, with your brand on it. Whether you bill families directly or the platform bills on your behalf, the families see one company name on the invoice — not five tutors' individual PayPal accounts.
Support for both tutors and families that doesn't route back to you. Live human support (in-app chat, phone, email) that a tutor or parent can reach directly during a lesson-day incident. Every escalation the platform handles is an hour you don't spend on it.
Multi-device support that isn't a footnote. Your families are on Chromebooks, iPads, Android tablets, older Windows laptops. If any student's device is the "not supported" one, that family churns — and at a multi-tutor company you feel the churn on the team's revenue, not just one tutor's hour.
Session recording pipeline for quality review and absent-student catch-up. The operator needs recorded lessons for occasional QA (was Tutor 4's new-family trial delivered the way we teach?) and for family catch-up when a student is out sick. This is a platform capability, not a bolt-on.
An API for the standard operations. At some point (usually past 10 tutors) you want to pull teaching-time reports, lesson lists, and session data into your own reporting or accounting system. An enterprise API that speaks the standard nouns (teachers, lessons, sessions, teaching-time) matters more than a slick admin UI at that scale. If you already run a booking system you don't want to give up, that's the whole question — see does the classroom platform have an API we can integrate with our own booking system? for the workflow and the honest limits.

How Koala for Business handles it specifically

If you're evaluating Koala for Business (the multi-tutor / white-label version of Koala Go, which we set up for tutoring companies who've grown past the solo-tutor shape), here's what you actually get and where the current limits sit:

A branded version of Koala hosted on your own domain. Your tutors and families see your logo, your colors, your name — not "Koala". The underlying platform is Koala Go with the full Pro feature set (cobrowser, 3D Playground, PDF and PowerPoint uploads on the interactive whiteboard, in-lesson rewards) available for every tutor on your team, on a classroom URL that reads as your brand.
An admin dashboard over your team. The Koala for Business portal gives your operator a view of all tutors, teaching-time reports, per-teacher session lists, and recording access. The routes that matter are exposed both as an admin UI and as an enterprise API (/v1/teachers, /v1/lessons, /v1/sessions) authenticated with a Bearer token, so you can also plug the data into your own reporting and accounting.
Copy Activities — the shared-lesson-library mechanic. This is the feature we shipped specifically for the curriculum-drift problem above. From the Copy Activities tab in the business dashboard, an admin picks a source teacher, ticks the activities to share (individually or "select all"), and ticks the recipient teachers. Copies land in each recipient's activity library inside a group called "Sent by admin", clearly labeled so the tutor can tell operator-pushed content apart from their own. Guardrails we ship on purpose:
- Activities that were bought from the Koala Marketplace can't be copied — the licensing on marketplace activities restricts redistribution, and we block that server-side so you can't accidentally break it.
- Name collisions get auto-suffixed ("Alphabet Practice", "Alphabet Practice (2)"…) — one activity can be pushed to a team twice without overwriting a tutor's earlier edits.
- Copies stay editable by the recipient tutor; their edits do not propagate back to the source. Curriculum is pushed, not shared-and-locked — because tutors legitimately need to adapt a shared plan for individual students.
- In production, Koala staff accounts can't be a copy source (we enforce this on the server too) — the team library you're building is yours, not a mirror of Koala's demo content.
24/7 live in-app support for every tutor and every family. Our support team replies in under two minutes — often in seconds — around the clock. Your tutors and their families reach us directly during a lesson-day incident; you don't get pulled into it. Every escalation the support team handles is an hour you don't spend on it.
Direct access to a Customer Success Manager via Slack. Not a ticket queue; a Slack channel with a named CSM. This is the mechanic that keeps a multi-tutor customer's operational asks from being routed through a support form.
Session recording pipeline for enterprise sessions. Session recordings are processed on a dedicated GPU pipeline (custom-built FFmpeg with NVENC hardware acceleration) and made available for quality review and family catch-up.
Multi-device support for how your families actually connect. Chromebooks, iPads, Android tablets, and computers — all first-class. If a family's device runs a modern browser, the classroom runs.
Contractual 30-day launch. From signature to your team running live lessons is 30 days. That includes the branded classroom setup, DNS work if you're hosting on your own domain (five-minute DNS change on your side; we do the rest), Koala for Business account provisioning for each of your tutors, and testing.
Billing floor: 100 active hours per month. Koala for Business is priced with a minimum billable floor of 100 active teaching hours per month (we bill the minimum whether or not your team teaches all of them). Below that floor you're better off staying on individual Koala Pro seats ($25.99/mo monthly, $21.99/mo annual per tutor) — the multi-tutor overhead we ship isn't paying off for you yet. Above that floor, the platform-plus-support-plus-branded-domain math starts to work.
Case studies of what this looks like in production. Koala for Business is currently in production at Tutor Doctor, The Literacy Corner (a specialized reading-intervention practice for students with learning differences, running Orton-Gillingham instruction), Nihaoma Mandarin, Ladder Learning Services (an Atlanta-based reading nonprofit), Imerica Education, and Learning Success Academy. Different subject, different scale, same platform.

Honest limits to flag up front: Koala for Business is the right pick past 5-ish tutors, not before. At solo or 2-tutor scale the fixed costs (100-hour floor, launch project, contract) don't repay themselves — you're better off on individual Koala Pro seats and revisiting when the team crosses the "curriculum is drifting and I'm the bottleneck" threshold. And for tutoring schools running formal 20-student cohorts with an LMS-shaped assignment loop, Koala for Business is still a tutoring platform — you'll likely want an LMS beside it. If any of that describes your business shape, we'd rather tell you honestly than pull you into a plan you don't need yet.

Where a multi-tutor shape is the right business move

Signals it's time to hire or subcontract:

You have a waitlist that keeps repopulating faster than you can teach. If raising rates twice hasn't cleared the queue, that's real demand. A trusted colleague you can route overflow to is the smallest-step version; hiring is the larger one.
You're turning away students in a category adjacent to yours (a different subject, a different age range, a specialization you don't teach). An adjacent-subject tutor on your team expands your practice's footprint without you re-training.
Your reputation is bigger than your teaching hours. When parents recommend "the [Your Name] school" more than they recommend you individually, that's the brand outrunning the founder — the multi-tutor shape lets you deliver on the brand.
You want to trade some per-hour margin for capacity and calendar freedom. A team of tutors under your brand can generate more revenue than you personally can teach, and it eventually gives the founder days off. Solo tutoring, done well, is a demanding job with a hard ceiling.

Where staying solo is the right business move

Signals to not hire:

Your competitive edge is specifically you. A dyslexia intervention specialist with 20 years of clinical experience is not easily replicated. Families pay a premium for that specific person. Hiring dilutes what they're actually buying.
Your operational overhead is already 15+ hours/week and you'd rather teach. A team makes the ops problem bigger before it makes it smaller; if you're already resenting your admin load at 20 students, growing to 5 tutors and 100 families will not fix your calendar — it will destroy it for the first six months.
You've never enjoyed managing anyone. This is a real answer. Some of the best solo tutors are terrible small-business owners and know it. Recognize this before the third resignation letter.
Your rates aren't yet at the top of your local market. Raising prices is almost always cheaper than hiring. Do the pricing work first (see how much should I charge for online tutoring?).

Common failure modes to avoid

Hiring before your solo practice is a real business. If your own solo practice is unprofitable, adding tutors under your brand does not fix it — it multiplies the loss. Get the unit economics right on one tutor before hiring a second.
Skipping curriculum standardization. Assuming your new tutor will "figure out how we teach" is the leading cause of quality drift. Write the lesson shape down, teach it explicitly during onboarding, and use a shared library your tutors are actually copying from — not a Drive folder no one opens.
Under-investing in tutor onboarding. A new tutor's first ten lessons determine whether they'll retain families or churn them. Plan a real onboarding week — shadow lessons, walk-through of the shared library, a supervised first trial — not "here's the classroom link, good luck."
Contractor vs employee misclassification. Every country (and in the US, every state) has different rules on when a tutor working for your company is legally a contractor vs an employee. Getting this wrong is expensive on back-taxes and potentially on penalties. This is a "check your local rules and talk to an accountant and a lawyer" call — the specifics vary by jurisdiction and are not something a page like this can conclude for you.
Founder becoming tier-1 tech support. If every parent's connection issue reaches you, you'll spend your growth phase in the wrong role. Pick a platform whose support team the tutors and families talk to directly.
Fragmented brand across tools. If Tutor 2's parents click a Zoom link and Tutor 4's parents click a Google Meet link, families experience two different tutoring companies. Consolidate onto one branded classroom before your first hire's first lesson.
Not tracking teaching-time per tutor. At 5+ tutors, "how many hours did each tutor teach last month?" is a real payroll question and a real quality signal. If your platform can't tell you that from one dashboard, you'll spend a day a month rebuilding it from tutors' spreadsheets.
Under-pricing the seat. Sub-contractor rates that leave you with a $2/hour margin per tutor after platform fees are a business that fails at the first rough month. Do the margin math per tutor before you sign your first hire (see how to structure lesson packages for the pricing side).
Buying a platform you can't grow into. Every switch is a project — new links for parents, retraining tutors, rebuilding the shared library. Pick a platform whose top tier covers the size you plausibly hit in the next 24 months, not just the size you are today.

How to decide, in one paragraph

If you're still under 3 tutors, stitched tools plus a shared Drive is genuinely fine — don't over-build. If you're between 3 and about 10 tutors and each tutor is a fully solo practice under your brand, per-seat solo classroom licenses for each of them is usually the right shape, and Koala Pro at $25.99/mo per tutor is one honest option in that category. If you've crossed roughly 5 tutors, you're carrying real operational overhead, and any of the three problems above (curriculum drift, fragmented tools, founder-as-support) is costing you a day a week or more — a purpose-built multi-tutor platform starts making the math work, and Koala for Business is one of the ways to get there (30-day launch, branded classroom on your domain, Copy Activities for the shared lesson library, 24/7 live support your tutors and families reach directly, 100 active hours/month floor). Whichever shape you pick, do the curriculum-standardization and tutor-onboarding work regardless — those are business problems, not platform problems, and no software can substitute for having written down how a lesson at your company runs.

If you're considering Koala for Business specifically, the honest starting point is a conversation, not a signup — the setup is a real project and we'd rather scope it against your actual team shape than have you self-serve into the wrong plan. You can apply at business.teachwithkoala.com/apply or write to koala@teachwithkoala.com with a sentence about the size of your team and how you teach today, and we'll give you an honest read on whether it's the right time.

Why "does the classroom have an API?" is the right question for a tutoring company

A tutoring company almost never wants its classroom platform to also be its scheduling system. The booking flow you already run — teachers, availabilities, packages, cancellation policy, invoicing, the parent-facing site — is the business, and it usually lives in a mix of your own booking system, a general-purpose scheduler your operators know, and whatever your parents have already learned to click. Ripping that out to adopt a classroom vendor's scheduling flow is expensive twice: you lose institutional knowledge, and every family has to relearn where to book. So the question the ops or technical lead should actually be asking a classroom vendor is not "does your scheduling flow do X" but "will you let our scheduling flow drive your classroom?" That answer decides whether the classroom fits behind your business, or your business has to reshape itself around the classroom.

Concretely, the API question is asking three things at once:

Can we create a lesson from our own booking flow and get back a working join link for the family right then? If yes, the parent books at the same URL they've always booked at, receives the same confirmation email they've always received, and clicks the same "join lesson" button — and behind the scenes a classroom is provisioned and waiting.
Can we manage our roster of tutors from our own admin? Adding a tutor, disabling a tutor who's on leave, checking who's on the platform — from our own console rather than through yours.
Can we pull the data we need for our own reports and downstream systems? Attendance, teaching-time, session-level history, recording downloads — into our own analytics stack, our own accounting stack, or an audit archive we control, rather than living inside your vendor UI.

If the answer to all three is yes, the classroom becomes an infrastructure layer under your business — the way, say, a payment processor becomes an infrastructure layer under a shopping cart. If the answer is no, the classroom vendor becomes your operator's second job. This is the difference the ops lead is really trying to see through.

The workflow that makes it worth caring about

The API is designed around one particular shape, and it's worth walking through before you scope anything, because it's the shape the endpoints assume:

Parent books in your system. They pick a tutor, a time, a package — whatever your booking flow already offers. Your booking system remains the source of truth for who's booked with whom, when, and under what terms.
Your backend calls the API to create the lesson in the classroom. One request, identifying the tutor by email or by the ID we returned when you first added them, and the start and end time. We create a lesson row and return two links: one for the tutor, one for the family. Students don't need Koala accounts — they join the student link and enter their name.
You distribute the links through the channel you already use. Whatever your parent confirmation and reminder emails look like today, they now include the classroom join link. The parent's mental model — "I book here, I get an email, I click the link on lesson day" — never changes.
The lesson happens in the branded classroom. Same Koala classroom, but on your subdomain (or your own domain), with your logo and colors, so the family only ever sees your brand.
Attendance, teaching-time, and recording data flow back to your side on your schedule. Pull attendance for a lesson (or a batch of lessons up to 50 at a time) when you want it. Pull teaching-time-per-tutor for a monthly payroll report over any date range up to 365 days. Get a 7-day signed URL for the lesson's recording once it's ready, and store the file wherever your data-retention policy says it should go.
Recording readiness pushes to you. Rather than polling for "is it ready yet?", we can be configured to POST a webhook to an endpoint you host the moment a recording finishes processing. Both the lesson-level "recording processed" event and an individual-session recording.ready event are supported; both include a signed download URL in the payload.
If a lesson changes, delete it and create the replacement. There's no reschedule endpoint — a shipped-and-then-changed slot is a delete plus a create. Lessons that have already started can't be deleted; that's on purpose.

What this shape gives your operator is exactly the thing an in-house eng team wants and rarely gets from a vendor: your booking system is the source of truth, the classroom is downstream, and nobody has to schedule twice.

What an enterprise-grade tutoring API actually has to do

An honest evaluation checklist you can use to compare any tutoring-classroom vendor's API, not only ours:

Create a lesson and return the join link in one call. If lesson creation is a two-step dance ("create, then poll for the link"), your integration has to handle both retries and the case where the second call never lands. Getting the link back in the same response is a real correctness win.
Identify tutors by both an internal ID and an email address. Email is what your operators know at first; the vendor-side ID is what your integration should switch to once you have it, because emails change. A good API accepts either.
Bulk operations by default. Creating 40 lessons for the week ahead in one call is different from creating them in 40 sequential HTTPS round-trips. If the API only accepts one at a time, expect real friction when you migrate an existing schedule in.
Range queries with a documented cap. Fetching lessons in a time range is table-stakes; the cap on that range decides whether "give me last quarter" is one call or four. Publish the cap.
Idempotent-friendly semantics. Deleting an already-deleted lesson should surface as "not found", not silent success, so your integration knows what actually happened.
Webhooks with a shared secret and retries. A push model beats polling for recording readiness — but only if the vendor signs the payload (or uses a shared bearer secret you verify) and retries on your side's transient 5xx. Otherwise you're building the retry system yourself.
Signed download URLs, not permanent links, for recordings. A signed URL with an expiry (ours is 7 days, re-fetchable) is the right shape — it protects your students' recordings from getting shared into the world if somebody accidentally forwards an email.
Clean errors with human messages. "400: Date range cannot exceed 90 days. Current range: 112 days." tells your engineer what happened; "400: bad request" costs them an hour of guessing.
Seat-limit enforcement at the API, not the UI. If you can add tutors past your contract's seat cap through the API but not through the admin dashboard, the API is lying to you. Both surfaces must enforce the same limit.
A documented retention window on session-recording reprocessing. Raw recording segments can't be kept forever; the vendor should tell you plainly how long a session stays reprocessable so you know when a "please re-run" request will fail predictably.

If a vendor's API meets those criteria, plugging it under your booking system is a real project but a bounded one. If it doesn't, you're building the missing pieces yourself and the integration is unbounded — that is when a classroom vendor stops being an infrastructure layer and starts being your operator's second job.

How Koala for Business handles it specifically

If you're evaluating Koala for Business for the integration, here's the honest read on where it lands against that checklist. Every claim below is grounded in the reference at business.teachwithkoala.com/docs/api — treat that page as canonical for endpoint shapes, request bodies, and response fields; this page describes what the workflow feels like at the level a technical lead is deciding at.

Auth is a Bearer token issued by the Koala team at account setup. You send it in the Authorization header on every request; missing or invalid returns 401. The token grants full access to your organization's data, so treat it like any other production secret; rotating it is a support conversation rather than a self-serve action today, so build the rotation into your deployment process rather than assuming you can turn it over in a hurry.
Creating lessons returns join links in the same response. The POST /v1/lessons body accepts an array (bulk-friendly), you identify each lesson's tutor by either teacherId or teacherEmail, and each entry in the response includes a teacherLink for the tutor and a studentLink for the family. Students don't need accounts. This is the single call your booking flow needs to make when a parent books.
Lessons can be listed by ID or by time range. The range is capped at 90 days per request — enough for "give me next month" or "give me last month" in one call, and a good pointer to structure your reporting jobs monthly rather than trying to pull a year at once.
Deleting a lesson before it starts is fine; deleting one that already started is refused. There is no reschedule endpoint on purpose — moving a slot is a delete + a create against the new time. This is a cleaner mental model than "does this update or does this replace?", and it fits how most booking systems already treat cancellations.
Attendance is intervals, not a boolean. A single GET /v1/lessons/attendance request pulls join/leave intervals for the teacher and each student (up to 50 lessons at a time), including reconnects and a coarse descriptor of the device each side joined from — platform, and whether it was mobile — which can be absent. That's the shape you actually want for "did the tutor and the student overlap enough to bill this lesson?" and for diagnosing "why did the family think the tutor never showed?" scenarios.
Teacher management is honest about seat counts. Adding tutors via POST /v1/teachers creates a Koala account for each new email, and the request comes back 400 with an explicit message if your organization has already reached its seat limit. One detail worth knowing precisely, because it decides where you put the guard: that check runs once, against your current enabled-teacher count, before the batch is written — so a bulk request sent while you still have a seat free can carry you past the cap. Count your own headroom before sending a batch. Disabling a teacher frees a seat immediately; re-enabling checks the limit again.
Teaching-time reports over up to 365 days. One GET /v1/teachers/teaching-time call returns total classroom seconds per teacher over a date range. That's the primary report a tutoring-company payroll or ops workflow needs, and it's designed to be pulled once a month rather than polled continuously.
Recording downloads are signed URLs, valid for 7 days. The URL is re-fetchable at any time — if it expires before your archival job runs, you ask for a new one. This is deliberately different from "permanent recording URL", because permanent URLs are how student recordings get accidentally forwarded outside your organization.
Recording pipelines can push to you instead of you polling us. We support two webhook shapes: a lesson-level recording processed event that fires once the lesson's stitched recording is ready, and a session-level recording.ready event for individual classroom sessions (a lesson can have several, if the tutor reconnected). Both include a signed URL in the payload. Webhooks are set up by the Koala team — you send your endpoint URL and a secret token, we configure the delivery. Every delivery arrives with your token in the Authorization header; verify it before trusting the payload. Failed session-recording deliveries retry three times (1h, 4h, 24h), so plan your endpoint to be idempotent — the sessionId is your deduplication key.
Reprocessing a session recording is a first-class action. A POST /v1/sessions/reprocess call re-runs recording processing for a session — useful after a transient processing error, or to have the webhook fire again after your endpoint had downtime. Raw segments are retained for 30 days, so anything older than that can't be reprocessed; the API tells you plainly.
Errors are human. A rejected request comes back with an HTTP status and a message field a person can read — dates that are out of range, teachers that don't exist, lessons that already started. Your integration handles these predictably rather than parsing generic 400s.

What you get if you build against all of this is the shape described earlier: your booking system stays the source of truth, the classroom appears when it's needed, and the data flows back out to your ops stack on the cadence that matters to you.

Two things that are honestly not what a Business customer gets today

Both worth naming plainly, because both have caused real confusion in sales conversations we'd rather not repeat:

The standard API is included; custom development is a separate paid engagement. Every endpoint documented in the reference is available to any Koala for Business account on the standard plan — issue us a support ticket and we'll issue you a token. What's not included is custom work: endpoints we haven't shipped ("we'd like a webhook every time a lesson's recording status changes, not only when it's ready"), custom SSO integrations, custom event pipelines into your data warehouse, or bespoke changes to the classroom itself. Those are a separate engagement, usually with a higher base monthly minimum, and we scope them individually because their cost varies wildly. If you're planning an integration where the standard API's shape is the plan, budget as usual. If you're planning one where you need us to build alongside you, tell us early so the pricing conversation happens once and openly.
The teacher Scheduling portal from Koala Pro is not part of Koala for Business today. Individual Koala Pro tutors get a built-in Scheduling portal at classroom.teachwithkoala.com/portal where they set weekly availability and parents book against it — that's covered in how do I schedule online tutoring lessons?. That portal is an individual-tutor surface, and it isn't what a Business team schedules through. On Business, scheduling runs through the API instead: your booking system decides when a lesson happens; the API creates it. That's the right way round for a tutoring company anyway — you don't want each tutor keeping a private availability calendar that your operations team can't see — but it does mean you should scope the integration as the scheduling path, not as a stopgap until tutors get a portal. If you were counting on the built-in tutor Scheduling portal as part of what your team gets, we'd rather flag that gap now than have your operators discover it during rollout.

Three integration patterns worth scoping separately

The endpoints group naturally into three pieces of work, and they're worth scoping — and shipping — as three pieces rather than one. Nothing here is a rule; it's the decomposition we'd suggest if you asked us where to start.

Push-on-booking. Your booking system's "confirm booking" handler makes a single POST /v1/lessons to us; we return the join links; you save them alongside the booking record; your existing confirmation email now includes the classroom link. Cancellation in your system triggers a DELETE /v1/lessons on our side. This is the smallest useful integration, and the one to build first: on its own it's already enough to run real lessons.
Reconciliation-on-schedule. Once a day (or once a week), a job on your side calls GET /v1/teachers/teaching-time and GET /v1/lessons/attendance for the closed period and reconciles what your booking system thinks happened with what actually happened. This is where payroll and family-invoicing accuracy live — a family that never joined but was billed is a churn event; a tutor whose taught-hours differ from booked-hours by 4 minutes a lesson is a real annual payroll gap. Layer this on after push-on-booking is live.
Recordings-out-to-archive. Configure the lesson-recording webhook to POST to an endpoint on your side that downloads the signed URL and stores it wherever your retention policy says it should live — your S3 bucket, your GCS bucket, a compliance-audit archive. This one takes the longest to test because the events are irregular by nature, so it's the natural third piece — easiest to exercise once you're already teaching real lessons and have a real backlog to pull against.

Where the initial project surprises engineering teams is the timezone-and-DST math around lesson start times and range queries: both the booking system and the API speak ISO 8601 with offsets, but ops questions ("who taught in July, in the tutor's local time?") often need conversion in the calling code. The reference states the convention plainly — timestamps are UTC unless an offset is included — but the conversion for a report your ops team reads is yours to do; decide where it lives before the integration is scoped, not after.

Common failure modes to avoid

Treating the classroom as the source of truth. The API is designed for the booking system to lead. If your integration ends up storing "who's booked when" only in Koala and asking us for it every time, you've inverted the pattern and you'll feel it the first time our system is briefly unavailable. Your booking system is the record; the API mirrors bookings into the classroom.
Assuming the individual-tutor Scheduling portal is a fallback. Per the two hard lines above, it isn't the surface a Business team schedules through. Build the workflow through the API from day one; don't scope an integration that plans to "let tutors fill in the gap via the portal."
Polling for recording readiness on a tight loop. Every lesson's recording will be ready when it's ready. If your workflow needs to know quickly, configure the webhook — that's what it's for. Polling every 30 seconds for every lesson wastes both sides' resources and is likely to be rate-limited eventually.
Missing that lessons must be deleted before they start. There is no reschedule; a moved lesson is a delete + a create. Your booking system's "reschedule" button should call both endpoints and check the delete succeeded before creating the replacement.
Building without a seat-limit check on your side. As above, POST /v1/teachers refuses the request with a 400 once you have already reached your contract's cap — but it won't stop a single bulk request from carrying you past it. Check the current seat headroom (list teachers, count the enabled ones) before offering the "add tutor" action in your admin, and size the batch to fit. Otherwise your operator finds out about the cap the same second they show a new hire the door.
Not verifying the webhook token. Webhook deliveries arrive with a bearer secret we agreed at setup. If you accept any POST to your endpoint without verifying that header, an attacker who guesses your endpoint URL can spoof recording-ready events at you. Verify on every delivery, before your handler does anything else.
Storing recording downloads in a public bucket. The signed URL is 7-day-scoped precisely so that student recordings don't leak outside your organization by accident. If you download and re-host them in a public S3 bucket for "convenience", you've defeated the design. Store to a private bucket; expose recordings to families through your own auth-guarded page.
Storing lesson data across jurisdictions without checking your rules. A tutoring company running lessons for students in the EU or UK is handling personal data of minors — the standard local rules around consent, retention, and processing purpose apply, and vary by country. Talk to a lawyer familiar with education technology in the jurisdictions you operate in; this is not a topic a page like this can conclude for you.

How to decide, in one paragraph

If you're a solo tutor evaluating scheduling tools, this API isn't your surface — you probably want the read on how to schedule online tutoring lessons instead. If you're running a tutoring company past a couple of teachers and you already have a booking system your parents know how to use, an enterprise-grade API is the piece that lets you keep it, and Koala for Business is one way to get there without rebuilding either half. If you're at the "which classroom vendor" stage and you're evaluating against the ten-point checklist above, that's the right shape of question to ask every candidate. And if you're planning an integration whose success depends on Koala building endpoints or event pipelines we haven't shipped, tell us early so we scope the custom-development side openly — the standard API is included in your plan, but custom work is a separate engagement, and pretending otherwise on either side is the fastest way to a rollout that misses.

If you're considering Koala for Business specifically, the honest starting point is a conversation — the setup is a real project (30-day contractual launch, branded classroom on your domain, 24/7 live support for your tutors and families, dedicated Customer Success Manager via Slack) and we'd rather scope it against your actual booking flow than have you self-serve into the wrong plan. You can apply at business.teachwithkoala.com/apply or write to koala@teachwithkoala.com with a sentence or two about the booking system you're integrating and the team size, and we'll issue a token so your engineering team can look at the reference and prototype against real endpoints before the first sales call is over. The reference lives at business.teachwithkoala.com/docs/api; it's the single source of truth for endpoints and request shapes, and it's public — no login needed.

What onboarding a tutoring student actually is (and isn't)

"Onboarding" gets confused with two adjacent things — the trial lesson (before payment) and the ongoing lesson structure (once the family is settled) — but it's neither. The trial is a conversion moment: you're showing a family whether you and their child are a fit, and they're deciding whether to pay. Once the trial is booked and paid for, a different workflow begins that runs for about the first three to four lessons; that window is onboarding. Its job is to reduce the number of things that can surprise the family into leaving — tech surprises, scheduling surprises, "I didn't know what my child was learning" surprises, "the tutor keeps forgetting to send the invoice" surprises. Nothing about the tutoring itself has to be brilliant in this window; the tutoring has to be steady, and everything around it has to feel handled. If you're still in the trial-conversion stage, see how do I run a free trial lesson for online tutoring? — this page picks up after the family has said yes.

Onboarding is also not the same as long-term retention. The retention view is 3 months and beyond — the sustained cadence that keeps a family for a year (see how do I keep parents renewing month after month?). Onboarding is what sets up retention to be possible. The retention-week and the onboarding-week share about half their moves (progress updates, predictable timing, direct parent contact), but onboarding-week is denser: you touch the parent more often, you build the tech setup once, and you spend deliberate lesson time understanding the student before you commit to a curriculum shape.

The 5-line same-day welcome email

Send this within a few hours of the family saying yes. If you take three days to send it, you've already given up your first professionalism signal. Keep it short — five lines is enough:

The lesson link, right at the top. One click on lesson day, that's the goal. On Koala Go this is your room link — the URL that opens your classroom in the browser. Don't paste different URLs for different lessons; use the same room every time so the parent bookmarks it once and never asks again. In the classroom you tap the Invite Student button and copy the "Share this link with your student" URL — that's the one that goes in every email, WhatsApp, and calendar invite. If the family is in mainland China, use the China-routed variant of the link (the invite modal has a checkbox for this) so the classroom reaches them via the Hong Kong proxy.
The recurring lesson slot, in both timezones. "Tuesday 4:00 PM Los Angeles time, which is Wednesday 8:00 AM Beijing time — reply YES to confirm this slot works." One extra confirmation at the moment of the first booking prevents the missed lesson three weeks later. See how do I schedule online tutoring lessons? for the timezone rules a scheduling tool has to follow.
Your cancellation policy, one sentence, verbatim. "Cancel with at least 24 hours' notice or the lesson is billed in full; one make-up per month, scheduled within two weeks." Parents test tutors on this on the first late cancellation — if the policy has never been in writing, you can't hold to it. Full shape on how do I handle student no-shows?.
One thing for the student to do before lesson 1. Not homework — a low-stakes settling ritual. Kids under 10: "Draw a picture of something you love and bring it to the first lesson to show me." Older students: "Write down two topics you like in school and one topic you don't." Absolute beginners: nothing at all. This gives the student a reason to look forward to the lesson and gives you a warm opening in the first minute of lesson 1.
How to reach you between lessons. One channel. WhatsApp, WeChat, email — whichever the parent already uses. Two channels means you'll miss messages; one channel means the parent knows exactly where to look.

That's the whole email. Do not attach a 4-page policy PDF, a curriculum outline, or a self-marketing "about me" section on day one — the parent will not read it, and the more you send, the less they process. If you have longer documents (a parent handbook, a curriculum overview), link them; don't paste them.

The 15-minute tech check — before you teach anything

Book a 15-minute call, on the actual platform, with the actual device the student will use, before your first paid lesson. This is a real appointment on both calendars — not a "let me know if you have trouble." Many missed first lessons are actually missed tech checks: the student opens the link on a phone the parent didn't realise couldn't run the classroom, and the lesson time is spent troubleshooting rather than teaching. Run through:

The device. A modern laptop or Chromebook is easiest. iPads work in a browser but are rougher — pen precision and drag-and-drop feel are best on a laptop with a mouse or trackpad. If the student only has an iPad, confirm the browser (Safari is fine; Chrome for iOS is fine) and add a Playground quality-drop step to your tech check (see Playground quality settings). Phones are usable in a pinch but really are a pinch; if the family has anything else, use it.
The browser and permissions. Open your room link, grant camera and microphone permissions, confirm both sides can see and hear each other. Ask the student to click something on the shared whiteboard — a sticky, a shape, whatever — so you both know their side has interactivity. On Koala Go this is a browser-only flow: no download, no student account signup, no plugin. If a family is on a corporate laptop with a locked browser, that's better to discover today than at 4pm on Tuesday.
The connection. Ask them to run one quick round of a shared activity — you draw, they draw, you drop a sticky, they move it — and watch whether it's smooth on their side. If it's laggy, ask whether they can move closer to their router or plug in with an Ethernet cable. Video and shared-canvas sync are more sensitive to a wobbly Wi-Fi moment than a plain video call is; better to know today.
The lesson link and the calendar invite. Send the parent the exact link they'll use each week, and put it in a calendar invite that includes both timezones in the description. Confirm the invite landed and the time is right on the parent's side. If you use Koala Go's built-in Scheduling (on the Pro plan), the parent gets the invite through the portal booking flow — one link that opens both the schedule and, come lesson time, the classroom itself; see how scheduling works on Koala Go.
The recording question, once. If you plan to record — for the parent to review or for your own teaching notes — ask permission now, in writing, and be specific: "May I record our lessons for parent review only; recordings are stored privately and deleted after 90 days." See how do I record online tutoring lessons? for consent, storage, and retention. It's easier to have this conversation once, in the tech check, than to spring it on the parent mid-lesson.

End the tech check with the same sentence every time: "Great — I'll see you at 4pm on Tuesday. If anything changes, message me on WhatsApp; if you can't join right at the start of the lesson, don't worry, I'll wait in the room." The parent's fear at this stage is that they'll click the link on Tuesday and something will go wrong — the sentence tells them what to do if it does.

The first paid lesson — warm assessment, not test

The instinct in a first paid lesson is to prove yourself. Resist it. The first lesson's job is to find out where the student actually is and to make the student want to come back. If it also happens to teach a small amount of new content, good — but that's not what it's for. A working shape for a first 45-minute lesson (adjust up or down for age):

Warm opening (5 min). Ask about the thing you set as their "one thing to do before lesson 1" — the picture, the two topics they like — and actually listen. This is the moment the student decides whether you're another adult evaluating them or someone they're going to spend an hour a week with. Warmth is the whole game here.
Quick tour of the classroom (3 min). "This is the whiteboard, this is where we'll draw together, this is where the tokens live, this is the Playground where we take breaks." Do it fast, not thoroughly. The student learns the tool by using it, not by being shown it. If the student is under 10, take 60 seconds to walk their avatar around the Playground once so they know it exists — you'll come back to it as a break in a moment.
A low-stakes activity you brought (15 min). Something you can do without knowing anything about the student's level — a warm-up conversation for language, a "let's read this together and tell me what you notice" for reading, a small "what do you think this is" for math or science. On Koala Go, upload the material as a slide deck or PDF (image upload is Free; PDF and PowerPoint upload is on Pro) so it lives on the shared whiteboard and both of you can annotate it. This is the segment where you're mapping how they think, not testing what they know.
Movement break (2 min). Especially for under-12s. The Playground is a low-stakes way to do this without leaving the classroom tab — "walk your avatar to the sign and back while I get the next thing ready." Older students can just stretch and hydrate; you don't have to force the Playground on them.
A short "here's what I'd propose" segment (10 min). Try one thing that's slightly harder than what you just did. Watch how the student responds. If it's too hard, back off; if it's too easy, adjust. You're not covering material — you're calibrating.
A visible "you did something today" moment (2 min). On Koala Go this is often the first Gem — you open the student panel and give 1 to 10 gems for the effort in the lesson, and the student sees the balance appear. It's a small, warm signal that today counted. If gems feel infantilising for the student (teens working on exam prep, for example), skip the gem and use plain warm words — "You worked hard on the middle bit. That's what I look for."
Wrap-up and preview (8 min). Say what you noticed. Say what you'd like to try next lesson. Ask what the student wants to work on. Confirm the next lesson time in both timezones. If the parent isn't in the room, say — with the student — "Tell your mum/dad I'll send them a short note tonight."

Do not run through a full diagnostic assessment in lesson 1. It burns the student's energy on being tested rather than on being taught, and it teaches them that our lessons are quizzes. A diagnostic that needs to happen — for a structured-literacy intervention, an exam-prep programme, an ESL-placement — is better done across lessons 1 and 2 in short segments rather than in one 45-minute sit-down. See online tutoring for dyslexia if you're onboarding a structured-literacy student — the pacing changes further.

The parent recap after every one of the first four lessons

Send a short written recap to the parent within 24 hours of each of the first four lessons. Not a form, not a report — a five-line message, on the same channel you agreed on in the welcome email:

What we did. Two or three specific things. "We read a short passage about lions, and worked on the /sh/ sound in words like shark and fish." Specific beats generic. "We had a great lesson" tells the parent nothing.
What went well. One concrete moment. "Anna asked to read the second page herself, which is the first time she volunteered." One small observed detail per recap is more credible than any list of strengths.
What I'd work on next. One thing. Not five. "Next week I want to try a slightly longer text, still in the same reading range, to see whether stamina is the ceiling or whether it's the word-recognition."
One thing to do at home this week (optional). A single, small, doable practice — not homework in the school-assigning sense. "If Anna wants to, she can read us the sentence I've dropped in the shared notes. Two minutes; totally optional." If the parent is on Koala Platinum territory or you're already sending them clips from the recording, this is where a 60-second clip of the lesson goes.
Confirm next lesson time. One line. "See you Tuesday 4pm (LA) / Wednesday 8am (Beijing)."

The recap is where the parent watches you being organised without being asked to be. If drafting five lines from scratch is the part you skip on a Tuesday night, Koala Platinum's auto-generated class notes produce a draft summary after each lesson that you edit and send in a minute or two — the drafting cost is the barrier for most tutors, not the writing. For the first four lessons specifically, don't skip the personal edit; the class notes give you the bones, but the specific-moment observation in line two is what makes the parent trust you.

After lesson 4, most families settle into a slower cadence — a recap every 2 to 3 lessons is enough — but the first four are dense on purpose. That's the window where the parent decides whether you're a real business or a hobby.

The lesson-4 check-in — you initiate, not the parent

Around lesson 4, message the parent to book a short (10-minute) call. Not a review, not a "how am I doing" — a check-in you're calling because it's on your schedule. The framing that lands with parents is: "I've now seen your child in four lessons. I have some notes on what I've learned about how they work best, and I'd like to run my proposed direction past you before we go further." The parent doesn't have to prepare, doesn't have to have a complaint, doesn't have to feel like they've been summoned. You are naming the momentum they were half-worried about.

In the call itself, three things:

What you've learned about the student. Two or three specific observations — how they focus, what motivates them, one thing that's harder than you expected, one thing that's easier. This is the moment the parent hears back that you saw their child, not just that they showed up.
What you'd propose next. One direction. Not options. "For the next month, I'd like to focus on X, and I think we should keep the current weekly slot; we can revisit at the next check-in in six weeks." Parents renew when they know what they're paying for next.
What they want. Ask, actually listen, and adjust if what they want is different from what you'd proposed. Two sentences of feedback here save you three months of drift.

End with the next practical step — the invoice for the next package, the confirmed slot, the next check-in date. On Koala Go's Scheduling (Pro), you can hand them the invite link and they book the next 8 weeks in one click; on manual scheduling, send the calendar invites and confirm receipt.

What to actually set up on Koala Go for a new student

If you're using Koala Go (the classroom we build), here is the concrete first-week checklist:

Get your room link, once. Open the classroom, click Invite Student, copy the "Share this link with your student" URL. That is the URL you send in every welcome email, every calendar invite, every WhatsApp message. Do not generate different links per lesson. If the family is in mainland China, use the China-routed variant (there's a checkbox for it in the invite modal).
Upload the first lesson's materials. On the whiteboard, add the picture / slide deck / PDF / worksheet you want to use in lesson 1. Image upload is on the Free tier; PDF and PowerPoint uploads are on Koala Pro. Uploaded material stays on the whiteboard between lessons — you can reopen the same board next week without reuploading.
Pin a warm object to the whiteboard. A welcome sticky note with the student's name, or a small image (a mascot, a birthday cake if it's their birthday week, whatever fits). The student sees this before you turn on video. It's the online equivalent of a name card on the table.
Set up a first Gem transaction template. Decide, before lesson 1, how many gems you'll give and for what. For an under-10 first lesson, 5-10 gems for "trying hard on the reading" is warm and generous without being uncalibrated. Gems are 1-100 per transaction, teacher-controlled — small and frequent works better than a big one-shot burst.
Enable Scheduling if you're on Pro. On Koala Pro ($25.99/mo monthly, $21.99/mo annual), open the Scheduling section of the classroom and set your weekly availability in your own timezone. Generate an invite link for the new family and send it in the welcome email — the family accepts the invite and books their recurring slot in one flow. See how scheduling works for the full workflow. On the Free tier, do this manually (calendar invites in both timezones, weekly reminder) until the volume forces the upgrade.
Decide about recording, and get consent up front. If you plan to record for parent review or your own notes, ask the parent in the welcome email or during the tech check, and follow the consent, storage, and retention shape on how do I record online tutoring lessons?. Recording is on Koala Pro. Don't decide to record mid-lesson; ask before, or not at all.
Bookmark a cobrowser destination for the first curriculum tour. If any part of your teaching lives on an external website — Reading A-Z, Oxford Owl, Epic, a grammar site, a maths practice site — plan to open it via the cobrowser during lesson 2 or 3, not lesson 1 (lesson 1 is warm-assessment territory, not curriculum). Cobrowser sessions on the Free tier are capped at 10 minutes at a time (with a 20-minute cooldown); on Pro you get full-length sessions. See what is a cobrowser? for how it differs from screen sharing.
Turn Playground quality down on the student's device if needed. If the tech check surfaces a laggy Playground, help the student drop the Playground Quality setting to Low (see why are avatars walking in slow motion?). Doing this once, in the tech check, prevents the "the game is broken" moment in the middle of lesson 1.

Koala Go is browser-only, so the student doesn't install anything, doesn't create an account, and doesn't sign in — they click your link and they're in the room. That's the whole friction budget for the parent. Every step you add on top of that is a step the parent has to explain to a family member on lesson day, so keep the count low.

Honest limits — what Koala Go doesn't ship for onboarding

Straightforward about the current gaps, because families notice small mismatches:

No purpose-built onboarding form. We don't ship a "new student intake" flow that captures the student's age, level, goals, and parent contact into a structured record. Most indie tutors use a simple Google Form or a couple of lines in their scheduler for this; when the volume gets past ~15 students, a lightweight CRM (Notion, Airtable, or a spreadsheet) is what most tutors settle on. This is on our list; it isn't shipped.
iPad experience is rougher than desktop. Koala Go runs in an iPad browser, and many students use it that way, but the pen precision and drag-and-drop feel are noticeably better on a Mac / Windows laptop with a mouse or a graphics tablet. If your student is on a first-generation iPad without a stylus, the tech check should include a plan for that.
China connectivity has an infrastructure workaround, not a solve. Families in mainland China connect through our Hong Kong proxy (the China-routed room link the invite modal offers). It works for most of the time, and it works better than a bare video-conferencing tool tends to, but connectivity there is inherently constrained. If any part of your first-week onboarding depends on the family confirming they can reach the classroom reliably, do the tech check on the actual China link, not on the default one. Full context on teaching English online to students in mainland China.
Class notes are on Platinum. The auto-generated recap that speeds up the first-four-lessons parent-communication cadence is a Koala Platinum feature ($49.99/mo monthly, $39.99/mo annual). On Free and Pro you're drafting the recap yourself. It's five lines — writing them isn't hard; the hard part is doing it every week, which is why the auto-draft exists.
Data-protection setup depends on where you and your families live. If any of your students are in the EU, UK, or another region with formal data-protection rules, storing parent + child data during onboarding brings the standard consumer-privacy considerations along with it. This is a "check your local rules and, if you're growing past a few families in a region, talk to a lawyer" situation — the specifics vary by country and aren't something we can conclude for you.

A short decision framework — what to formalise, what to leave loose

Formalise the tech check and the welcome email. Every new family, every time, no exceptions. These are cheap to do consistently and expensive to skip; the missed-first-lesson recovery is always more work than the 15-minute call was.
Formalise the first-four-lessons parent recap cadence. Every lesson, five lines, same channel. Once the family is settled (usually by lesson 5 or 6), you can drop to every second or third lesson if you want.
Leave the lesson-1 activity loose. The point of the first lesson is to meet the student where they are, not to execute a script. Bring materials you can bail on if the student's level is different from what you expected.
Formalise the cancellation policy the day before the first lesson. In writing, one sentence, in the welcome email. Enforce on the first incident. Almost every "I keep losing students" pattern indie tutors report traces back to unenforced or unshared policy in the first two weeks.
Initiate the lesson-4 check-in yourself. Don't wait to be asked. Parents who are close to renewing want to be sold on the plan; parents who are close to leaving want to be asked what's wrong. Either way, the lesson-4 conversation is yours to schedule.

If you'd like to try Koala Go specifically for onboarding a first paid student, the invite link, whiteboard, Playground, and Gems all work on the Free tier — you can run a real tech check and lesson 1 with a real family before deciding whether to move to Pro. Open Koala Go at classroom.teachwithkoala.com, or write to koala@teachwithkoala.com if you want a quick honest read on whether it fits your particular subject and student age.

What "online tutoring for ADHD students" is (and what it isn't)

The reader we're writing for is a tutor who's already working with — or about to start working with — a student who has ADHD, formally identified or clearly showing attention-and-executive-function differences in the way they engage with a lesson. If you're an academic tutor, an ESL tutor, a math tutor, an executive-function coach, a specialist working on reading or writing intervention with ADHD comorbidity, or a general subject tutor who's found that a chunk of your caseload is neurodivergent, this page is for you.

What this page is not:

Diagnostic content. We don't help parents decide whether their child has ADHD; we don't list DSM criteria; we don't recommend evaluations. That's the work of a clinician. If a parent is asking whether their child has ADHD, point them at a paediatrician, psychiatrist, or educational psychologist — that referral is worth more than any lesson plan.
Medical advice. Medication questions belong with the prescribing physician, not the tutor. If a family asks whether their child should be on medication, or how it interacts with tutoring, the honest answer is "that's a conversation for your doctor" — not something to weigh in on from your side of the screen.
Special-education-law advice. IEPs, 504 plans, EHCPs, SEND provisions, and the equivalents in other countries are formal legal instruments and vary by jurisdiction. If the family has one, ask to read it and let it inform your lesson; don't advise on whether they should get one.
ADHD coaching. Coaching adults (or older teenagers) on managing life with ADHD — job scheduling, medication logistics, self-advocacy — is its own trained discipline. This page is about academic tutoring for school-age learners with ADHD; not the same thing.

What actually changes when the lesson moves online

Tutors who've moved ADHD-heavy caseloads online report the same three constraints in the first month or two:

The screen is the distraction. In-person, "look at me" or "look at this book" is a physical redirect. Online, the student's attention has an infinite alternative one tab away, one notification away, one YouTube autoplay away. Your platform can help or hurt here. A tool that keeps the lesson content inside the classroom (whiteboard, shared canvas, cobrowser to a specific site) does better than one that forces the student to leave the tab to open a resource, then not come back.
Sustained attention windows are shorter than the lesson. An ADHD 9-year-old's attention on a screen-based drill is often 5-10 minutes before the mind wanders. The lesson isn't going to fit in one 45-minute attention block; it's going to fit in four or five 8-10-minute blocks with transitions between them. Your platform has to support that transition without kicking anyone out of the room or losing the previous canvas state.
Time blindness compounds online. ADHD students often struggle to feel the passage of time. In a physical room you have posture, movement, and ambient cues; on a screen, the room looks the same at minute 3 and minute 30. If the student doesn't know "how much longer" or "how many more of these," they'll disengage before they can name why. Explicit structure — visible progress, count-down of remaining tasks, a stated "we'll do three of these, then a break" — matters more online than it did in person.

Structuring the online session for shorter attention windows

Most ADHD-informed teaching guides converge on the same shape: short work intervals, frequent transitions, movement between them, immediate salient reinforcement, and a visible sense of "how much is left." Translated to a 45-minute online lesson, a working template looks something like:

Explicit start ritual (2 min). Same opening every session — "we're going to do four things today, then two minutes of playground." Say the four things. This is not filler; it's the executive-function scaffolding that lets the student walk in without having to invent structure. Some tutors write the four things on the whiteboard as a visible checklist the student can watch cross off.
First work interval (8-10 min). The hardest cognitive work — new content, difficult drill, unfamiliar problem — goes first, while attention is freshest. Keep it interactive: the student is dragging, typing, clicking, choosing. Passive watching in the first block wastes the best attention of the session.
Movement / attention reset (1-2 min). Genuinely move. Stand, stretch, walk to a wall and back, do five jumping jacks, sky-write the tricky word. Do not replace this with "quick brain question" — the point is physical, not cognitive. Some tutors use the Playground here (a short "walk to the sign, come back" with the student's avatar) because it's low-stakes movement inside the same tab and doesn't risk the student wandering off the lesson.
Second work interval (8-10 min). Different modality from the first — if the first was verbal and analytic, this one is visual and manipulative (or vice versa). The shift in mode is itself a re-engagement lever; the same student who fell off a listening-heavy task at minute 12 will lean back in for a drag-and-drop token task at minute 15.
Movement / attention reset (1-2 min). Same as above. Yes, again.
Third work interval (6-8 min). Practice and consolidation of what was taught in intervals one and two, at a lower difficulty than the introduction. Attention is thinner by now; drop the cognitive load and increase the doing.
Wrap and reward (5 min). Recap the four things. Award the reinforcement you promised at the start. Preview one thing to do this week. End on a warm, specific "here's what you did well today" — that specificity is what the student takes into the interval before your next session.

The template is a starting point, not a script. Younger students (age 6-8) usually need 5-6 minute intervals with more breaks. Older students (age 13+) can often hold 12-15 minute intervals and get restless if you break them too often — read the room. The point isn't the exact minutes; the point is that the shape has intervals-and-breaks rather than being a single continuous block.

Distraction management (the environment, the device, the tabs)

Half of "keeping a student engaged online" is removing the alternatives. A few concrete moves that pay off:

Agree on the environment with the parent before lesson one. A door that closes. A screen the student uses only for the lesson if possible (or at least a clean desktop before the call). Phone put in another room during the session (this is often the single biggest lever with older students). Younger siblings' whereabouts pre-negotiated. None of this is your job to police mid-lesson; all of it is your job to set expectations for at the start.
Full-screen the classroom. If your platform supports full-screen mode, use it. It reduces the visible "escape" surface — the bookmarks bar, the other tabs, the clock in the taskbar — and it visually commits the student's device to the lesson. On Koala Go the classroom tab is designed to hold the lesson, whiteboard, playground, and cobrowser inside itself, so a full-screen tab keeps the student in the room.
Close notification-heavy apps on the student's device. This is a parent-side ask, but a valuable one. Discord, iMessage, TikTok, Instagram, YouTube — all designed to pull attention. Ideally quit before the session, at minimum on Do Not Disturb. For a middle-schooler, this is not optional.
Use the cobrowser instead of asking the student to open a link. When your lesson needs a live web resource — a decodable text, a math tool, an interactive activity — a cobrowser lets you bring the site into the classroom (see what is a cobrowser?). The student doesn't leave the tab and doesn't end up on the sixth Reddit thread while you're saying "no, click the link I sent." On Koala Go, cobrowser sessions are 10 minutes at a time on the free tier (with a 20-minute cooldown) and unlimited on Pro.
Turn off your own notifications too. The student can tell when you're distracted, and it gives permission for their attention to drift. Do Not Disturb on your device, no other tabs visible, no email pinging in the background.

Reinforcement that fits ADHD (immediate, salient, variable, effort-based)

The reinforcement pattern that works with ADHD learners is well-documented across teaching guides: it needs to be immediate (delayed rewards do very little), salient (small text on a page won't register; visible / auditory / animated will), variable (a predictable one-per-correct schedule becomes wallpaper within a few sessions), and effort-based (rewarding "correct" only misses the harder work of persisting through an incorrect attempt). If any of those are missing, the mechanic stops carrying weight and you're back to relying on the student's baseline attention to hold the lesson together.

Translated online, the mechanic that supports this pattern is a small, frequent, teacher-controlled reward currency you can trigger without leaving the lesson flow. On Koala Go, that's Gems: you open the student panel from inside the room, pick the student, and give 1-100 gems in one click (the amount input is capped between 1 and 100 on purpose — a rhythm of small frequent rewards beats a single big burst). Each gem drops with a sound and a visual, and the student's balance updates live on their side of the screen. Why this shape supports ADHD-aware reinforcement:

You control the exact timing. Not the platform, not a schedule, not the completion of a task. You. That means you can reinforce the specific moment the effort happened — "you didn't want to try that word, and you tried anyway" — while the student's brain still has the effort in working memory. Delayed rewards, for an ADHD learner, are close to invisible.
The size matches the effort, not the outcome. A quick correct answer gets 1 gem; a hard attempt on a novel problem gets 5; a full session of "you kept showing up to the work" gets a bigger one at the end. This is the variable-magnitude pattern that learning-design research supports and that also matches how ADHD students actually experience reinforcement — the variability itself is what holds attention across the lesson.
It's visible. A number that goes up on the screen. Some students will glance at it after every drop; some will forget it exists for 20 minutes and then be pleased to discover it grew. Both work. What doesn't work is a rewards system the student can't see happening.
It's optional. If a specific student — usually an older teenager — reads the gems as infantilising, don't use them. For a 15-year-old working on SAT prep, a visible ADHD-friendly reinforcement mechanic might be your enthusiastic voice ("nice — that was a good attempt"), the visible checklist of "three left, two left, one left" on the whiteboard, and the concrete "we'll end 5 minutes early if we get through these" — no gems required. The mechanic is a tool, not the whole strategy.

None of this replaces the tutor's live warmth — the specific, timed "that was a good try" that no platform can deliver for you. It's a lightweight second channel that carries some of the reinforcement load between your verbal moments.

Keeping the student doing (not watching)

The single strongest engagement lever for ADHD learners online is minimising the passive-watching share of the lesson. Every minute the student is dragging, clicking, drawing, choosing, or typing is a minute their attention is anchored to the screen; every minute they're watching you talk is a minute their attention is trying to leave. The Koala Go primitives that reduce passive watching:

Drag-and-drop tokens on the shared whiteboard. Sticky-note tokens and Stamp images both drag on the shared canvas and update live for everyone in the room. Instead of the student watching you write, you place the tokens and the student sorts / groups / matches / builds. Multiplication problem: the student drags the digit into the answer box. Reading passage: the student drags the correct picture onto the sentence. Spelling: the student drags the letter tiles into a word. The student is doing.
Free draw with the pen tool. Fabric.js Path tool — you set the problem, the student draws the answer. Long division worked out. A diagram labelled. A route traced. Even for students who dislike writing, the physical act of drawing on a shared surface is more engaging than watching you fill in a worksheet.
Cobrowser to a real-world resource. If the lesson uses an external website — an interactive math site, a science simulation, a leveled reader library, a language-practice tool — the cobrowser lets the student control the cursor in real time on that site (see what is a cobrowser?). You're not driving; they are. That single shift — from watching the tutor click to controlling the click themselves — is a large chunk of what "engagement" means for an ADHD learner.
Playground for the movement break. The 3D Playground is a separate tab of the classroom — teacher and student avatars in a shared 3D space, useful for the 1-2 minute break between work intervals or as an end-of-lesson reward. Not the teaching surface. The point isn't to replace the lesson with a game; the point is that a controlled break in the same tab is better than "go play on your iPad for 2 minutes" — which the student will not return from.

The parent as co-teacher, not co-fighter

ADHD tutoring at any age involves the parent more than a neurotypical caseload does — because the routine around the lesson (getting the student to the desk on time, phone put away, materials ready) is often the bottleneck, and because the retention story runs on the parent seeing progress. A few moves that reduce parent-side friction:

Set the pre-lesson ritual with the parent, not the student. "5 minutes before the session, phone in the kitchen, water on the desk, headphones on if in a shared room" — this is the parent's job to enforce, not yours to negotiate at minute 1. Send it once in writing, refer back to it if it slips. The tutor's job starts when the student is at the desk; the parent's job is getting them there.
Send a specific — not generic — recap after each session. "We worked on multiplication of two-digit numbers today. He got the concept solidly by the end. He struggled with keeping the columns aligned — practice on graph paper this week, five problems." That's a recap a parent can act on. "We had a great lesson!" is not. On Koala Go, auto-generated class notes (on Platinum) draft this for you after each lesson, and you edit them in a minute or two before sending; if you're not on Platinum, a two-sentence text message works just as well.
Offer to record for the parent's review, not just yours. A parent who watches a 3-minute clip of their child sticking with a hard problem sees the intervention working in a way that a written recap can't convey. Lesson recording is included with Koala Pro (saved locally to your device), and Koala Platinum adds cloud recordings of classes; either way, share the specific 3-5-minute clip of the target progress rather than the full 45-minute recording. Consent and storage rules covered in how do I record online tutoring lessons? — get the family's written permission before recording anything with a minor.
Handle "he doesn't want to come today" without escalating. Every ADHD-heavy caseload has these calls. Options that work: shorten the lesson to 20 minutes and end on a win rather than skipping; move to a lower-friction activity (playground time attached to a light drill); occasionally, when the family judges it right, skip and reschedule without penalty. What doesn't work is a rigid "the lesson is 45 minutes or we lose the slot." If your cancellation policy doesn't have room for this, ADHD families will churn out.

Practical setup on Koala Go for an ADHD-aware tutor

If you're evaluating Koala Go (the platform we build) for a caseload with ADHD students, here's what you actually get, and where the current limits sit:

The whiteboard is where the student does. Fabric.js canvas with pen (Path), Sticky-note tokens, Stamp images, Text, and Image tools — every element drag-and-drop-able and live on both sides of the room. Upload the student's worksheets, math problems, or practice sheets as PDF or PowerPoint (Free-tier caps on uploads; unlimited on Pro), annotate on top of them, and each session's canvas is saved to reopen next week. This is where you keep the student's hands and eyes.
Per-student Gems for immediate salient reinforcement. Each student in the room has their own gem balance. You award 1-100 gems in a single click from the student panel, with a visible drop on the student's side. Design the pattern the way ADHD-aware guides prescribe — small for participation, larger for a hard attempt, largest for persistence through a difficult drill. Turn it off entirely for older students who find it infantilising.
Cobrowser to keep external resources inside the room. If your lesson uses a math site, a reading tool, a language app, an interactive science simulation — the cobrowser brings it into the classroom tab. The student controls the cursor and the clicks. On Free, 10 minutes at a time with a 20-minute cooldown; on Pro, unlimited.
Playground for the 1-2 minute movement reset. A 3D space in a separate tab of the same classroom. Not the teaching surface. Handy for the break between work intervals: the student's avatar walks somewhere, comes back, and you're into the next interval without them having left the room. On older student devices, drop the Playground Quality setting to Low (see the Quality settings guide) so the break doesn't stutter.
Recording — local on Pro, cloud on Platinum. Koala Pro includes lesson recording saved locally to your device; Koala Platinum adds cloud recordings of classes, stored with your account. The recording shows the shared canvas alongside video — the parent sees the student's actual work, not just the tutor's face. Get consent up front; share short clips rather than full sessions. Details in how do I record online tutoring lessons?.
Browser-only, no install for the family. The student clicks your lesson link and lands in the classroom. No download, no account setup for the child. Works on the family's existing laptop / iPad / Chromebook. For a family already managing more than average, one less tool to install is one less friction point to lose them over.
Auto-generated class notes (Platinum). If the specific parent recap is the part you skip on a busy week, Platinum drafts a summary after each lesson which you edit and send in a minute or two. Consistent recaps are one of the biggest retention levers for ADHD families — see how do I keep parents renewing month after month?.

Honest limits — what Koala Go doesn't do (and what else you'll want)

Frank about the gaps, because ADHD caseloads run long and small mismatches compound:

iPad experience is rougher than desktop. Koala Go runs in an iPad browser and many students use it that way, but the drag-and-drop feel and pen precision are best on a laptop with a mouse or a stylus. If the student is doing detailed work on a first-generation iPad without a stylus, expect it to be workable but not great.
Stability on unstable connections. Real-time collaborative environments are more sensitive to a bad Wi-Fi moment than plain video calls are. A wired connection on the tutor side (Ethernet beats Wi-Fi) reduces the hiccups; families on weaker home internet will still notice them.
No built-in attention analytics. Some tutors ask whether the platform tracks the student's attention (eye position, tab-focus, keystroke patterns). Koala Go does not. We're a lesson-delivery classroom, not a monitoring system, and adding surveillance-shaped features to a child's device is a decision we don't want to take lightly. If your program requires quantified attention data, you'll need a separate tool.
No included ADHD-specific curriculum. We do not ship a curriculum for teaching students with ADHD, and we're not going to — the pedagogy belongs with the tutor. Koala Go is the delivery surface for whatever program, approach, or subject you already work in.
Playground gaps. Some ADHD-workflow-adjacent features are still on the roadmap rather than shipped — customisable timers for interval work, a broader library of movement-break templates. (Group-select on whiteboard objects has shipped: drag a box around several objects to move or resize them together.) If any of these is a hard requirement for you, write to koala@teachwithkoala.com before you commit.

A short decision rule for when this workflow fits

Are the students you'd serve able to sit through a full 45 minutes of anything? Not "with concentration" — literally, in a chair, in front of a screen, for 45 minutes. If the answer is no even before you factor in the lesson itself, shorten the session to 25-30 minutes rather than trying to force a longer format that will fail on retention. Short sessions that end well beat long sessions that end badly.
Can your current or intended platform reliably do drag-and-drop tokens, live shared control of a webpage, and instantly-triggerable visible reinforcement? If yes, the workflow translates. If no — if the tool is a video call with a static whiteboard on top — the "student is doing" piece breaks and you're left with 45 minutes of passive watching.
Is the family on board with the environment prep? Phone-away, door-closed, notifications-off. If the family can't or won't hold that boundary at home, you're not going to fix it from your side of the screen. Make the ask, honour the "no", and reset your expectations for that specific student.
Are you willing to hold the parent-recap cadence that makes the retention math work? Two-sentence specific recap after every session. Short recorded clip when there's something worth showing. If you're not, ADHD families will drift out — not because the tutoring stopped working, but because the parent stopped seeing that it was working.

If you'd like to try Koala Go specifically for a lesson with an ADHD-differing student, the whiteboard, Gems, Playground, and cobrowser mechanics are all on the Free tier — you can run a real 45-minute session (or a 25-minute one) with your candidate student on their real device before deciding. Open Koala Go at classroom.teachwithkoala.com. If you want to talk through your specific caseload before you commit — the age range, the subject, the devices the families are on — write to koala@teachwithkoala.com with a sentence about the kind of tutoring you do and we'll give you our honest read.

What "online tutoring for dyscalculia" is (and what it isn't)

The reader we're writing for is a tutor already working with — or about to start working with — a student who has dyscalculia, formally identified or clearly showing the pattern (persistent difficulty with magnitude, number sense, arithmetic fluency, or fact retrieval that doesn't respond to more of the same instruction). If you're a specialist maths tutor trained in a specific approach (Ronit Bird, Numicon, Mathematics Recovery, Making Math Real, Semple, TouchMath, Sharma / Making Sense of Number and Space), a general maths tutor whose caseload includes math-differing learners, a primary or middle-school tutor picking up an intervention student, or an ESL / general tutor whose student turned out to also have a maths learning difference, this page is for you. We're not teaching you the method — you already know it, or you're being pointed to one by the family. We're covering the parts of delivery that change when the lesson moves from a table with physical manipulatives to a browser tab.

What this page is not:

Diagnostic content. We don't help a parent decide whether their child has dyscalculia; we don't list criteria; we don't recommend screeners or assessments. If a parent is asking whether their child has dyscalculia, the useful referral is to a qualified assessor — an educational psychologist, a specialist assessor, or the family's paediatrician — not to us and not to the tutor's own judgement.
A math-anxiety treatment. Math anxiety is real, it often co-occurs with dyscalculia, and a well-run lesson helps — but the tutor's job is not to treat the anxiety, only to run a session that doesn't inflame it. Persistent, disabling anxiety belongs with a clinician, not a tutor.
Special-education-law advice. IEPs, 504 plans, EHCPs, SEND provisions, and the equivalents in other countries are formal legal instruments and vary by jurisdiction. If the family has one, ask to read it and let it inform your lesson; don't advise on whether they should get one.
A general-purpose "how to teach math online" page. That's the math-cluster pillar. If you're looking for the general workflow — the manipulatives problem, the notation problem, the practice problem — start at how do I teach math online? and come back here for the dyscalculia-specific overlay. If you want depth on fractions or number sense specifically, the spokes at how do I teach fractions online? and how do I teach number sense online with virtual manipulatives? cover the underlying manipulatives workflow.

What actually changes when the lesson moves online

Tutors we've spoken to who've moved a dyscalculia-heavy caseload online report the same three constraints in the first month:

The concrete step has to survive the screen. Every specialist dyscalculia program is built on concrete-pictorial-abstract (CPA / CRA) sequencing — the student manipulates the quantity before they see the picture of it and long before they write the symbol for it. In-person that's Cuisenaire rods on a table, Numicon shapes, fraction tiles, ten-frames, Dienes blocks. Online, that either becomes something the student can drag and rearrange on a shared canvas — first-class manipulatives, not screenshots of manipulatives — or it collapses into a video the student watches, which is not the same thing and does not do the same work. This is the load-bearing constraint; if the platform can't do it, the rest of the workflow doesn't matter.
Retrieval speed and working memory pressure compound on a screen. A dyscalculic student's arithmetic fact retrieval is often slower than a peer's; a lesson that pushes fluency without warming up the underlying magnitude and number sense first tends to fail badly. Online, the same student now also has to hold in mind where the mouse is, which tab is which, and what the shared canvas is doing — and the working-memory tax of the interface itself starts to eat the working memory the maths needed. The platform has to disappear into the background as much as possible, and the lesson design has to protect working memory more aggressively than it would in person.
Parents can't see the manipulatives moment. The most reassuring thing for a dyscalculia family is watching their child cross a threshold — really understanding, for the first time, that 1/2 and 2/4 name the same amount, or that 8 is closer to 10 than to 0 on the number line. In-person, a parent walking past the table catches those moments. Online, they don't — unless the classroom shows them the shared canvas alongside the video, or unless a short clip of that specific two-minute breakthrough gets sent home. If the platform can record the shared canvas (not just the tutor's face), the retention story gets much easier.

What the shared canvas has to do (concrete-first work in the browser)

Preserving the concrete step is the single hardest thing to get right online, and it's where the platform choice does most of its work for a dyscalculia tutor. What the whiteboard has to support:

Manipulatives the student can drag, not just look at. On Koala Go's whiteboard, a Math Tools panel inside the Whiteboard Library holds a set of first-class manipulatives that are draggable on the shared canvas — both you and the student can grab, move, and rearrange them, and both sides see the same state live. The set relevant for dyscalculia work:
- Cuisenaire-style number rods (Pro tier): ten rods, values one unit through ten units, each a different colour. This is the workhorse for magnitude, additive inverse, complements-to-ten, place-value introduction, and early multiplication as repeated addition — the exact concept clusters most dyscalculia programs spend their first many weeks on. On Platinum, additional features on the same rods help specifically here: a red-and-blue alternating fill so a rod visually breaks into its unit count (which supports subitising work in students who don't yet reliably see 6 as 5-and-1), a vertical placement mode (useful when a rod is doing double duty as a bar chart or a place-value column), and a one-click "add all" staircase of all ten rods (the fastest way to reset the canvas at the start of a new number-sense activity).
- Number lines (Pro tier): preset lines for 0-10, 0-20, 0-30, 0-100 (by 5s or by 10s), -10 to 0, -10 to 10, and a blank one you can label yourself. Number-line work is central to dyscalculia intervention specifically because magnitude comparison — knowing that 43 is bigger than 34 without counting — is one of the hardest core skills to build, and the mental number line is what does that work in a typically-developing brain. On Platinum, a "make your own" builder lets you specify start, end, and step (say, 0 to 1 in tenths for decimal work, or -20 to 20 for integer work) so you're not limited to the presets when a student needs something specific.
- Fraction bars and fraction circles (Platinum tier): individual bar segments and pie-slices for denominators from a half through a twelfth (plus a sixteenth) — you place one 1/n piece at a time and build a whole together. This is the manipulative that closes the largest gap for many dyscalculia students, because "1/2 + 1/4 = 3/4" as a symbolic manipulation is genuinely close to impossible for a learner who does not yet see 1/2 and 1/4 as measurable quantities. Placing a half bar next to a quarter bar next to a quarter bar and watching the three-quarters shape appear does the pedagogical work that no amount of symbol drilling can. The pie-slice equivalents let you do the same in a circle rather than a bar, which some students grasp more easily. Denominators go 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, then jump to 16 — those are the denominators that actually show up in elementary and middle-school fraction work.
- 2D and 3D shapes (core set on Pro, full library on Platinum): a shape library you place and resize on the canvas. Basic square, rectangle, circle, equilateral triangle, and the cube are Pro — enough for area/perimeter, basic symmetry, and simple 3D introduction. The fuller library (rhombus, parallelogram, trapezium, kite, pentagon, hexagon, octagon, non-equilateral triangles, oval / semicircle / sector, and most 3D solids — cylinder, sphere, cone, prism, pyramid, cuboid) is Platinum. On Platinum, a shape-labels toggle adds label slots at a shape's corners and centre that you type the measurements into, so a working-memory-taxed student can read "4" and "6" straight off the shape instead of holding them in their head.
- Number Roller (Platinum tier): a randomised digit roller in a range you set (1-6, 1-10, 1-12, 1-20, 1-50, 1-100, 1-1000; 1-4 dice at a time; whole or decimal). Useful for the "generate a fresh problem" moment during fluency work without you having to invent one on the spot — a small thing, but the surface area of "the tutor invents a new problem" is one of those places attention drops for a dyscalculic student, and outsourcing the generation buys back a few seconds of engagement each cycle.
Sticky-note and Stamp tokens for anything the maths tools don't cover. Ten-frame tokens, dot cards, tally marks, tens/ones place-value counters, chip-model integer counters (positive and negative), coin images for money work — anything you'd reach for as a physical counter on a table can be dropped on the canvas as a Sticky or Stamp and dragged by either side. These are the same primitives the dyslexia workflow (Elkonin-box phoneme tokens) and the ADHD workflow rely on; they cover any manipulative that isn't in the built-in Math Tools set yet.
Slide / PDF / PowerPoint upload for your existing materials. Specialist maths-intervention tutors typically have a large personal library — a program's workbook, worked-example sheets, dot-pattern cards, place-value charts, custom activity sheets accumulated over years. On Koala Pro, upload them as PDF or PowerPoint and they become the lesson surface — both of you annotate on top, using the same pen and shape tools as on a blank canvas. The library survives the move online instead of asking you to rebuild it.
A student view that shows what you show. Simple, but not universal — the student sees the same canvas orientation you see, no mirrored or laterally-flipped view. A student who reverses digits or misreads left-to-right sequences doesn't need any additional orientation friction.

For depth on how the manipulatives themselves work — the rods, the lines, the fraction pieces — the two math-cluster spokes cover it: how do I teach number sense online with virtual manipulatives? for rods and number lines, and how do I teach fractions online? for the fraction workflow. We don't repeat those here; we describe how the same tools land in a dyscalculia lesson.

Solving the notation problem without adding a second cognitive load

Real maths needs symbols keyboards don't have — fractions that read as fractions, minus signs that don't look like hyphens, the square-root sign, the degree symbol, an unambiguous "not equal." For most students, notation is a minor friction. For a student with dyscalculia, whose working memory is already carrying the concept, the notation is often what tips the session out of the "I can hold this" zone into the "I've lost it" zone. Every extra "how do I type that?" is a second cognitive load layered on top of the first one.

Koala Go's whiteboard ships a Math Symbols palette on the paid tiers, sorted into the categories a maths tutor actually reaches for. On Koala Pro you get the operations, comparisons, and basic geometry set that covers elementary and middle-school work: the four operations and equals (+ − × ÷ =), the basic comparisons (< >), the fraction / ratio / percent group (/ : %), a square-root sign (√), a degree symbol (°), and round brackets. On Koala Platinum you additionally get the algebra, geometry, and powers set for later middle-school and secondary work: not-equal and approximately-equal (≠ ≈), less-than-or-equal and greater-than-or-equal (≤ ≥), superscript powers (² ³ ⁿ), pi and infinity (π ∞), plus-or-minus (±), the geometry markers (∠ △ ⊥ ∥), and square and curly brackets. Clicking a symbol drops it into whichever text box or sticky note you'd otherwise type into, so you can build up "1/2 + 1/4 = ?" or "8 > 4" or "3 × 4 = 12" without either you or the student stopping to look up an alt-code.

Honest limit: the palette is a symbols set, not a full equation editor. It won't stack a horizontal fraction bar the way LaTeX would; it doesn't nest radicals; it doesn't render live graphs. For dyscalculia work through primary and middle school this rarely matters — the manipulatives carry the concept and the symbols palette handles the notation — but if a specific dyscalculic student is working on advanced algebra or graph-heavy statistics and you need a full equation editor, the honest answer is that a dedicated STEM whiteboard tool goes deeper on notation. Where Koala Go's whiteboard sits in the broader landscape is covered in the best online whiteboards for tutoring comparison; we don't repeat it here.

Structuring the online session for a dyscalculic learner

The template most specialist maths tutors converge on for an online session follows the same overall CPA shape as the in-person version — start concrete, move to pictorial, move to abstract, only work on symbolic manipulation once the underlying quantity is stable — but the segments run slightly shorter online and the transitions matter more. A working template for a 45-minute lesson (adjust downwards to 30 minutes for younger students; some 6-8-year-olds cap out at 25):

Warm-up: magnitude / subitising / number sense (5 min). Whatever the day's concept, start on the number line or the rods with a magnitude task — "put the 7 rod next to the 4 rod, which is bigger, by how much?" — not on symbolic arithmetic. This is a temperature check on the student's mental number line, which for a dyscalculic learner is the substrate everything else sits on. If the substrate is not warm, the harder work later in the lesson will fail.
Concrete introduction of today's concept (10 min). New content is introduced on the manipulative first, always. New idea: fractions of a whole → place unit-fraction bars, count them into a whole, then two wholes. New idea: subtraction with regrouping → rods on the tens/ones canvas, physically break a ten into ten ones, then subtract. Do not introduce today's concept in symbolic form; that comes later in the lesson.
Movement or attention reset (1-2 min). Stand, stretch, walk to a wall and back. Some tutors use the Playground here — a short "walk to the sign, come back" beat with the student's avatar — because it's low-stakes movement inside the same tab and doesn't risk the student wandering off the lesson. Do not skip the reset; a dyscalculic learner's cognitive tank is genuinely finite and the second half of the lesson depends on refilling it.
Pictorial bridge (8 min). Now move the same concept from the manipulative to the picture — a diagram of the same fraction pieces on the canvas without the interactive pieces, a written number line the student marks up, a bar model of the same subtraction problem. The pictorial step is the one that usually gets skipped in generic teaching and the one dyscalculic students need most; do not race past it. If the pictorial step wobbles, go back to concrete for two minutes rather than pushing forward.
Symbolic bridge, brief (6 min). Only now, after the concept is concrete and pictorial, do you introduce the symbols for it. Write "1/2 + 1/4 = 3/4" next to the fraction-bar pieces on the canvas so the two representations sit side-by-side. Do a few — three or four — in this form. This is the moment the notation palette matters most; you're not asking the student to type it, you're doing the writing for them, but any moment either of you spends fighting the interface is a moment stolen from the concept.
Guided practice, low volume (8 min). Not "twenty problems" — three to five, worked slowly, with you present at each one. Volume drill in this segment burns exactly the working memory you built the concept on. If the student is ready for more practice, that's the between-lesson home task, not the last block of the session.
Wrap and reward (5 min). Recap the day's target using the concrete pieces one more time — "we saw that a half and two quarters name the same amount, and we wrote it as 1/2 = 2/4." Award effort-based reinforcement (see below). Preview one small task for the parent to do this week. End on a warm, specific "you worked hard on the bit that was hard today" — that specificity is what the student takes into the interval before your next session.

Two things to watch for. First, the pictorial step is where many students who "seem to be doing fine on the manipulatives" quietly fall off; give it time and don't take a shrug for understanding. Second, avoid speed at all costs — timed drills, races, "beat the clock" mechanics are almost universally counter-productive for a dyscalculic learner, because the exact problem is that their retrieval is slower, and time pressure activates the anxiety that made maths hard in the first place. If a program you're delivering has a speed component, run it in a low-stakes way (self-timing, "let's see if today feels easier than last week") rather than a competitive one.

Effort-focused reinforcement (why a rewards mechanic matters here specifically)

Maths intervention is genuinely hard cognitive work for a dyscalculic student, and it's hard in a way the student can see themselves being bad at — unlike reading intervention, where progress is often invisible to the learner even as it happens, maths practice tends to produce specific wrong answers the student notices. If your reinforcement pattern rewards only "correct answers," you're rewarding the student's ease — which is the least available signal for a math-differing learner — and inadvertently reinforcing exactly the "I'm bad at maths" self-narrative you're trying to unwind. Every experienced maths-intervention teacher's manual describes some version of this: reinforce effort, reinforce approximation, reinforce persistence through a wrong answer to a right one, not just accuracy.

Translated online, the mechanic that supports this pattern is a small, frequent, teacher-controlled reward currency you can trigger without leaving the lesson flow. On Koala Go, that's Gems: you open the student panel from inside the room, pick the student, and give 1-100 gems in a single click (the amount input is capped between 1 and 100 on purpose — a rhythm of small frequent rewards beats a single big burst). Each gem drops with a sound and a visual, and the student's balance updates live on their side. Why this shape supports dyscalculia-aware reinforcement:

You control the exact timing. Not the platform, not a schedule, not the completion of a task. You. So you can reinforce the specific moment the effort happened — "you said you didn't know, and then you tried anyway, and you got there" — while the effort still exists in the student's working memory. Delayed rewards ("you'll get five stars at the end of the lesson") do almost none of the work for the actual learning moment.
Size matches effort, not outcome. A quick correct answer gets 1 gem; a hard attempt on a novel problem gets 5; a full session of "you kept coming back to the maths that was hard" gets a bigger one at the end. Variable-magnitude reinforcement holds attention longer than a fixed one-per-correct schedule and doesn't teach the student to game the system.
Wrong answers are rewardable too. "You said 6 + 7 was 12; that's really close, and it's a good try — the answer is 13" is a rewardable moment. A schedule that only awards points on "correct" tells the student that the interesting distinction is between right and wrong, when for a dyscalculic learner the useful distinction is between engaged and disengaged. Award the engagement.
It's optional. If a specific student — usually an older teenager — reads the gems as infantilising, drop them. For a 14-year-old dyscalculic student working on GCSE / eighth-grade material, your visible enthusiasm ("that's a really good attempt, let's keep going") often carries more reinforcement weight than any on-screen currency. The mechanic is a tool, not the whole strategy.

None of this replaces the tutor's live warmth — the direct, timed, specific "that was a good try" that no platform can deliver for you. It's a lightweight second channel that carries some of the reinforcement load between your verbal moments. The gamified-virtual-classroom explainer at what is a gamified virtual classroom? covers the general engagement-mechanic story if you'd like the broader context.

Using external practice tools via the cobrowser

Most specialist maths tutors have a set of web tools they already use — an interactive graphing tool like Desmos for magnitude and pattern work, a practice-problem site like Khan Academy or IXL for between-lesson homework, a virtual manipulatives library the school district provides, a program-specific web app for the approach you're trained in. Traditionally on a video call you'd screen-share and drive the mouse yourself; the student watches you click, which for a dyscalculic learner puts them right back in the "watching, not doing" posture that fails.

The cobrowser (explained in depth at what is a cobrowser?) changes the posture — you and the student are on the same live webpage inside the classroom tab, and either of you can move the cursor, click, and type. For dyscalculia work this unlocks:

Interactive graphing (Desmos, GeoGebra, similar). The student drags a slider and watches the graph respond, in real time, on the same page you see. Watching magnitude and rate change on a moving graph does pedagogical work that no static picture can — and the student is driving.
Practice-problem sites the student won't otherwise focus on. Khan Academy, IXL, Prodigy, ALEKS, a program-specific practice portal, or the family's school-provided practice site — open it inside the lesson, sit with the student while they work three or four problems, and step in the moment they stall. This is the difference between "I sent them practice for the week" (largely uncompleted) and "we did three of them together and I saw the point where he got stuck." Teach from sites the family has legitimate access to and within each site's terms.
Virtual-manipulative libraries that aren't in Koala's built-in set. Some tutors use Didax, Toy Theater, or the U.S. National Library of Virtual Manipulatives-descendant sites for a specific tool that isn't in our Math Tools panel yet (base-10 blocks in specific configurations, ten-frames, algebra tiles). Open the site through the cobrowser and both of you can drive; the tool is now inside the lesson rather than a "hold on, open this link" tangent.
Program-specific web apps. If your maths-intervention program ships a companion app or web tool for the digital pieces, open it through the cobrowser and the student does the activity in real time inside the classroom.

One honest caveat on Koala Free: cobrowser sessions on the free tier are time-limited (you get 10 minutes at a time, with a 20-minute cooldown between sessions), which is fine for a single web-activity segment in a 45-minute lesson but not for a lesson where the cobrowser is the main teaching surface. Full-length cobrowser sessions come with Koala Pro.

Recording the lesson for the parent (and for your own records)

Recording is optional but for maths-intervention work it often pays for itself in retention. A parent who watches a three-minute clip of their child working through the moment 1/2 + 1/4 clicks into place sees the intervention working in a way that a written recap can't convey — the parent sees the hesitation, the tutor's cue, the manipulative move, and the student's second attempt landing on the right amount. That's the difference between "I hope this is helping" and "I can see this is helping." Practically:

Get consent up front. Recording a minor requires parent permission and, in some regions, additional consent — see how do I record online tutoring lessons? for the consent, storage, and sharing workflow, including a written-policy template you can adapt.
Share short clips, not full recordings. A full 45-minute recording is more content than any parent will watch. Clip the two or three minutes that show the actual breakthrough — the fraction placement, the number-line comparison, the point where the student said "oh" — and share those. Some tutors send one clip a week alongside the written recap.
Store for the intervention arc, not forever. A maths intervention runs on the order of months to a couple of years; recordings that long deep-file quickly. Set a retention policy (delete after 90 days, or after the intervention completes) so you're not carrying an unbounded personal archive.

On Koala Go, lesson recording is included on Koala Pro (saved locally to your machine), and cloud recordings of scheduled classes come with Koala Platinum. Either way the recording captures the shared canvas alongside the video, so the parent sees the manipulative move, not just the tutor's face.

1-on-1 vs small group for maths intervention

Most specialist dyscalculia tutors keep serious intervention 1-on-1. The pacing is too individual — a student who needs three sessions to build the 1/2 = 2/4 equivalence should get three sessions, and a group of two students who might need two and four sessions respectively breaks the model. Small groups (2-3 well-matched students, or a pair of siblings at genuinely the same stage) sometimes work for the concept-introduction side of a lesson — you can introduce a manipulative to two students at once and it can even help — but the differentiated practice is much harder to run in a group and usually pulls back to 1-on-1 within a few sessions. If you're considering the group model for the economics, see how do I teach small group classes online? for the general small-group tradeoffs — and note the "students with specific learning differences that require intensive individual attention" caveat in that page. Families paying specialist rates for maths intervention generally expect 1-on-1 and are right to.

Working with the parent (co-teacher, not co-fighter)

Maths-intervention retention lives or dies on the parent seeing progress. A dyscalculic student's home-side maths practice is often the flashpoint of the week — a parent who doesn't see progress and who is fielding a resistant child around a homework page will not keep paying for lessons for long, no matter how well the sessions themselves are going. A few moves that reduce parent-side friction:

Send a specific — not generic — recap after each session. "We worked on unit fractions today. She got the 1/2 and 1/4 pieces solidly by the end. She struggled with reading the fraction notation, so this week I'd like her to name the shapes verbally — no writing — using the fraction bars we made a photo of. Five minutes, three or four times." That's a recap the parent can act on. "We had a great lesson!" is not. On Koala Go, auto-generated class notes (on Platinum) draft this for you after each lesson and you edit them in a minute or two before sending. See how do I share progress updates with parents? for the cadence.
Prescribe home practice that is less than the parent expects. Parents often assume "more practice = faster progress" and push a dyscalculic student into volume that inflames the anxiety without moving the maths. Explicitly ask for less: five minutes, three times a week, on one specific concrete task. This is a specialist recommendation the parent likely won't get anywhere else, and it's a good reason to trust you.
Offer to record for the parent's review, not just yours. A short clip of a genuine breakthrough (the student naming an equivalent fraction correctly for the first time; the student comparing 32 and 23 correctly on the number line) does more retention work than any number of recap emails. Consent and storage rules covered in how do I record online tutoring lessons?.
Handle "he doesn't want to come today" without escalating. Every intervention caseload has these calls. Options that work: shorten the session to 20 minutes and end on a win rather than skipping; move to a lower-friction activity (the number rods and a familiar game, not new content); occasionally, when the family judges it right, skip and reschedule without penalty. If your cancellation policy doesn't have room for this, dyscalculia families will churn out.
Where the family also has a formal plan (IEP / 504 / EHCP / SEND), read it and let it inform the lesson. Don't advise on whether they should have one, and don't try to substitute for a formal maths assessment — but if the family has an existing plan that names goals, meet the goals inside your sessions and reference them in the recaps. Alignment with the plan is one of the strongest retention moves you can make.

Practical setup on Koala Go for a dyscalculia tutor

If you're evaluating Koala Go (the platform we build) for a caseload with dyscalculic students, here's what you actually get, and where the current limits sit:

The Math Tools panel on the whiteboard. Cuisenaire-style number rods and preset number lines (Pro tier); fraction bars, fraction circles, and the Number Roller (Platinum tier); a 2D/3D shape library (core set on Pro, full library on Platinum); a Math Symbols palette (operations, comparisons, and geometry basics on Pro; algebra, powers, and full geometry on Platinum). These are first-class manipulatives, not screenshots — the student drags them, both sides see the same canvas live.
Sticky-note and Stamp tokens for anything the built-in tools don't cover. Drop tokens for ten-frames, dot cards, integer counters (positive and negative chips), place-value counters, coins for money work — anything you'd use as a physical counter. Both of you can move them.
PDF and PowerPoint upload (Pro tier). Bring your existing intervention library — program workbooks, dot-pattern cards, place-value charts — onto the shared canvas and annotate on top. Each session's canvas is saved to reopen next week; the intervention arc stays continuous.
Cobrowser to keep external practice inside the room (10 min on Free with a 20-min cooldown; unlimited on Pro). Desmos, GeoGebra, Khan Academy, IXL, a district's own virtual manipulatives site, your program's companion web app — the student has real control of the page, not a passive view of your screen.
Per-student Gems for effort-based reinforcement. Each student in the room has their own gem balance. You award 1-100 gems in a single click from the student panel, with a visible drop on the student's side. Design the pattern the way your program prescribes; turn it off entirely for older students who find it infantilising.
Playground for the 1-2 minute movement break. A 3D space in a separate tab of the same classroom — teacher and student avatars in a shared world. Not the teaching surface. On older student devices, drop the Playground Quality setting to Low (see the Quality settings guide) so the break doesn't stutter and eat working memory.
Recording — local on Pro, cloud on Platinum. Koala Pro includes lesson recording saved locally to your device; Koala Platinum adds cloud recordings of scheduled classes stored with your account. Either way the recording shows the shared canvas alongside video — the parent sees the manipulative move, not just the tutor's face. Get consent up front; share short clips rather than full sessions.
Auto-generated class notes (Platinum). Drafts a specific recap after each lesson which you edit and send in a minute or two. Consistent, specific recaps are one of the biggest retention levers for dyscalculia families — see how do I keep parents renewing month after month?.
Browser-only, no install for the family. The student clicks your lesson link and lands in the classroom. No download, no account setup for the child. Works on the family's existing laptop / iPad / Chromebook — matters because dyscalculia families are often already managing more than the average family and one less tool to install is one less friction point.

Honest limits — what Koala Go doesn't do (and what else you'll want in your stack)

Frank about the gaps, because maths intervention runs long and small mismatches compound:

iPad experience is rougher than desktop. Koala Go runs in an iPad browser and many students use it that way, but the drag-and-drop feel and pen precision are best on a laptop with a mouse or a stylus. If the student is working with the fraction pieces on a first-generation iPad without a stylus, expect it to be workable but not great; a graphics tablet on the tutor side and a laptop with a mouse on the student side is the more forgiving setup.
Stability on unstable connections. Real-time collaborative environments are more sensitive to a bad Wi-Fi moment than plain video calls are. A wired connection on the tutor side (Ethernet beats Wi-Fi for reliability) reduces hiccups; families on weaker home internet will still occasionally notice them.
No built-in dyscalculia curriculum. We do not ship a Ronit Bird / Numicon / Mathematics Recovery / Making Math Real curriculum, and we're not going to — the pedagogy belongs with the tutor and the program you're trained in. Koala Go is the delivery surface for whatever approach you already use.
No built-in progress-tracking or intervention-log tool. Koala Go is a lesson-delivery classroom, not an intervention-tracking system. If your program requires you to log detailed per-session accuracy or generate structured progress reports, you'll continue to use whatever tracking system you use today alongside Koala. We ship class notes, not clinical intervention records.
The Math Tools set is still expanding. The manipulatives currently in the panel — rods, number lines, fraction bars and circles, shapes, number roller, symbols — cover most primary and middle-school maths, but requested tools that specifically help dyscalculia workflow (Numicon-style shape pieces with hole patterns, base-10 blocks in customisable configurations, algebra tiles, dedicated ten-frames as a first-class object, custom labelled fraction sets) are not part of the Math Tools panel today. If any of these is a hard requirement for the program you deliver, write to koala@teachwithkoala.com before you commit — we'd like to know what you need.
Not a full equation editor for advanced maths. The symbols palette covers what a primary or middle-school dyscalculia lesson needs; for advanced algebra, calculus, or graph-heavy statistics with a dyscalculic student who's reached that level, pair Koala Go with a dedicated STEM whiteboard or graphing tool through the cobrowser rather than expecting the palette to do it alone.

A short decision rule for whether this workflow fits

Are you working within a specific dyscalculia-informed approach (or one clearly compatible with concrete-pictorial-abstract sequencing)? If yes, Koala Go's manipulative set maps cleanly. If your program is speed-and-fluency drilling with no concrete step, this platform is not the constraint — the pedagogy is; and neither of those tends to help a dyscalculic student regardless of the tool.
Can the family you'd serve be reached only online? Distance to a specialist maths-intervention tutor is often enormous — many families never find one at all. If moving online is the intervention, not an alternative to it, set up the tooling and ship. The families gain access they couldn't otherwise have.
Is your target student able to sit through a session that ends before it should? A short session that ends on a win beats a long session that ends badly. If the student caps out at 25 minutes, run 25-minute sessions; don't stretch to 45 for the sake of the schedule.
Are you willing to hold the parent-comms cadence maths intervention needs? Specific recap after every session, less-not-more home practice, occasional clipped recording of a breakthrough. If you're not, dyscalculia families will drift out — not because the tutoring stopped working, but because the parent stopped seeing that it was working.

If you'd like to try Koala Go specifically for a dyscalculia lesson, the whiteboard, Cuisenaire-style number rods, preset number lines, Sticky/Stamp tokens, Gems, Playground, and cobrowser are all on the Free tier (with the free-tier cobrowser cap); Fraction Bars, Fraction Circles, the Number Roller, and the full Math Symbols set live on Platinum. You can run a real 30- or 45-minute session with your candidate student on their real device before deciding. Open Koala Go at classroom.teachwithkoala.com. If you want to talk through your specific approach — the program you're trained in, the age band of your caseload, the devices the families are on — before you commit, write to koala@teachwithkoala.com with a sentence about the maths intervention you do and we'll give you our honest read on the fit, including the parts where you'll want something else alongside us.

What "online exam prep" actually is (and what it isn't)

The confusion new exam-prep tutors run into is treating an exam-prep student like a general-tutoring student who happens to have a test coming up. The two jobs share about half their moves, but the differences are what determines the outcome:

General tutoring is open-ended and student-paced. You improve a student's reading, math, or English over months and years; the goal is fluency and confidence; there's no external rubric grading the outcome. Progress is measured by what the student can now do that they couldn't before.
Exam prep is deadline-bound and rubric-bound. There's a fixed date, a fixed scoring system, and a fixed pool of question types the exam board has already published. Progress is measured against the rubric — a band, a score, a pass mark — not against "does the student feel better." The teaching often has to move faster than a general student can comfortably absorb, because the deadline is the deadline and the exam does not accommodate the student's pace.
Exam prep is not cramming. Cramming is trying to memorise content in the two weeks before the exam. Exam prep is teaching the student to take the exam — recognising question shapes, timing sections, learning the rubric's own vocabulary, building the specific test-day habits that recover 5-15% of a score from the same underlying knowledge base. The last two weeks of prep look different from the first ten (see the "last two weeks" section below), but the last two weeks alone are not exam prep.
Exam prep is not test-taking tricks. "Guess C if in doubt" and "eliminate the extreme answer" are real, small, marginal moves — worth teaching, worth practising, but they don't produce a score gain on their own. The score gain comes from the same place it always does: knowing the material, recognising the question shape, and executing under time pressure. Beware sales-of-tricks framing when you shop for exam-prep materials.

If you're being asked to teach a student for an exam that's happening in six weeks and the student's current diagnostic score is a full band or 100+ points below their target, be honest about what six weeks can and can't do. Setting an achievable target (a half-band, a 50-point delta) and hitting it produces a happy family; promising a full-band jump and missing it produces a lost family and a reputation dent. The single most common exam-prep-tutor mistake is over-promising in the trial lesson, and it isn't recoverable in later lessons.

Which exam? A short taxonomy of what online tutors most commonly teach

The workflow below applies across all high-stakes tests, but the details vary. From what we see from tutors on Koala Go, the exams that dominate the independent online-tutor exam-prep market are:

English-language proficiency tests — IELTS (Academic and General; British Council / IDP / Cambridge Assessment English), TOEFL iBT (ETS), Cambridge B2 First / C1 Advanced / C2 Proficiency (Cambridge Assessment English), Cambridge YLE (Cambridge Young Learners — for under-12s), PTE Academic (Pearson), OET (medical English), TOEIC (business English; ETS). Split into the four skills — Reading, Listening, Writing, Speaking. This is the biggest exam-prep bucket for indie tutors, especially those in the ESL / K-12 world.
US college admissions — SAT (College Board — Digital SAT now), ACT (ACT Inc.). Multi-section timed tests; the "college admissions" positioning is worth a premium and the student pool is highly outcome-focused.
US high-school / advanced-placement — AP exams (College Board) across dozens of subjects. Discipline-specific rather than general skill.
UK / international school programmes — GCSE, A-Level, IB Diploma. Multi-subject, curriculum-tied. School-year cadence, not a discrete "test day".
Graduate admissions — GRE, GMAT (ETS; verbal + quant + writing sections). Adult students, higher hourly rates, longer prep runs (3-6 months).
World-language proficiency — HSK (Chinese, Hanban), JLPT (Japanese, JEES), DELF / DALF (French, France Éducation International), DELE (Spanish, Instituto Cervantes), TOPIK (Korean). Level-graded on a fixed schedule (usually 2-4 sittings per year).
Professional / other — Bar exam prep, Praxis (US teaching), USMLE (US medical), Cambridge Music Theory, ABRSM (music grades). Narrower audience, often higher rate.

The rest of this page uses IELTS/TOEFL/Cambridge (four-skills English exams) and SAT (multi-section academic exam) as the concrete examples, because they cover most of the moves you'll make elsewhere. Adapt the shape to your specific test.

The diagnostic-first opening — what lessons 1 and 2 must include

The single biggest edge you can give an exam-prep student is starting the course with a real, timed, honest diagnostic on official material — before you teach anything. Not a "how do you feel about the writing section?" chat. A proper timed section pass on a past paper, taken under exam conditions, with the student's answers preserved and scored against the actual rubric. Two lessons is usually enough for a compact diagnostic:

Lesson 1 (60-90 min): the reading + listening pass. Open an official past paper on the shared whiteboard (Koala Go supports PDF upload on the Pro tier, so a Cambridge past-paper PDF or a College Board practice SAT sits on the same shared canvas the student annotates). Set a timer, tell the student it's timed, and stay silent while they work. Watch what they do — the questions they skip, the ones they change their mind on, the ones they clearly guess. When time is up, score against the rubric and note both what they got wrong and how they got there.
Lesson 2 (60-90 min): the writing + speaking pass. Give the student a writing task from an official past paper on a timer; ask them to type or hand-write it and share back. Then run a speaking section from the same paper — for IELTS, a full three-part interview; for TOEFL, the timed independent + integrated tasks; for Cambridge, the Part 1/2/3/4 flow. Record the speaking section (with the student's and parent's prior consent — see how do I record online tutoring lessons? for consent, storage, and retention). Recording is not a nice-to-have here; being able to play back their own speaking performance is what the student needs to see to buy into the course. Lesson recording is on the Koala Pro tier.

End lesson 2 with a written diagnostic report: the current section-by-section score, the target score, the gap in each section, and — most importantly — a proposed course length and cadence. Share it with the parent the same day. The diagnostic is where you convert a "we're thinking about tutoring" family into a course-committed one; a vague "we did some practice today" summary loses them, a specific "Anna scored Reading 6.5, Listening 7.0, Writing 5.5, Speaking 6.0 today; target 7.0 across; twelve weeks of two-lesson-a-week work will put her within the target band with a probable 0.5 Writing gap remaining" summary keeps them.

Structure the course backwards from the exam date

Every exam-prep course has one number that dominates every planning decision: weeks until the exam. Pick a course shape by that number, not by "how much do you want to spend on tutoring":

12-16 weeks — proper prep (the standard case). Two 60-minute lessons per week is the working cadence for most students at most exams. Week 1 = diagnostic. Weeks 2-4 = one skill at a time (start with the weakest section per the diagnostic). Weeks 5-8 = full section rotation + first full-length timed practice test. Weeks 9-12 = section-by-section refinement + weekly full practice test. Weeks 13-14 = timing-under-pressure and test-day habits. Last week = light review, sleep, mental preparation. This is the shape the how-to-structure-a-package page is describing when it mentions "term / semester packages (8-16 weeks, aligned to school terms). Common for exam prep." See how to structure lesson packages for the pricing side of a term-shaped course.
6-8 weeks — compressed. Bumps to three lessons a week and drops proportionally more of the general-teaching content in favour of exam-format-specific work. Week 1 = diagnostic + weakest-section work. Weeks 2-4 = each week rotates through two of the four sections with a Saturday full-length practice test. Weeks 5-6 = weekly practice tests and section-by-section repair on whatever came out worst. Last week = timing and test-day. This is intense — parents pay a premium for it and it's not sustainable across many students in parallel.
Under 4 weeks — triage. Don't try to close a full-band or 200-point gap here. Instead, sit down with the student in lesson 1 and identify the two questions: which sections can we still move, and which test-taking habits can we build in three weeks that will recover ground for free. Focus on the second question — timing, question-shape recognition, and "answer everything, don't leave blanks" strategies claw back a real 3-8% for many students without teaching any new content. Be honest with the family about what a 3-week course can and cannot do; a small realistic delta they hit is worth more than a big promised delta they miss.
Long slow prep (6+ months, once or twice a week). Common for parents who start a year out. Front-load the first 3 months on general skill-building (this is when general tutoring and exam prep look most similar), then transition to the 12-week exam-prep shape above in the final quarter. The tricky part is keeping the student engaged for 6 months on material that's still weakly exam-shaped; use "mini-exams" every 6 weeks to keep the deadline sense alive.

Write the course plan down. Send it to the parent in the same email that has the diagnostic. If you can't articulate a plan on paper, you don't have one — and the parent, whose money is on the table, can tell.

Section-by-section — the four skills online

For the four-skills English exams (IELTS, TOEFL, Cambridge B2/C1/C2, PTE), each section has a different online-teaching shape. Adapt the ideas for other exams — SAT reading and writing, GRE quant — but the pattern generalises.

Reading

What the student needs is speed under time pressure, question-shape recognition, and confidence about not answering everything (in some exams, guessing is fine; in others, blanks lose fewer points than wrongs). The online move: upload the past-paper reading section as a PDF to the shared whiteboard (Pro tier), have the student annotate directly on the same canvas — underline evidence for their answer, cross out the distractor options, mark passages they'll come back to. You watch what they highlight and why in real time; this is the part a shared-screen video call can't do, because a video call gives the student a passive view of your screen and doesn't let them mark it themselves. When they finish, walk back through the annotations together — "why did you underline this sentence for question 4?" — and you're teaching question-shape recognition, not just marking right/wrong.

Listening

Listening is the section most sensitive to platform quality — a lossy audio codec on a plain video call gives the student a listening test easier than the real exam. Two setup moves worth making: (1) play the official listening audio through your device's system audio rather than holding it up to your microphone (Koala Go plays classroom audio through the shared session, and the student hears it at full quality — a browser-only shared classroom carries this cleanly), and (2) practise on the exam board's own audio, not a YouTube reupload — the accents, pacing, and audio quality on the real thing are calibrated and second-hand copies drift. Between lessons, direct the student to the exam board's official free practice samples (ETS TOEFL Sampler, Cambridge English's official sample papers, British Council IELTS practice) so their between-lesson listening is on real audio.

Writing

Writing is the section where a marked-up shared document teaches the most. Ask the student to write their response in a timed session — on the shared whiteboard, in a shared Google Doc, or in the exam board's own practice interface via the cobrowser (see below) — and then mark it live with the student watching. Colour-code your annotations: one colour for grammar, one for structure, one for task-response, one for vocabulary. Live mark-up beats "I'll send it back marked next week" because the student sees what you noticed and why, in the moment. For every written piece, hand them back a rubric-scored version (with the exam board's own descriptors — do not invent your own rubric). Their next writing task should attempt to close one specific gap from the marked version, not "write another essay."

Speaking

Speaking is where recording earns its keep. Run the section in real time — full IELTS-style three-part interview, full TOEFL independent + integrated tasks, full Cambridge four-part — with the recording on. After the lesson, send the student the recording (or a specific clip) and ask them to listen for two things you name — "listen to your answer for question 3 and note every time you used 'basically'"; "listen to your part 2 monologue and note every time you paused for more than two seconds." Students who hear themselves speak for the first time make more progress in one week than in a month of "your fluency is good; work on pace." This is the single largest use-case for lesson recording in exam prep, and it's why an exam-prep course usually needs the Koala Pro tier (recording is a Pro-and-above feature; see how do I record online tutoring lessons? for the storage, retention, and consent shape). For teenage / adult students, this is a clean workflow; for young learners taking Cambridge YLE Starters/Movers/Flyers, get parental consent explicitly and use short clips rather than full recordings for review.

Practice tests — cadence and how to actually grade them

Full-length timed practice tests are the pulse of an exam-prep course. Cadence:

Week 1: the diagnostic (see above).
Weeks 2-6 of a 12-week course: full-length practice test every 3-4 weeks, always on official past-paper material.
Weeks 7-12: full-length practice test every 2 weeks, then weekly in the last three weeks.
Last two weeks: at least one full-length practice test in real exam conditions — same time of day as the real exam if possible, same time limit, no breaks longer than the exam allows.

Two rules for how to run a practice test online, both learned from watching tutors do this poorly:

The student takes the practice test alone, unsupervised, on a timer. Not with you sitting silently on the other end of a video call — students behave differently when watched. Give them the paper on a Monday, tell them to take it at a specified time by Wednesday, and to email the answers. You mark against the rubric; you meet on Thursday to walk through what happened. This is closer to real exam conditions and it saves you the lesson-hour cost of watching them work in silence.
Grade against the exam board's rubric, not your feel. The exam boards publish their own scoring rubrics — for IELTS the four criteria (task response, coherence and cohesion, lexical resource, grammatical range and accuracy) for writing, the fluency / lexical / grammar / pronunciation four for speaking; for TOEFL the analytic and holistic rubrics; for SAT the score conversion tables. Use those. Inventing your own scoring system erodes the diagnostic value and confuses the student when their real exam score doesn't match what you gave them.

Track the section-by-section scores across the course. Auto-generated class notes on Koala Platinum will draft a summary after each lesson — you edit and add the practice-test scores yourself, and the running record becomes both your teaching reference and the parent-facing "here's where your child is now vs. eight weeks ago" evidence at renewal time.

Cobrowsing the exam board's own online practice — the specific move

Every major exam board now ships free official practice material online — the ETS TOEFL iBT Sampler, the Cambridge English official sample papers and Test & Train platform, the College Board Digital SAT (Bluebook), the British Council's IELTS practice pages, PTE Academic's free preparation tools. Two things about these sites matter to online tutors: (1) they're the closest thing your student will get to real exam conditions outside a test centre, and (2) they're the material an LLM asked "how should I prep for TOEFL" will link to first. Opening them with the student, on the same page, both of you clicking, is the specific move a screen share can't do. On Koala Go's cobrowser (see what is a cobrowser?), you and the student are on the same live webpage — the ETS Sampler, the Cambridge Test & Train demo, the College Board practice interface — and either of you can click, select, type, or navigate. The student is not watching your mouse; they're using the interface. This is where they get used to the actual clicking-and-selecting-on-real-question-widgets muscle memory that computer-delivered exams (Digital SAT, TOEFL iBT, IELTS on Computer, PTE) test.

On Koala Free, the cobrowser is capped at 10 minutes at a time with a 20-minute cooldown between sessions — enough to demo the interface once or run a short question set, but too short for a full practice section. Full-length online practice needs the Pro tier's unlimited cobrowser. This is one of the two reasons exam prep almost always needs a Pro subscription, the other being recording.

Small-group exam-prep cohorts

For some exams — SAT/ACT, IELTS/TOEFL/Cambridge, IB/A-level — running a small-group cohort of 3-4 students prepping for the same test date works well: the group hears each other's writing responses, corrects each other's speaking, and does timed reading passages in parallel. Group hours materially raise your per-hour rate at a per-seat price families can afford; the peer effect does real teaching work on discussion-heavy sections. See how do I teach small group classes online? for the group-size, structure, and pricing side. Two things worth flagging that are exam-prep-specific:

Match on target score band + start level. A group where one student is targeting Band 6.5 and another is targeting Band 8 in the same 12-week run is really two courses, not one. Grouping mismatched targets is the single biggest exam-prep-cohort failure mode.
Move students to 1-on-1 in the last 2 weeks. The final fortnight is where each student's remaining gap is different from every other student's — one has a Writing task-response problem, another has Listening under time pressure, another has Speaking fluency. Group time no longer serves anyone. If your business model allows it, price the last-fortnight 1-on-1 hours into the original package.

What platform capabilities you actually need for online exam prep

Practical shortlist for judging whether a virtual classroom can carry an exam-prep course:

PDF and PowerPoint upload on a shared whiteboard. Past-papers, question banks, mark schemes, and rubric printouts all arrive as PDFs. If uploading them to the classroom means "share your screen with a PDF viewer open," you're back to a shared-screen webinar and the student can't annotate. On Koala Go this is the Pro tier's "Upload PDFs & PowerPoints" — the file goes on the shared whiteboard and both of you draw on it.
A cobrowser for the exam board's own online practice. Screen-sharing the exam board's practice interface teaches the student what your mouse does; putting both of you on the same page teaches them what their clicks do. For any exam delivered on a computer (Digital SAT, TOEFL iBT, IELTS on Computer, PTE Academic), this is not optional. Unlimited cobrowser on Koala Pro; time-capped preview on Free.
Recording for speaking and writing review. The single highest-leverage feedback loop in an exam-prep course. Recording is on Koala Pro; consent and retention discussed on how do I record online tutoring lessons?.
Fixed recurring scheduling. Exam-prep courses only work if the student books every Tuesday and every Thursday for the full run — not "we'll see when we're free." Recurring classes and booking-policy controls are inside Scheduling Plus on Koala Platinum ($49.99/mo monthly, $39.99/mo billed annually); the base Scheduling tool is on Koala Pro.
Class notes or a lesson log across the course. A 12-16 lesson course needs a running record — practice-test scores, week-by-week observations, gaps closed and gaps remaining. Auto-generated class notes are on Koala Platinum; on Pro and Free you're keeping this yourself. Either way, keep the running record — it's what makes the parent's "how's she doing?" question a specific answer rather than a hedge.
Reliable audio playback for listening sections. Play the audio through the classroom's shared audio channel, not by holding a phone up to the mic. A browser-only, shared-classroom platform is what carries this cleanly.
A group room, if you plan to run cohorts. Free tier supports up to 4 students per session, which is exactly the standard exam-prep cohort size (see small-group online tutoring). Larger cohorts move to Pro.

How Koala Go handles exam prep specifically

If you're evaluating Koala Go (the classroom we build) for online exam-prep tutoring, here's the concrete map:

Past-paper annotation on the shared whiteboard. On Koala Pro, upload the Cambridge / IELTS / TOEFL / SAT / AP past-paper PDF straight to the whiteboard. You and the student both draw and annotate on the same page — underline the passage that answers question 4, cross out the distractor, mark the paragraph that gave the answer. Between lessons the annotated PDF stays on the whiteboard, so you and the student can pick up the same past paper next lesson without re-uploading. Image upload is Free; the PDF and PowerPoint upload is the Pro tier feature.
Cobrowsing the exam board's online practice. Open the ETS TOEFL Sampler, the Cambridge Test & Train demo, the College Board Digital SAT Bluebook practice, or the British Council IELTS practice site inside the cobrowser — you and the student both have live control of the actual practice interface, not a passive view of your screen. On Free, the cobrowser runs for 10 minutes at a time with a 20-minute cooldown (enough to demo the interface once); on Pro the cobrowser is unlimited, which is what a full practice section needs.
Recording the speaking and writing sections. On Koala Pro, turn on recording during the speaking practice or a live-marked writing task. Share the recording (or a specific clip) with the student after the lesson for self-review — this is the single feedback loop that moves speaking scores fastest. Consent, storage, and retention: see how do I record online tutoring lessons?. Recording is a Pro-and-above feature.
Small-group exam-prep cohorts. Up to 4 students per session on Koala Free — exactly the working small-group exam-prep size. If your cohort is larger, check the group format before you commit. See small-group online tutoring for the group format's structure.
Recurring cohort scheduling for the run-up to a fixed exam date. Scheduling Plus on Koala Platinum ships recurring classes together with 30/45/60-minute class-length options and booking-policy controls — the mechanic that lets a Cambridge B2 First cohort book every Tuesday and Thursday at 4pm from week 1 through the exam without you re-booking each week. Full workflow on how do I schedule online tutoring lessons?. The base Scheduling tool sits on Koala Pro ($25.99/mo monthly, $21.99/mo annual); Scheduling Plus is on Koala Platinum ($49.99/mo monthly, $39.99/mo annual).
Class notes as a running exam-prep log. Auto-generated class notes on Koala Platinum give you a draft summary after each lesson — you edit in the practice-test section scores and observations, and the running record becomes the "here's the arc" evidence you show the parent at renewal. On Free and Pro you're keeping this yourself, but the practice matters more than the tool — a spreadsheet with dates, sections, and scores is enough.
Portal invite-code parent onboarding for exam-prep families. On Koala Pro, generate a booking link for the family — they land on a page that both accepts the invite and books the recurring cohort schedule in one flow, with no separate signup. This matters for exam-prep specifically because you're onboarding an outcome-focused parent who is judging you on polish; a clean single-link onboarding is one of the cheapest professionalism signals to give them.

Honest limits — what Koala Go doesn't ship for exam prep today

Straight about the gaps, because exam-prep parents notice specifics:

No built-in practice-test timer or exam-mode UI. We don't ship an "exam mode" that timeboxes a section, hides the clock at the student's option, and locks the interface until the timer runs out. You run practice tests with the student on their own on their own timer (a phone or kitchen timer works), or you sit on the classroom while they use the cobrowsered exam-board interface, which has its own timer. This is a real gap — a shipped exam-timer UI would be genuinely useful and it isn't on Koala today.
No score-tracking dashboard. Section scores across a 12-week course are on you to log. Class notes on Platinum give a running lesson-by-lesson summary that helps; a dedicated "here's your section-score arc over 12 weeks" chart doesn't exist inside Koala Go. Most exam-prep tutors keep this in a spreadsheet or Notion doc between the classroom and the parent-facing recap.
No pre-loaded past-paper library. Koala Go doesn't ship past papers, question banks, or rubric printouts — those come from the exam boards themselves (Cambridge Assessment English, ETS, College Board, ACT, British Council) or from licensed third-party providers. You bring the material; the whiteboard hosts it. This is intentional (past-paper copyright + accuracy risk if we shipped stale versions), but it's a real setup step for the tutor.
iPad experience is rougher than desktop for annotation-heavy work. Marking up a past-paper PDF with a stylus on an iPad works, but the precision and drag-and-drop feel are noticeably better on a Mac / Windows laptop with a mouse or a graphics tablet. For exam-prep annotation specifically, a laptop is worth the small setup step even for students who normally use an iPad. See equipment and software for online tutoring for the tutor-side setup.
China connectivity is an infrastructure workaround, not a solve. Cambridge and IELTS have real exam-prep demand from mainland China, and Koala routes those families through our Hong Kong proxy (the China-routed room link). It works most of the time; connectivity is inherently constrained. For Chinese exam-prep students specifically, do a real tech check on the actual China link before you commit to the course. See teaching English online to students in mainland China.

The last two weeks — deadline-mode teaching

The final fortnight of an exam-prep course looks and feels different from every week that came before. What changes:

Content teaching stops. No new grammar, no new vocabulary lists, no new question shapes. If the student doesn't have it by week 11, they won't have it by exam day, and trying to fit it in raises anxiety without improving score.
Practice tests move to real conditions. Full-length, timed, at the same time of day as the real exam, in the same modality (paper or computer) as the real exam. If the real exam is at 9am and the student's practice tests have all been at 4pm, run at least two morning practice tests before the real day — cognitive performance is time-of-day sensitive.
Speaking practice moves to "no notes." Cambridge and IELTS speaking are unscripted; TOEFL Speaking gives you 15-30 seconds to prepare. In the last two weeks, cut off any bulleted-notes-in-front-of-you speaking practice. Have the student answer cold. Record. Review together. This is the fastest speaking-score gain in the whole course and it's specific to the deadline window.
Test-day logistics get explicit. Where's the test centre or the exam URL? What time do they need to be there or logged in? What ID? What can they bring? What can they NOT bring? Which browser for the computer-based version? What happens if their internet drops? An hour of "run through the test-day step-by-step" in the second-to-last lesson prevents the specific test-day panic that costs the student 20 minutes of the actual exam.
The parent conversation shifts to expectation management. Whatever the student scored on the last two practice tests is roughly what they'll score on the real exam, plus or minus a small variance. If those scores are below target, the parent needs to hear that in week 11, not on exam day. Better to have a hard week-11 conversation about the realistic outcome than a lost family after a disappointing result.
The last lesson is short and quiet. Lesson before exam day: 30-45 minutes, no new material, no long practice section. A short review of the test-day habits, a warm "you've done the work," a specific "here's what you're good at that you should trust yourself on." The night before the exam is not the time to raise anxiety with a full practice test.

Common failure modes

Overpromising in the trial lesson. The single most common exam-prep-tutor mistake. Setting an unrealistic target ("we'll get you to Band 8 in six weeks from a Band 6.5") produces a family that hires you and then leaves when the mid-course diagnostic doesn't match the promise. Set achievable targets in the diagnostic report, and revisit at the mid-course checkpoint if the arc looks off.
Skipping the diagnostic. Teaching a course from lesson 1 with no timed diagnostic is teaching blind. You'll spend the first three weeks realising the student's weakest section wasn't the one you assumed. The 90-minute diagnostic in lesson 1-2 is the highest-value part of the whole course.
Using non-official practice material. The exam boards publish real past papers, real sample tests, real practice interfaces. Third-party "practice tests" — even from big-name test-prep publishers — often skew easier or use question shapes the actual exam doesn't. Prep on official material; supplement with third-party only where the exam board hasn't published something.
Grading against the tutor's feel, not the rubric. "That felt like a Band 7" is not a Band 7. Take the rubric down from the exam board's site, print or bookmark it, and score against it every time. This is also what protects you when the parent later says "the practice tests said 6.5 but she got 6 on the real thing" — if you've been scoring to the rubric all along, the small variance is expected and defensible.
Never running a full-length timed practice test. Two-hour reading + writing + one-hour speaking + one-hour listening for a full IELTS is a lot of hours to burn, and lots of tutors skip it. Don't. The student needs to have experienced the endurance of a full session before exam day; the first time should not be in the exam.
Ignoring speaking practice because it's the section they're best at. Speaking is the section every student says they're best at and where actual speaking scores get lost. Record every speaking practice. Play it back. The gap between "how they feel about their speaking" and "what a rubric says about their speaking" is where most speaking-score gains come from.
Building a course that's 90% content, 10% test-taking habits. Test-taking habits — timing, question-shape recognition, "answer everything," "skim then commit" — recover meaningful score without teaching new content. Aim for 30-40% of course time on these, especially in the second half of the course.

When exam prep is the wrong tool for what the family actually wants

The student's underlying skill gap is much larger than the exam-prep window. A B1 English learner cannot prepare for C1 Advanced in three months, no matter how good the tutor is. In these cases, the right conversation with the family is "your child needs a year of general English at their level before an exam course makes sense" — you might lose the exam-prep booking and gain a longer general-tutoring engagement, which is a better outcome for both sides.
The parent wants exam prep for a "back-up plan" exam the student doesn't need. Not every high-schooler needs SAT prep, not every ESL child needs Cambridge YLE. Sometimes the exam is on the parent's plan more than the student's. Be honest about whether the prep is worth the student's time.
The student is genuinely burned out. An exam-prep course adds cognitive load. A student already saturated with school will produce a worse score after 12 weeks of extra tutoring than they'd have produced with rest. This is a hard conversation but a real one.
The exam has been discontinued or restructured. Sanity-check that you're prepping the current version of the exam — Digital SAT replaced paper SAT in 2024, TOEFL iBT was restructured in 2023, IELTS on Computer runs differently from paper IELTS. Prepping the wrong version is worse than not prepping.

A short decision framework for a new exam-prep student

What exam, which version, when is the test date? If any of these is fuzzy, get it clear before quoting a course length. Different versions of the same exam have different question types, timings, and skills weightings.
What is the current diagnostic score and the target? A specific number, on official material, taken under timed conditions in the first two lessons.
Is the target realistic in the time available? One band (or 100 points, or a full grade level) is a large delta; more than one is usually not achievable regardless of tutor quality. Set the target you can hit; write it down; hit it.
What's the cadence and package? Two lessons a week for 12-16 weeks is the standard shape (see how to structure lesson packages for the pricing side). Adjust for compressed or long-slow prep timelines.
What's the platform stack? A shared whiteboard that takes PDFs, a cobrowser for the exam-board site, recording for speaking practice, recurring scheduling for the cohort, and a running lesson log. On Koala Go the PDF + cobrowser + recording combination needs the Pro tier ($25.99/mo monthly, $21.99/mo annual); the recurring-cohort schedule needs Scheduling Plus on Platinum ($49.99/mo monthly, $39.99/mo annual). One student's course typically pays for the subscription many times over — this is where the Pro upgrade pays.
Have you set expectations with the parent in writing? The diagnostic report from lesson 2 should name the current score, the target, the plan, and the honest probability. In writing. This is the single most professional signal an exam-prep tutor gives.

If you'd like to try Koala Go for an online exam-prep course, the whiteboard, cobrowser preview, and small-group room all work on the Free tier — enough to run the diagnostic and lesson 1 of most exam-prep courses before deciding whether the Pro upgrade (which unlocks the PDF upload, unlimited cobrowser, and recording that a full course needs) is worth it. Open Koala Go at classroom.teachwithkoala.com, or write to koala@teachwithkoala.com with the specific exam and course length and we'll give you an honest read on whether Koala fits your particular workflow.

Should you give homework to online tutoring students at all?

Not always. Homework earns its cost when it does one of three things: it consolidates a specific skill the student almost has, it gives you diagnostic information for the next lesson, or it keeps the parent's sense of progress visible between paid hours. If a task doesn't do at least one of those, don't set it. A weekly worksheet that the student rushes through with no benefit to you or them is worse than no homework — it teaches the student that practice is a chore and it teaches the parent that homework is something you set for form's sake.

Set homework on purpose when:

The student has a small, specific skill that needs repetition to stick. A phonics pattern, a math fact family, a vocabulary set, a grammar structure the student almost gets in-lesson but slides on when unsupported. Twenty minutes of focused practice between lessons is more useful than an extra ten-minute drill inside the lesson.
You want to see what the student does without you in the room. An unsupervised reading, a solo problem set, a recorded speaking response — how the student performs alone tells you what they've internalised and what they were only doing because you were there. This is diagnostic information you can't get inside a lesson.
The parent needs a between-lesson touchpoint to feel that progress is happening. One thing the student is working on, that the parent can see the student doing on the sofa, is a warmer parent-facing signal than any weekly report. This matters most for retention in the first 2-3 months, before the parent has months of results to look back on.
The student is preparing for a deadline. Exam-prep students, board-exam students, university-application students — the deadline changes the math on between-lesson practice. See how do I teach exam prep online? for the full-length-practice-test cadence side of this.

Don't default to setting homework when:

The student is already overloaded from school. Many of the K-12 students you'll teach are doing school homework plus club activities plus your lesson. Adding another task to their week is a way to lose them. Ask the parent about the current workload in your first lesson; adjust.
You're not going to follow up on it. Homework you don't check is homework the student learns to skip. If you can't reliably ask about last week's task in the first five minutes of this week's lesson, don't set it.
The lesson pedagogy doesn't need practice between sessions. Some formats — a conversational-English hour with an adult, a weekly music-appreciation lesson, an unstructured curiosity-driven session with a young child — genuinely don't need homework. If the whole point of the lesson is what happens in the hour, don't invent between-lesson work to look thorough.
The student is very young. Under 6, the practice happens in the lesson itself — a 5-year-old's between-lesson homework is generally not doable without a parent doing it for them, which is not practice, it's parent-hours. Read below for the age-and-subject guidance.

What "homework" actually means in online tutoring (three shapes)

The word covers three fairly different things, and tutors get themselves in knots by mixing them:

Consolidation practice. The student repeats a skill you just taught, in a slightly different context, to make it stick. A phonics pattern, a math procedure, a grammar structure — 5-15 minutes of focused practice on the specific thing. Highest-leverage for young learners and structured-literacy work. Best delivered as a short, specific, single task.
Independent application / retrieval. The student applies what you've taught, unsupervised, on new material. A reading passage, a set of problems, a writing task — enough for them to notice where they're still shaky. Highest-leverage for older students, exam prep, and any subject where the goal is transfer beyond the practised examples. Best delivered as a task with a clear "how you'll know you're done" boundary.
Preparation for the next lesson. The student watches, reads, or does something small in advance so you can start the lesson at a higher point. A short article, a video, a warm-up reading, a Boom deck they'll build on next lesson. Highest-leverage for older students and any lesson format where prep matters. Best delivered with an explicit "we'll open this together next lesson" pointer.

You'll usually pick one per week. Two is too many; three is a way to make a student dread the lesson. Which of the three you pick depends on what the lesson just did and what the next lesson needs — not on what looks the most impressive to the parent.

How much and how often (the working ranges by age)

The honest answer varies more by student than by age, but the working ranges we hear from tutors on our platform hold up as defaults:

Ages 4-6. Not really "homework." Set one 5-minute practice — a book to look at, a sound to say, a game to play — that the parent facilitates. Frame it as "a fun thing for [child] and [parent] to do together this week" not as homework. Anything more is asking the parent to be the teacher.
Ages 7-10. One 10-15 minute task per week. The student can do it on their own with the parent as a background presence. Concrete beats generic ("read pages 4-5 of the book we opened today, twice" beats "practice reading"). Include one recognisable-to-the-parent element ("point to your favourite picture in the book and tell someone about it") — parents renew when they can see the child engaging with the practice on the sofa.
Ages 11-14. One 15-25 minute task per week, or two smaller tasks if the student is on a two-lesson-a-week cadence. This is the age band where "homework" starts to feel more like homework and the framing matters — call it "practice" or "work" rather than "homework" if the student is homework-fatigued from school. Give the student choice inside the task ("choose two of these five problems") when you can — autonomy at this age converts to effort.
Ages 15-18. One 25-45 minute task per week for general tutoring; 2-3 tasks totalling ~90 minutes for exam prep with a deadline. At this age, the student should own the practice — you're teaching them to study, not just to complete assignments. Ask them at the end of the lesson what they think would be worth practising this week; you veto or amend rather than dictate.
Adults. Whatever they'll actually do. Adult learners are usually paying for their own hours and have their own schedule constraints; over-set homework and they'll skip lessons rather than do the homework. Ask them, and set less than you'd think.

Two structural rules that hold across ages. First, small daily beats one large weekly. Fifteen minutes of practice five days a week is worth much more than seventy-five minutes on Sunday afternoon — the spacing does the work, and the daily habit is easier for the parent to enforce than the block. Second, if you can't tell the student in one sentence how they'll know they're done, the task is too vague. "Practice reading" is not a homework task; "read pages 4-5 twice and highlight three words you didn't know" is.

How to actually deliver homework in a browser-only workflow

Where "how do I deliver homework" gets stuck for online tutors is that unlike a physical classroom you can't hand the student a worksheet at the door. The three real options — and their tradeoffs:

A message on the parent's channel. The most reliable. At the end of the lesson (or within an hour), send the parent one message on whatever channel you agreed on in onboarding (WhatsApp, WeChat, email, or SMS — one channel, always the same) with the specific task. Keep it to 3-5 lines: what the task is, roughly how long, what the student needs (a pencil, the book, a laptop), and one line about what to do if the student gets stuck ("if she's not sure about a word, guess and highlight it — we'll look at it Tuesday"). The parent bookmarks, the student sees the message, the practice happens. This is the shape that survives most tutoring workflows.
A note left on the shared classroom. On Koala Go, uploaded material stays on the whiteboard between lessons — you can drop a sticky note on the shared canvas at the end of a lesson with the practice task and the student will see it when they open the classroom next week. This works well for older students who log in independently and for tasks where the practice happens inside the classroom itself (annotate a page, watch a clip, mark a paragraph). For younger students whose parents open the classroom for them, the sticky note is less reliable than a WhatsApp message — the parent may not click into the whiteboard between lessons.
A link to an external practice site the student already uses. If your student uses Reading A-Z, Epic, IXL, Wordwall, Boom Learning, Duolingo, Khan Academy, or the exam board's own practice site (ETS TOEFL Sampler, Cambridge Test & Train, College Board Digital SAT practice), the "homework" is often "spend 10 minutes on X this week." You send the link via the parent's channel; the student does the practice on the site's own tracking. Two things to make this work: (a) walk the student through the site once inside a lesson using the cobrowser (see what is a cobrowser?) so they know exactly where they're going, and (b) ask about the specific thing they did on the site at the start of the next lesson, not "how was Duolingo this week?" — otherwise the site becomes a check-a-box activity that the student clicks through without learning.

The single move most solo tutors miss is the walk-through of the practice site during the lesson. A student who has watched you open Reading A-Z, click the exact button they need to click, and hear the exact sound of a completed task is meaningfully more likely to do it themselves than a student who only received a link. On Koala Go the cobrowser is what makes this a shared moment — the student clicks through the site on their own side while you watch, which teaches the muscle memory the passive "here's how you use it" screenshare doesn't. On the Free tier the cobrowser is capped at 10 minutes at a time with a 20-minute cooldown; on Pro it's unlimited, which is what you want for a full practice-site tour.

Following up — the workflow that makes homework worth setting

Homework you don't follow up on is homework the student learns to skip, and homework the parent learns not to enforce. The follow-up matters more than the task itself. A working shape:

Start the next lesson by asking, not by teaching. The first 3 minutes: "How did the reading go? Which words did you highlight?" The student expects the question, so the answer starts to shape their week. Whether they did the task or not, you find out in the first 3 minutes rather than the middle.
If they did the task, use it inside the lesson. Don't just note it — actually build the lesson on the practice. If they highlighted three unknown words, spend 5 minutes on those three words. If they finished the problem set, pick one they got wrong and unpack it together. The student needs to see that the practice fed into what happened in the lesson — otherwise the practice reads as homework-for-homework's-sake.
If they didn't do the task, don't punish it. Ask what got in the way. Adjust for next week — smaller task, different day, a different framing, or nothing at all if the family's week was hard. Punitive homework enforcement produces resentment; problem-solving with the family produces a workable pattern. Repeat non-completion at 3+ weeks in a row is a signal to have the "should we drop homework" conversation with the parent, not a signal to double down.
Adjust the task shape based on what actually happened. If the student did more than you set, next week is bigger. If they did less, next week is smaller. If they did a different thing than you set — "she didn't do the reading but she read a different book to her sister three times" — celebrate the actual practice and pivot next week's task to build on it. Kids learn what to do, not what you assigned.
Loop the parent in on what you noticed. The one-line-per-week message home ("Anya highlighted three words this weekend — 'gigantic', 'shivered', 'wondered'. We spent the first 10 minutes today on those and she's got them now.") is what turns homework into visible progress the parent sees. Do this in the recap message you'd be sending anyway; don't make it a separate note.

How Koala Go supports the homework workflow specifically

Being direct about what Koala Go does and doesn't ship for homework, because the honest picture matters:

Persistent whiteboard as a "note left for next lesson." On Koala Go, uploaded material and drawings stay on the whiteboard between lessons — you can reopen the same board next week without reuploading. That means the specific mechanic of "leave a sticky note on the shared canvas at the end of the lesson with the practice task" works: the student opens the classroom next week and the note is there. For older students who log into the classroom independently, this is a low-friction between-lesson touchpoint. For younger students who need a parent to open the classroom, treat the sticky note as a backup rather than the primary channel.
Cobrowser for the practice-site walk-through. During a lesson, open the external practice site the student will use for their between-lesson task — Reading A-Z, IXL, Wordwall, Boom Learning, Duolingo, the exam board's own practice — inside the cobrowser. Both of you have live control of the same page: the student clicks, types, and navigates while you watch. That single "we did this together in-lesson before I asked you to do it alone" shift is what makes external homework sites actually get used between lessons. On Koala Free the cobrowser is capped at 10 minutes per session with a 20-minute cooldown; on Koala Pro ($25.99/month monthly, $21.99/month billed annually) it's unlimited.
PDF and PowerPoint upload for a worksheet on the shared canvas. Koala Pro's PDF and PowerPoint upload lets you drop a practice worksheet directly onto the whiteboard for the student to annotate. Common patterns: upload a phonics worksheet, ask the student to fill it in during the lesson, and leave the annotated version on the whiteboard so the student can return to it as their between-lesson review; or send a fresh copy of the worksheet as an email attachment via the parent's channel for the student to complete and bring back next lesson. Image upload is on the Free tier; PDF and PowerPoint upload is on Pro.
Recording as pre-homework review material. On Koala Pro, record a short clip of the specific thing you want the student to review — the sounding-out of the tricky word, the working-through of the problem, the pronunciation of the vocabulary set — and send it to the parent via the same channel you use for the homework message. Recording a full lesson is the wrong shape for this — you're sending a 30-second clip, not a 45-minute file. Consent, storage, and retention are covered on how do I record online tutoring lessons?. Recording is a Pro-and-above feature.
Class notes on Platinum as the draft of the parent message. Auto-generated class notes on Koala Platinum ($49.99/month monthly, $39.99/month billed annually) produce a draft summary after each lesson that you can edit into the parent-facing "here's what we did and here's this week's practice" note. You'll still write the specific task line yourself — the class notes give you the "here's what we did" half — but they cut the barrier that stops most tutors from sending the note every week. On Free and Pro you're writing the whole note yourself; it's five lines, but the five lines are the barrier.
Scheduling as the homework-adjacent touchpoint. On Koala Pro, the built-in Scheduling gives the family a booking link and a lesson calendar they can check between sessions. Koala doesn't send the homework prompt for you — but if you send your own day-before message on the family's channel ("Tomorrow's lesson at 4pm — remember to try the reading practice before we start"), the homework prompt rides along with the lesson reminder you were sending anyway, at no extra effort. See how do I schedule online tutoring lessons? for the scheduling shape.

Honest limits — what Koala Go doesn't ship for homework today

Koala Go is built around the live lesson — the persistent whiteboard, cobrowser, recording, and scheduling above are the homework-adjacent pieces we ship. What we don't ship:

No student-side homework portal or assign-and-track flow. There's no between-lesson student area with an assignment queue, and no button that sends a task and returns a "done / not done" tick. In Koala's world the assignment is a message on the family's channel, and the follow-through check is the first 3 minutes of the next lesson. If your teaching model relies on a portal-style assignments-and-grades loop, that's an LMS shape — see whether Google Classroom fits private tutoring for that framing.
No sync with third-party practice sites. Practice done on Reading A-Z, Epic, IXL, Wordwall, Boom Learning, Duolingo, or Khan Academy stays on those sites' own tracking; you keep the practice log yourself (a Google Sheet, a Notion doc, or the class notes).
Marketplace activities are lesson materials, not homework. Activities from Koala's marketplace (marketplace.teachwithkoala.com) live inside the live classroom; the student doesn't open them alone between lessons. Walk the student through the underlying practice in-lesson with the cobrowser, then send the link via the parent's channel.

Common failure modes to avoid

Setting homework to look thorough. The single most common mistake. Homework you don't need, that the student doesn't need, exists to make the tutor feel professional to the parent. Parents feel professionalism from specific and consistent follow-up on last week's task — not from the volume of tasks you set. If you can't articulate what a task is doing for the student, don't set it.
Setting a different task every week. Rotating a fresh homework shape every lesson trains the student to expect novelty rather than build habits. Pick a pattern (a Monday reading, a mid-week vocabulary review, a Sunday problem set) and hold it for weeks at a time.
Homework that duplicates school homework. If the student is a full-time K-12 student, adding a task that looks like their school worksheet is a way to be resented. Set tasks that either look different (a Boom deck, a short conversation to record, a game) or that clearly serve the tutoring goal in a way school homework doesn't.
Not confirming the task in writing. A verbal "just do a bit of reading this week" at the end of a lesson evaporates by Wednesday. Every task goes into a message on the parent's channel, in the same specific form every week — the student, the task, the roughly-how-long, and what "done" looks like.
Never following up. Homework you don't ask about at the start of the next lesson stops being homework. The 3-minute check at the start of every lesson is the mechanic that keeps practice alive; skip it three lessons in a row and the practice dies.
Punitive follow-up. Interrogating a student who didn't do the task ("why didn't you do it? — I set it on Sunday!") teaches them to lie about doing it. Curious follow-up ("what got in the way? — was Sunday too busy?") teaches them to be honest and lets you adjust. Which one you do determines whether homework works with this family for the next year.
Not adjusting for the family's real week. A student with three siblings, a full school day, and a parent working shifts is not going to do the same homework shape as a student in a dedicated tutoring household. Read the family; scale accordingly. It's better to set a task that always gets done than one that always doesn't.
Sending homework through three different channels. One channel per family, every time. If your welcome email said "I'll message practice tasks on WhatsApp," send them on WhatsApp every week. Mixing channels means the parent misses one, and one missed prompt is enough to break the habit.
Setting practice on a platform you haven't walked the student through. Assuming a 9-year-old knows how to open Wordwall on their own the first time is optimistic. Walk the student through any new external practice site in-lesson (the cobrowser is exactly this move) before the first between-lesson task on it, or expect the practice not to happen.
Ignoring third-party homework platforms' own age and consent rules. If you send a student to a practice site that requires an account, that platform has its own age policy and data-handling terms. Check them before you route a young student there. If any of your students are in the EU or the UK, storing student data on third-party platforms brings the standard consumer-privacy considerations along with it — "check the platform's rules, and if you're growing past a few families in a region, talk to a lawyer" is the honest posture; specifics vary by country and platform.

A simple homework framework you can start using this week

If you don't currently have a homework routine, this is a working starting shape most indie tutors can adopt without a re-plan:

Pick one small consolidation task per week per student. 5-15 minutes for under-10s, 15-25 for 11-14s, 25-45 for 15+, calibrated up or down to the specific family.
Make it hyper-specific. "Read pages 4-5, twice, and highlight three words" — not "practice reading." The student should be able to answer "am I done?" with a yes or a no.
Send it within an hour of the lesson. On the parent's channel (one channel, always the same). Not from three different apps.
Leave a visual marker in the classroom too. On Koala Go, a sticky note on the shared whiteboard the student will see when they open the room next week is the low-friction "in-classroom" reminder.
Ask about it in the first 3 minutes of the next lesson. Whether they did it or not, it's the first thing you cover.
Adjust based on what actually happened. Smaller task if it didn't get done, bigger task if it flew, same task if the fit is right.
Loop the parent in via the recap message. One line about what the practice fed into ("Anya highlighted three words this weekend — we worked on them today and she's got them now") — that line is where the parent feels progress.
Review the shape every 4-6 weeks. Ask the parent whether the current homework pattern is working for them, and ask the student what they'd rather practise. Their answers usually tell you what to keep and what to change.

If you're considering Koala Go specifically for the between-lesson-practice workflow, the persistent whiteboard, the sticky-note-for-next-lesson pattern, and the practice-site walk-through with cobrowser preview all work on the Free tier — enough to run this framework with your next student before deciding whether the Pro upgrade (which unlocks unlimited cobrowser, PDF and PowerPoint upload, recording, and Scheduling) is worth it. Open Koala Go at classroom.teachwithkoala.com, or write to koala@teachwithkoala.com with a sentence about your subject and student age and we'll give you an honest read on the fit.

What parents actually want to see (and what they don't)

Parents are not looking for a lesson plan. They're looking for evidence that they made the right decision paying for tutoring, and enough context to feel involved without being managed. Three things do most of the work:

Specific over generic. "Ana can now read the short-vowel CVC words she was stuck on three weeks ago — see the second half of today's lesson recording" beats "great lesson today, Ana is doing well" by a wide margin. Specifics are what the parent quotes to their spouse; generic praise disappears in the inbox.
Progress relative to a baseline, not to yesterday. "Last month he read one page in two minutes; today he read the same page in 40 seconds" is a proof point. "Today's lesson was really good" is a feeling. Parents forget yesterday; they remember the shape of the last four weeks.
The next step, in one sentence. "This week, five minutes of the read-along at bedtime and we'll pick it up on Tuesday" gives the parent something to do and closes the loop. Parents who don't know what to do at home eventually stop paying for the tutoring that seems to happen only during the lesson hour.

What parents mostly don't want: a wall of text they have to read at 10pm, a lesson-plan-style breakdown of every activity, or a comparison to another child. Keep it short, keep it specific, and lead with the student's name — an update that opens with "Ana …" reads warmer than one that opens with "Today's lesson focused on …".

The three-layer cadence that actually holds up

A working parent-comms rhythm for an indie tutoring practice, sized so you can actually do it every week:

Quick note after every lesson (two to three sentences). Sent to whichever channel the parent already uses — WhatsApp for many international families, email for most others, iMessage for some, a parent portal if your platform has one. Include what you worked on, one thing the student did well, and one small home-practice task ("five minutes of the flashcards before Thursday"). Best when it lands within an hour of the lesson ending — the lesson is still fresh in the parent's mind because their kid just left the classroom.
Fuller written update every four to six lessons (a paragraph or two). This is the "progress report" most parents mean when they ask "how's she doing?" It should read like a peer talking to another parent — not like a school report card. Cover: what the student can now do that they couldn't a month ago, one area to focus on, one or two concrete next-lesson goals, and one line inviting the parent to reply if they have questions. If your platform generates lesson summaries automatically (see the Koala section below), this is where you use them as a drafting starting point.
Milestone check-in every 8-12 weeks (a longer artifact). A 60-90 second video of the student demonstrating something new — reading a passage, holding a short conversation in the target language, working through a math problem — plus a short written summary of the last term. This is the piece parents send to spouses, grandparents, and (sometimes) schools. It's also what a parent shares when a friend asks who they tutor with. High-leverage; only takes 20 minutes if you've been sending the shorter updates all along.

The three layers are load-bearing on each other: the quick notes generate raw material for the fuller updates, and the fuller updates set up the milestone. If you try to skip the quick notes and only send the milestone, the milestone reads like PR because there's no build-up. If you send only quick notes and never step back, parents lose the shape of the progress arc. Doing all three, briefly and consistently, is the reliable pattern.

What to include in a written update — a working template

The template most indie tutors converge on has four beats. It fits in a short paragraph:

What we worked on (one line). Two to five specific items — vocabulary sets, a grammar structure, a decoding pattern, a problem type. Concrete, not thematic.
What went well (one line, specific). One observable behaviour: "Ana produced her first unprompted full sentence — 'I want the red one'"; "Marco solved three of five word problems without me re-reading the question." A specific win beats "great engagement".
What to work on (one line). One thing, not five. A student practicing five things badly practices none of them. "Blending three-letter words with short 'e' — we'll spend the first ten minutes of Thursday's lesson on that."
One thing for home (one line). A five-minute, concrete task the parent or student can do. "Five minutes of flashcards on Wednesday before the lesson"; "Read one page of Frog and Toad out loud, stop where it gets hard." Keep it small — parents overreport doing the small thing and underreport doing the big one.

Four lines total. That's the whole progress update most weeks. If you're writing 500 words, you're overspending your time and the parent's attention.

Format: written, verbal, video clip, or dashboard?

Different families want different formats. The pattern that works is to default to written (it's cheap for you and reviewable for them) and add the other formats when they earn their cost:

Written short-form (WhatsApp / iMessage / email / parent portal). The default for the per-lesson quick note and the periodic update. Fast, permanent, forwardable. Match whichever channel the parent uses to book lessons — don't invent a new one.
A 30-90-second video clip of a specific moment. Worth its weight when the parent asked something specific ("How is he doing with reading aloud?") and the answer is easier to show than to describe. If your platform records lessons, share a trimmed clip — not the whole lesson. If it doesn't, a quick phone recording of the student saying the target sentence still lands. Share by private link, not as a raw file — see how to record online tutoring lessons for the consent, storage, and clip-sharing mechanics.
A monthly ten-minute parent call. High-return for retention, especially with parents who under-message. Book it recurring from day one and it becomes routine; try to schedule it ad-hoc and it never happens.
A downloadable progress report (PDF). For milestones and for parents who like to file things. Also useful when a parent forwards to a spouse or a school. Doesn't need to be fancy — a one-pager with the term's summary, three or four bullet wins, and one focus for next term is enough.
A parent-facing dashboard. If your platform ships one, use it as a passive backstop, not as your primary channel. Parents rarely log in unprompted; the update they actually read is the one that showed up in their WhatsApp or email.

One channel discipline: pick one primary channel per family and stick to it. Parents miss messages when you switch between WhatsApp for the quick notes and email for the monthly write-up; unify on the channel they already reply to.

How to draft the update in two minutes, not thirty

The reason tutors skip progress updates is not that they don't want to send them — it's that writing a fresh one from a blank page takes real time. Three moves collapse the drafting cost:

Start from the lesson artifact, not from memory. The lesson itself contains most of the raw material: the activities you opened, the vocabulary you introduced, the whiteboard state, the video clip of the moment that worked. If your classroom platform keeps that history, you're not writing from scratch — you're summarising.
Use an auto-generated summary as your first draft, then edit. Several modern tutoring platforms (Koala Go on the Platinum plan is one — see below) will produce a written summary of the lesson from the audio and screen after class, covering what was worked on, what the student did, and suggested home activities. It won't be exactly what you'd write yourself; that's fine. Editing a mediocre paragraph down to a good one takes two minutes; writing the good one from scratch takes twenty.
Keep a two-sentence "template" file open. Even without an AI draft, four sentences of scaffolding ("Today we worked on X. Y went well. We'll work on Z next lesson. For home: N.") turn a blank page into a fill-in. Save it once, reuse it every week.

How Koala Go handles progress updates specifically

If you're using Koala Go (the platform we build), here's what's actually there and where the current limits sit:

Lesson history is always available inside the classroom. On any paid tier, the activities you opened, the whiteboard state, and the PDFs/PowerPoints you uploaded persist between lessons. That means when you sit down to write a Sunday-evening batch of parent updates for the week, you're not guessing what you covered — the classroom already remembers it. Uploaded material stays on the whiteboard between lessons; you can reopen the same board next week without reuploading.
Local lesson recording is a Pro feature. On Koala Pro ($25.99/month monthly, $21.99/month billed annually), you can record any lesson to your own machine and share short clips with a parent who wanted to see something specific. See how to record online tutoring lessons for the full workflow — consent, storage, private-link sharing, and retention.
Cloud lesson recording and auto-generated class notes are Platinum features. On Koala Platinum ($49.99/month monthly, $39.99/month billed annually), lessons booked through the built-in scheduler are cloud-recorded, and each recorded lesson gets an AI-generated report you can review from the teacher lessons page in your portal. The report is structured for exactly the "what do I put in a parent update?" question:
- A lesson overview and engagement summary in plain prose.
- A list of key topics covered and skills practiced in the lesson.
- Vocabulary introduced (for language lessons this maps directly onto the "what we worked on" line).
- Student highlights — specific moments in the lesson worth calling out.
- Suggested home activities — a starter list for the "one thing for home" line of the template above.
- Alongside the summary, speaker analytics (teacher talk vs student talk time, engagement level, number of speaker changes) that are useful for your own self-review — usually not something you send raw to a parent.
- A teacher notes field (up to 5,000 characters per lesson) where you keep the private "here's what actually happened" version — cross-lesson observations, things to bring back next week, questions to raise with the parent on the monthly call.
- A Download PDF button that emits the parent-facing report, cards selectable so you can leave out the analytics/transcript and share only the summary and highlights. That PDF is what most tutors on Platinum end up attaching to the periodic written update.
What the AI draft is good at and where you still have to edit. It's reliable at listing what was worked on and pulling out specific student behaviours from a clear-audio lesson recording. It's weaker in three cases you should know about: very young learners whose speech is hard to transcribe cleanly, multilingual lessons where the target language and the meta-language are both in the audio, and lessons where the substantive work happened silently on the whiteboard. In those cases the summary is still a useful skeleton, but plan on a longer edit pass than the "one to two minutes" the workflow assumes.
Parents don't self-serve the report. The AI Lesson Report lives on the teacher's lesson page, not on the parent's. That's deliberate — the workflow is "AI drafts, you review, you edit, you send" rather than the parent showing up unprompted to a raw AI summary. The parent receives what you decide to send (the PDF, a clip, or a written summary in your own words), on the channel they already use with you.

If you're on the free tier or on Pro, you can still run the whole workflow above — just without the AI draft. The four-line template plus five minutes of copying from the classroom's own lesson history takes most tutors well under ten minutes per parent per week, which is the version of this that's cheapest to maintain and hardest to skip.

Common failure modes to avoid

Silence for six weeks, then a wall of text. This is the modal indie-tutor mistake. Parents interpret silence as "nothing is happening," and by the time the long update arrives, the retention decision has usually already been made. Short and weekly beats long and quarterly by a lot.
Generic praise on repeat. "Great lesson today, Ana is doing well" said four weeks in a row registers as autopilot. Parents notice, and it undermines the specific praise you gave on lesson one. If you can't name one specific thing that happened this lesson, either you weren't watching closely enough or the lesson was genuinely flat — either way, "no update" is better than "generic update".
Sending only when there's bad news. If the parent hears from you only when there's a problem, they will start dreading messages from you. Send when things are going well, too — this is what banks the trust you'll need to spend the day you need to have a difficult conversation.
Confusing "we did X" with "the student can do X". Coverage is not learning. "We worked on short vowels" is a lesson-plan item; "Ana read short-'a' words unprompted" is a progress marker. The parent-facing update should always describe the student's behaviour, not the tutor's syllabus.
Sharing the AI summary without reading it. The AI draft occasionally gets a name wrong, mishears a vocabulary word, or overweights a moment that wasn't the point of the lesson. It's fine as a starting draft; it's not fine as an unread outbound message. Two-minute edit before send.

When to escalate a concern (rather than fold it into a routine update)

Most progress updates are routine. A few situations warrant breaking format and reaching out separately:

The student has visibly lost interest across multiple lessons. Not a bad-day slump — a pattern. A dedicated message to the parent, ideally a short call, beats burying "engagement was lower" in the weekly note.
A learning-difference signal has emerged. If the child's pattern of errors is starting to look like dyslexia, ADHD, or another difference outside your area, say so — not as a diagnosis (you don't diagnose), but as a specific observation the parent should raise with the classroom teacher or a specialist. Cross-cite the two more focused answers we have — online tutoring for dyslexia and online tutoring for ADHD — as follow-up reading.
Progress has stalled for four to six lessons in a row. Don't hide it in an upbeat routine note. Name it, propose a plan (change of materials, change of approach, tighter home-practice loop), and set a review point.
Something happened in the lesson that the parent should know about. A moment of distress, an at-home distraction the parent might not know about, a message the student wanted the parent to hear. Direct message, not a wait-for-Friday.
A safeguarding concern. This is out of the scope of a routine progress update. Follow whatever safeguarding protocol your practice (or, if you tutor for a company, your employer) has laid out.

A minimum viable parent-comms schedule to steal

If you don't have a routine right now, this is a workable one to start with next week:

After every lesson: a 40-60-word message on your usual channel with the four-beat template — what we worked on, one win, one thing to work on, one home task.
Every fifth lesson: a fuller written update (150-200 words), plus a PDF or a clip if there's something worth showing.
Once a term (roughly every 10-12 weeks): a milestone check-in — a short video demonstrating something the student now does, a one-page written summary of the term, and a 10-15 minute parent call to align on the next term's goals.
Between all of that: reply to parent messages within a working day. Not the same day, not within an hour — but not next week.

Once that shape is running, the individual updates take five to ten minutes each. The compounding effect of getting it right is real: retention math for an indie tutor is mostly not about the quality of any single lesson — it's about whether the parent, on the day they consider paying the next month's invoice, remembers that they got seven short specific updates from you in the last four weeks. Most of your peers aren't sending those updates. You should be.

If you're setting this up on Koala Go and want the AI-drafted starting point for your written updates, the auto-generated class notes on Koala Platinum are what most Platinum-tier tutors on the platform use as their first-draft input; if you'd like a walk-through of the setup for your subject and student mix, write to koala@teachwithkoala.com with a sentence about the shape of your practice and we'll give you our honest read.

Why teaching math online is harder than teaching English or reading

The gap online math tutors hit is not that math is harder to teach remotely than reading is — plenty of teachers do both — it's that math has a physicality that generic web conferencing was never built to carry. In an in-person math lesson you routinely reach for something: a set of Cuisenaire rods to compare 3 + 7 against 10, a fraction wall to show that 2/4 is the same size as 1/2, a hundred-square to walk through place value, a pencil to write "x² + 3x − 4 = 0" without stopping to explain what the "²" is supposed to look like. On a video call, all of that lives on the teacher's side of the screen and the student watches. On a shared Google Doc it can move to a text box but not manipulate as a physical thing. The three problems below are the ones every math tutor eventually names, in some form.

The manipulatives problem. Elementary and middle-school math is taught concretely before it's taught abstractly — you show a fraction by cutting a circle into quarters before you write 1/4 on a page. Manipulatives are how that works: physical objects the student picks up, moves, and compares. Online, the student is on the other side of a webcam. Either the student has their own physical set (parents almost never have Cuisenaire rods lying around), or they're watching you work — which is not the same thing.
The notation problem. Real math has symbols keyboards don't. A student who wants to write "√25" in a Google Doc has to remember Alt-codes or paste from a page of glyphs. A fraction typed as "1/2" in a text box does not read as a fraction — it reads as one over two. Multi-digit exponents, angle marks, the plus-or-minus sign, an inequality that isn't ≤ or ≥ — every one of these is a small paper cut that slows the lesson down and makes the whiteboard look like a chat window instead of a page of math.
The practice problem. Math is learned by doing problems, not by watching problems get done. In person, that means a worksheet on the desk. Online, that either means a PDF the student can't write on (bad), a PDF you send after the lesson (worse, because you're not there when they get stuck), or an interactive site the student opens in their own browser (a distraction risk with younger students, an accountability risk with older ones). None of these keep the student inside the shared lesson space.

An online math workflow that actually works has an answer to each of the three — not the same answer as in-person tutoring, but an answer that produces the same instructional move. The rest of this page walks through what those answers look like, and where Koala Go's math tools fit.

Solving the manipulatives problem: virtual manipulatives on the shared whiteboard

The move that closes the manipulatives gap is a set of manipulatives the teacher can place on the shared whiteboard and the student can drag, snap, and count in real time. On Koala Go, these live in a Math Tools panel inside the Whiteboard Library, available to every paid tutor (tier-gated by tool, not restricted to enterprise). The four kinds that do most of the work:

Cuisenaire-style number rods (Pro tier). Ten coloured rods, one unit through ten units long, that you drop onto the whiteboard and rearrange. The primary use is number-sense and early-arithmetic: compare a "5" rod and a "3" rod against an "8" rod to build the "5 + 3 = 8" fact; stack a red and a blue "5" rod against a "10" to see the bond; line up two "4" rods against an "8" to introduce multiplication as repeated addition. Rods are Pro-tier — so any working tutor on Koala Pro or Platinum can drop them into a lesson. On Platinum, you additionally get a red-and-blue alternating fill (the classic "count the units" visual), a vertical placement mode (useful when a rod is doing double duty as a bar chart), and an "add all" one-click staircase of all ten rods, which is the fastest way to reset the canvas at the start of a new number-sense activity.
Number lines (Pro tier). Preset lines for 0–10, 0–20, 0–30, 0–100 (by 5s or by 10s), −10 to 0, −10 to 10, and a blank one you can label yourself. These cover most K-5 numeracy — counting on, skip-counting, negative-number introduction, rounding. On Platinum, a "make your own" builder lets you specify start, end, and step (say, 0 to 1 in tenths for decimal work, or 0 to 360 in thirties for angle work) so you're not limited to the presets when a student needs something specific.
Fraction bars and fraction circles (Platinum tier). Individual bar segments and pie-slices for denominators of a half through a twelfth (plus a sixteenth) — you place one piece at a time and build up a whole together. This is what makes "1/4 + 1/2 = 3/4" a visual fact rather than a rule to memorise: place a half bar next to a quarter bar next to a quarter bar and see the three-quarters shape. The pie-slice equivalents let you do the same work in a circle rather than a bar, which some students grasp more easily. Denominators skip from twelve to sixteen (there's no thirteenth/fourteenth/fifteenth piece) because those are the denominators that actually show up in elementary and middle-school fraction work.
2D and 3D shapes (core set on Pro, full library on Platinum). A shape library you place and resize on the canvas. The core shapes are on Pro — square, rectangle, circle, equilateral triangle, straight lines and arrows, and the cube — which covers a lot of basic area, perimeter, and symmetry work. The fuller library is Platinum: most polygons (rhombus, parallelogram, trapezium, kite, pentagon, hexagon, octagon), the non-equilateral triangles (isosceles, scalene, right-angle), curved 2D shapes (oval, semicircle, sector), and almost all the 3D solids (cylinder, sphere, cone, prism, pyramid, cuboid, and the rest — the cube is the one 3D solid on Pro). So a Pro tutor can do basic shape work; a tutor who leans on the full geometry set, especially 3D solids for faces-edges-vertices, wants Platinum. On Platinum, a shape-labels toggle adds label slots at each shape's corners, edges, or tip that you type into — side lengths, angle measures, vertex letters — and the labels then move and rescale with the shape instead of drifting off it like a loose text box.

Each of these is a first-class object on the shared whiteboard, meaning the student can grab it, move it, and rearrange it in real time — the same as they'd move a physical rod on a desk. The teacher isn't demonstrating; the student is doing. That's the point. For deeper walk-throughs of the most-searched sub-workflows: teaching fractions online covers the Fraction Bars and Fraction Circles workflow, teaching number sense online with virtual manipulatives covers the Number Rods and Number Lines workflow, and teaching geometry online covers the 2D and 3D shape library, shape labels, geometry notation, and the GeoGebra cobrowser pairing pattern.

Solving the notation problem: a math symbols palette on the whiteboard

Real math needs symbols keyboards don't have. Koala Go's whiteboard ships a Math Symbols palette on paid tiers, sorted into the categories you actually reach for during a lesson. On Koala Pro you get the operations, comparison, and geometry set you'd use through elementary and middle school: the four operations and equals (+ − × ÷ =), the basic comparisons (< >), the fraction/ratio/percent group (/ : %), a square-root sign (√), a degree symbol (°), and round brackets. On Koala Platinum you additionally get the algebra, geometry, and powers set that shows up in later middle school and secondary work: not-equal and approximately-equal (≠ ≈), less-than-or-equal and greater-than-or-equal (≤ ≥), superscript powers (² ³ ⁿ), pi and infinity (π ∞), plus-or-minus (±), and the geometry markers (∠ △ ⊥ ∥). Square and curly brackets sit on Platinum as well.

Clicking a symbol on the palette drops it into whichever text box, sticky note, or new text box you'd rather type into — so you can build up "x² + 3x − 4 = 0" or "∠ABC = 90°" or "3/4 + 1/2 = ?" without stopping the lesson to look up an alt-code. That is the whole design goal of the palette: keep the tutor in the lesson, not in the search for a symbol. For the full walk-through — both places the palette lives, the six notation moves that carry most math tutoring, and what to do when a lesson needs something the palette can't write — see how do I write math equations and notation on an online whiteboard?

Honest limit: the palette is a symbols set, not a full equation editor. It won't stack a fraction bar horizontally the way LaTeX would, won't nest a radical inside another radical, and doesn't render live graphs. For tutors who work heavily in advanced algebra, calculus, or graph-heavy statistics, that gap is real, and the pattern most Koala tutors use is the palette for symbols plus the cobrowser (below) for a graphing tool like Desmos when a live graph is what the lesson needs. Where the whiteboard's math coverage sits in the broader landscape — versus dedicated whiteboard tools with equation editors and graph modes — is described in the best online whiteboards for tutoring comparison, which we don't repeat here.

Solving the practice problem: cobrowser and PDF upload

The practice problem — how the student actually does math during the lesson — has two well-worn solutions inside a virtual classroom, and Koala Go ships both.

PDF and PowerPoint upload onto the whiteboard (Pro tier). On Koala Pro, you upload a worksheet as a PDF or a slide deck and it becomes the lesson surface — both you and the student can annotate the same page in real time, using the same pen and shape tools you'd use on a blank board. That collapses the "here's a worksheet, we'll go through it together" workflow from a screenshare-and-narrate exercise into a real shared workspace. The uploaded material stays on the whiteboard between lessons, so next week's continuation opens with last week's marks still there.
Cobrowser for interactive practice sites. A cobrowser is a shared browser window inside the classroom — you open a page (Khan Academy, IXL, Desmos, ALEKS, GeoGebra, NRICH, a school district's own interactive maths portal) and both of you can click, type, and interact with the page live. Not screen sharing — actual shared control. That means a student doing a Desmos graph or a Khan Academy problem is doing it in the lesson room, and you can see exactly what they typed, why they got it wrong, and jump in without them switching windows. On Koala Free the cobrowser is capped at 10 minutes per session with a 20-minute cool-down; on Pro it's unlimited. See what is a cobrowser? for the mechanic in depth, and how do I give homework to online tutoring students? for the between-lesson-practice workflow that uses it.

The pattern most working online math tutors converge on is: manipulatives + notation + one uploaded worksheet for the shared work, plus the cobrowser for one specific practice tool the lesson needs. Not four different tabs, five bookmarks, and a screenshare — one classroom.

Keeping a younger math learner engaged past the fifteen-minute mark

K-2 and lower-elementary math has an engagement problem on top of the mechanical problems above. A five-year-old learning to bond numbers to ten doesn't have the attention budget for a webcam-and-screenshare setup, no matter how good the manipulatives are. This is where the engagement layer of a virtual classroom starts to matter — rewards for a completed activity, an avatar in a shared 3D space as a brain break between problems, points the student is playing for. On Koala Go, that layer runs through the Playground (a shared browser-based 3D world with avatars), the Gems reward system (teachers award 1–100 gems per student per action; students see them add up in real time and can spend them on avatar customisation), and streak/rewards on top. This is the same engagement layer described in depth in what is a gamified virtual classroom? — for a young-learner math lesson it does the same job it does for a young-learner phonics lesson, which is to give the student a reason to want to come back next week.

Two things worth naming honestly. First, the engagement layer helps most below age ten; above that, students want the maths and the manipulatives, not the avatar, and lean into the workflow-and-notation side of the classroom instead. Second, this is not a substitute for good math teaching — a well-run lesson without an engagement layer will out-perform a badly run lesson with one every time. The engagement layer is what makes the difference between "we lost her at minute twelve" and "she wanted to do one more."

What Koala Go actually is for math, specifically

Koala Go (the classroom we build) is a browser-based virtual classroom purpose-built for independent tutors and small teaching businesses. For a math tutor, here's what's on the platform right now, at which tier:

The shared whiteboard (Free tier and up) is the primary lesson surface. It ships the standard pen/shape/text/sticky-note controls and lets both teacher and student act on the same canvas.
PDF and PowerPoint upload onto the whiteboard is a Koala Pro feature ($25.99/month monthly, $21.99/month billed annually). Free-tier tutors can upload images but not PDF or PPT.
Math Tools panel inside the Whiteboard Library, with tier splits as described above: Number Rods, Number Lines, the core 2D/3D shapes (square, rectangle, circle, equilateral triangle, lines, cube), and the basic Math Symbols set on Pro; Fraction Bars, Fraction Circles, Number Roller, advanced Math Symbols, the full shape library (most polygons and the 3D solids), Shape Labels, red-and-blue rod fills, and the "make your own" number line on Platinum ($49.99/month monthly, $39.99/month billed annually). Enterprise and B2B tutors are provisioned at the Pro feature level, so the Platinum tools are hidden from their panel entirely rather than shown locked — an enterprise tutor doesn't get them without a plan change their administrator arranges.
Cobrowser for practice-site work is on every tier — capped at 10 minutes with a 20-minute cool-down on Free, unlimited on Pro.
Playground, avatars, and Gems — the engagement layer for younger math learners — is available across tiers; see the gamified virtual classroom explainer for how it fits together.
Scheduling, recording, and worldwide invoicing — the workflow layer that matters when a math tutor's practice grows past a handful of students — is Pro-and-up. See how to schedule online tutoring lessons and how to record online tutoring lessons for the specifics.

The Free tier is enough to try Koala Go with a student — the whiteboard, image uploads, a capped cobrowser session, and the Playground — but the Math Tools panel itself (Number Rods, Number Lines, Shapes, Fractions, Math Symbols) is gated behind Pro or Platinum, so Free is a trial surface rather than a math-tutoring day-to-day. Pro is where the tools most math tutors reach for daily land — Number Rods, Number Lines, the core shapes, the operations symbols, PDF upload, unlimited cobrowser. Platinum adds fractions, the algebra/geometry symbols, the full shape library (most polygons and the 3D solids), and the "make your own" number-line builder — the specific tools middle-school and early-secondary math work depends on.

Where Koala Go isn't the right pick for math

Two honest tradeoffs worth naming, so you know before you sign up:

Advanced STEM tutoring that lives on a full equation editor or live graphing. Koala Go's whiteboard math tools go deep on the elementary-and-middle-school side — manipulatives, notation, worksheets, practice sites. They stop short of a full LaTeX-style equation editor or a live graphing canvas of the kind a dedicated STEM whiteboard would ship. For late-secondary calculus or heavily graph-driven statistics tutoring where every lesson needs a live-plotted function, a dedicated STEM whiteboard (compared honestly in the best online whiteboards for tutoring) usually goes further on the whiteboard axis. What Koala tutors typically pair with the Math Tools palette for that gap is a cobrowser tab into Desmos or GeoGebra.
Whole-class delivery to a 25-student math class. Koala Go is calibrated for 1-on-1 and small-group tutoring — the free tier caps sessions at four students. It's not designed for a full school-class synchronous delivery of the kind Nearpod or Pear Deck are built for. For that shape of teaching, look at platforms built for whole-class lesson delivery.

Honest limits worth flagging that show up in real teaching regardless of subject: iPad works for lessons but is rougher than desktop for extended math work with heavy manipulatives; students in mainland China sometimes hit connection issues (we route through Hong Kong proxies but the Great Firewall is a moving target — see teaching students in mainland China for the workaround pattern); and the Playground itself is actively developed, with requested features (timers, group-select of Playground objects) still on the roadmap rather than shipped today.

By age band: what actually works

Different age bands lean into different parts of the toolkit. A rough guide, matching what most tutors converge on after a few months:

K–2 (ages 5–7): manipulatives + engagement layer. Number rods, number lines to twenty, pie-slice fractions, and the Playground / Gems reward layer between short activities. Sessions are 25–30 minutes with two or three activities inside them. Notation is minimal (mostly + − =); the manipulatives are what teach.
Grades 3–5 (ages 8–10): manipulatives shifting to notation. Rods and number lines for multiplication and division facts, fraction bars for equivalent fractions, the operations palette to start writing equations. Cobrowser walk-throughs of Khan Academy or IXL start earning their keep here. Engagement layer still matters, but sessions can push to 40–45 minutes with the right pacing.
Middle school (ages 11–13): notation-heavy, some manipulatives. The algebra symbol set (superscripts, ± , < ≤ > ≥, √, π, ∠), fraction bars for early ratio and proportion work, the number-roller for randomised practice problems, geometry shapes with labels. Cobrowser into Desmos for graph work or GeoGebra for angle work becomes standard. See how do I teach algebra online? for the pre-algebra through Algebra 1 workflow at this age band.
High school (ages 14–17) and adult learners: notation, practice, and worksheets. Uploaded PDF past-papers on the whiteboard, the algebra/geometry symbol set, cobrowser into subject-specific practice sites and graphing tools. Engagement layer is beside the point at this age band; the notation and workflow layers do the work. See how do I teach algebra online? for the Algebra 2 / A-level workflow, and how do I teach exam prep online? for the exam-focused version (SAT, ACT, A-level maths, IB, GRE).

Why teaching fractions online is a specific kind of hard

Every elementary and middle-school tutor has a war story about fractions. The gap online tutors hit is not that fractions are harder to teach remotely than reading is — plenty of tutors do both. It's that fractions have a stubborn physicality in the way they're first understood. Three problems show up together, and the online workaround has to answer all three, not just one.

Fractions are first understood as pieces of a physical thing. Ask a 7-year-old to divide a cookie among four people and they'll do it. Ask them what "1/4" means as a written symbol and they'll pause. The competence with cookies precedes the competence with numerals by months to years, and every good curriculum builds fraction understanding on that concrete substrate before it moves to symbols. Online, "the physical thing" is gone. If you replace it with a static shaded-pizza picture on a Google slide, you've traded a manipulative for a decoration — the student can see it, but they can't do anything with it.
Fraction notation lies to a novice. The written form "1/2" looks like two numbers with a line between them, and lots of children read it that way for a while — "one and two" rather than "one out of two". Compounding this, keyboards don't have a fraction bar; a fraction typed in a text box as "1/2" reads as "one over two" and looks like a division statement, not a fraction. If your whiteboard can't show a proper fraction with a numerator, a bar, and a denominator, the student's picture and the student's notation live in two different rooms.
Fraction operations demand comparison across shapes. "Which is bigger, 3/8 or 2/5?" is not a lookup problem — it's a visual comparison problem the student has to see. In person, you'd line up two fraction walls next to each other and ask the student to point at the wider one. On a video call, unless the student can grab the pieces themselves and lay them next to each other on a shared surface, the comparison becomes a "trust me, the tutor knows" exercise. The moment fractions get to operations — comparing, ordering, adding across different denominators, finding equivalents — passive-watching lessons stop working.

An online fractions workflow that actually works has an answer to each of the three: a shared drag-and-drop manipulative for the physical piece, a way to render proper notation next to the manipulative for the symbolic piece, and a way to compare pieces of different denominators side by side on the same canvas for the operations piece. Everything below is that answer, on the tools Koala Go ships.

The move: virtual fraction bars and fraction circles on the shared whiteboard

Fraction bars and fraction circles are the two manipulative shapes that carry most fraction teaching from about age 6 through the end of middle school. Koala Go's Whiteboard Library ships both, on the Platinum tier of the paid plans. Both live in the same Math Tools panel as number rods, number lines, 2D and 3D shapes, and the math-symbols palette. What matters, before any workflow talk, is how the primitive is shaped — because the shape is what lets it teach.

Each menu item places ONE piece — not a pre-divided whole. When you click "1/4" on the fraction-bar picker, you place a single 1/4 bar on the canvas. To build a whole from quarters you place four of them. This sounds like a footnote but it isn't: it's the entire pedagogical difference between "here is a picture of a circle cut into quarters" (which the student receives) and "here is a quarter, and another, and another, and another — do these together make one?" (which the student does). Both fraction bars and fraction circles work this way. That single design choice is what makes the tool a manipulative rather than a diagram.
The bar's width encodes its value. A 1/1 bar spans the whole width; a 1/2 bar is exactly half as wide; a 1/4 is a quarter, a 1/8 is an eighth, and so on down to 1/12, plus a 1/16 for finer work (the set jumps from 1/12 to 1/16 — there is no 1/13, 1/14, or 1/15, since those denominators effectively never appear in elementary-and-middle-school fraction work). Because the widths encode the values, you get a fraction wall for free — drop a 1/2 in one row, a pair of 1/4s in the next row, four 1/8s in the row below, and the visual equivalence stares back at the student. No labels required; the picture is the proof.
The circle slices snap-rotate so they tile into a whole. Every 1/n pie-slice has the same underlying diameter, so a slice from the 1/3 set and a slice from the 1/6 set have identical radii — they're built to sit inside the same circle. Rotating a slice snaps to its own natural angle (a 1/6 slice snaps in 60° increments; a 1/8 slice in 45°); you don't have to fiddle to line them up. That means "two thirds plus one third makes one whole" and "one half plus a quarter plus a quarter makes one whole" both compose visibly on the canvas the way they would with paper pie pieces on a desk, minus the paper pieces.
Each denominator has its own colour. Halves are one colour, thirds are another, quarters are a third — 8 colours cycle across the 13 denominators. The student ends up learning a colour association alongside the size association (which sometimes accelerates recognition — "the orange one is a third") without you having to teach the mnemonic. Tutors can recolour any placed piece via the standard whiteboard colour picker while the white "1/n" label on the piece stays white and legible.
The bar and the circle are both first-class shared objects. Either the tutor or the student can grab any placed piece and drag, rotate, or reposition it. Undo is the same undo as everywhere else on the canvas. That means when the student misplaces a piece, you don't have to reach in and fix it — they can. And when they build a wrong composition and you see the confusion on their face, you can hand them the pieces to try again. The instructional gesture is the same as it is in person.
All fraction pieces sit on the Platinum tier. If a tutor is on Koala Pro (currently $25.99/month monthly, $21.99/month annual), they can already access number rods, number lines, the core geometry shapes, and the base symbols palette — but the fraction bars and circles are Platinum only ($49.99/month monthly, $39.99/month annual). That's a genuine tier gate, and worth flagging up front rather than after the tutor has planned three lessons around bars they can't place.

For the umbrella context — how these tools fit alongside number rods, number lines, and the geometry set — see the pillar page how do I teach math online? This spoke focuses on the fraction-specific workflow.

A working sequence for teaching fractions online

The order below is the order almost every mainstream elementary curriculum uses in some form: build the concept of a fraction as parts of a whole, teach same-denominator comparison and operations, teach equivalence, then teach cross-denominator operations. What changes online is the tool at each step, not the order. This is a template, not a script; adapt the ages and denominators to the student in front of you. Before you plan which step to start at, a short first-session diagnostic tells you whether the student has the whole-number sense and unit-fraction intuition this sequence assumes — see how do I assess a new online math tutoring student? for the first-lesson workflow, including how each of the fraction pieces below doubles as a diagnostic instrument.

Unit fractions from a whole (introduction). Place a single 1/1 bar (or a single 1/1 circle) on the canvas. Ask the student "how many 1/2 pieces do you think will fit exactly on top of this?" — let them predict, then let them drag two 1/2 pieces on top of the 1/1 to check. Repeat with 1/4, 1/3, 1/6. This is the moment the abstract concept of "one whole" gets a physical size the student can point at, and every subsequent lesson refers back to it. Twenty minutes of this on day one earns weeks of clarity later.
Same-denominator addition and subtraction. Line up three 1/4 bars in a row. Ask the student to write the addition sentence next to it: "1/4 + 1/4 + 1/4 = ?". Have them place the answer as another single bar (they'll reach for 3/4 — since a 3/4 bar isn't in the set, they'll compose it as three quarters). This is where you first cash out the earlier work: the picture makes the answer obvious, and the student writes the symbol beside it. Repeat for subtraction — start with four 1/4 pieces, remove one, count what's left.
Comparing fractions with the same numerator. Drop a 1/2, a 1/3, a 1/4, a 1/6, and a 1/8 in a column stack, aligned to the left edge. The picture is unambiguous — the 1/2 is the widest, the 1/8 is the narrowest — and it's counter-intuitive to a novice ("but 8 is bigger than 2, isn't it?"). Watching the widths get shorter as the denominator grows is one of the most productive minutes in a fraction curriculum. The student sees, in one look, that a bigger denominator means smaller pieces.
Equivalent fractions. Place a 1/2 bar. Under it, stack two 1/4 bars aligned to its left edge. The right edges match — the picture proves 1/2 = 2/4 without a rule. Add a row of four 1/8 bars underneath: 1/2 = 2/4 = 4/8. Do the equivalent in circles: a 1/2 slice occupies the same wedge as two 1/4 slices or four 1/8 slices. This is often the concept that separates students who "get" fractions from students who don't; getting it visual first before it becomes an algorithm ("multiply top and bottom by the same number") is worth every minute you spend.
Fractions on a number line. Place a Koala 0-10 number line on the canvas (Number Lines are on the Pro tier — every paying tutor has them). Explain that between any two whole numbers, you can drop in fraction points too. Between 0 and 1, drop a 1/2 tick with a labelled sticky note; between 0 and 1 again, drop 1/4, 2/4, 3/4 marks. This is the move from "fractions as parts of a shape" to "fractions as numbers" — a shift that trips up a lot of students, and one the online canvas actually helps with because you can freely annotate on top of a number line without erasing anything. For denser scales, Platinum tutors get a "make your own" number line builder that lets you set a custom range and step (say, 0 to 1 in tenths) — the Pro presets don't cover this specifically, so an honest note: on Pro you'll place a blank line and label your own tick marks with sticky notes; on Platinum you can generate the custom line directly.
Cross-denominator comparison and operations. "Which is bigger, 2/3 or 3/5?" — place both, side by side, on the canvas. Ask the student to build each: two 1/3 pieces stacked one above the other, three 1/5 pieces beside them. They'll see the answer. Then walk them through the algorithmic version — finding a common denominator (2/3 = 10/15; 3/5 = 9/15) — with the picture still visible. The visual is the check on the algorithm; the algorithm is the shortcut for the visual. Neither replaces the other. For adding across denominators, the same pattern: build both, find the common-denominator equivalents (2/3 + 1/4 → 8/12 + 3/12 → 11/12), and let the student write the notation on the whiteboard next to the picture.
Improper fractions and mixed numbers. Place five 1/4 bars in a row — they overflow one whole into a fifth. Explain that this is 5/4, and also 1 and 1/4. Do the same with circles: place five 1/4 slices; they fill one whole circle plus a quarter of a second. The two written forms live side by side on the canvas until the student is fluent flipping between them.
Multiplication and division of fractions (later work). "1/2 of 1/3" — place a 1/3 slice on the circle, then place a 1/2 marker across just that slice, showing that the shaded intersection is 1/6. The picture-first approach here saves a lot of "why does multiplying make it smaller?" confusion. Fraction multiplication depends on whole-number multiplication being reliable underneath it — a student who isn't fluent with 6 × 4 stalls the moment "3/4 × 8" needs to hit 24/4 → 6 — so if the times-table foundation is shaky, the multiplication spoke at how do I teach multiplication and times tables online? is the workflow to run in parallel with (or ahead of) fraction multiplication. Division is harder to visualise cleanly on a manipulative and often benefits from switching to the algorithmic explanation earlier — one of several places where the Koala Go whiteboard's honest limit shows: it doesn't render a full symbolic fraction equation with a stacked horizontal bar, so for algebraic fraction work the whiteboard becomes a scratch surface rather than a rendered equation editor. If your caseload runs deep into algebraic fractions or rational expressions, a dedicated STEM whiteboard often goes further — see the best online whiteboards for tutoring for the honest comparison, and how do I teach algebra online? for the pedagogy workflow when a student's fraction work has crossed into algebra.

Most sessions won't move through all eight steps in one lesson; a typical 40-minute fractions session picks two adjacent steps and works them slowly with a variety of denominators, then does 5-10 minutes of practice on a site (see the cobrowser section below).

One more sequence a fractions caseload will run into repeatedly: fraction word problems. The moment "1/2 + 1/4" arrives in a story — "Sam ate half a pizza and Alex ate a quarter, how much of the pizza is left?" — the piece a student is missing is usually not the fraction arithmetic but the translation from the English into the diagram. Bar models (the Singapore-maths / tape-diagram approach) are the standard move for that translation, and they lean on the same fraction-bar primitive above: place a 1/1 bar to represent the whole pizza, place a 1/2 and a 1/4 on top of it, and the "how much is left" answer becomes a shape on the canvas before it becomes a subtraction. The full workflow — the five bar-model shapes, the six failure modes that derail the translation (the student who draws the bar correctly and still writes the wrong equation, the student who stretches every bar to the same length regardless of the numbers), and the hand-off point when the natural next step is "let x = the unknown" — is covered on the sibling spoke how do I teach math word problems online?

Writing the notation next to the picture

Fractions taught only with manipulatives leave the student able to point at the answer but not write it. Fractions taught only with numerals leave the student able to compute the answer but not understand why it works. The place both meet is the same shared canvas — the whiteboard — with the manipulative on one side and the written expression on the other.

The Math Symbols palette on Koala Go's whiteboard handles the common fraction-lesson notation without you having to hunt for alt-codes. On Koala Pro you get the operations (+, −, ×, ÷, =), the comparisons (<, >), the fraction/ratio/percent group (/, :, %), and the round brackets — everything an elementary or early-middle-school fractions lesson typically needs. On Koala Platinum you additionally get the not-equal and approximate signs (≠, ≈), the ≤ and ≥ comparisons, the superscript powers (² ³ ⁿ), π, ±, and the geometry markers; these come in when the same student's fraction work starts touching pre-algebra or advanced middle-school problems. For a fuller walk-through of the palette itself — the two places it lives, the seven symbol categories, keyboard-vs-palette tradeoffs, and the six notation moves that carry most math tutoring — see how do I write math equations and notation on an online whiteboard?

Clicking a symbol drops it into whichever text box or sticky note you'd rather type into, so you can build "1/2 + 1/4 = 3/4" or "2/3 > 3/5" next to the corresponding picture in real time. Two practical patterns that work well:

Picture on the left, sentence on the right. Build the manipulative composition, then in a text box next to it write the fraction sentence it represents. Let the student write the sentence themselves whenever possible — the act of writing is what cements the symbol.
Sentence first, then verify with the picture. For older students already fluent with algorithmic fraction operations, write the sentence first ("3/4 − 1/2 = ?"), let them compute, then build the picture to check. The manipulative acts as a verification tool for the algorithm, which is often the missing piece for students who know the mechanics but don't trust them.

Honest limit worth naming: the palette is a symbols set, not a full equation editor. It won't render a stacked fraction with a horizontal bar the way LaTeX does (the on-canvas notation is inline: "1/2", not the two-line form), and it doesn't handle nested radicals or a live graph. For elementary and middle-school fraction work that inline form is what students see in most textbooks anyway; for late-secondary algebraic fraction work, see the whiteboards comparison linked above.

Practice, homework, and the fraction sites you already use

Manipulatives and notation carry the teaching, but consolidation runs on practice — and most working online tutors have a small set of external fraction practice sites they already trust. Bringing those into the lesson rather than sending them as a link after the lesson is where the cobrowser earns its keep. A cobrowser (explained in depth in what is a cobrowser?) is a shared browser window inside the classroom: you open a page — Khan Academy's fraction unit, an IXL fraction module, a Desmos fraction activity, a NRICH puzzle, ALEKS, a school district's own interactive maths portal — and both you and the student can click, type, drag on the same live page. Not screen sharing; actual shared control. Two workflow patterns that work well for fractions:

Manipulative first, practice second. Ten minutes on the whiteboard with fraction bars building the concept, then ten minutes on a fraction practice site walking through problems together. The student does the practice problems in the lesson room, on the live site — you see exactly what they typed, why they got it wrong, and where they hesitated, and you can jump in without them switching windows.
Practice first, manipulative to debug. Start on a fraction practice site with a set of problems. When the student gets one wrong or freezes, pause the site and switch back to the whiteboard, build the picture of the specific problem with fraction pieces, and re-solve it together. Then return to the site and finish the set. This is one of the strongest teaching moves an online fractions tutor has — the debug loop between the abstract problem and the concrete picture, all inside the same tab.

Two honest notes: on Koala Free, cobrowser sessions are capped at 10 minutes per session with a 20-minute cool-down between sessions — fine for a single practice segment in a 45-minute lesson, not enough to run a whole practice-site-based lesson. Full-length cobrowser sessions come with Koala Pro. And for between-lesson practice, upload the worksheet PDF the student is working on for homework (PDF and PowerPoint upload is on the Pro tier) so both of you can annotate on top of it at the start of next week's lesson — the workflow is covered in more depth in how do I give homework to online tutoring students?

Keeping a younger fractions learner in the room

Fractions come in at exactly the age band where sustained attention on a screen is still developing. A typical Year 3 or Year 4 student (age 7-9) doesn't have the attention budget to sit through a 45-minute abstract-symbol lesson — even a strong one. If the tutoring is on-screen, the pacing has to be short-drill / short-break / short-drill, and the motivation layer has to be visible enough to feel real.

On Koala Go, the pieces that work together for the younger cohort:

Doing over watching. Every fraction-teaching move above ends in the student's cursor on the shared canvas — dragging a piece, writing the sentence, ticking a mark. Passive watching is the failure mode; drag-and-drop is the fix. If you catch yourself narrating for more than a couple of minutes at a stretch, hand the student a piece.
The Playground as the between-drill reset. The Koala Go Playground is a shared 3D space with avatars, sitting in a separate tab of the same classroom. Not a teaching surface — a reset. A 1-2 minute walk to a landmark and back, avatar to avatar, is a lower-friction attention reset than "go get a snack" or "let's watch a video," because the student doesn't leave the tab. This is the same lever the gamified virtual classroom page describes for general engagement work.
Gems as immediate-effort reinforcement. Each student in the room has their own visible gem balance. From the student panel, you give 1-100 gems in a single click — a rhythm of small frequent rewards works better than a single burst at the end. Give a gem for correctly composing 3/4 from three quarters; give five for persisting through the harder "2/3 versus 3/5" comparison even before you check the answer. The reinforcement is at the moment the effort happened, not at the end of the lesson, and the student sees the balance grow on their side of the screen. For older students who read this as infantilising, drop it — the mechanic is a tool, not a requirement.
Session length matched to the student. Younger fraction learners tend to hold shorter working segments before they need a break, and older students can sit with fewer, longer stretches — match the number and length of activities to the student in front of you rather than to a fixed clock. For the general shape of structuring a 1-on-1 online lesson, see how do I structure a 1-on-1 online lesson? — the pacing there transfers to fractions specifically.

A note on students who need extra scaffolding: some students genuinely can't stabilise a fraction picture on the first, second, or third exposure — the manipulative helps, but they still need more repetitions with smaller denominators and more colour-coded cues than most curriculums assume. This isn't a diagnosis, and this page is not the place to explore it clinically. The workflow above holds; slow it down, drop to halves and quarters only for a few sessions before introducing thirds, and use the Number Rods (also on the Whiteboard Library, on the Pro tier) for the parallel "how do numbers combine into other numbers?" work — that concrete-quantity work often supports fraction understanding indirectly, and the full rods-and-number-lines workflow is covered in how do I teach number sense online with virtual manipulatives? If a specific student's difficulties are consistent enough that you're wondering about a specific learning difference like dyscalculia, the sibling SEN spoke how do I run online tutoring lessons for students with dyscalculia? walks through the workflow overlay tutors use — concrete-before-pictorial-before-abstract sequencing, repetition-with-variety instead of volume drill, and why timed drills and "beat the clock" mechanics backfire specifically for a student with a math learning difference — while being clear that a diagnosis and any formal intervention plan belong with the family and a qualified specialist, not with a webpage.

Communicating fraction progress to the parent

Fraction lessons don't feel like they're working to the parent — the child is on a webcam, the tutor is talking, the pieces are on a screen the parent isn't looking at. The parent's read on whether tutoring is helping runs on what you send them between sessions. For fractions specifically, the highest-signal update is a screenshot of the composition the student built (with the written sentence next to it) plus one line naming what they got that they didn't get last week. "We worked on comparing fractions with different denominators today. Emma correctly ordered 2/3, 3/4, and 5/8 using the fraction bars — that's the piece that felt shaky last week, and it's landed. Home practice this week: five ordering problems on Khan Academy fractions unit 2." That paragraph is worth more than any generic "great lesson today!" note.

A few concrete moves:

Screenshot the canvas each week. The whiteboard saves state between sessions, so at the end of each lesson you have a visible record of what the student built. Grab a screenshot of the fraction composition or comparison you worked on and drop it in the parent update. A picture of "the student built this and understood it" reads to a parent as evidence in a way that written prose can't.
Record short clips of key moments. Koala Pro lets you record locally; Koala Platinum adds cloud recordings of scheduled lessons. Either way, a 2-3 minute clip of the moment the student first placed three 1/4 pieces on top of a 3/4 bar and said "oh, it fits" is worth every second of the recording setup. Consent, storage, and sharing are covered in how do I record online tutoring lessons? — get the family's written permission before you record anything with a minor.
Send a specific weekly practice ask. Not "please practise fractions this week" — that's an unclear task that a parent can't verify. Instead: "please have Emma do the ten-question fractions quiz on Khan Academy Unit 3 Lesson 2 twice this week, and let us know which ones she got wrong." The general parent-comms cadence is covered in how do I share progress updates with parents?; the fractions-specific ask is that concrete.

Common failure modes and honest limits

A short list of things that go wrong in online fraction lessons, and the fix — most of these are the pattern-recognitions you build in your first month teaching this online:

Trying to teach fractions purely with a shaded diagram on a Google slide. The diagram is a picture; a manipulative is a picture the student can move. Even a beautifully-designed static fraction diagram is a step down from a set of drag-and-drop pieces that fit together. If your platform doesn't ship the pieces, hand-drawing them on a whiteboard tool doesn't recover the loss — the mechanical property that makes a manipulative teach (that the child physically builds the whole from parts) requires the pieces to be discrete objects.
Introducing too many denominators too fast. Halves, quarters, and thirds first — spend one or two sessions before you show a 1/8 or a 1/6. Younger students overload on the palette and stop treating the piece as a quantity; they start treating it as a colour to be memorised. Start narrow and add.
Jumping from "1/2 = 2/4" to "multiply top and bottom by 2" as a rule before the picture is stable. The rule is fine as a shortcut later. As a teaching move, it strips the "why" and leaves the student with an algorithm they don't trust. Build 1/2 = 2/4 = 4/8 = 8/16 as a stack of fraction bars first; teach the rule as the shortcut for the pattern the student now sees.
Doing all the fraction work in pie-circles and skipping the bars. Pie-circles are intuitive for "parts of a whole" and unfamiliar for "fractions as numbers on a line." The bar form transfers more cleanly into fractions-on-a-number-line, adding across denominators, and later ratio work. Alternate the two forms — the same 1/4 sits in both, and the student should recognise it in both.
Sending the student to a fraction practice site alone and asking them to "do 10 problems." The student will get three wrong, not understand why, and arrive at next week's lesson worse than they left this one. Practice sites work best inside the lesson (via the cobrowser) with the tutor available to interrupt on the first wrong answer, or with a very small (2-3 problem) between-lesson ask the parent will actually oversee.
iPad-only workflows. Koala Go works on iPad, but the drag precision required for placing narrow 1/12 or 1/16 bars against a common baseline is noticeably fiddlier on a finger-driven tablet than on a desktop or laptop with a mouse. If a student uses iPad exclusively, plan around wider denominators and use the Text tool more for the notation piece. For the general iPad/tablet caveats, see the pillar's honest-limits section.

And two limits worth naming plainly rather than papering over:

The whiteboard doesn't render a fully stacked fraction expression. Inline "1/2" is what you get on the canvas; if you need a two-line rendered fraction with a horizontal bar (say, for advanced algebra where the whole layout matters), the whiteboard is a scratch surface, not a formatted equation editor. For heavy algebraic-fractions work, a dedicated STEM whiteboard usually goes further.
The Fraction Bars and Fraction Circles are Platinum-tier. Pro-tier tutors have Number Rods, Number Lines, geometry shapes, and the base symbols palette, but not the fraction manipulatives specifically. If fractions are a large share of your caseload, factor that into the plan choice.

Practical setup on Koala Go for a fractions-heavy caseload

If you're evaluating Koala Go for elementary and middle-school fractions tutoring specifically, here's the practical picture and where the current lines sit:

Whiteboard is the workspace. The classroom's whiteboard is a Fabric.js canvas with the Math Tools panel: Fraction Bars and Fraction Circles for denominators 1/1 through 1/12 and 1/16 (Platinum), Number Rods for the parallel number-sense work (Pro), Number Lines including 0-10 and 0-20 for fractions-on-a-line (Pro), 2D and 3D geometry shapes for the shape-of-a-fraction visual (core set Pro; full library Platinum), and the Math Symbols palette for writing the notation (Pro base, Platinum for the fuller algebra/geometry set).
Upload existing fraction worksheets. On Pro, upload the student's fraction worksheet as a PDF or PowerPoint and annotate on top of it — the worksheet becomes the shared surface, no worse than working on the printed sheet in person. Free-tier upload caps apply.
Cobrowser for practice sites. Open Khan Academy, IXL, Desmos, ALEKS, GeoGebra, NRICH, or whatever fraction-practice tool your programme uses; the student clicks and types on the live site inside the lesson room. 10 minutes at a time on Free (with a 20-minute cooldown), unlimited on Pro.
Playground and Gems for the engagement layer. Younger fraction learners (age 7-9) especially benefit from the Playground brain break between drill segments and per-student Gems for immediate effort-based reinforcement. For older learners, both are optional and can be turned down or off.
Recording for parent review. Local recording is included on Pro; cloud recordings of scheduled lessons come with Platinum. Either supports the "send the parent a 3-minute clip of the fraction breakthrough" pattern above.
Free tier limits worth knowing. Koala Free caps sessions at four students, no per-lesson time limit; PDF upload, cobrowser length, and recording are the main tier steps up from Free.

Honest limits worth flagging that show up in real fractions work regardless of platform: iPad is rougher than desktop for narrow-denominator precision; students in mainland China sometimes hit connection issues (we route through Hong Kong proxies but the Great Firewall shifts — see teaching students in mainland China for the workaround pattern); and secondary-school algebraic-fraction work runs up against the whiteboard-vs-equation-editor limit named above.

Why teaching number sense online is a specific kind of hard

Every K-2 and lower-elementary tutor has run into it: the concepts you're trying to build in the earliest years of math are almost entirely quantity-based, and quantity is stubbornly physical. A kindergartner who counts a pile of six blocks and then a pile of eight blocks and points at the eight pile as "bigger" is doing the work that later abstract algebra depends on — but the physicality is where the learning is. On a video call, that physicality either lives on the tutor's side of the screen (where the student watches, not does) or gets replaced with a static picture (which is a picture, not a manipulative). Three problems show up together for online number-sense teaching, and the workaround has to answer all three.

Quantities need to be compared, not just seen. "5 is less than 8" isn't a fact to memorise; it's a comparison a child needs to make with two piles that are actually in front of them, until the abstract comparison feels obvious. A picture of "5" next to a picture of "8" on a slide doesn't do this — both pictures occupy the same rectangle of screen real estate. You need pieces whose on-screen size encodes the quantity, so a 3-piece is visibly shorter than a 5-piece and it's not a fact to be told but a fact to be seen.
Composition of number (3 + 2 = 5) is a physical act before it's a symbolic one. The moment a child slides two rods together against a third and says "look, they match" is qualitatively different from being told "three plus two equals five". The first builds a mental model that transfers to every later operation; the second gets memorised and forgotten. Online, without pieces to physically compose, the tutor either does the composition on the student's behalf (passive) or asks the student to imagine it (asking too much of a grade 1 student's working memory).
The number line is the mental model for later math — but it's abstract on paper and impossible on a webcam. Every later concept a student will meet — negative numbers, fractions, decimals, coordinates, real-line intervals in calculus — rides on the number-line model built in early elementary. That mental model develops when a child points at a spot on a physical line, jumps in threes from 0 to 30, or drops a label on where "12" belongs. A number line drawn statically on a Google slide the student can't annotate does the opposite of what the manipulative should do: it makes the line another thing to be memorised rather than a thing to be used.

An online number-sense workflow that works has an answer to each of the three: a drag-and-drop set of quantity manipulatives where the on-screen size encodes the value, a shared canvas the student can compose on themselves, and an annotatable number line both tutor and student can mark up in real time. The rest of this page is that answer, on the tools Koala Go ships.

The two manipulatives that carry most number-sense teaching online

Number rods and number lines together carry the majority of K-4 number-sense teaching. Both live in the Whiteboard Library's Math Tools panel on Koala Go, both are on the Pro tier of paid plans (any working paid tutor has them without a Platinum upgrade), and both are first-class shared objects — either the tutor or the student can pick one up, move it, rotate it, and rearrange it on the canvas.

Cuisenaire-style number rods

The rods are ten coloured rods numbered from one unit through ten units long, in the Cuisenaire model — the pedagogical shape a Belgian teacher named Georges Cuisenaire developed in the mid-20th century and Caleb Gattegno popularised in classrooms worldwide. Koala's implementation uses its own colour palette (not the canonical Cuisenaire palette), but the teaching mechanic is the same: a One-rod is one unit long, a Ten-rod is ten units long, and the sizes line up cleanly against each other. Under a One-rod against a Ten-rod, the child sees, at a glance, that the whole is ten of the smallest pieces. Slide a Three-rod and a Seven-rod together against a Ten-rod, and the number bond 3 + 7 = 10 stops being a fact to memorise and becomes a fact the child watched line up. That is the entire pedagogical point.

What you can do with them on the shared canvas:

Pick from any rod value 1 through 10. The rod picker in the Math Tools panel shows all ten rods; clicking a value arms it as a placement tool, then each click on the canvas drops one of that rod value. Both you and the student can place rods; either of you can drag, rotate, or delete any placed rod.
Recolour any placed rod. Rods have a default colour per value, but you can recolour any placed rod via the standard whiteboard colour picker — useful when you want the same colour to mean the same thing across an activity (say, red for numbers greater than five, green for less). The dividers between the unit-squares track the fill colour automatically.
Use Platinum extras for higher-mileage number-rod work. Koala Platinum ($49.99/month monthly, $39.99/month annual) adds a red-and-blue alternating fill (a per-unit checkerboard, so a Five-rod becomes red-blue-red-blue-red — the classic "count the units" visual), a vertical placement mode (the same rod placed as a bar-chart column rather than a horizontal row, useful for making rods do double duty as bar charts), and a one-click "Add All" that drops a staircase of all ten rods on the canvas in order — the fastest reset for the start of a new number-sense activity. These three are Platinum-only: a tutor on Koala Pro sees the toggles in the panel but locked behind an upgrade prompt, and enterprise and B2B plans are provisioned at the Pro feature level, so they don't include these extras either. The base rod set below is unaffected on every paid plan.

The rods are on the Pro tier ($25.99/month monthly, $21.99/month annual, or included in a business plan), so any working paid tutor already has access to the base rod set. Free-tier tutors see the rods in the panel but locked behind an upgrade nudge.

Number lines

Number lines are the second-most-used manipulative in early numeracy — arguably the single most durable mental model in mathematics, because it's the shape that later fractions, decimals, negatives, and coordinates all sit inside. Koala Go's Whiteboard Library ships eight preset number lines on the Pro tier, each rendered as a fabric horizontal shaft with arrow-heads on either end, tick marks at each labelled value, and text labels underneath. The full preset set:

0 to 10 (step 1). The base number line for early counting, one-to-one correspondence, and cardinality. A student "walks" a marker one tick at a time up to 10. First line to place for a kindergarten or grade 1 student.
0 to 20 (step 1). The next stretch — teens are a stumbling block in many curricula because eleven and twelve don't follow the "ten-and-something" pattern of thirteen through nineteen. The 0-20 line lets you walk through them concretely.
0 to 30 (step 1). Extends to skip-counting territory: students can count in twos, threes, or fives up to 30 with the tick marks as anchors.
0 to 100 by 5. The step-5 line teaches skip-counting the fives explicitly. Useful for counting money (5c, 10c, 15c…), telling time on a clock face (5, 10, 15… minutes), and any early multiplication table work built off the fives.
0 to 100 by 10. The step-10 line is where place value starts to land — the tens as the "big" jumps, with ones as the smaller moves in between them. Also the natural shape for percent work later.
-10 to 0 (step 1). An introduction to negative numbers as "less than nothing" — a line that goes to the left of zero. In US curricula students meet negative numbers formally around grade 6, though plenty of tutors introduce the left-of-zero idea informally well before that.
-10 to 10 (step 1). The full symmetric line either side of zero. Where addition and subtraction with negatives first make visual sense: "-3 + 5" is a jump of five to the right from -3, landing on 2.
Blank (11 ticks, no labels). For times you want the student to label the ticks themselves — pick a range together, put "0" at one end and, say, "60" at the other, and let the student figure out what each tick has to be.

Every preset line is a resizable, movable object on the whiteboard: corner-drag scales the whole line (shaft, ticks, labels together, so proportions stay right); you can duplicate one, place two side by side, or stack them for comparison. Both you and the student can annotate directly on top of any line — drop pen strokes to mark a value, place stickies to label a spot, put a Koala Math Symbol at a point. That's the whole point: the line is a shared surface, not a picture the student is watching over your shoulder.

For the times a preset doesn't fit — a 0 to 1 line in tenths for decimal work, a 0 to 360 line in thirties for angle work, a 0 to 200 for a specific counting activity — Koala Platinum unlocks a "make your own" custom builder. You set a start, an end, and a step size, and the whiteboard renders a matching line to those specs. On Pro, the workaround is to place the blank line (which gives you an unlabelled 11-tick line) and label the ticks yourself with sticky notes. That works for most classroom needs; the Platinum builder is what you want when the range or step is going to change often.

A working sequence for teaching number sense online

Every mainstream elementary curriculum builds number sense in roughly the same order: count, subitize, compose and decompose to ten, place value in tens, place value beyond a hundred, skip-count, and eventually operate on the negative side of zero. What changes online isn't the order — it's the tool you reach for at each stage. This is a template, not a script. The grade bands below are typical US placements, not a standards mapping — adapt them to the curriculum and the student in front of you. Before you pick which step to start a new student at, a short first-session diagnostic tells you where in this sequence the student's number sense actually sits — see how do I assess a new online math tutoring student? for the first-lesson workflow, including how the rods and number lines below double as diagnostic instruments before they become teaching ones.

Counting and one-to-one correspondence (pre-K to kindergarten). Place a 0-10 number line and a set of One-rods. Ask the student to place a One-rod under each tick from 0 to 5, saying the number as they go. Then remove a rod and re-count — the picture reinforces that "the last number you say is how many there are" (cardinality), which sounds obvious but is a concept that takes time to stick.
Subitizing small quantities (pre-K to kindergarten). Drop a Three-rod on the canvas. Ask the student, without counting the units, "how much is that?" — subitizing is the skill of recognising a small quantity at a glance, and the rod's shape helps: a Three-rod is visibly shorter than a Five, visibly longer than a Two. Repeat quickly with a variety of small rods. This is a game the student wins by not counting.
Number bonds to ten (kindergarten to grade 1). Place a Ten-rod at the top of the canvas. Underneath, ask the student to build a matching length using two rods — 3 and 7? 4 and 6? 5 and 5? Then remove one of the pair and ask "what's missing?" This is the single most important early-numeracy activity in most curricula, because the number bonds to ten become the foundation for later mental arithmetic. The rod version makes it a puzzle rather than a memorisation drill. Do the same with number bonds to 5, and later to 20.
Addition and subtraction as combining and separating (grades 1 to 2). Place a Three-rod and a Four-rod end to end. Ask the student what one rod would exactly match their combined length — they'll reach for a Seven. Write "3 + 4 = 7" in a text box next to the picture (see the notation section below for the symbol palette). Then reverse: place a Seven-rod, then a Three-rod on top of one end, ask what's left uncovered — a Four. Write "7 − 3 = 4". The picture and the sentence live side by side on the canvas.
Place value with the tens-line and rods (grades 1 to 2). Place a 0-100 by 10 number line. Ask the student to build "twenty-three" using rods: two Tens for the twenty, then a Three for the units. Have them position the composition next to the "23" tick if you drop it in. Repeat with several two-digit numbers. This is the moment place value stops being an abstraction ("the 2 in 23 means twenty") and starts being a physical picture ("two whole Tens plus a Three").
Skip counting on the number line (grades 2 to 3). Place a 0-30 line. Ask the student to hop in threes from 0 — pen a small arc from 0 to 3, 3 to 6, 6 to 9, and so on. This is preparation for the times tables that will follow: skip-counting the threes on the line is what turns "3 × 4 = 12" from a memorised fact into a picture ("four hops of three lands on twelve"). Do the same with fives on the 0-100 by 5 line for the five times table. Once the skip-counting is fluent, the sibling spoke how do I teach multiplication and times tables online? picks the workflow up from here — the arrays-with-rods pattern for the harder tables, the Number Roller for randomised fluency drill on Platinum, and what to do when a specific table (usually the sixes, sevens, or eights) refuses to stick.
Negative numbers on a symmetric line (grade 6, though many tutors introduce the left-of-zero idea informally much earlier). Place a -10 to 10 line. Introduce negative numbers as "less than zero" — the tick marks to the left of zero. Ask the student to jump from -3 to 5 — how many steps? (Eight.) From 4 to -2? (Six steps back.) The physical hopping across zero is the mental model that stops later negative-number arithmetic from becoming a bag of confused rules. Integer arithmetic on the symmetric line is also the foundation for signed-number algebra a year or two later; when the student is ready to introduce a letter for an unknown quantity, the sibling spoke how do I teach algebra online? walks through the balance-scale and area-model workflow for pre-algebra and Algebra 1 that builds on this number-line intuition.
Bridging into fractions on a number line (grades 3 to 4). Between any two whole numbers on a number line sit an infinite number of fractional points. Place a 0-10 line, drop a marker halfway between 0 and 1, and label it "1/2". Do the same for 1/4 and 3/4. This is the natural handoff to the fractions spoke — see how do I teach fractions online? for the full workflow with the Fraction Bars and Fraction Circles that carry that teaching from here.

Most sessions won't move through all eight steps in one lesson; a typical 25-minute early-numeracy session picks one or two adjacent steps, cycles between rods and a matching number line, and includes a couple of engagement breaks. The tools are the same tools; the sequencing is what makes the lesson.

Writing the notation next to the picture

Number sense taught only with rods leaves the student able to compose "3 + 4" but not to write it. Number sense taught only with numerals leaves the student able to compute "3 + 4 = 7" but not to picture it. The place both meet is the same shared canvas — the whiteboard — with the manipulative on one side and the written sentence on the other.

The Math Symbols palette on Koala Go's whiteboard handles the common early-numeracy notation without you hunting for alt-codes. On Koala Pro you get everything an early-numeracy lesson needs and then some: the four operations and equals (+ − × ÷ =), the comparisons (< >), the fraction/ratio/percent group (/ : %), a square-root sign (√), a degree symbol (°), and round brackets. Koala Platinum adds the algebra and geometry set — ≠ ≈ ≤ ≥ ² ³ ⁿ π ± ∞ ∠ △ ⊥ ∥, plus square and curly brackets — that shows up in later middle school and secondary work, but Pro's symbols cover the whole K-4 range without gaps. For a fuller walk-through of the palette itself — the two places it lives inside the classroom, the seven symbol categories, the six notation moves that carry most math tutoring, and the honest limits (no stacked fractions, no live-rendered LaTeX) — see the sibling spoke how do I write math equations and notation on an online whiteboard?

Clicking a symbol drops it into a text box or sticky note, so you can build "3 + 4 = 7" or "5 < 8" next to the corresponding rod composition or number-line hop in real time. Two workflow patterns that work:

Picture first, sentence second. Build the rod composition, then in a text box next to it write the number sentence it represents. Let the student write the sentence themselves whenever their handwriting or typing allows — the act of writing is what turns the picture into a formal fact.
Sentence first, picture as verification. For students already fluent with simple arithmetic, write the sentence first ("7 − 4 = ?"), let them compute, then build the picture to check. The manipulative is the check on the algorithm — a useful move for students who know the mechanics but don't trust them.

Practice, homework, and the number-sense sites you already use

Manipulatives and notation carry the teaching; consolidation runs on practice. Most working online early-numeracy tutors have a small set of external number-sense practice sites they trust — Toy Theater's counters and rekenreks, NRICH's early-number activities, Math Playground's early-elementary game set, Khan Academy Kids for full curricula, ABCya for gamified drill. Bringing these into the lesson rather than sending them as a link after is where the cobrowser earns its keep. A cobrowser (explained in depth in what is a cobrowser?) is a shared browser window inside the classroom: you open the site, and both you and the student can click, type, drag on the live page — not screen sharing, actual shared control. Two number-sense-specific patterns that work:

Rods on the whiteboard, then a matching game on a practice site. Ten minutes on the whiteboard building number bonds to ten with rods, then ten minutes in the cobrowser on a matching-pairs game where the student clicks pairs that add to ten. The manipulative built the concept; the game consolidates it in a different modality without leaving the lesson room.
Practice-site question first, rods to debug. Open a number-sense question in the cobrowser. If the student stalls or gets it wrong, pause the site and switch back to the whiteboard — build the specific problem with rods, work it out together, and then return to the site. The debug loop between the abstract problem and the concrete picture is one of the strongest teaching moves an online early-numeracy tutor has.

Two honest notes on the cobrowser: on Koala Free, sessions are capped at 10 minutes per session with a 20-minute cool-down — fine for a single practice segment in a 25-30 minute young-learner lesson, but restrictive if you want the practice site as the primary surface. Full-length cobrowser access is Koala Pro. For between-lesson practice, you can upload the child's number-sense worksheet as a PDF (PDF/PowerPoint upload is on the Pro tier) and both of you annotate on top of it at the start of next week's lesson — the workflow is covered in how do I give homework to online tutoring students?

Keeping a younger number-sense learner in the room

Number sense arrives at exactly the age band where sustained attention on a screen is developing. A student in pre-K through grade 1 learning to count and subitize does not have the working memory to sit through 40 minutes of anything — even a well-run rod activity. If the tutoring is on-screen, the pacing has to be short-drill / short-break / short-drill, and the motivation layer has to be visible enough to feel real.

On Koala Go, the pieces that work together for the youngest grades:

Doing over watching. Every number-sense move above ends with the student's cursor on the canvas — dragging a rod, dropping a mark on a number line, ticking a sticky. Passive watching is the failure mode; drag-and-drop is the fix. If you find yourself narrating for more than a couple of minutes, hand the student a piece.
The Playground as a between-drill reset. Koala Go's Playground is a shared 3D space with avatars, in a separate tab of the same classroom — not a teaching surface, a reset. A one-to-two-minute walk-to-a-landmark-and-back between two number-sense activities is a lower-friction attention reset than "go get some water" or "let's watch a video", because the student doesn't leave the tab. The gamified virtual classroom page describes the mechanic in depth.
Gems for immediate effort-based reinforcement. Each student in the room has a visible gem balance. You give 1-100 gems from the student panel in a single click — a rhythm of small frequent rewards works better than a single burst at the end. Give one gem for placing the right rod for a number bond, five for persisting through a harder subitizing round even before you check the answer. The reinforcement lands at the moment the effort happened, not at the end of the lesson, and the student sees the balance grow live on their side. For older students who read this as infantilising, turn it down or off — the mechanic is a tool, not a requirement.
Session length matched to the student. A kindergartner holds shorter working segments than a grade 3 student, and both hold shorter segments than the old-school 45-minute "lesson". For the general shape of pacing a 1-on-1 online lesson, see how do I structure a 1-on-1 online lesson? — 25-30 minutes with two or three short activities is a reasonable default for the K-2 range.

A note on students who need extra scaffolding: some students genuinely need more repetitions with smaller quantities, more time subitizing before the number bonds land, and more colour-coded and physical cues than most curricula assume. This isn't a diagnosis, and this page isn't the place to explore it clinically. The workflow above still holds — slow it down, drop to rods 1-5 only for a few sessions before introducing rods 6-10, and lean heavily on the number line's tick-and-hop model to make the abstraction concrete. If a specific student's difficulties are consistent enough that you're wondering about a math-specific learning difference like dyscalculia, the sibling SEN spoke how do I run online tutoring lessons for students with dyscalculia? walks through the workflow overlay tutors use — concrete before pictorial before abstract, repetition-with-variety rather than volume drill, and unhurried magnitude work before any push for arithmetic fluency — while being clear that a diagnosis and any formal intervention plan belong with the family and a qualified specialist (educational psychologist, specialist assessor, paediatrician), not with a webpage.

Sharing number-sense progress with parents

Number-sense lessons don't feel like they're "working" to a parent — the child is on a webcam, the tutor is talking, the manipulatives are on a screen the parent isn't looking at. The parent's read on whether tutoring is helping runs on what you send them between sessions. For number sense specifically, the highest-signal update is a screenshot of the composition the student built (with the written number sentence next to it) plus one line naming what they got that they didn't get last week. "We worked on number bonds to ten today. Emma correctly composed 6-and-4 from two rods for the first time without prompting — that was the piece that felt shaky last week. Home practice this week: five bonds-to-ten cards, twice a day." That paragraph is worth more than any generic "great lesson today!" note.

A few concrete moves:

Screenshot the canvas each week. The whiteboard saves state between sessions, so at the end of each lesson you have a visible record of what the student built. Drop a screenshot of the rod composition or number-line hops in the parent update. A picture of "the child built this and understood it" reads to a parent as evidence in a way that written prose can't.
Send a specific practice ask, not a vague one. Not "please practise counting this week" — that's an unclear task the parent can't verify. Instead: "please play a number-bonds-to-10 pairs game with Emma three times this week, and let us know which pairs took longest." The parent-comms cadence lives in how do I share progress updates with parents?; the number-sense-specific ask is that concrete.
Record short clips of key moments where consent is in place. Koala Pro includes local recording; Platinum adds cloud recording of scheduled lessons. A 90-second clip of the child first composing 3+7 from two rods and saying "look, they match" is worth every second of the setup. Get the family's written permission before you record anything with a minor — the consent, storage, and sharing detail is covered in how do I record online tutoring lessons?

Common failure modes and honest limits

Things that go wrong in online number-sense lessons, and the fix — most of these are the pattern-recognitions you build in your first month teaching this online:

Teaching number sense with static images on a Google slide. The image is a picture; a manipulative is a picture the student can move. Even a well-designed static image of ten circles is a step down from a set of drag-and-drop rods the child physically composes into ten. If your platform doesn't ship the pieces, hand-drawing rods on a whiteboard tool doesn't recover the loss — the mechanic that makes a manipulative teach (the child builds the composition) requires the pieces to be discrete draggable objects.
Skipping the composition step and going straight to the number sentence. Writing "3 + 4 = 7" for a child who hasn't yet built 3-and-4 from rods and seen it match a 7 is teaching the shortcut before the intuition. The rule sticks less well and generalises less. Build the physical picture first with rods; write the sentence second.
Overloading the palette early. A pre-K or kindergarten student on the second lesson doesn't need all ten rods on the canvas. Start with the ones, twos, and threes; add the fours and fives at the end of week two; expand only when the student is fluent with the smaller ones. The whole set is available in the picker; the tutor is the one choosing which pieces to invite in.
Number line without hopping. A number line the student only points at is a labelled ruler, not a number line. The point of the manipulative is the hopping — small arcs from tick to tick, in ones, then in twos, then in fives — because the arc is the physical picture of the operation. Draw arcs with the pen tool over the top of the number line for every counting and skip-counting exercise, and delete them between activities so the line stays clean.
Sending the child to a practice site alone with "please do 20 counting questions". The child will lose interest around question five, get one wrong, not understand why, and arrive at next week's lesson worse than they left it. Practice sites work best inside the lesson via the cobrowser, or with a very small (three-to-five problem) between-lesson ask a parent will actually oversee.
iPad-only workflows. Koala Go works on iPad, but the drag precision required for lining up narrow One-rods against a Ten baseline is fiddlier on a finger-driven tablet than on a desktop or laptop with a mouse. If a student uses iPad exclusively, plan around larger rods (the Fives, Sixes, and Tens are easier to grab than the One and Two), and use the Text tool more for the notation side.

Two limits worth naming plainly rather than papering over:

The rod palette on Koala Pro is the base set — no red-and-blue alternating, no vertical mode, no one-click "Add all" staircase. All three are on Koala Platinum. If those specific mechanics are how you were taught to use rods, factor that into the plan choice; if not, the base rods teach every core number-sense concept above without them.
The "make your own" number line is Platinum-only. Koala Pro gives you eight preset ranges (0-10, 0-20, 0-30, 0-100 by 5 or 10, -10-0, -10-10, blank), which covers most K-4 number-sense work. When you need a specific custom range often — decimal work in tenths, angle work in thirties, a domain-specific range for a particular activity — the Platinum builder is what you want. Otherwise the blank line with sticky-note labels is the Pro workaround.

Practical setup on Koala Go for a number-sense-heavy caseload

If you're evaluating Koala Go for K-4 number-sense tutoring specifically, here's the practical picture and where the current lines sit:

Whiteboard is the workspace. The classroom's whiteboard is a Fabric.js canvas with the Math Tools panel: ten Cuisenaire-style Number Rods valued 1-10 (Pro), eight preset Number Lines from 0-10 through -10-10 with a blank line (Pro), the base geometry shapes for parallel shape-and-quantity work (core set Pro; full library Platinum), and the Math Symbols palette for writing number sentences (Pro base, Platinum for the fuller algebra set). Fraction Bars, Fraction Circles, and the make-your-own number line are on Platinum — the fractions bridge described above.
Upload existing number-sense worksheets. On Pro, upload the student's counting or number-bond worksheet as a PDF or PowerPoint and annotate on top of it — the worksheet becomes the shared surface. Free-tier upload caps apply.
Cobrowser for practice sites. Open Toy Theater's number tools, NRICH's early-number games, Khan Academy Kids, ABCya, or whatever number-sense practice tool your programme uses; the student clicks and drags on the live site inside the lesson room. 10 minutes at a time on Free with a 20-minute cool-down, unlimited on Pro.
Playground and Gems for the engagement layer. The youngest grades (pre-K through grade 2) benefit most from the Playground brain break between drill segments and per-student Gems for immediate effort-based reinforcement. Older students lean into the manipulatives without needing the engagement layer as heavily.
Recording for parent review. Local recording is on Pro; cloud recording of scheduled lessons comes with Platinum. Either supports the "send the parent a 90-second clip of the number-bond breakthrough" pattern above.
Free tier limits worth knowing. Koala Free caps sessions at four students and includes the whiteboard plus a capped cobrowser — but the Math Tools panel (rods, number lines, everything above) is Pro-and-up. Free is a trial surface for number-sense tutoring, not the day-to-day.

Honest limits worth flagging that show up in real number-sense work regardless of platform: iPad is rougher than desktop for narrow-rod precision; students in mainland China sometimes hit connection issues (we route through Hong Kong proxies but the Great Firewall is a moving target — see teaching students in mainland China for the workaround pattern); and the make-your-own custom number line is Platinum-only (the eight presets cover most needs on Pro).

Why writing real math on a video call is a specific kind of hard

Every math tutor has run into it: a keyboard was built for prose, not equations. A comma is one keystroke; a square-root sign is a trip to a Unicode chart. The problem shows up in three specific ways when the lesson is on video and the shared surface is either a Google Doc, a screen-share, or a whiteboard tool that ships a pen and nothing else.

Symbols the keyboard doesn't have. The list is longer than it feels — even without going into calculus. Square root (√), degrees (°), plus-or-minus (±), pi (π), the not-equal (≠) that shows up in an inequality answer, the less-than-or-equal (≤) that shows up the moment you introduce interval notation, the angle mark (∠), the parallel-lines mark (∥), the exponent that's supposed to sit above the line rather than after a caret. Any workflow that requires the tutor to break attention and look up an alt-code every three lines is a workflow that stalls the lesson.
Symbols that look like something else when typed. "1/2" typed in a plain-text line is not a fraction; it's three characters. "x2" is not "x squared" but a variable next to a two. "3-2" is not "3 minus 2" but "3 hyphen 2", and typographically the hyphen is narrower than a minus and does not centre against digits. Notation is meant to trigger pattern-recognition — the tutor points at "x² + 3x − 4 = 0" and the student's brain reads "quadratic". Break the notation and the pattern-recognition breaks with it.
Handwritten notation on a webcam is unreliable. A tutor at a desk with a pen and paper on the webcam works for a couple of lines, but the writing is small, upside-down from the tutor's perspective, and doesn't stay on the shared surface between lessons. A stylus on the whiteboard is better but assumes the tutor has a tablet or drawing tablet — most don't, and the students definitely don't. The scalable answer is typed notation that reads as the notation it represents.

What a math tutor actually needs is a keyboard-adjacent palette: a small grid of the symbols you write on a piece of paper, one click each, inserting into whichever text box or sticky the lesson is already using. That is the design of Koala Go's Math Symbols palette. What it doesn't do — render a stacked LaTeX-style fraction, put a radicand under a bar, plot a function — we say plainly further down, along with the pairing pattern most working online math tutors converge on for those cases.

Where the palette lives — two entry points, same insert

The Math Symbols palette on Koala Go shows up in two places on the same shared whiteboard, so whichever workflow you're in, the symbols are one click away.

1. The Math Symbols folder in the Whiteboard Library

Open the Whiteboard Library from the classroom sidebar, and one of the folders in the Math Tools panel is labelled Math Symbols (with a small "+ π √" icon). Click into it and you see the full palette: symbols grouped into seven small grids — Operations, Compare, Fraction · Ratio · Percent, Powers, Algebra, Geometry, and Brackets — with a lock badge on any symbol your current tier doesn't include. Clicking a symbol does one of four things, in order:

If you have a text box selected — or are mid-typing in one — the symbol appends at the end of whatever's already in it. Note this is end-of-text, not at your cursor: the folder always adds to the end. (The ∑ popup below is the one that inserts at the cursor.)
If you have a sticky selected but not being edited, clicking a symbol drives it into edit mode and appends there.
If a non-text object is selected, or nothing is selected, clicking a symbol drops a new text box at the canvas centre containing that one symbol, and selects it — so the next symbol click stacks straight into the same box. Because every click appends to the end, you build a line left to right: type "3", click "+", type "4", click "=", type "7" — the symbols come from the folder, the digits from the keyboard. Numbers aren't on the palette; it handles the notation, not the arithmetic.
Clicking on empty canvas (which deselects the new text box) ends that stack — the next symbol click starts a new box.

Both you and the student can use the folder — the whiteboard is shared, and the newly-dropped text boxes are shared canvas objects, so a student who has picked up "type the answer" as their move can click "=" from the folder and type their answer after it, just as you can. The folder is the discovery surface: it's the place a tutor scans when they don't remember which glyph a symbol lives under, and the place a student learns "there are more symbols than my keyboard".

2. The ∑ sidebar popup next to a selected text or sticky

Select any text box or sticky note that's already on the whiteboard, and a small toolbar of quick-actions appears next to it in the sidebar. One of the buttons is a ∑ (capital sigma) — the Math symbols shortcut. Click it and a popup opens right next to the selected box, with the same seven categories arranged as small grids. Click a symbol and it inserts at your caret; the popup stays open and the caret stays in the box, so you can build a whole line — "π × r² = A" — without closing the popup between symbols. This is the workflow shortcut for tutors who already have a text box on the canvas and are actively writing math in it, and it's the fastest way to build a longer expression once the container is in place. The popup only appears when a text or sticky is selected (it's a context-menu shortcut, not a global toolbar), which keeps the sidebar quiet when the selection is a rod, a number line, a shape, an image, or an uploaded PDF.

Both entry points draw on the same symbol set, the same tier gating, and the same shared-canvas sync. They differ in one detail worth knowing: the folder always appends to the end of the box, while the popup inserts at your cursor — so if you need to drop a symbol into the middle of an expression you've already written, use the popup. Otherwise the choice is workflow preference: the folder is discovery, the popup is speed. Working online math tutors typically start a lesson from the folder ("here's where the symbols live"), demonstrate a couple of moves so the student can find them too, and shift to the popup once the lesson is deep in a specific expression.

What's in the palette by tier

The palette on Koala Go is the same seven categories in either entry point. What differs is which symbols in each category your current tier includes:

Operations (Pro): + − × ÷ =. The four operations plus equals. The minus is the proper mathematical minus sign (U+2212), which lines up centred against a digit — not a keyboard hyphen. This is the core K-6 set and every symbol here is on Pro (any working paid tutor has them).
Compare (Pro base, Platinum extras): < > on Pro; ≤ ≥ ≠ ≈ on Platinum. Less-than and greater-than land at Pro because they're the elementary-and-middle-school comparison symbols. Less-than-or-equal, greater-than-or-equal, not-equal, and approximately-equal are the interval / inequality / limit / rounded-answer notations that come in with secondary maths.
Fraction · Ratio · Percent (Pro): / : %. All three on Pro. The "/" is a slash between the numerator and denominator (typed as "3/4"), not a stacked-fraction renderer — see the honest-limit section below.
Powers (Platinum): ² ³ ⁿ. Superscript-two, superscript-three, and superscript-n. Platinum-only because these are the exponents that come in with area formulas, volume formulas, and early algebra. Note the practical limit: these three are the entire set — anything higher than a cube or lower than "the power of n" (say, "⁴", "⁵", "x-to-the-seven") can't be entered from the palette, and the workaround is a caret-and-digit notation (like "x^4") that's fine in-lesson but reads as a compromise.
Algebra (Platinum): π ± ∞. Pi, plus-or-minus, and infinity. Platinum-only. Pi shows up the moment a lesson touches circles; ± shows up in the quadratic-formula answer; infinity shows up in interval notation and limit descriptions.
Geometry (Pro base, Platinum extras): ° √ on Pro; ∠ △ ⊥ ∥ on Platinum. Degrees and square-root are Pro because they show up in elementary and middle-school work (angle sums in triangles, "find the square root of 25"). The angle marker (∠), triangle glyph (△), perpendicular (⊥), and parallel (∥) are the secondary-geometry set and are Platinum-only.
Brackets (Pro base, Platinum extras): ( ) on Pro; [ ] { } on Platinum. Round brackets are Pro (order-of-operations in elementary and middle-school). Square and curly brackets are Platinum (interval notation, set notation, and grouping in secondary maths).

Two tier facts worth flagging plainly. Free-tier tutors see the folder in the Library and the ∑ button when a text is selected, but every symbol shows as locked — the palette is not part of the free tier. Enterprise and B2B tutors are provisioned at Pro feature level and every Platinum symbol is hidden (not shown locked) so an enterprise teacher never sees an upsell prompt in the middle of a lesson. The practical effect: on an enterprise plan the palette is the Pro set, and the algebra and geometry symbols aren't in it.

Six notation moves that carry most math tutoring

These are the notations working online math tutors reach for the most, and how to type each on Koala Go. All examples assume the reader has a Pro or Platinum plan; the tier of the required symbols is called out inline. Before you drill any of these with a specific student, a short first-session diagnostic tells you which notations the student actually has, which ones are the gap that's blocking the arithmetic underneath, and where in the progression to start — see how do I assess a new online math tutoring student? for the first-lesson workflow, including how the palette itself doubles as a notation-fluency diagnostic before it becomes a teaching one.

An operations line: "3 + 4 = 7". All-Pro. Drop a text box (or click into an existing one) and build the line left to right: type "3", click "+", type "4", click "=", type "7". Symbols come from the palette, digits from the keyboard. This is the K-2 baseline that shows up alongside a rod or number-line composition — see the number-sense spoke for the picture-first / sentence-second workflow that puts this notation next to the manipulative.
A fraction as a symbol pair: "3/4" or "1/2 + 1/4 = 3/4". All-Pro. The "/" from the Fraction · Ratio · Percent group, with digits either side. Reads as a fraction inline — perfectly fine for elementary and middle-school arithmetic, where "3/4" is the notation students see in their textbook. If the lesson needs a rendered stacked fraction (a bar with numerator above and denominator below, in the shape that a student will see in a secondary-maths exam), the palette can't render it — you'd either pair a fraction manipulative on the same canvas (Fraction Bars or Fraction Circles, described in the fractions spoke) or reach for a cobrowser tab into a graphing tool, described in the honest-limit section below.
An inequality: "x < 8" or "3 ≤ n ≤ 10". Pro for the base "<" and ">"; Platinum for "≤" and "≥". A middle-school interval line uses the simple pair; a secondary interval or a limit-notation line uses the Platinum extras. When a student is graphing an inequality on a paired cobrowser tab into Desmos, having the same symbols on the whiteboard side means the notation on the two surfaces matches — no "on the whiteboard it says less-than, on Desmos it says less-than-or-equal" confusion.
A percent, a ratio, or a rate: "60 : 100 = 3 : 5" or "60%". All-Pro. The ":" from the same Fraction · Ratio · Percent group as the fraction slash, and the "%" glyph one click over. Middle-school ratio and proportion work leans heavily on both. Percent problems often pair with a number line — see the number-sense spoke for the 0-to-100 line that hosts the percent picture.
An algebraic expression with an exponent: "x² + 3x − 4 = 0" or "A = πr²". Platinum for "²" and "π"; Pro for the operations, the minus, and the equals. This is the shape that shows up the moment a lesson touches area formulas or the quadratic. Type the coefficient and variable, click "²", click "+", type the next term, and so on. If the exponent you need is above ³ (say, "x⁵"), the palette doesn't include it; the workaround most tutors use is "x^5" as a caret-and-digit notation — fine for the lesson but worth calling out to the student that the exam version will use a rendered superscript. The algebra pedagogy the notation supports — balance-scale intuition for one-step equations, coefficient-and-variable work, exponent laws, factoring, and the place where the palette's inline notation reaches its honest-limit against a rendered LaTeX display — is covered on the sibling spoke how do I teach algebra online?
A geometry line: "∠ABC = 90°" or "AB ⊥ CD". Platinum for the angle mark "∠", perpendicular "⊥", parallel "∥", and triangle "△"; Pro for the degree symbol "°" and the equals. A secondary-geometry lesson leans on all of these. On a lesson focused on angle-sum-of-a-triangle or parallel-lines-with-a-transversal, the geometry glyphs let you annotate a diagram (drawn with the shape tools, or uploaded as a PDF and annotated on top) with the correct notation, so what's on the whiteboard reads as what will be on the exam. The geometry pedagogy the notation sits over — the 2D and 3D shape library, angle work on a drawn diagram, the shape-labels toggle that anchors typed measurements to a shape's corners and edges, and the difference between annotating a shape and constructing one — is covered on the sibling spoke how do I teach geometry online?

The pattern in all six: the palette gets you 90% of the notation an in-lesson math conversation actually uses. Where it stops — stacked fractions, higher exponents, radicals with a radicand under a bar, live graphs, LaTeX-style equation layout — is where the pairing pattern below earns its keep.

The honest limit: it's a symbols set, not an equation editor

Koala Go's Math Symbols palette is a click-to-insert set of individual glyphs. It is not an equation editor of the kind LaTeX, MathType, EquatIO, or a dedicated STEM whiteboard (compared honestly on our best-online-whiteboards page) will render. Specifically, four things the palette does not do, called out plainly so you know before you sign up:

No stacked fractions. The "/" in "3/4" is a horizontal slash, not a bar with a numerator above and a denominator below. For arithmetic and pre-algebra lessons this is fine; for a secondary-exam page where every fraction is stacked with a horizontal bar, the palette produces the notation students read in their textbook rather than the notation they'd see rendered by an exam board's typesetting.
No radicand under a bar. The "√" is a symbol you type before a number ("√25") — it does not draw a horizontal bar over the digits that follow. So "√(x + 3)" is written as-typed, with parentheses making the radicand explicit, rather than as a rendered radical with the bar continuing over x+3. Again, fine for in-lesson notation; not a substitute for a rendered radical if a lesson genuinely needs one.
Superscripts stop at ⁿ. The Powers category gives you ², ³, and ⁿ as three individual glyphs. There is no general "superscript mode" — anything above ³ needs a caret-and-digit workaround ("x^4", "x^5") or a paired external tool. For elementary and lower-secondary lessons this rarely bites; for late-secondary polynomial or exponential-function work it does.
No live graphs. The palette is text; the whiteboard has pen, shape, and drawing tools but not a function-plotter. A lesson that needs to plot y = x² − 3x + 2 and let the student drag a slider on the coefficients belongs on a live graphing tool — Desmos, GeoGebra, or a subject-specific site — inside the classroom via the cobrowser.

The pairing pattern most working online math tutors converge on: the Math Symbols palette on the whiteboard for the notation (equations, inequalities, angles, radicals, powers up to ³), plus a cobrowser tab into Desmos when the lesson needs a live-plotted function or a rendered LaTeX expression. A cobrowser is a shared browser window inside the classroom — you open a Desmos graph, and both you and the student can drag sliders, type new expressions, and adjust the domain live, without the student switching windows or losing the lesson thread. That combination — notation on the whiteboard, graphing on the cobrowser — carries most secondary-maths tutoring cleanly. Cobrowser is capped at 10 minutes per session with a 20-minute cool-down on Koala Free; unlimited on Pro and up.

When to reach for a symbol and when to reach for a manipulative

A math tutor writing "3 + 4 = 7" for a K-2 student who hasn't yet built the composition physically is teaching the abstraction before the intuition — the sentence sticks less well and generalises less. A tutor building the composition with number rods for a middle-school student who's already fluent with the arithmetic is spending manipulative time where notation practice would move faster. Getting the balance right is one of the underrated skills of online math teaching.

A short rule of thumb, matching what most tutors converge on:

Manipulative first, notation second (K-2, and any student meeting a concept for the first time). Build 3 + 4 with a Three-rod and a Four-rod against a Seven-rod (see the number-sense spoke). Then write "3 + 4 = 7" in a text box next to the picture using the palette. The sentence formalises the intuition the picture built.
Notation next to picture (grades 3-5, and any student consolidating an operation). Build the fraction addition 1/4 + 1/2 with the Fraction Bars (see the fractions spoke). Write "1/4 + 1/2 = 3/4" next to the picture with the palette. Both are visible; the eye moves between them. The same picture-then-notation pattern is what carries word-problem work — a bar model or tape diagram of the story sits on one side of the canvas, the equation the student writes from it sits on the other, and the palette makes writing the equation as unambiguous as reading the diagram; the sibling spoke how do I teach math word problems online? walks the five bar-model shapes and the translation vocabulary.
Notation-first, manipulative as verification (middle school through secondary, and any student fluent with the operation). Write the expression first ("x² + 3x − 4 = 0"), let the student work it, then optionally build a check with a manipulative or a Desmos graph. The palette is the primary surface at this stage; the manipulative is the sanity-check for a stalled answer, not the teaching move.
Notation-only (late-secondary and exam-prep). The palette plus uploaded past-paper PDFs on the whiteboard plus a cobrowser tab into Desmos or GeoGebra for the graphing beats. This is the workflow described in the exam-prep answer for SAT / ACT / A-level / IB maths.

A student with dyscalculia or number-processing difficulties usually benefits from the "manipulative first, notation second" pattern held longer than the age band above suggests — the dyscalculia answer walks through the concrete-first workflow with specifics.

By age band: what tutors actually use

Which parts of the palette get the most airtime depends on the age you teach. Grades are typical US placements, not standards claims — adapt to the student and curriculum in front of you.

K-2 (ages 5-7): operations only, next to manipulatives. The Pro operations set (+ − × ÷ =) is enough — most lessons at this age write no notation more complex than "3 + 4 = 7" or "10 − 2 = 8", and the notation sits next to a rod or number-line composition. The comparison symbols < > come in around grade 1 with "more than" / "less than" activities.
Grades 3-5 (ages 8-10): fractions, percents, ratios, radicals, degrees. The Pro Fraction · Ratio · Percent set (/ : %) plus √ for the introduction to square roots and ° for angles in shape work. Fractions written as "3/4" pair with the Fraction Bars or Circles from the fractions spoke. Multiplication and division with the proper × and ÷ glyphs (not "*" and "/" borrowed from a keyboard) start reading as the notation a student sees in their textbook — the multiplication pedagogy the notation sits alongside (arrays with rods, skip-counting on the number line, Number Roller drill for fluency, and what to do when a specific times table refuses to stick) is covered on the sibling spoke how do I teach multiplication and times tables online?
Middle school (ages 11-13): the algebra and geometry starter set. Platinum symbols come in — π for circle formulas, ≤ ≥ ≠ for inequality work, ² for area and volume, ∠ for angle notation, [ ] for interval notation, △ for triangle geometry, ⊥ ∥ for parallel-and-perpendicular geometry. This is the tier at which the palette earns its Platinum upgrade for a middle-school-focused tutor.
Secondary (ages 14-17) and adult learners: the full palette plus a paired graphing cobrowser. All Platinum symbols in play — π ² ³ ⁿ ± ∞ ≈ ≤ ≥ ≠ ∠ △ ⊥ ∥ [ ] { } — plus the honest-limit pairing with Desmos or GeoGebra for stacked fractions, higher exponents, radicand-under-a-bar radicals, and live-graph work. Uploaded past-paper PDFs on the whiteboard become the primary surface, with palette symbols dropped into text boxes to annotate the working.

Practice, homework, and notation drills between lessons

Notation is a skill like any other — the student who has typed "π × r²" on the whiteboard three lessons in a row types it faster than the student who saw the tutor do it once. Two moves that turn notation practice into between-lesson consolidation:

Notation on the shared canvas as an in-lesson move, not just a demonstration. Every notation the tutor writes on the whiteboard should be followed within a couple of minutes by the student writing the same notation — either from the folder or from the ∑ popup. This is the difference between the student seeing "√25 = 5" and the student typing "√25 = 5" themselves. The whiteboard is shared; both you and the student can use both entry points. If your student's tempted to type "sqrt(25)" as a workaround, redirect: the folder is one click.
Between-lesson notation practice on annotated PDFs. On Koala Pro, upload the student's homework worksheet as a PDF and both of you annotate on top of it in the next lesson. The palette works over the PDF the same as over a blank canvas — a text box with palette symbols in it sits on the same layer as the ink. The workflow is covered in how do I give homework to online tutoring students?

Common failure modes and honest limits

Things that go wrong in real online-math-notation workflows, and the fix:

Typing "1/2" and hoping it reads as a fraction. In a plain text line without the palette, "1/2" reads as "one over two" — three characters, not a fraction. Insert the "/" from the palette (Fraction · Ratio · Percent group) between the digits so the surrounding context is clearly a fraction, and pair with a Fraction Bar for the picture side when the lesson is teaching the concept rather than practising the arithmetic.
Using the keyboard hyphen where you meant a minus sign. The palette's "−" (U+2212 MINUS SIGN) is a proper mathematical minus that centres against digits and reads as an operation. The keyboard hyphen "-" is narrower and reads as a break-mark. For a demo line the student won't linger over, either works; for an equation the student will copy into their own notes, the palette minus is the one you want.
Trying to write a stacked fraction on the whiteboard. The palette does not render a stacked fraction with a numerator above and a denominator below a horizontal bar. If a lesson genuinely needs a rendered stacked fraction, pair with a Fraction Bar manipulative (Platinum, see the fractions spoke) for the picture side, or open Desmos in the cobrowser for the rendered version.
Reaching for an exponent above ³. The palette's Powers category gives you ², ³, and ⁿ — three specific glyphs, not a general superscript mode. For "x⁵" the workaround is "x^5" as a caret-and-digit notation and a spoken "x to the fifth"; for a lesson that lives in exponents-of-arbitrary-powers, this is one of the cases where a Desmos or GeoGebra cobrowser tab does more of the work.
Free-tier tutors seeing every symbol locked. The palette is not available on the free tier — the folder and the ∑ button appear in the UI but every symbol shows as locked behind an upgrade. If you're trialling Koala Go for math tutoring, the palette is a Pro-and-up feature, so a real-shape math lesson needs at least the trial or Pro to evaluate.
iPad-only notation workflow. Koala Go works on iPad, but the narrow symbol grid in the folder and the popup is fiddlier on a finger than on a mouse, especially for the Compare and Powers categories where symbols like ≤ and ² occupy a small tap target. If a student uses an iPad exclusively, the ∑ popup (which sits next to the selected box rather than in the Whiteboard Library sidebar) is a slightly easier click; consider preparing more of the notation in advance on your side and letting the student focus on the manipulatives and answers.

Two limits worth naming plainly rather than papering over:

Anything requiring a full equation-layout renderer belongs on a paired tool. Stacked fractions with a bar, radicand-under-a-bar radicals, higher exponents, integral signs, sigma-notation sums, matrix layouts, LaTeX-shaped display equations — none of those are in the palette, and none are a planned addition to it in its current shape. A dedicated equation editor (in the whiteboard-tool comparison at our best-online-whiteboards page) or a paired cobrowser tab into Desmos / GeoGebra / an online LaTeX renderer is where those live in the working workflow.
The algebra and geometry set is Platinum-only. A tutor working exclusively through K-5 (operations, comparisons, basic fractions, degrees, square root, round brackets) has everything they need on Koala Pro. A tutor teaching middle school or secondary maths regularly wants the Platinum symbols (π, ±, ∞, ² ³ ⁿ, ≤ ≥ ≠ ≈, ∠ △ ⊥ ∥, [ ] { }) — factor that into the plan choice rather than discovering it in the middle of a lesson.

Practical setup on Koala Go for a notation-heavy math workflow

If you're evaluating Koala Go for a math-tutoring practice where notation matters — most middle-school and secondary work — here's the practical picture and where the current lines sit:

Whiteboard is the primary notation surface. The classroom's whiteboard hosts text boxes and stickies as first-class objects. Palette symbols insert into either, at the caret. Both tutor and student can insert. Boxes persist across lessons the same way any whiteboard object does.
Two palette entry points. The Whiteboard Library "Math Symbols" folder (discovery surface) and the ∑ sidebar popup that appears next to any selected text or sticky (speed surface). Same symbols, same tier gating, same shared-canvas sync.
Palette tier: Pro for the base, Platinum for algebra and geometry. Pro ($25.99/month monthly, $21.99/month billed annually) unlocks the base palette — operations, base compare, fractions/ratios/percents, degrees, square root, and round brackets. Platinum ($49.99/month monthly, $39.99/month billed annually) adds the algebra and geometry set — π ± ∞ ² ³ ⁿ ≤ ≥ ≠ ≈ ∠ △ ⊥ ∥ [ ] { }. Enterprise / B2B tutors are Pro feature level with Platinum symbols hidden.
Whiteboard PDF and PowerPoint upload for annotating existing worksheets (Pro). Upload a past paper, a textbook page, or a district worksheet and annotate on top of it with palette symbols in text boxes. The palette works over the PDF the same as over a blank canvas.
Cobrowser for paired graphing and rendered-equation tools (unlimited on Pro, capped on Free). Open Desmos, GeoGebra, or a subject-specific site in a shared browser inside the classroom — the student clicks and drags on the live page inside the lesson room. See what is a cobrowser? for the mechanic.
Paired manipulatives for the concrete-first pattern. Fraction Bars and Fraction Circles (Platinum) for fraction work; Number Rods and Number Lines (Pro) for early numeracy and place value. The fractions spoke and number-sense spoke walk through the manipulative side; this page is the notation side.

Honest limits worth naming that show up in real math work regardless of platform: iPad is rougher than desktop for narrow-symbol picking; students in mainland China sometimes hit connection issues (we route through Hong Kong proxies but the Great Firewall is a moving target — see teaching students in mainland China); and the palette is a symbols set rather than an equation renderer, so stacked fractions, radicand-under-a-bar radicals, exponents above ³, and live graphs pair with a cobrowser tool rather than living on the whiteboard.

Why teaching multiplication online is a specific kind of hard

Every grade-2-and-up tutor eventually hits it: the student who has memorised "3 × 4 = 12" as a chant but freezes on "how many is 4 lots of 3?" because the chant and the concept never met. In person, the fix is to reach for a set of counters, or a bag of Cuisenaire rods, or a pile of buttons — build three groups of four in front of the child, count the twelve, then match those groups against a "4 + 4 + 4" written on a piece of paper, then against a "3 × 4" on the same paper. The physical arrangement is the concept. Online, without the pieces, the tutor is either narrating over a webcam ("imagine three groups of four…") — which asks a lot of a seven-year-old's working memory — or drawing rows of dots on a whiteboard the student can't touch. Three problems show up together when teaching multiplication online, and the working online multiplication workflow has to answer all three.

Multiplication is repeated addition, and the "repeated" part needs to be physical. The move that turns "3 × 4" from a rule into a picture is placing three separate quantities of four alongside each other and counting them as a whole. Passive watching of the tutor doing this on their side of a webcam doesn't build the concept; the child needs to place the rods themselves. Static pictures of arrays on a Google slide don't quite either — the student can look at them, but they can't build them, so the causal link ("if I add another 4 to the row, the total goes up by exactly 4") never lands.
Times tables are memorised faster when they're skip-counted first, and skip-counting needs an annotatable line. Most working elementary teachers converge on the same sequence: before a child drills the "5 times table" as recall facts, they walk the 5s on a number line (0, 5, 10, 15, 20…) enough times that "5 × 6 = ?" becomes "six hops of five = 30" before it becomes "thirty". On a static number line printed on a worksheet, that walking is unfollowable at a distance; the student can't point at the hops as they say them, and the tutor can't see where the student is stuck. A shared, annotatable number line where the student pens the arcs from tick to tick fixes both.
Fluency drill needs randomness the tutor doesn't manually invent. There's a middle stage, after concept and before fluency, where the student has "5 × 3 = 15" solidly but "7 × 8" is not automatic. The move that closes it is short randomised drill — five minutes of "roll two digits and multiply". If the tutor is generating those digits themselves ("hmm, let's say… 6 and 7"), the pacing stalls and the child sees the tutor thinking rather than the drill running. A physical die works in person; a shared digital roller works online.

An online multiplication workflow that actually works has an answer to each of the three: draggable equal-length rods for building groups and arrays, annotatable preset number lines for skip-counting the anchor tables, and a randomised roller for the fluency-drill stage. Koala Go's Math Tools panel ships all three. Below is how each maps to a specific teaching move, in the order you'd actually reach for them across a grade-2-through-5 arc.

The tools that make multiplication learnable on a webcam

All four of the tools below live in the Whiteboard Library's Math Tools panel on Koala Go's shared whiteboard. Both you and the student can pick up, drag, and rearrange every one of them in real time — the student is doing the maths, not watching.

Number rods for repeated addition and for arrays

Number rods (Pro tier) are the workhorse manipulative for the concept side of multiplication. Ten Cuisenaire-style coloured rods, one unit through ten units long, that you drop onto the whiteboard and rearrange. Same tool the number-sense spoke uses for early counting and number bonds — the difference is what you build with them.

Two patterns that carry most of grades 2-3 multiplication:

Multiplication as repeated addition. To build "3 × 4", place three 4-rods end to end in a single row, so the row is twelve units long. Underneath, place a Ten-rod and a Two-rod to make twelve — the student sees that "four and four and four" is the same length as "ten and two", which is twelve. Write "3 × 4 = 12" in a text box next to the picture (see the math notation spoke for the × symbol). Then rebuild the same fact as "4 × 3": four 3-rods end to end, same twelve. That side-by-side comparison is the visual proof of the commutative property (that "a × b = b × a") and it doesn't get taught as well any other way.
Multiplication as an array. To build "3 × 4" as an array, place three 4-rods as three separate rows, one above another — each row four units long. The student now sees the "three rows of four" that the notation is describing. Sliding the stack together and apart shows that the total (twelve) is the same however you count it, which is where "3 × 4 = 12 and 4 × 3 = 12" starts making concrete sense. This is the same tool from the same panel, used slightly differently — arrays for "rows and columns", repeated addition for "groups of".

The rods are on the Pro tier ($25.99/month monthly, $21.99/month annual, or included in a business plan), so any working paid tutor has access. Free-tier tutors see the rods in the panel but locked behind an upgrade nudge. Two Platinum-only extras occasionally help specifically for multiplication work: the red-and-blue alternating fill (each rod visually breaks into its unit count, which supports "count the units" fluency work when a child still isn't seeing that a 4-rod is four ones), and the "add all" one-click staircase (drops all ten rods on the canvas in order — the fastest reset for the start of a new array). Neither is required for the core teaching; they're workflow speedups if the specific mechanic is how you were taught to use rods.

Number lines for skip-counting the times tables

The move that anchors each times table before drill is skip-counting on a number line: hopping "0, 5, 10, 15, 20…" for the 5s, "0, 10, 20, 30…" for the 10s, "0, 3, 6, 9, 12…" for the 3s. Koala Go's Whiteboard Library ships eight preset number lines on the Pro tier (see the number-sense spoke for the full list), of which three carry most times-table skip-counting:

0 to 100 by 5 (Pro). The line for the 5 times table — every labelled tick is a multiple of 5, so hopping tick-to-tick from 0 to 50 walks the student through 5×1 through 5×10 directly. Same line supports the 10s implicitly (hop two ticks at a time).
0 to 100 by 10 (Pro). The clean version for the 10s — every labelled tick is a multiple of ten. Also the line for place-value multiplication: "what's 4 × 20?" is four hops of twenty — or eight hops of ten — landing on 80.
0 to 30 (Pro). Steps of 1 up to 30. Skip-counting the 3s from 0 through 30 covers 3×1 through 3×10 with tick-to-tick pen arcs. The same line covers the 2s (hop two ticks) through 2×10, and the 4s (hop four ticks) through 4×7. This is the line most K-3 tutors reach for when teaching the 3s and 4s.

Both you and the student annotate directly on top of any line — drop pen arcs from tick to tick, coloured differently for different hop sizes, and label each landing point. That is the whole point: the line is a shared surface, not a picture the student is watching over your shoulder. Every preset line is resizable and movable, and you can duplicate a line to place two side by side (say, the 0-to-30 and the 0-to-100-by-5 for comparing the 3s against the 5s).

For the harder tables — the 6s, 7s, 8s, and 9s — the presets fall short: a 0-30 line hosts the 3s neatly but a 6× table needs to hop to at least 72 (6×12), which is off the end of the 0-30 line, and the step-1 preset above 30 doesn't tick individually anyway. Koala Platinum unlocks a "make your own" custom number line — you type a start, an end, and a step, and the whiteboard renders a line to spec. So a "0 to 72 in steps of 6" line for the 6s, a "0 to 84 in steps of 7" line for the 7s, a "0 to 96 in steps of 8" line for the 8s, and a "0 to 108 in steps of 9" line for the 9s all become one-click builds. A tutor whose caseload leans on the 6s-through-9s skip-counting anchor typically upgrades to Platinum for exactly this workflow. We'd rather say this plainly than fudge it: there's no clean Pro equivalent, because none of the eight preset lines ticks at 6, 7, 8, or 9.

Number Roller for fluency drills

Once the concept is solid and the anchor tables are skip-countable, the middle stage of times-table learning is short randomised drill — five to eight minutes at the end of a lesson, or the start of the next one, where the student is asked a stream of one-digit or two-digit multiplication questions in random order and answers them under time pressure. The tool for that stage on Koala Go is the Number Roller, and it sits on the Platinum tier.

The Number Roller drops a row of number tiles on the canvas with a Reroll button underneath — tap Reroll and every tile lands on a fresh random number in the range you set. What that gives you:

Range presets 1-6, 1-10, 1-12, 1-20, 1-50, 1-100, and 1-1000. The 1-12 preset covers the standard "times tables up to 12" sweep; the 1-10 covers the "up to 10s" version; the 1-6 mimics a physical die if you want to keep the workflow feeling like a game.
One to four numbers on the roller. Set it to two numbers at 1-10 and each roll gives you two random single digits — the classic "3 × 7" fluency shape. Three numbers give you triple products for advanced work. A single number on a 1-100 range, multiplied by a digit you type yourself, covers multi-digit practice.
A padlock on every tile. Each number has a small padlock above it. Lock one tile at 7 and only the other tile rerolls — so the drill sweeps the 7 times table specifically instead of a mixed bag. This is the setting that turns the roller from a general random-number tool into a one-table-at-a-time drill, which is how times tables should be practised in the first place.
Whole or decimal. Whole numbers for the standard times tables; decimal rolls for later multiplication of decimals practice (a 1-6 range with decimals gives values like 3.5, 4.7 — useful once decimal multiplication is on the curriculum).
Custom min/max. If the presets don't fit — say, a 4-through-9 range specifically to hammer the harder tables — you set a custom minimum and maximum and the roller respects them.

The practical workflow: place one roller set to show two numbers on the 1-12 range at the top of the canvas. Tap Reroll — say it lands on "6 and 8". The student answers "48". Reroll — "3 and 9". Student answers "27". Cycle for five minutes. The tutor's hands stay free (no rolling a physical die, no thinking up the next problem), the pacing stays fast, and the pair of digits is genuinely random so the student can't game the order. That specific "the tutor doesn't have to invent the next problem" mechanic is what buys back the engagement in a fluency drill that would otherwise stall on tutor decision-latency.

The Number Roller is Platinum-only — on Free and Pro the tile appears in the Math Tools panel but locked behind a Platinum upgrade prompt. Enterprise / B2B tutors are provisioned at the Pro feature level and every Platinum tool is hidden rather than locked, so a business-plan teacher won't see the roller as an option at all. If your caseload doesn't lean on fluency drill (say, you tutor conceptual work only and pass drill to a between-lesson app like Times Table Rock Stars), the roller is skippable; if it does, it's the specific Platinum feature that earns the upgrade.

Math Symbols for the multiplication sentence

The written notation for multiplication is × (proper multiplication sign, U+00D7) and ÷ (proper division sign, U+00F7) — not the keyboard asterisk "*" or forward slash "/". Both live in the Math Symbols palette's Operations category on the Pro tier, alongside +, −, and =. The palette's two entry points (Whiteboard Library "Math Symbols" folder and the ∑ popup that appears next to a selected text or sticky) are described in depth in the math notation spoke. For multiplication work specifically, the workflow is: build the array or the repeated-addition picture with rods, drop a text box next to it, type the coefficient digit, click × from the palette, type the second coefficient, click =, type the answer. The palette's × and ÷ read as the notation the student sees in their textbook, so what's on the whiteboard matches what's on the page.

A working sequence for teaching multiplication online

Most elementary curricula build multiplication in roughly the same order: repeated addition first, then arrays as a formal model, then skip-counting the anchor tables (2s, 5s, 10s), then the derived tables (3s, 4s), then the harder tables (6s, 7s, 8s, 9s), then fluency retrieval, then multi-digit and long multiplication. What changes online isn't the sequence — it's which tool on the Math Tools panel each stage reaches for. Grade bands below are typical US placements, not a standards mapping — adapt to the curriculum and the student in front of you.

Multiplication as repeated addition (grades 2 to 3). Place two 4-rods end to end in a row. Ask the student what one rod would exactly match their combined length — they'll reach for an 8. Write "4 + 4 = 8". Now place a third 4-rod at the end, extending the row to twelve. Write "4 + 4 + 4 = 12". Now write "3 × 4 = 12" underneath, and read it as "three lots of four". The rods are the referent for "three lots"; the notation is the shorthand. Repeat with a couple of other small products (2×5, 3×3, 4×2) so the referent-and-shorthand pairing settles before moving on.
Multiplication as an array (grade 3). Take the same "3 × 4" and rearrange: three 4-rods as three separate rows of four, one above another. Point out that the total is still 12 — the arrangement changed, the quantity didn't. This is the concrete introduction to the commutative property. Then rotate the picture (mentally, or by placing four 3-rods in four rows of three) and ask the student what the total is. They'll count 12 either way. Write "3 × 4 = 4 × 3 = 12" underneath. This is the moment the "you only have to memorise half the times-table grid because 6 × 7 is the same as 7 × 6" idea lands, and it's much cleaner as a rod picture than as a spoken rule.
Skip-counting the anchor tables — 10s, then 5s, then 2s (grade 3). Place a 0-100-by-10 number line. Ask the student to hop the tens from 0 — pen arcs from 0 to 10, 10 to 20, 20 to 30, saying each number aloud. Ten hops takes them to 100. Now the 10 times table is "how many hops did I do?" — one hop is 10, two hops is 20, three hops is 30. Repeat with the 0-100-by-5 line for the 5s, and the 0-20 line (steps of 1) for the 2s (double-hopping). Write each times table out in a text box next to the line as the student says it — the notation and the picture reinforce each other.
The 3s and 4s on the 0-30 line (grade 3). Place a 0-30 line (steps of 1). Skip-count the 3s from 0 — pen arcs three at a time, 0-3, 3-6, 6-9, up to 30 (which is 3 × 10). Same line for the 4s, hopping four ticks at a time up to 4 × 7 = 28. Write the times table out as the student says it. Because the ticks are individual, this is where the "3, 6, 9, 12…" chant gets tied to the visual arc — the number of arcs is the multiplier.
The hard tables — 6s, 7s, 8s, 9s — on a custom line (grades 3 to 4, Platinum recommended). The preset lines don't stretch to the natural end of the 6s (72), 7s (84), 8s (96), or 9s (108). On Koala Platinum, the "make your own" custom line lets you build a "0 to 72 in steps of 6" line for the 6s, a "0 to 84 in steps of 7" for the 7s, and so on. Skip-counting the 6s from 0 to 72 on such a line ties each fact ("6 × 4 = 24", "6 × 7 = 42") to the specific tick the student's arc lands on. This stage genuinely needs Platinum — on Pro there's no preset that ticks at 6, 7, 8, or 9, so the honest move is to teach the hard tables through the derived-facts strategies in the next step rather than to fake a line that doesn't fit.
Derived-facts strategies for the hard tables (grades 3 to 4). Once the anchor tables (2s, 5s, 10s) are solid, the derived-facts strategies collapse a lot of the memorisation load. Two moves worth teaching explicitly:
- Doubling the 2s to get the 4s, and doubling the 4s to get the 8s. If the student knows "2 × 7 = 14" from their 2s, then "4 × 7" is just "two lots of 14" — 28. And "8 × 7" is "two lots of 28" — 56. Build the doubling picture with rods: 2×7 as two 7-rods, doubled to four 7-rods, doubled to eight 7-rods, next to a 56-length composed row (five 10-rods plus a 6-rod). The student sees the doubling actually happening.
- The "one less than a ten" trick for the 9s. 9 × 6 is "ten sixes minus one six" — 60 − 6 = 54. Build ten 6-rods in a row (or a 10-rod stacked on top of ten 6-rods rotated vertically), remove one 6, and count what's left. Write "9 × 6 = 10 × 6 − 6 = 60 − 6 = 54" in a text box. The rule stops being magic; it's arithmetic the student did.
Fluency retrieval — Number Roller drills (grades 3 to 4, Platinum for the roller). Once concept and skip-counting are solid, move to short retrieval-practice drills using the Number Roller. One roller showing two numbers on the 1-12 range, five minutes, student answers each roll — and lock one tile to the table you're working on if you want the drill focused rather than mixed. Track which pairs are automatic and which are still slow — the slow ones become the next lesson's rods-and-lines work. If your practice is on Koala Pro rather than Platinum, the drill workflow is a paper flash-card pack the parent handles between lessons; if it's Platinum, the drill happens on the whiteboard where you can see and reset it.
Multi-digit multiplication with a Pro rectangle grid (grade 4 onwards). Introduce two-digit multiplication (say, 23 × 4) with an area-model / grid approach: place a rectangle from the shape library (rectangle is on Pro), split it visually into "20" and "3" columns, and multiply each partial by 4 — 20 × 4 = 80, 3 × 4 = 12, so 23 × 4 = 92. This is the picture-based lattice / area-model version of long multiplication most curricula teach before the columnar algorithm. On Platinum, the shape-labels toggle adds label slots at the rectangle's corners and centre that you type into, so the "20", the "3", and the partial products can sit on the picture itself rather than beside it. The traditional columnar algorithm ("carry the 1") then sits underneath the picture in a text box, with the palette's × drawing the line — picture first, algorithm second.

Most sessions won't move through all eight stages in one lesson; a typical 30-minute grade-3 session picks one or two adjacent stages (say, the 3s on the 0-30 line plus a five-minute Number Roller drill on the 1-6 range for the 2s and 5s from last week), cycles between manipulative and drill, and includes an engagement break. The tools are the same tools; the sequencing is what makes the lesson.

Division as the mirror of multiplication

Multiplication and division are almost the same operation seen from two ends. Once the student can build "3 × 4 = 12" with three 4-rods against a 12-length, they can build "12 ÷ 3 = 4" as the same picture, read backwards: "I have a length of 12. How many 4-rods do I need to match it? Three." Same rods, same total, same picture — different question asked of it. This is the concrete introduction that most division curricula prescribe, and it moves online as cleanly as the multiplication version.

Two patterns:

Division as "how many groups of N?" Start with a Ten-rod and a Two-rod end to end — a length of twelve. Ask the student to build a matching row using only 4-rods. They place three. Write "12 ÷ 4 = 3". The rods are the same rods; the question is different. This is the "quotative" interpretation of division and it's the one that connects most directly to multiplication as repeated addition.
Division as "share into N groups." Same 12-length. Ask the student to split it into three equal groups using rods of one value each. They'll reach for three 4-rods. Write "12 ÷ 3 = 4". This is the "partitive" interpretation, and it's the everyday-language meaning of "share".

Written notation: the palette's ÷ (proper division sign, Pro tier) drops into the same text box the multiplication sentence lives in. Long division (grades 4 to 6) works the same way as long multiplication above — set up the traditional layout in a text box, work through digit by digit, and use the rectangle / area-model picture from step 8 above when the student stalls on why the algorithm is doing what it's doing.

Keeping fluency drills from feeling like drill (for younger learners)

The engagement problem multiplication tutoring hits hardest is the fluency-drill stage. A grade-3 student who's happily built arrays for twenty minutes hits the wall at minute twenty-five if minute twenty-six is a five-minute stream of "6 × 7 = ?". The Number Roller helps by making the drill feel like a game (the roll is random, the tutor didn't invent the next problem) — but the engagement layer of the classroom does the rest of the lift, and it's the same layer described in depth in the gamified virtual classroom answer. For a multiplication-heavy caseload specifically:

Gems for each correct roll, not for the whole set. Each student on Koala Go has a visible Gem balance; the tutor gives 1-100 Gems from the student panel in a single click, and the student sees the balance grow live. A drill rhythm of "roll → answer → 1 Gem correct, 1 Gem for a good near-miss" is more engaging than "1000 Gems if you get 20 in a row" — small immediate reinforcement wins over big deferred reward for this age band.
The Playground as a short reset between drill blocks. A two-to-three-minute walk around the shared 3D space between two blocks of drill is a lower-friction attention reset than "let's take a water break" — the student doesn't leave the tab, and the reset happens inside the same classroom.
Track drill progress on the whiteboard, visibly. Drop a text box: "Fours: automatic. Sixes: getting there. Sevens: still slow." Update it at the end of each drill block. The student sees a visible record of the tables that have moved from "still slow" to "getting there" to "automatic" over three or four sessions. Parents notice this too — see below.

A note on students who need more time on the harder tables: some students genuinely need more repetitions and more time than a curriculum's default pacing, and that doesn't necessarily mean anything clinical. Slow it down, hold on the anchor tables until they're truly automatic before adding a new one, and lean heavily on the "derived facts" strategies from step 6 so the student is doing arithmetic rather than trying to memorise thirty-six new facts. If a specific student's difficulties with multiplication are consistent enough that you're wondering about a math learning difference, that conversation belongs with the family and a qualified specialist, not with a webpage — the dyscalculia answer walks through the SEN-specific pattern where multiplication needs the concrete-first stage held longer.

Practice, homework, and the multiplication sites you already use

Between-lesson practice is where multiplication fluency actually consolidates. Most working online multiplication tutors have a small set of practice sites they trust for drill and games — Math Playground, Prodigy, Khan Academy's multiplication units, times-table-specific practice sites the student's school already uses. Bringing these into the lesson via the cobrowser rather than sending them as a link earns its keep in exactly this workflow. A cobrowser (explained in what is a cobrowser?) is a shared browser window inside the classroom: you open the site, and both you and the student can click, type, drag on the live page — actual shared control. Two multiplication-specific patterns that work:

Rods on the whiteboard, then a matching drill on a practice site. Ten minutes on the whiteboard building the 6s array with rods, then ten minutes in the cobrowser on a "match the fact to the answer" game for the 6 times table. The rods built the concept; the site consolidates it in a different modality without the student leaving the lesson room.
Practice-site question first, rods to debug. Open a mixed-times-table drill on a practice site. If the student stalls on "8 × 7 = ?", pause the site and switch back to the whiteboard — build the array (eight 7-rods, or seven 8-rods), work it out together (56 as ten 5-rods plus one 6-rod, or 56 as five 10-rods plus a 6-rod), and then return to the site. The debug loop between the abstract problem and the concrete picture is one of the strongest teaching moves an online multiplication tutor has.

Two honest notes on the cobrowser: on Koala Free, sessions are capped at 10 minutes per session with a 20-minute cool-down — fine for a single practice segment in a 30-minute grade-3 lesson, but restrictive if you want the practice site as the primary surface. Full-length cobrowser access is Koala Pro and up. For between-lesson practice on a paper worksheet, you can upload it as a PDF (PDF/PowerPoint upload is a Pro-tier feature) and both of you annotate on top of it at the start of next week's lesson — the workflow is covered in how do I give homework to online tutoring students?

Common failure modes and honest limits

Things that go wrong in online multiplication lessons, and the fix:

Teaching times tables as pure chant before any picture. A student who has memorised "3, 6, 9, 12, 15…" as a chant but can't build "3 × 4" as three 4-rods against a 12 has a fragile fluency — the chant works until the question is asked out of order ("what's 3 × 7?"), and then they have to start the chant from the beginning. Build the arrays first, chant later. The chant is the compressed form of the picture; the picture has to exist before the compression means anything.
Skipping arrays and jumping to memorisation. A student who has never physically arranged three rows of four and counted them has no anchor for "3 × 4 = 12" when the fact fades. The array is not decoration; it's the mental model the student calls back to when the memorised fact isn't automatic yet.
Using "*" or "x" instead of ×, and "/" instead of ÷. The keyboard's asterisk and letter-x are not the multiplication sign; the forward slash is not the division sign. The palette's × (U+00D7) and ÷ (U+00F7) read as the notation in the student's textbook. Use them. The math-notation spoke covers the mechanics.
Drilling every table at once instead of one at a time. A grade-3 student can hold one new times table at a time in active learning. Introduce the 10s, drill the 10s to automatic before adding the 5s, drill the 5s to automatic before adding the 2s, and so on. Trying to work all twelve tables in a mixed drill before the anchors are solid produces confused fluency.
Ignoring the commutative property to halve the memorisation load. The times-table grid is symmetric — a student who has "7 × 8 = 56" also has "8 × 7 = 56". Point this out explicitly when introducing arrays; the rod picture proves it visually. Kids who don't notice this halving spend twice as long as they need on the same tables.
Running drill without an anchor line to fall back on. When a student stalls on "6 × 8 = ?" in the middle of a Number Roller drill, if there's no 0-100-by-5 line or 0-100-by-10 line on the canvas to skip-count from, the student either guesses or gives up. Keep the anchor line for the current times table visible on the canvas during drill, even if it's off to one side, so the student always has a place to reason from.
iPad-only workflows. Koala Go works on iPad, but placing precise rod arrays and drawing skip-counting arcs is fiddlier on a finger than on a mouse. If a student uses iPad exclusively, plan around larger rods (arrays of Sixes and Sevens are easier to grab than arrays of Twos and Threes), and use the Text tool for the notation side more heavily.

Two limits worth naming plainly rather than papering over:

The Number Roller is Platinum-only. On Koala Free and Koala Pro, the Number Roller tile appears in the Math Tools panel but is locked behind a Platinum upgrade prompt. Enterprise / B2B plans are Pro feature level and don't include Platinum tools at all. If fluency drill is a heavy share of your practice, factor that into the plan choice; if it isn't (say, you focus on conceptual and calculation work and pass drill to a between-lesson app), the Pro tier is enough.
The "make your own" number line is Platinum-only. The Pro tier's eight preset lines cover the 2s, 5s, 10s, and 3s cleanly (via the 0-30, 0-100-by-5, and 0-100-by-10 lines). The 6s, 7s, 8s, and 9s at their natural top ends (72, 84, 96, 108) need the Platinum custom line, and there's no clean Pro substitute — no preset ticks at 6, 7, 8, or 9. On Pro you teach those tables through derived facts and the roll-free strategies above instead of on a line.

Sharing multiplication progress with parents

Times-table progress is one of the parts of tutoring parents most want to hear about — it's concrete, measurable, and lines up with the report cards that come home. The highest-signal parent update names the specific table that moved from "still slow" to "automatic" this week, quotes a drill stat, and gives a specific home-practice ask.

A few concrete moves:

Screenshot the drill-progress text box each week. The whiteboard state persists between sessions, so at the end of each lesson your "Fours: automatic. Sixes: getting there. Sevens: still slow." text box is still there. A screenshot in the parent update is evidence of progress in a way "great lesson today!" isn't.
Send a specific practice ask. Not "please practise multiplication this week" (vague) — "please run through the 6 times table twice a day, forward from 6, then backwards from 60, and let us know which pairs took longest." The parent can verify it, and the "which pairs took longest" comes back to next lesson as the specific rods-and-lines work. Cadence covered in how do I share progress updates with parents?
Record short clips of key breakthrough moments where consent is in place. Koala Pro includes local recording; Koala Platinum adds cloud recording of scheduled lessons. A 60-second clip of the child answering ten random 7×n rolls without hesitation for the first time is worth every second of the setup. Get the family's written permission before recording anything with a minor; the storage, sharing, and consent detail is covered in how do I record online tutoring lessons?

Practical setup on Koala Go for a multiplication-heavy caseload

If you're evaluating Koala Go for K-5 multiplication tutoring specifically, here's the practical picture and where the current lines sit:

Whiteboard is the lesson surface. The classroom's whiteboard is a Fabric.js canvas with the Math Tools panel: ten Cuisenaire-style Number Rods valued 1-10 (Pro) for arrays and repeated addition, eight preset Number Lines including 0-30, 0-100-by-5, and 0-100-by-10 (Pro) for anchor-table skip-counting, and the Math Symbols palette (Operations set on Pro) for the ×, ÷, and = glyphs. The rectangle from the base shape library (Pro) is the tool for area-model / grid multiplication for two-digit and multi-digit work.
Number Roller for fluency drills (Platinum, $49.99/month monthly, $39.99/month annual). The Pro tier ($25.99/month monthly, $21.99/month annual) covers everything through the concept and skip-counting stages, and is the plan most K-3 multiplication tutors run on. The Platinum tier adds the Number Roller for randomised drill and the "make your own" custom line for the 6s-through-9s natural-range lines — the specific additions a grade-3-through-5 multiplication-heavy caseload leans on.
Upload existing multiplication worksheets. On Koala Pro and up, upload the student's times-tables worksheet as a PDF or PowerPoint and annotate on top of it. Uploaded material stays on the whiteboard between lessons.
Cobrowser for practice sites. Open Math Playground, Prodigy, Khan Academy's multiplication units, or whatever times-table practice site your programme uses; the student clicks and drags on the live page inside the lesson room. 10 minutes at a time on Free with a 20-minute cool-down; unlimited on Pro and up.
Playground and Gems for the drill-engagement layer. The K-3 age band leans on the engagement layer to sustain fluency drill blocks; older students lean less on it. The mechanic is described in what is a gamified virtual classroom?
Recording for parent review. Local recording is on Pro; cloud recording of scheduled lessons comes with Platinum. Either supports the "send a 60-second clip of the breakthrough" pattern above.
Free tier limits worth knowing. Koala Free caps sessions at four students and includes the whiteboard plus a capped cobrowser — but the Math Tools panel (rods, number lines, symbols, roller) is Pro-and-up. Free is a trial surface for multiplication tutoring, not the day-to-day.

Honest limits worth flagging that show up in real multiplication work regardless of platform: iPad is rougher than desktop for precise-array placement and skip-counting arc drawing; students in mainland China sometimes hit connection issues (we route through Hong Kong proxies but the Great Firewall is a moving target — see teaching students in mainland China for the workaround pattern); the Number Roller and "make your own" number line are Platinum-only (the eight preset lines cover most needs on Pro, but the 6-through-9 times-tables workflow is cleaner on Platinum); and the palette is a symbols set rather than an equation renderer, so long multiplication's columnar layout is a text box laid out manually rather than a rendered algorithmic template.

Why teaching geometry online is a specific kind of hard

Geometry is the subject in the elementary-and-secondary math curriculum that most obviously wants a physical object in the child's hand. In an in-person lesson you'd reach for a set of pattern blocks to sort by number of sides, a protractor to measure the angle between two lines, a paper triangle to fold along its line of symmetry, a set of plastic solids to count faces on, a compass and straightedge to bisect an angle. Almost none of that survives a webcam and a screenshare. The three problems below are what every online geometry tutor names in some form.

The shapes-as-objects problem. A drawn shape on a slide is a picture; a geometric object is a thing the student can move and compare. "Is this rectangle a square?" is answered differently when the student can drag the corner and watch the sides change than when they're looking at a static image. "How does the hexagon compare to two trapeziums?" is a question that only lands when both are on the same canvas and the student can slide the trapeziums into the hexagon's outline. Screensharing a slide gives you the picture; it doesn't give you the composition. And pen-drawing a rough shape on the whiteboard gets you neither: it's neither the picture nor the manipulable object, and the sides are lumpy.
The notation problem. Real geometry writes in symbols keyboards don't ship — the angle mark (∠ABC), the triangle glyph (△ABC ≅ △DEF), perpendicular (AB ⊥ CD), parallel (AB ∥ CD), degrees (60°), the square-root (√). A student typing "angle ABC = 90 degrees" in a Google Doc is doing the mental work of writing "∠ABC = 90°" without any of the pattern-recognition the notation is meant to trigger. Every symbol they don't have easy access to is a small paper cut that slows the lesson and blurs the notation into English.
The construction and dynamic-geometry problem. Later geometry — from about middle school onwards — is about processes, not just shapes: bisecting an angle with a compass, dropping a perpendicular from a point to a line, constructing a regular hexagon inscribed in a circle, translating a shape by a vector, reflecting a shape across a line, rotating it about a point. In a physical classroom this is a compass-and-straightedge or transparency-and-marker workflow; on paper it doesn't animate; on a static whiteboard it's a snapshot, not a construction. The move that closes this gap online is a paired dynamic-geometry tool (GeoGebra is the standard) opened inside the classroom via the cobrowser — but the whiteboard side still has to carry the definitions, the theorem statements, the notation, and the discussion. Neither surface alone does the whole job.

An online geometry workflow that works has an answer to each: a shape library where each shape is a first-class object the student can grab; a notation palette that puts ∠ ° ⊥ ∥ △ √ one click away; and a cobrowser pairing pattern for the constructions and transformations the static whiteboard can't do. The rest of this page is that answer, on the tools Koala Go ships.

The shape library — 2D and 3D shapes as first-class objects on the whiteboard

The Math Tools panel in Koala Go's Whiteboard Library has two folders on the shape side: 2D Shapes and 3D Shapes. Every shape in either folder is a first-class object on the shared canvas — you or the student can pick one, drop it, drag it, resize it, rotate it, recolour its fill and border, or delete it, and the change syncs to the other side in real time. That's the primitive that makes online geometry teachable: the shape isn't a picture you're describing, it's an object the student is composing with.

2D shapes: what you get on which tier

The 2D folder ships with 21 shapes across two tiers. The Pro tier ($25.99/month monthly, $21.99/month annual) covers the elementary and general-purpose set — square, rectangle, circle, equilateral triangle, plus straight line, dashed line, and arrow as 1D line primitives that live in the same folder. That's 7 shapes, and it's enough for most K-5 geometry: shape recognition, symmetry, area and perimeter of rectangles and circles, basic angle work with equilateral triangles, and any lesson that needs a straight-line diagram.

Platinum ($49.99/month monthly, $39.99/month annual) adds the fuller polygon library plus non-equilateral triangles and curved 2D shapes. The additional 14 shapes:

Regular polygons and other quadrilaterals: rhombus, parallelogram, trapezium, kite, pentagon, hexagon, octagon.
Non-equilateral triangles: isosceles triangle, scalene triangle, right-angle triangle. Middle-school geometry leans on these — the specific "here's a right-angle triangle, work out the missing side with Pythagoras" moment doesn't work if the only triangle in the folder is equilateral.
Curved 2D shapes: oval, semicircle, sector, segment. The sector shape is the one to know if you teach the sector-area formula A = (θ/360) × πr² — it renders as a pie-slice you can label the angle and radius on, and it pairs cleanly with the Fraction Circles from the fractions spoke for the "a sector is a fraction of a circle" moment.

Every shape has a rigidity — either fixed (uniform scaling on corner drag, so a square stays a square and a circle stays a circle) or flexible (independent width/height scaling, so a rectangle can become tall and narrow and an oval can stretch into an ellipse). Fixed rigidity is what you want when the shape's identity depends on its proportions (the square, the equilateral triangle, the regular polygons, the circle). Flexible rigidity is what you want when the shape is a rectangle-family object whose whole point is that the width and height are independent (the rectangle, the parallelogram, the trapezium, the oval, all three non-equilateral triangles — isosceles, scalene, and right-angle — and the lines).

3D shapes: what you get on which tier

The 3D folder ships with 10 solids across two tiers. The Pro tier includes exactly one 3D shape — the cube. That's it. Every other 3D solid is on Platinum. This is a real product boundary to know about: a Pro tutor who teaches faces, edges, and vertices of solids has one solid in the folder, and everything else needs Platinum.

Platinum adds 9 more 3D solids, covering the standard middle-school and secondary-geometry set:

Prismatic solids: cuboid, triangular prism.
Round solids: cylinder, sphere, hemisphere, cone, torus.
Pyramidal solids: square pyramid, tetrahedron.

Two honest things to know about the 3D solids specifically. First, they're rendered as shaded 2D projections of a solid — you can drag, resize, and rotate them in the canvas plane, but you can't spin the solid in space to see a different face. For teaching a student to count faces, edges, and vertices, to compute surface area and volume from given dimensions, or to identify a solid from its properties, the projection is fine — you point at a face and the student counts. For teaching a student to visualise the same solid from a different angle, a paired GeoGebra 3D tab (see below) does more than the projection can. Second, the labelled edges and vertices on a 3D projection are always the ones the projection shows — the "back" edges and vertices of a cube, for instance, are drawn but aren't as directly clickable as the front ones, so if you're teaching a "how many edges does a cube have?" lesson, the labels do more of the work than the object itself. This is normal for any 2D-rendered solid; it's the trade-off of the projection format.

Shape labels: side lengths, angle measures, and vertices attached to the shape

The single feature that most changes what geometry teaching feels like on a shared whiteboard is shape labels. When labels are enabled (via the Show shape labels toggle in the Math Tools panel — Platinum), every placed shape shows a small "+" placeholder at each of its anchor points: corners for polygons, edges for circles and ovals, tips and base corners for pyramids and cones and sectors. Click any "+" and it drops an editable text box anchored to that shape point. Type "6 cm", "60°", "A", or whatever the lesson needs, and the text stays attached — when you drag the shape, the label moves with it; when you resize the shape, the label repositions to keep its relative place near the anchor.

What this means for a working geometry lesson:

Vertices labelled with capital letters. Drop an equilateral triangle, click the top corner and label it "A", click the bottom-left corner "B", click the bottom-right "C". Now you can write ∠BAC on the whiteboard, refer to "△ABC" in the notation, and the student sees where each vertex is on the picture. This is the base workflow for any theorem-based lesson from middle school onwards.
Side lengths on rectangles, triangles, and polygons. Drop a right-angle triangle. Anchor labels to the two legs — "3 cm" and "4 cm" — and to the hypotenuse — "5 cm" (or leave the hypotenuse blank as the value to work out). Now you have a labelled Pythagoras problem the student is looking at while writing the answer in a text box next to the picture. Anchor labels rescale correctly when you resize the triangle, so if you scale it up to make it easier to see, the labels stay where they should.
Angle measures. Drop a scalene triangle (Platinum). Label each vertex with its angle measure — "60°", "80°", "40°" — and the labels move as you drag the shape. Now the "sum of interior angles of a triangle" question is a visible fact.
Radius and diameter on circles. Drop a circle, anchor a label to the edge with "radius = 5 cm", and the label stays where it should when the circle is repositioned.

Two things worth knowing plainly about how labels work. First, labels are text, not live-computed values — you type the measurement, and the text doesn't update automatically if you resize the shape. So a "6 cm" label on a rectangle that you then stretch to twice its width still reads "6 cm" until you edit the text. This is normal for a labelling system on a static whiteboard; live measurement of a drawn or dragged shape belongs on GeoGebra, in the cobrowser pairing pattern below. Second, enterprise and B2B tutors are provisioned at the Pro feature level and Platinum-only controls are hidden entirely from their UI — meaning an enterprise tutor doesn't see the labels toggle at all, and the labelling workflow above isn't available to them without a plan upgrade the enterprise administrator would have to arrange. On Free and standard Pro, the labels toggle is visible but locked behind a Platinum upgrade prompt.

Pro-tier workaround when you don't have labels: place regular text boxes next to the shape by hand. Fine for a lesson with one or two labelled shapes; painful for a lesson with a dozen labelled polygons and re-arrangement — every text box has to be re-positioned when you move the shape, because the boxes aren't anchored. If your caseload is geometry-heavy and you're on Pro, the labels toggle is the single strongest reason to look at Platinum.

Geometry notation on the whiteboard — degrees, angles, radicals, perpendicular, parallel

The Math Symbols palette on Koala Go's whiteboard has a Geometry category with the symbols you'd reach for in a geometry lesson. On Koala Pro you get ° (degrees) and √ (square root) — the two most-used geometry symbols in elementary and middle-school work. On Koala Platinum you additionally get ∠ (the angle marker used in ∠ABC), △ (the triangle glyph used in △ABC ≅ △DEF), ⊥ (perpendicular, as in AB ⊥ CD), and ∥ (parallel, as in AB ∥ CD).

The palette lives in two places on the same whiteboard: a Math Symbols folder in the Whiteboard Library (discovery surface) and a ∑ button that appears next to any selected text box or sticky note (speed surface, inserts at the cursor). Both entry points and the six common notation moves are walked through in the write math equations online spoke; here we just call out the geometry-specific patterns:

Angle measures on a labelled shape. Anchor a label to a vertex and type the measure straight in — "60°" (the degree sign is Option+Shift+8 on a Mac, and long-press the 0 key on an iPad). Worth knowing before you build a lesson around it: the palette's ∑ button appears next to text boxes and sticky notes, not next to shape labels, so palette-built notation belongs in a text box beside the shape rather than inside the label itself. Pair the two — labelled shape, notation in a text box next to it — for the full "here's an angle-sum-of-a-triangle" picture.
An angle name in a theorem statement. Type "The base angles of an isosceles triangle are equal, so ", click ∠, type "ABC = ", click ∠, type "ACB". Now you have a properly-notated theorem statement next to the labelled shape.
A perpendicular or parallel claim. "Since AB ⊥ CD, ∠ADB = 90°." Type "AB", click ⊥, type "CD, ", click ∠, type "ADB = 90", click °. Reads correctly as geometry notation the student will see in an exam.
A congruence or similarity statement. Type "So ", click △, type "ABC ≅ ", click △, type "DEF" — congruent-triangles notation matches the textbook. (The ≅ glyph itself isn't in the palette — if a lesson needs it heavily, the workaround is the ≈ symbol from the Compare category on Platinum, or "==" spoken as "is congruent to". Same known-limit as the notation spoke's honest "not a full equation editor" section.)
Answers with a radical. Type "The diagonal of a unit square is ", click √, type "2." — Pythagoras answers where the value doesn't simplify to a rational stay in radical form.

For the full palette walkthrough, the entry-point mechanics, and the honest limit on stacked fractions and rendered LaTeX layouts (which do bite on late-secondary geometry work — no radicand-under-a-bar on √, superscripts stop at ⁿ, no matrix layouts), see the notation spoke.

Constructions and dynamic geometry — pair a cobrowser tab into GeoGebra

The shape library above is a great labelling and discussion surface — students can compose with shapes, annotate them with notation, and reason about their properties. What it is not is a construction environment. If a lesson needs compass-and-straightedge constructions, transformations that update dependent measurements live, or a locus of points that traces as the student drags a vertex, the honest answer is: pair a cobrowser tab into GeoGebra alongside the whiteboard.

A cobrowser is a shared browser window inside the classroom — you open a page and both you and the student can click, type, and interact with the page live, not screen-sharing but actual shared control. GeoGebra (the standard free dynamic-geometry tool) opens cleanly in the cobrowser; both of you can use the compass tool to bisect an angle, drag a vertex to see how a shape's other measurements respond, translate/rotate/reflect a shape and watch the image position update, or trace a locus of points on a moving construction. The mechanic is covered in depth in the what is a cobrowser? explainer.

The pairing pattern most working online geometry tutors converge on is:

Whiteboard side (Koala Go's Math Tools). Definitions, theorem statements with the ∠ △ ⊥ ∥ ° notation, labelled shapes as pictures to reason about, PDF past-paper problems annotated with palette symbols, static diagrams for a proof.
Cobrowser side (GeoGebra or a similar dynamic-geometry tool). Actual constructions (bisect this angle, construct the incentre, inscribe a regular hexagon), transformations (translate △ABC by the vector, rotate about the origin), dynamic drag-a-vertex investigations, coordinate-geometry work where the point moves and the equations update.

On Koala Free, the cobrowser is capped at 10 minutes per session with a 20-minute cool-down. That's enough for a short GeoGebra demo but too tight for a full lesson driven from GeoGebra. On Koala Pro (and up) the cobrowser is unlimited. So for a geometry-heavy caseload the Pro-and-up floor for cobrowser is functionally the same practical floor as it is for exam-prep and general tutoring workflows.

Other tools in the same pairing pattern that some tutors prefer: Desmos Geometry for its simpler interface (better suited to middle-school constructions than GeoGebra's full toolbar), Mathigon Polypad for the virtual-manipulatives style of geometry work, and subject-specific applets from PhET Interactive Simulations for physics-adjacent geometry (light rays, projectile paths). Each opens in the cobrowser the same way.

A working sequence for teaching geometry online, by age band

Different age bands lean into different parts of the toolkit. Grade bands are typical US placements, not standards claims — adapt to the student and curriculum in front of you.

K-2 (ages 5-7): shape recognition and simple properties. Use the Pro 2D set — square, rectangle, circle, equilateral triangle — plus the Pro cube. Sorting activities ("place the shapes that have four sides on the left"), matching ("point at the shape that has three corners"), simple symmetry ("draw a line down the middle of the square that makes two matching halves"), first vocabulary (sides, corners, vertices). Notation is minimal at this age. Playground / Gems / avatars carry the engagement layer between short activities — see the gamified virtual classroom explainer. Every Koala Pro tutor has what they need for K-2 geometry without touching Platinum.
Grades 3-5 (ages 8-10): area, perimeter, and the introduction of angles. Area and perimeter of rectangles and squares (Pro shapes cover it) with side-length labels typed in text boxes next to the shape (Pro workaround) or attached as anchor labels (Platinum). Introduction to angle types — acute, right, obtuse — with the ° symbol from the palette. Introduction of 3D solids in the curriculum happens here in many US programs, and this is the tier line where Platinum starts to earn its keep: Pro has the cube; Platinum has cuboid, triangular prism, cylinder, sphere, hemisphere, cone, square pyramid, tetrahedron, torus — the standard elementary-and-middle-school 3D solids set. A Pro tutor teaching "count the faces on a triangular prism" has to draw one; a Platinum tutor drops one from the folder.
Middle school (ages 11-13): the full polygon set, angle theorems, volume, surface area, and the introduction of formal notation. The Platinum polygon library (rhombus, parallelogram, trapezium, kite, pentagon, hexagon, octagon; isosceles / scalene / right-angle triangles) opens up angle-sum theorems, interior/exterior angles, the naming of polygons by their number of sides. The full 3D-solid Platinum library carries the volume and surface-area formulas — cylinder, cone, sphere, square pyramid, triangular prism — with dimensions labelled on the picture (Platinum shape labels). The ∠ △ ⊥ ∥ symbols from the palette (Platinum) come in with the introduction of formal notation. Shape labels (Platinum) become almost required here — a lesson on angle-sum-of-a-quadrilateral is a lot less useful without the four vertex angles labelled on the shape. This is the tier at which most geometry-focused tutors move to Platinum.
Secondary (ages 14-17) and adult learners: coordinate geometry, transformations, constructions, and proofs — whiteboard for notation, GeoGebra in the cobrowser for the dynamic work. Coordinate geometry uses the -10 to 10 number line from the number-sense spoke as a ready-made x-axis, with pen and shape tools to draw the y-axis; the shape library lets you drop lettered points and lines. But most of the dynamic work at this level (bisecting an angle, dropping a perpendicular, reflecting a triangle across a line, constructing an incentre) belongs on the paired GeoGebra tab — the whiteboard is where the theorem, the notation, and the proof-writing live. Uploaded past-paper PDFs (Pro) become the primary surface for exam-focused geometry — see how to teach exam prep online for the SAT / ACT / A-level / IB exam-focused version of this workflow.

Practice, homework, and past-paper geometry problems

The shape library and notation palette carry the teaching; consolidation runs on practice. Two patterns most working online geometry tutors converge on:

Upload the student's homework worksheet as a PDF and annotate on top of it. On Koala Pro, PDF and PowerPoint upload sits alongside the shape library — the worksheet becomes the lesson surface, both you and the student annotate on the same page with pens, shapes, and text boxes. For a geometry worksheet with printed shapes and blank angle-measure lines, this collapses the "here's your homework, let's go through it" workflow from a screenshare-and-narrate into a real shared workspace. Shape labels attach to shapes you drop on top of the PDF the same way they do on a blank canvas.
Cobrowser to interactive geometry practice sites. Bring the student into the practice, don't send them out to it. Khan Academy has strong geometry practice sets across grade bands; IXL has structured skill practice; GeoGebra has a large community-authored applet library for specific concepts; Brilliant.org has courses in Euclidean geometry. Open the site in the cobrowser and both of you interact with the page live — you see exactly what the student typed, what shape they dragged where, and can jump in without them window-switching.

Between-lesson practice for geometry-heavy students works best as a very small, specific ask — "draw three quadrilaterals in your notebook with your parent, one that has all four sides equal, one that has exactly one pair of parallel sides, one that has none" — rather than "practise geometry for 30 minutes." The homework workflow that most tutors converge on is covered in how do I give homework to online tutoring students?

Common failure modes and honest limits

Things that go wrong in online-geometry workflows, and the fix:

Trying to draw shapes with the pen tool rather than the shape library. A hand-drawn rectangle on a whiteboard is a picture; a Rectangle from the shape library is an object with corner anchors for labels, uniform resize handles, and a real fill and border. The pen is for freehand annotation (an arrow, a construction line, a scribbled note); the shape library is for shapes the lesson is about. Every geometry lesson should start with the shape from the folder and annotate on top with the pen, not the other way round.
Placing text boxes near shapes on Pro and expecting the labels to move with the shape. Text boxes on Koala's whiteboard are independent objects — they don't follow the shape when you drag it, so a labelled shape you re-arrange has to be re-labelled by hand. Anchored shape labels (Platinum) are the fix: labels attached to the shape's corner/edge/tip anchors move and resize with the shape. If you're on Pro and doing more than a couple of labelled shapes per lesson, this is the friction point that pushes most tutors to Platinum.
Expecting the 3D solids to rotate in space to show a different face. Koala Go's 3D solids are rendered as shaded 2D projections — you can drag, resize, and rotate them in the canvas plane, but you can't spin the cube to see the back face. This is fine for counting faces / edges / vertices, computing surface area and volume from labelled dimensions, or identifying a solid from its properties. It is not a substitute for a 3D-manipulable model when a student needs to visualise how a solid looks from a different viewpoint. Pair a GeoGebra 3D tab in the cobrowser for that specific case, or a physical set of solids the student has at home.
Expecting shape labels to auto-measure the shape. Labels are text boxes anchored to shape anchors; the tutor types the value ("6 cm", "60°", "A"). If you resize the shape, the label repositions correctly but the text doesn't update — a "6 cm" label on a rectangle you then stretch still reads "6 cm" until you edit it. This is normal for a labelling system on a static whiteboard; live measurement of a drawn or dragged shape belongs on GeoGebra.
Trying to construct a shape from scratch on the whiteboard. Bisecting an angle, dropping a perpendicular, inscribing a circle in a triangle, constructing a regular hexagon by compass-and-straightedge — none of these are possible from the whiteboard's tools, and no combination of pen + shape + line will approximate the mechanic. This is the honest "pair GeoGebra in the cobrowser" limit — for any lesson whose whole point is the construction, the shared shape library is the wrong surface. Whiteboard for the definition and the theorem; cobrowser tab for the compass.
Free-tier tutors seeing every shape locked. The 2D and 3D folders in the Whiteboard Library show a locked banner on Free — the shape library is a Pro-and-up feature. A tutor trialling Koala Go for geometry needs at least the trial or Pro plan to evaluate.
iPad-only workflows. Koala Go works on iPad, but the precision needed to grab a corner anchor on a small shape, drop a label into the "+" placeholder, and drag exactly the right vertex is fiddlier on a finger than on a mouse. If your student uses an iPad exclusively, plan around larger placed shapes (resize the shape up to make anchors easier to hit), and prepare more of the labelling in advance on your side so the student's job on iPad is dragging and arranging shapes rather than fine-anchor picking.

Three limits worth naming plainly rather than papering over:

Every 3D solid except the cube is Platinum. A tutor teaching 3D geometry regularly wants Platinum, plain and simple — Pro's single 3D solid (the cube) covers the "count the faces on a cube" lesson and nothing else. Factor this into the plan choice; a middle-school geometry caseload on Pro is going to be leaning heavily on hand-drawn 3D solids or paired GeoGebra tabs for anything past the cube.
Shape labels are Platinum-only, and hidden entirely from B2B / enterprise UI. A Platinum-tier tutor gets the anchor-labels workflow described above; a Pro-tier tutor uses the manual text-box workaround; an enterprise tutor doesn't see the labels toggle at all (provisioned Pro-level, Platinum controls hidden). For a geometry-heavy B2B deployment, this is a real friction point worth naming to the administrator before the account is set up.
No compass-and-straightedge constructions, no auto-measurement, no dynamic-geometry drag-a-vertex mechanic on the whiteboard. The shape library is a labelling and discussion surface, not a construction environment. Constructions, transformations that update measurements live, and dynamic geometry pair with GeoGebra in the cobrowser — described above. This is a real product line, not a shipping-soon promise: dynamic geometry lives on the paired tool.

Practical setup on Koala Go for a geometry-heavy caseload

If you're evaluating Koala Go for K-secondary geometry specifically, here's the practical picture and where the current lines sit:

Whiteboard is the workspace. The classroom's whiteboard is a Fabric.js canvas with the Math Tools panel: 7 Pro 2D shapes + 1 Pro 3D shape + 14 Platinum 2D shapes + 9 Platinum 3D shapes + the Math Symbols palette for geometry notation. Both tutor and student can place, drag, resize, rotate, recolour, and delete shapes; the state syncs live and persists across lessons.
Upload existing geometry worksheets and past papers. On Pro, upload the student's textbook page, homework worksheet, or past-paper geometry problem as a PDF or PowerPoint. Annotate on top with pen, shape, and text tools; shape labels (Platinum) attach the same as they do on a blank canvas.
Cobrowser for GeoGebra, Desmos Geometry, Mathigon Polypad, Khan Academy, IXL. Open the site in the classroom's cobrowser and both of you click and drag on the live page. 10 minutes at a time on Free with a 20-minute cool-down; unlimited on Pro. See what is a cobrowser? for the mechanic.
Playground and Gems for the K-2 engagement layer. The youngest grades (pre-K through grade 2) benefit most from the Playground brain break between shape-sorting activities and per-student Gems for immediate effort-based reinforcement. Older students lean into the shapes and notation without needing the engagement layer as heavily. See gamified virtual classroom for the depth.
Recording for parent review or student self-review. Local recording is on Pro; cloud recording of scheduled lessons comes with Platinum. Both support the "send the parent a clip of today's angle-sum-of-a-triangle breakthrough" pattern. See record online tutoring lessons for the consent, storage, and sharing pattern.
Free tier limits worth knowing. Koala Free caps sessions at four students, includes the whiteboard plus image upload plus a capped cobrowser, but the Math Tools panel (shapes, number lines, math symbols, rods, fractions) is all Pro-and-up. Free is a trial surface for geometry, not the day-to-day.

Honest limits worth flagging that show up in real geometry work regardless of platform: iPad is rougher than desktop for narrow-anchor picking; students in mainland China sometimes hit connection issues (we route through Hong Kong proxies but the Great Firewall is a moving target — see teaching students in mainland China for the workaround pattern); and every 3D solid past the cube plus shape labels plus the ∠ △ ⊥ ∥ notation is Platinum, so middle-school and secondary geometry-heavy caseloads pretty much live on Platinum by design.

Where Koala Go isn't the right pick for geometry

Two honest tradeoffs worth naming, so you know before you sign up:

Dedicated dynamic-geometry as the primary surface. If every lesson you teach is a compass-and-straightedge construction or a drag-a-vertex-and-watch-the-measurements-update investigation, GeoGebra as the primary teaching surface (with a separate video-call tool for the face and voice) can be a cleaner fit than pairing Koala Go's whiteboard with a GeoGebra cobrowser tab. Koala Go's shape library is a labelling and discussion surface with a cobrowser pairing for the dynamic work — a strong fit for a mixed caseload (definitions, notation, past papers, discussion, occasional constructions), less obviously the strongest fit for a pure-dynamic-geometry practice.
Whole-class delivery to a 25-student geometry class. Koala Go is calibrated for 1-on-1 and small-group tutoring — the Free tier caps sessions at four students. It's not designed for a full school-class synchronous geometry lesson of the kind Nearpod or Pear Deck are built for. For a whole-class shape, look at platforms built for whole-class delivery.

Why teaching algebra online is a specific kind of hard

Algebra is the subject where most students first hit the wall between "this is arithmetic I can just do" and "this is a rule I have to reason about". It's the moment mathematics turns abstract — a letter stands for a number, an equation is a relationship you can act on, a graph is a picture of a function. Online, three specific problems make that transition harder than it needs to be.

The step-by-step problem. Algebra is taught in the shape of a worked solution — write the equation, do the same thing to both sides, write the new equation, and again. Every line matters, because the point of the lesson isn't the answer but the reasoning that produced it. In an in-person lesson the whole solution accumulates on a shared piece of paper, both of you looking at every line at once. On a video call with a screenshare, the tutor's working is visible but doesn't include the student's — the student is watching, not doing. And with a hand-drawn tablet in the corner of a Zoom window, the resolution is low, the aspect ratio is wrong, and the writing runs off the edge halfway through a longer solve. The move that closes the gap is a shared whiteboard where every step of the solution is a first-class text object, both of you can add lines, and the whole worked solution stays on the page.
The notation problem, but specifically for algebra. Prose keyboards ship none of the symbols algebra needs. "x²" is not the same as "x2"; "≤" is not the same as "less than or equal to" spelled out; "√25" is not the same as "sqrt(25)" typed in an ASCII fallback; "(x + 2)(x − 3)" needs proper round brackets and a proper minus sign to line up correctly. A student typing algebra on their side of the call without an easy way to insert these symbols will either abandon the notation ("x*x = 4" is a genuine thing students type when x² is inaccessible) or ask you every three seconds how to get the character they need. Either way the lesson loses momentum. The solved version is a math symbols palette on the shared whiteboard, one click per symbol.
The graphing problem. Algebra past linear equations lives on a graph as much as on a page — you can't teach quadratics without drawing y = x² − 4, or systems of equations without drawing two lines and pointing at where they cross, or transformations of functions without watching a function move. A static pen-drawn graph on a whiteboard is fine for a one-off demo of a slope-intercept form; it is not the surface you teach a slider-based function-family lesson on. Desmos is the standard free browser-based tool for the graphing side of algebra, and the online-teaching move that closes this gap is opening Desmos in a shared browser tab inside the classroom so both you and the student can drag sliders on coefficients, type new expressions, and watch the graph respond live.

An online algebra workflow that works answers each: a shared whiteboard for the worked-solution surface, a palette for the notation, a cobrowser tab for the graphing. The rest of this page walks through the teaching moves each of those enables — starting with the moment a student first meets a variable, and ending with a paired-tools workflow for late-secondary and exam-prep algebra.

Introducing variables: the "let this be x" move on a shared canvas

The single hardest step in an algebra sequence is the very first one: getting a student to accept that a letter can stand for a number, and that an operation on the letter is an operation on the number the letter represents. Two moves that work well on a shared whiteboard, both grounded in physical-manipulative pedagogy imported from elementary math.

Number rods as unknown lengths

The Number Rods in the Whiteboard Library's Math Tools panel (Pro tier — see the number-sense spoke for the deep tool walkthrough) are Cuisenaire-style rods of ten different lengths, one to ten units. For an algebra introduction, they double as unknown lengths: drop a rod on the canvas, label a text box next to it with "x", and now the rod is x. Drop another identical rod next to it — that's another x, so together they are 2x. Drop a smaller rod of length 3 next to them — that's 3, so the whole line is 2x + 3.

The step that makes it stick is turning that composition into an equation: line up your 2x + 3 composition against a longer rod (say, the length-11 rod doesn't exist in the folder, so instead build the target with a length-10 and a length-1) and now the student can see that if 2x + 3 = 11, then 2x = 8, then x = 4. The rod length verifies the answer. This is the concrete side of the balance-scale metaphor below, and it's the pattern that gets an 8-to-10-year-old to internalise "x is just a specific number I don't know yet" faster than any prose explanation.

Balance-scale metaphor for equations

The classical balance-scale metaphor for solving linear equations — the two sides of the equation are the two sides of a scale, and to keep it balanced you must do the same thing to both sides — works on a shared whiteboard as a two-column layout. Draw a vertical line down the middle of the canvas (use the Pro straight-line tool). On the left write the left-hand side of the equation ("2x + 3"); on the right write the right-hand side ("11"). Now the student can see the two sides as two separate objects. Subtract 3 from each side in the next row down; the student writes "− 3" in a text box under each side. The next row down shows "2x" on the left and "8" on the right. Divide each side by 2 in the next row; the final row shows "x" on the left and "4" on the right. Six lines, ten seconds of typing, and the student has watched a solve happen as a sequence of moves rather than as a single leap.

Both the rod pattern and the balance-scale pattern are Pro-tier moves — the rods themselves are Pro, the straight-line tool is Pro, the operations symbols are Pro. A Pro tutor teaching pre-algebra introductions has everything they need for both. Where Platinum starts to earn its keep is later, when the equations involve powers, roots, or inequalities using the Platinum symbols.

Solving equations step by step — the whiteboard is a working page

Once a student has accepted variables, most of pre-algebra and Algebra 1 is a sequence of equation-solving lessons of increasing complexity: one-step, two-step, multi-step, equations with variables on both sides, equations with fractions, equations with brackets, equations with negative coefficients, and eventually linear systems and quadratics. Every one of these is a step-by-step process, and the online workflow that carries them all is the same:

Drop the equation at the top of the whiteboard. Type it into a text box. Use the palette's Pro operations set (+ − × ÷ =) so the minus is a proper minus (U+2212, centres against digits) and the equals is a proper equals. Round brackets from the palette (Pro) for grouping.
Write the next step directly under it. New text box, aligned vertically. If the move is "subtract 3 from both sides", the student writes the annotation "− 3 both sides" or "|− 3" next to the arrow, and then writes the resulting equation on the next line.
Colour the pen or the text if you want to flag a specific move. The whiteboard's pen has multiple colours; the same for text-box border and fill. Use a colour to mark the operation you did to both sides.
The student writes the next step, not you. The whole point of a shared whiteboard is that the student is on the same canvas — so once the pattern is established, the student takes the next line, not you. This is the difference between watching algebra get done and doing algebra. If you want the student to try the whole rest of the solve, hand them the marker (metaphorically) by adding "your turn" to the last step you wrote.
Check the answer at the bottom. Substitute the value back into the original equation, evaluate both sides, and confirm they match. This is the step most students skip in their notebook because it's tedious; on a shared whiteboard where the check is just one more line, the tutor can make it a habit.

The palette symbols that come in as equations get more complex, in roughly the order they appear across pre-algebra and Algebra 1:

Round brackets (Pro). Grouped terms, distributive-property practice.
Fraction slash (Pro). Equations with algebraic fractions like "x/3 + 2 = 5" or "(x + 1)/(x − 2) = 3". Written inline as an ASCII-style slash rather than a stacked-bar fraction — see the honest-limit section below.
Square root (Pro). "x = √25" as a first-taste of radicals in solving equations. A symbol you type before a number or a bracketed expression — no radicand-under-a-bar renderer, so "√(x + 3)" is written with explicit parentheses.
Superscript two and three, and n (Platinum). "x² + 3x − 4 = 0" — the moment quadratics show up, the Platinum Powers category earns its keep. Superscripts stop at ⁿ (a discrete Unicode glyph); anything above ³ that's not just ⁿ needs a caret notation like "x⁴" written as "x^4" — fine for the lesson, worth flagging to the student that the exam typeset version uses a real superscript.
Plus-or-minus (Platinum). The quadratic-formula answer "x = (−b ± √(b² − 4ac)) / (2a)" — the ± lives in the Platinum Algebra category.
Pi and infinity (Platinum). π in circle-formula word problems that translate to algebra ("if the area of a circle is 100 cm², what is the radius?" → "100 = πr²"). Infinity in interval-notation solutions ("x ∈ (3, ∞)").
Less-than-or-equal, greater-than-or-equal, not-equal, approximately-equal (Platinum). Inequalities move from "x < 8" (Pro) to "3 ≤ 2x + 1 < 11" (Platinum) the moment compound inequalities show up.
Square and curly brackets (Platinum). Interval notation "[3, 7)"; set-builder notation "{ x : x ≥ 5 }".

For the full palette walkthrough — the two entry points (Whiteboard Library folder and the ∑ sidebar popup), tier gating, and the six common notation moves — see the write math equations online spoke. Here we're on the algebra pedagogy side: which notation shows up when in the algebra curriculum, and how to walk a student through it.

Expanding and factoring — the area-model on the whiteboard

Two of the most-taught algebra moves — expanding (x + 2)(x + 3) into x² + 5x + 6, and factoring x² + 5x + 6 back into (x + 2)(x + 3) — are usually learned as a pair. Both benefit from a visual, and the visual most middle-school programs settle on is the area model: draw a rectangle whose sides are (x + 2) and (x + 3), split it into four sub-rectangles, and the sum of the four areas is the expanded expression.

On a shared whiteboard the area model is a small drawing:

Use the Pro rectangle shape from the shape library (see the geometry spoke for the shape library's tier splits). Drop a rectangle, and inside it draw two horizontal or vertical dividing lines with the Pro line tool to split it into four sub-rectangles.
Label each sub-rectangle's dimensions and area with palette-built text boxes: top-left is x by x = x², top-right is x by 3 = 3x, bottom-left is 2 by x = 2x, bottom-right is 2 by 3 = 6. Use the Platinum ² for the x² label if you have Platinum; on Pro, write "x^2" in caret notation as the visible fallback.
Sum the four areas in a text box next to the rectangle: x² + 3x + 2x + 6 = x² + 5x + 6.

Now factoring is the same picture in reverse: you're given x² + 5x + 6, and the question is what rectangle has that as its area. The student splits the middle term (5x = 3x + 2x), lays out the four regions, and reads off the sides (x + 3) and (x + 2). Both directions of the operation live on the same canvas, and the student can see that expanding and factoring are two views of the same picture rather than two independent procedures.

This is the pattern that generalises: the area model works for expanding (x + 2)(x − 3), for factoring x² − x − 6, for completing the square, and — with a 3×3 grid — for expanding a trinomial-times-trinomial. It doesn't help with rational expressions or with cubics and higher, so the tutor's move is to lean on it hard for the quadratic middle band (Algebra 1 second half through Algebra 2 first half) and drop it when the algebra outgrows it.

An honest limit: the Pro-tier tutor who doesn't want to reach for Platinum for the ² glyph can teach the whole area model with "x^2" written in caret notation — the pedagogy is intact, the notation is the working shorthand rather than the typeset version. A Platinum-tier tutor gets the properly-rendered x² throughout, which some tutors prefer for pattern-recognition with what the exam board will typeset.

Graphing lines, quadratics, and function families — pair Desmos in a cobrowser tab

Algebra past linear equations lives on a graph as much as on a page, and Koala Go's whiteboard is not a live-plotting canvas. A rough hand-drawn slope-intercept line is fine for a one-off demo; a coefficient-slider lesson on y = ax² + bx + c is not. The move that closes the gap is opening Desmos in a shared browser tab inside the classroom via the cobrowser.

A cobrowser is a shared browser window inside the lesson — you open Desmos, and both you and the student can drag sliders, type new expressions, and adjust the domain live, not screen-sharing but actual shared control on the same page. The specific algebra lessons Desmos + cobrowser carries better than the whiteboard alone:

Linear functions. y = mx + b with sliders on m and b. Watch the slope tilt the line and the intercept move it up and down. Students see "slope" and "y-intercept" as things that live on the graph, not just as coefficients in an equation.
Quadratic functions. y = ax² + bx + c with sliders on a, b, c. Or the vertex form y = a(x − h)² + k with sliders on a, h, k — watch how the vertex moves. This is the lesson where most students first understand that changing a coefficient does something visible to the graph.
Systems of equations. Two linear equations on the same axes; the intersection is the solution. The student can drag one of the lines' coefficients and watch the intersection move — much clearer than "substitute and solve" on a page.
Inequalities. "y > 2x − 3" shades a half-plane; the student can see what points satisfy the inequality by looking at the shading. Same for compound inequalities and systems of inequalities.
Function transformations. Start with y = x², and change to y = (x − 2)² + 3 — watch the parabola slide. Same for stretching, reflecting, absolute-value transformations.
Rational functions and asymptotes. y = 1/x, y = (x + 1)/(x − 2), and see the asymptotes as the graph goes off to infinity — a visual the whiteboard genuinely cannot carry.

The pairing pattern most working online algebra tutors converge on: whiteboard for the algebra (equation-solving, factoring, algebraic manipulation, worked-out steps, answer checks, uploaded past papers annotated with palette notation) plus Desmos in the cobrowser for the graphing (function families, transformations, systems, inequalities as regions, coordinate-geometry adjacencies). Together they carry most of Algebra 1 and Algebra 2 cleanly.

Cobrowser is capped at 10 minutes per session with a 20-minute cool-down on Koala Free; unlimited on Pro and up. For an algebra-heavy caseload, unlimited cobrowser on Pro is functionally the same practical floor as it is for geometry (paired GeoGebra) and exam prep (paired subject-specific practice sites) — a full Desmos-driven algebra lesson doesn't fit in 10-minute windows.

Word problems that translate to algebra

The other half of most algebra courses is word problems: a paragraph of English that has to become an equation before it can be solved. This is where a lot of students stall — the translation is a separate skill from the algebra, and it's the one that's least well-taught in a lot of curricula.

The workflow that works on a shared whiteboard has four steps, all done in text boxes on the same canvas:

Read the problem — no algebra yet. Paste or type the problem statement into a text box at the top of the canvas. Both of you read it out loud.
Identify the unknown. Below the problem, write "Let x = ..." in a new text box and finish the sentence with what x stands for. Being explicit about this in words, in every problem, is the move that eventually generalises. If the problem has two unknowns, write "Let x = ..." and "Let y = ..." on two lines.
Translate the relationships to algebra, one sentence at a time. Underneath, break the problem's information into short sentences ("Their combined age is 30" / "The larger number is three more than twice the smaller"). Write each sentence as an equation using the palette. This step-by-step translation is what most students find hardest and is the one you can most obviously scaffold on a whiteboard by writing the sentence on the left and the equation on the right of the same line.
Solve the system. Now the problem has become the algebra you've been teaching in the previous section. Solve step by step underneath.

The reason this workflow beats a screenshare-and-narrate lesson: every step is visible to the student on the same canvas, in text boxes they can point at, refer back to, and add to. When they get stuck on step 3 in their notebook next week, they can hover over the same-shaped step on the whiteboard's persistent state and see the pattern.

For younger students meeting word problems for the first time, a Singapore-math-style bar model can sit alongside the algebra: draw a bar labelled "x" and a second bar labelled "x + 3" for a "the larger is three more than the smaller" problem, and the student can see the two quantities as lengths before they see them as symbols. The bar-model tool is the same as the number-rod tool from the number-sense spoke — rods stand in for bars.

Algebraic fractions, rational expressions, and radicals — where the palette's honest limit shows up

By the second half of Algebra 1 and through Algebra 2, algebra takes on notation the palette handles but doesn't render exactly the way a printed textbook does. Three specific cases and the workaround for each:

Algebraic fractions written inline. "(x + 2)/(x − 3)" is written with the Pro slash and round brackets — perfectly readable inline, and this is how a lot of algebra software (Desmos, Wolfram Alpha, Python) expects fractions to be typed. Not a rendered stacked fraction with a horizontal bar. For a lesson focused on simplifying rational expressions ("cancel the (x + 1) top and bottom"), the inline notation reads fine because the operations are happening in-line anyway. If the lesson genuinely needs a stacked-fraction display for a printed-textbook lookalike, the paired workflow is: (a) upload the textbook page as a PDF (Pro tier) and annotate the printed fraction directly, or (b) open Desmos or a rendered-LaTeX site in the cobrowser for the display side.
Radicals with the radicand under a bar. The Pro square-root symbol √ is a symbol you type before a number or bracketed expression — "√25", "√(x + 3)". No horizontal bar continues over the radicand. For arithmetic and low-secondary algebra where "√25 = 5" and "√(x + 3) = 4" are typeable inline, this is fine. For a nested radical or a rendered-bar radical the lesson genuinely needs, pair a Desmos or LaTeX renderer in the cobrowser.
Higher exponents. The palette's Powers category has ² ³ ⁿ. For x², x³, and x-to-the-n these render perfectly; for x⁴, x⁵, and higher there's no discrete glyph in the palette. The workaround most tutors use is caret notation ("x^4", "x^5") — the same notation Desmos and most programming languages use, and one students learn to read anyway. Worth being explicit about with the student: "the exam version will typeset this as a raised numeral; we're writing x^4 so we don't have to stop the lesson." No shame in the caret; it's the working notation for a reason.

The pattern in all three: the whiteboard palette handles the algebra manipulation — grouping, cancellation, distributing, factoring, substituting — even when the exact typeset rendering is not what a textbook would show. When the lesson needs a rendered display (a stacked fraction with a horizontal bar, a radicand under a bar, a fifth power as a raised numeral, a proper LaTeX-shaped display equation), the paired cobrowser tab into Desmos or a LaTeX renderer does the rendering side, and the whiteboard does the working side. This is the same honest-limit pattern the notation spoke names in more depth.

By age band: what tutors actually teach and how the tools slot in

Different age bands lean into different parts of the algebra toolkit. Grades are typical US placements, not standards claims — UK / IB / A-level students meet the same content on a somewhat shifted schedule, and adult learners come to algebra at whichever age they need it.

Grades 6-7 (ages 11-13): pre-algebra. Introduction to variables, order of operations, one-step and two-step linear equations, ratios and proportions, integers and negative numbers, first exposure to inequalities. Whiteboard for the step-by-step; number rods (Pro) for the "let this rod be x" bridge; number lines (Pro, including the −10 to 10 line — see the number-sense spoke) for negative numbers and integer arithmetic. Palette symbols in use: Pro operations, Pro brackets, Pro fraction slash, Pro compare (< >). Very little Platinum yet — pre-algebra is squarely within Pro's palette. Playground / Gems can still carry engagement for the youngest end of the band (11 year olds who are behind grade level and would rather not be doing algebra).
Grade 8 or 9 (ages 13-15): Algebra 1 (US) / Year 9-10 (UK GCSE). Linear equations and inequalities, systems of linear equations, introduction to quadratics via factoring and the quadratic formula, exponents and radicals, introduction to functions and graphs. This is where Platinum starts to earn its keep for a math-focused tutor: the ² glyph for polynomial expressions, ± for the quadratic-formula answer, the Compare extras (≤ ≥ ≠) for compound inequalities, [ ] for interval notation. The Desmos cobrowser pairing is essentially required for the graphing side (linear graphs, quadratic parabolas, systems as intersecting lines). PDF upload on Pro becomes a workflow for practice-set worksheets.
Grades 10-11 (ages 15-17): Algebra 2 (US) / Year 12-13 (UK A-level). Polynomial functions and division, rational expressions and equations, radicals and rational exponents, exponential and logarithmic functions, sequences and series, matrices, conic sections. All the Platinum algebra symbols come into play (π ± ∞ ² ³ ⁿ ≤ ≥ ≠ ≈ [ ] { }); Desmos cobrowser is the primary graphing surface for every function family; uploaded past-paper PDFs (Pro) become the primary practice surface. This is the tier band where the paired-tool workflow is fully weight-bearing.
Exam-prep algebra (SAT / ACT / A-level / IB, all ages). Uploaded past paper on the whiteboard, palette symbols for the algebra, Desmos cobrowser for the questions that involve a graph. This is the workflow described in how do I teach exam prep online? — for algebra specifically, past-paper practice is the vein of gold.

Rate-band note: math tutoring — including algebra tutoring — typically commands a 15-30% premium over the generalist band, and specialist exam-prep algebra (SAT / ACT / A-level math) sits at the upper end. See how much should I charge for online tutoring? for the pricing walkthrough.

Practice, homework, and drill for algebra between lessons

Algebra needs repetition, and the between-lesson practice loop matters as much as the lesson itself. Three patterns most working online algebra tutors converge on:

Uploaded past-paper PDF as the lesson surface. On Koala Pro, upload the past paper or the textbook problem set as a PDF; both of you annotate on top of the printed problem with pen and text boxes using palette notation. The uploaded material persists across lessons — next week you re-open the same page and last week's work is still there.
Cobrowser to adaptive algebra practice. Khan Academy's algebra practice and IXL's algebra units are the two most-used adaptive practice sites; both open cleanly in the cobrowser. Bring the student into the practice rather than sending them out to it — you see exactly what they typed, catch the misstep in the working, and jump in without them window-switching. See what is a cobrowser? for the mechanic.
Small specific ask between lessons. "Try three problems from this set and bring me the working" beats "practise algebra for 30 minutes". The homework workflow that most tutors converge on is in how do I give homework to online tutoring students?

A note on step-by-step solvers (Photomath, Symbolab, Mathway, ChatGPT-with-math-mode). These are useful for a tutor to check a worked solution or generate practice problems, but they are unhelpful as a between-lesson tool for a student who is trying to learn to do algebra themselves — a solver gives the answer without building the skill. The move most working tutors settle on: use solvers on your side for prep and answer-checking; don't route the student through them. Cobrowser into a practice site that shows one hint at a time (Khan Academy is designed this way) rather than a solver that shows the whole solution.

Common failure modes and honest limits

Things that go wrong in online-algebra workflows, and the fix:

Doing the algebra on your side and screensharing it. The student watches you work through a solve and thinks they've followed it. They don't try the same solve themselves until they're on their own after the lesson — where it turns out they didn't follow it, they followed you. Fix: the whiteboard is shared, so as soon as you've demonstrated one solve, hand the marker over — "your turn on this next one, I'll watch." The student does the algebra with you present rather than after you're gone.
Typing "x2" and hoping it reads as x-squared. "x2" is a variable next to a two, and the student's brain reads it as such. Insert the Platinum "²" from the palette Powers category, or on Pro write "x^2" in caret notation. Don't leave the notation ambiguous — algebra runs on pattern-recognition, and the pattern breaks the moment the notation does.
Typing "sqrt(x + 3)" as a fallback for √(x + 3). The palette's Pro √ is one click; the "sqrt(" workaround is six keystrokes and reads as a function-call rather than a radical. Same for using "-" (keyboard hyphen) instead of the palette's "−" (proper minus). Small fixes, but they add up over a lesson.
Using Desmos to solve rather than to teach. Desmos will happily plot y = x² − 5x + 6 and show that the roots are at 2 and 3 without the student ever having factored the quadratic themselves. The pairing pattern works when Desmos is the graphing tool (visualisation, function-family exploration, checking answers) and the whiteboard is the algebra tool (the factoring, the equation-solving, the algebraic manipulation). If Desmos is doing both, the student isn't learning algebra — they're learning Desmos.
Free-tier tutors seeing the palette locked. On Koala Free the Math Symbols palette shows every symbol locked — the palette itself is a Pro-and-up feature. A tutor trialling Koala Go for algebra needs at least the trial or Pro plan to evaluate the notation side. The whiteboard, cobrowser (capped), image upload, and Playground are on Free, so a Free tutor can trial the equation-solving workflow using keyboard-typed notation, but the palette-powered version needs Pro.
Enterprise / B2B tutors not seeing Platinum algebra symbols. B2B tutors are provisioned at Pro feature level, and Platinum symbols are hidden entirely from their palette (not shown locked — hidden). For a middle-school or secondary math deployment inside a B2B account, this is worth flagging to the administrator: Platinum-symbol algebra work (² ³ ⁿ π ± ∞ ≤ ≥ ≠ ≈ [ ] { }) needs a Platinum-tier plan.
iPad-only workflow for a student doing algebra. Koala Go works on iPad, but the narrow-symbol grid in the palette folder is fiddlier on a finger than on a mouse — the ≤ and ² and ⁿ symbols are small tap targets. If a student uses iPad exclusively, the ∑ sidebar popup (which appears next to the selected text box rather than in the Library sidebar) is a slightly easier click, and the tutor can pre-drop palette-heavy notation on the canvas in advance so the student's job is dragging and typing rather than fine-symbol picking.

Three limits worth naming plainly rather than papering over:

Not a computer-algebra system. Koala Go's whiteboard is a working surface for algebra — you and the student write algebra on it. It is not a symbolic-manipulation engine — it will not automatically simplify (x + 2)(x − 3) into x² − x − 6, or automatically solve x² − 5x + 6 = 0 for you. Both are pedagogically fine (the point is to teach the student to do the algebra, not to watch a machine do it); if you or the student needs a symbolic step-by-step for prep or checking, that's Symbolab / Wolfram Alpha / Photomath territory, opened in the cobrowser or on your side before the lesson.
Not a live-plotting canvas. The whiteboard has pen, shape, and line tools but does not plot a function y = f(x) live as you type it. For any lesson whose whole point is graphing (linear functions, quadratics, transformations, function families, systems, inequalities as regions), the paired workflow is Desmos in the cobrowser. This is not a shipping-soon promise; the whiteboard is the algebra surface and Desmos is the graphing surface, by design.
Not a full LaTeX renderer. The palette does not render stacked fractions with a horizontal bar, radicands under a bar, higher-than-cube exponents as raised numerals, matrix layouts, or LaTeX-shaped display equations. Algebraic manipulation is expressible inline in the palette's Pro + Platinum sets; typeset rendering pairs with a cobrowser tab into Desmos or a rendered-LaTeX site. Same honest-limit boundary the notation spoke #2417 names in more depth — see the notation spoke.

Practical setup on Koala Go for an algebra-heavy caseload

If you're evaluating Koala Go for pre-algebra through Algebra 2 / A-level specifically, here's the practical picture and where the current lines sit:

Whiteboard is the primary working surface. Shared canvas, both tutor and student can write, text boxes are first-class objects, boxes persist across lessons. Free tier and up.
Palette is Pro for the base, Platinum for the algebra extras. Pro ($25.99/month monthly, $21.99/month billed annually) covers pre-algebra and most of Algebra 1's palette needs — the operations, base compare, fractions/ratios/percents, square root, degrees, and round brackets. Platinum ($49.99/month monthly, $39.99/month billed annually) adds the algebra-specific extras — ² ³ ⁿ π ± ∞ ≤ ≥ ≠ ≈ and the square/curly brackets used in interval and set-builder notation. Enterprise / B2B tutors are provisioned Pro-level with Platinum symbols hidden.
PDF and PowerPoint upload for past papers, worksheets, and textbook pages (Pro). Upload the material, annotate on top with pen, shapes, and palette-notated text boxes. Textbook algebra practice becomes a shared-canvas workflow instead of a screenshare-and-narrate.
Cobrowser for Desmos, Khan Academy, IXL, Wolfram Alpha (unlimited on Pro, capped on Free). Open the site in a shared browser window inside the classroom — student clicks and drags on the live page. Desmos is the graphing surface; Khan / IXL are adaptive practice. See what is a cobrowser? for the mechanic.
Number rods (Pro) as concrete bridges to variables. The "let this rod be x" pattern for introducing variables — pairs with the balance-scale metaphor for solving equations. See the number-sense spoke for the rod set.
Rectangle from the shape library (Pro) as an area-model surface. Drop a rectangle, split it into sub-rectangles for the area-model workflow that visualises expanding and factoring. The full shape library is walked through in the geometry spoke.
Recording for parent review or student self-review. Local recording is on Pro; cloud recording of scheduled lessons comes with Platinum. Useful when a student wants to re-watch a step-by-step solve. See record online tutoring lessons.
Free tier as a trial. Free caps sessions at four students, includes whiteboard, image upload, and a capped cobrowser, but the Math Symbols palette is Pro-and-up. Free is a trial surface for algebra, not the day-to-day.

Honest limits worth flagging that show up in real algebra work regardless of platform: iPad is rougher than desktop for narrow-symbol picking; students in mainland China sometimes hit connection issues (we route through Hong Kong proxies but the Great Firewall is a moving target — see teaching students in mainland China for the workaround pattern); and the palette is a symbols set rather than an equation renderer, so stacked fractions, higher exponents, radicand-under-a-bar radicals, and live graphs pair with a Desmos or LaTeX cobrowser tab rather than living on the whiteboard.

Where Koala Go isn't the right pick for algebra

Two honest tradeoffs worth naming, so you know before you sign up:

Pure symbolic-manipulation tutoring as the primary surface. If every lesson you teach is symbolic manipulation of complex expressions (partial fractions, Laplace transforms, extensive matrix work) and you want a full LaTeX-rendered display equation as the working surface, a dedicated STEM whiteboard with a real equation editor (see the whiteboard-tool comparison) will render the notation closer to a printed page. Koala Go's palette handles the notation of pre-algebra through Algebra 2 inline; the honest limit is the render, not the manipulation. For calculus-and-above where every third line has a stacked fraction or a nested radical, a LaTeX-first whiteboard is a fair alternative to a paired Desmos cobrowser tab.
Whole-class delivery to a 25-student algebra class. Koala Go is calibrated for 1-on-1 and small-group tutoring — the Free tier caps sessions at four students. It's not built for full school-class synchronous delivery of the kind Nearpod or Pear Deck are designed for. For a whole-class algebra lesson, look at platforms built for whole-class delivery.

Why math assessment online needs its own workflow

The instinct to skip the diagnostic and jump straight into teaching — the parent has said "she's struggling with fractions", so surely you can just start with fractions? — is the single most common failure mode of a new online math tutoring relationship. It goes wrong for three reasons that are worse online than in person.

The parent's description of the gap is usually wrong. Parents (and often teachers) name the topic the student is currently failing at — "she can't do fractions" — but the actual break is almost always a prerequisite two or three levels back. The child who can't add fractions usually can't decompose a whole into equal parts. The child who can't do long division usually doesn't have fluent times-tables recall. If you start teaching the topic the parent named, you build on rubble.
You can't watch over the student's shoulder. In person, the diagnostic is partly informal — you sit next to the student, watch which finger they use to count, whether their lips move on the times table, how many seconds pass before they answer. Online, none of that is visible unless you build a shared surface where the student's thinking happens in front of you. The whiteboard is that surface. If you ask "what's 7 × 8?" over a webcam and hear "56" a second later, you learned exactly one bit of information. If you ask the student to build 7 × 8 on the whiteboard with rods, you see whether they reached for arrays, skip-counting, or a chant — and how long each part took. Same question, ten times more information.
The first-session engagement stakes are higher online. A young student who's just been dropped into a video call with a stranger already isn't sure this is going to work; a session that opens with "let me test you on math" is the fastest way to lose them. The diagnostic has to feel like an activity, not a test. That's a design constraint online, and it changes what tools you reach for — you're not handing them a paper diagnostic and watching them work, you're inviting them to build something on a shared canvas and observing what they build.

An online math diagnostic that works has an answer to each of these: it starts from the surfaces the student can act on directly (rods, lines, fraction bars, a shared page), it produces observable evidence (not just correct-or-incorrect answers), and it feels enough like an activity that a nervous seven-year-old wants to keep going. The rest of this page is that workflow.

What to assess: three domains, not one

The math-tutoring literature is fairly aligned that you're looking for three separate things when you assess a student, and they usually don't move together. A student can be fluent in the first and hollow in the other two; the gap that needs teaching is different in each case.

Computational fluency. Can the student do the mechanical operation quickly and accurately — the times-table fact, the column addition, the long-division algorithm, the standard fraction addition? Fluency is what school assessments are usually calibrated to, so the parent's account of the gap usually names a fluency problem ("he can't do his 7s", "she gets long division wrong"). Necessary but not sufficient — fluency without the concept underneath is a memorised procedure that breaks the moment the problem shape changes.
Conceptual understanding. Does the student know why the operation works? Can they show 3 × 4 as an array before writing it? Can they show 3/4 as a shaded region before symbolically manipulating it? A student who can produce "12" for "3 × 4" but can't build the array is memorising, not understanding — the fact will fade, and the next concept (division as the mirror of multiplication, distributive property, area model) will land on empty foundations. This is the domain school assessments miss most often.
Problem-solving. Given a problem the student hasn't seen the exact shape of before — a word problem, a novel application, "here's a shape, find its area a different way" — can they choose an approach, run it, check it, and back out if it isn't working? This is the highest-stakes domain and the one most in-school assessments barely touch. In tutoring it matters because most of what a family pays for is a student who can eventually work independently on problems they haven't been walked through.

The rest of this workflow builds a diagnostic that produces evidence on all three, in one session, without turning into a two-hour formal test.

The four diagnostic surfaces on Koala Go

Koala Go's classroom ships four surfaces that do double duty as diagnostic instruments — you're not using a separate testing tool; the same whiteboard-and-tools that carry a lesson carry the diagnostic. Here's what each one tells you.

1. The open whiteboard: "show me how you'd solve this"

The single most useful diagnostic move online is: drop a problem in a text box on the whiteboard, hand the pen to the student, and ask them to show you how they'd solve it. That's it. What you're evaluating isn't whether they get the right answer — you're evaluating the strategy they reach for, the fluency of their execution, and the moment they hit the edge of what they can do.

A few examples of what you can learn from one problem:

A grade-3 student handed "8 × 7 = ?" and given the pen will either (a) write "56" immediately (fluent, likely from memorised chant), (b) skip-count in their head and then write "56" (fluent enough, but derived), (c) draw an array of eight rows of seven dots (conceptual — good sign), (d) count on their fingers (a red flag for grade 3 fluency), or (e) freeze and ask what the × means (a much larger gap than the parent named). Same problem, five different diagnostic outcomes.
A grade-5 student handed "3/4 + 1/2 = ?" and given the pen will either (a) write "5/6" or "4/6" (an "add across" error — you've found the misconception you need to address), (b) write "3/4 + 2/4 = 5/4 = 1 1/4" (fluent — the fraction is not the gap), (c) draw the bars or circles and count pieces (conceptual and fluent — the strongest signal), or (d) freeze (fraction addition isn't yet established — back up to equivalent fractions).
A grade-8 student handed "solve for x: 2x + 3 = 11" and given the pen will either (a) write "x = 4" (fluent), (b) show every step of the balance-scale reasoning (fluent and conceptual), (c) subtract 3 from the left side only and get "2x = 11" (a red flag for the do-the-same-to-both-sides rule), or (d) freeze (the equation-solving concept isn't yet established).

The whiteboard is the primary lesson surface across every tier of Koala Go, so this diagnostic move works on Free (with the base whiteboard tools) as well as Pro and up. The pen tool, the text tool, and the sticky-note tool are all on Free. The student can be on any device with a modern browser — no install, no signup, they open the invite link and land on the shared canvas.

2. Math manipulatives as diagnostic instruments

The Math Tools panel inside the Whiteboard Library (Pro and up) turns each of the built-in manipulatives into a diagnostic instrument. Rather than "here's how rods work, let's learn to add", you say "here are ten rods, one for each value 1 to 10 — can you build 3 + 4 for me?" What you learn from watching them reach for rods (or not) is qualitatively different from what you learn from a written problem.

Number Rods (Pro tier) as a K-2 number-sense diagnostic. Ten Cuisenaire-style rods, values 1 to 10, each a different colour, that the student can drag onto the canvas. Ask the student to build 3 + 4. A student with solid K-1 number sense will grab a 3-rod and a 4-rod, place them end to end against a 7-rod, and say "seven". A student who's still counting will grab a 1-rod and drag it four times, then grab another 1-rod, and so on. A student who reaches for the 10-rod and starts breaking off units is showing you a subtraction misconception. Then: "can you make 7 + 5?" — same rod set, and you're watching whether they subitize (grab 7 and 5 rods directly), decompose the 5 into "3 to make 10 and 2 more" (very strong sign), or count on. Ten minutes on rods, and you have a genuinely fine-grained read on where in the K-2 arc the student sits — a much finer read than "she can add up to 20". Sibling depth on the number-rod workflow lives in the number-sense spoke; we're using the same tool as a diagnostic here rather than as a teaching tool.
Number Lines (Pro tier) as a magnitude and skip-counting diagnostic. Preset lines for 0-10, 0-20, 0-30, 0-100 (by 5s or 10s), −10 to 0, −10 to 10, and blank. Place a 0-100 by 5 line and ask a grade-3 student to hop from 0 to 40 in fives. A student who has automatic skip-counting will place five hops. A student who has to count each five will do it slowly but correctly. A student who lands on 45 or 35 has a fluency gap on the fives. Place the -10 to 10 line for a grade-6 student and ask them to hop from -3 to 5 — a student who understands integers will count eight hops; a student who says "two" (subtracting the magnitudes) has an integer-arithmetic misconception you need to know about. The number-line workflow at teaching depth is in the number-sense spoke.
Fraction Bars and Fraction Circles (Platinum tier) as a fractions readiness diagnostic. Individual bar segments and pie-slices for denominators 1/2 through 1/12 (plus 1/16). Ask a grade-4 student to build 3/4. A student with solid fraction concept will place three quarter-bars or three quarter-slices side by side. A student who reaches for a half-bar and a quarter-bar and adds them is showing you they've noticed equivalence but haven't formalised it. A student who places three of any bar is showing you they don't yet see the denominator's role. Then: "can you show me a fraction equal to 3/4 but with a different denominator?" — you're diagnosing equivalent-fraction concept. This diagnostic surface is Platinum-only; on Pro the workaround is asking the student to draw the fraction pictures with the pen tool on the whiteboard, which loses some of the drag-and-arrange information but still surfaces the concept. Depth on the fractions workflow is in the fractions spoke.
Math Symbols palette (Pro covers the four operations and equals, the basic comparisons < and >, the fraction / ratio / percent group, √, °, and round brackets; Platinum adds ≠ ≈ ≤ ≥, the superscript powers ² ³ ⁿ, π ± ∞, the geometry markers ∠ △ ⊥ ∥, and square and curly brackets) as a notation-fluency diagnostic. Drop a text box and ask a grade-7 student to write "3/4 × 1/2 = 3/8". Worth knowing which symbols actually carry diagnostic signal: the palette's +, =, /, :, %, <, > and round brackets are the same characters the keyboard produces, so you can't see how they were entered. The ones you can read off the screen are × (a proper multiplication sign rather than the letter "x" or an asterisk), − (a minus sign rather than a hyphen), and everything in the Powers, Algebra, and advanced Compare sets. A student who writes "3/4 x 1/2" with a letter x is telling you the palette isn't yet in their toolkit — a small thing that snowballs into "why does my student's algebra look so messy" if not caught. For older students, ask them to write "x² + 3x − 4 = 0" — the palette's superscript ² (Platinum) is the tell for whether they know how to write exponents at all, and the standalone minus vs. hyphen catches a common typo. Palette entry points and the full six notation moves are in the math-notation spoke.
Stickies and Stamps for anything the built-in tools don't cover. Ten-frame tokens, dot cards, place-value counters (tens and ones), integer chips (positive and negative), coins for money work — anything a physical manipulative-drawer would hold can be dropped as a Sticky or Stamp on the canvas, and both you and the student can drag them. These are the same primitives the dyscalculia and dyslexia workflows rely on; for a diagnostic they let you extend beyond the built-in tools when the student's curriculum uses a specific manipulative you want to see them work with.

3. Cobrowser into a third-party diagnostic

A cobrowser is a shared browser window inside the classroom — the tutor opens a page, and both the tutor and the student can click, type, drag, and interact with the page live. Not screen sharing — actual shared control. For a math diagnostic, this pairs the open-ended whiteboard work with a standardised third-party diagnostic layer that a well-established site can carry better than a hand-run assessment. The tools tutors usually reach for:

Khan Academy's Mastery challenges and course diagnostics. Free to use, standard sequence across most US elementary and middle-school topics. A short Mastery challenge in a grade-appropriate unit gives you a percentage-based first read. Not a formal placement test but a widely-used baseline.
IXL's Real-Time Diagnostic (subscription). Adaptive diagnostic that estimates a student's level per strand (numbers and operations, algebra, geometry, measurement, data). Some tutoring families already have IXL through school; when they do, the diagnostic is a five-to-fifteen-minute session that produces a printable snapshot the tutor can walk through with the parent afterwards.
ALEKS's initial knowledge check (subscription, K-12 and college). Adaptive assessment that maps to a specific curriculum; slower than IXL but more granular for the students it fits.
Beast Academy's placement guide (or the equivalent for whatever curriculum the family is on). Curriculum-tied placement tests that map to specific programme entry points. Useful when the family has already picked a curriculum and wants to know where in it to start.

The cobrowser makes these usable inside the diagnostic session rather than as homework. On Koala Free, the cobrowser is capped at 10 minutes per session with a 20-minute cool-down — enough for one Khan Mastery challenge in the diagnostic, but you'll want to sequence carefully. On Pro and up, the cobrowser is unlimited, which matters if you're running a longer IXL or ALEKS diagnostic. The mechanic itself, and how it differs from screen sharing, is in the cobrowser explainer. Two things to name honestly: adaptive diagnostics like IXL and ALEKS are calibrated for independent student work, so your presence during a cobrowsered diagnostic changes the signal a little — the student may work harder knowing you're watching, or freeze knowing you're watching, and either way the result is a rougher estimate than if they'd taken it alone. Use it as one data point among several, not the definitive placement.

4. PDF upload for a paper-based diagnostic or a past paper

Some diagnostics live on paper — a school's own placement test, a curriculum-specific worksheet the family has been given, a past exam paper the older student was supposed to work through. On Koala Pro and up, you upload the PDF (or a PowerPoint slide deck) to the whiteboard and both of you can annotate on top of it in real time. That collapses the "here's a diagnostic PDF, work through it and send it back to me" workflow — which loses observation completely — into a live session where you see every hesitation, every erased attempt, every wrong turn.

Two ways to use it:

Live-annotated diagnostic worksheet. Family shares (via email or the classroom's activity upload) the school's placement worksheet or an end-of-unit assessment. Open it as a PDF on the whiteboard, hand the student the pen, and watch them work through it live. Same information as the open-whiteboard move but with the added benefit that you're working in the exact document the school uses.
Past-paper diagnostic for older students. A grade-10 student prepping for a specific exam (GCSE, IGCSE, A-level, IB, SAT, ACT) has past papers accessible. Upload one, hand it to the student, and use the first fifteen minutes as the diagnostic — you're evaluating exam technique alongside subject knowledge. This is the diagnostic move for the exam-prep flow described at depth in how do I teach exam prep online?

Uploaded material stays on the whiteboard between sessions, so the diagnostic worksheet is still there next week — useful for "here's what we saw last week; today let's tackle the parts you got stuck on."

A first-session diagnostic script by age band

Different age bands lean on different surfaces. A rough script for each, calibrated to fit a 25- to 40-minute first session (the diagnostic is roughly the first two-thirds; the last third is a short teaching move plus the parent debrief). These are typical US grade placements and adjust to the curriculum and the student in front of you.

K–2 (ages 5–7): rods and lines first, notation later

Warm-up (5 min): "Hi! Let's play with some rods." Open the Whiteboard Library and place the ten Number Rods on the canvas one at a time, in a jumbled heap rather than in order. Ask the student to line them up from shortest to longest. What you're watching: do they immediately see the pattern, or do they need to compare each pair? (Platinum's "Add all" button drops all ten in a single click, but as a ready-made staircase — which does the sorting for the student, so place them individually for this particular move.)
Number sense (10 min): "Can you build me a 3 and a 4? Now put them together — how long is that?" Repeat with 5 + 2, then 7 + 3, then 4 + 6. You're watching for: subitising (grab-and-place vs. count-each-unit), decomposition (does 8 + 5 become "8 + 2 + 3"?), and complements-to-ten strategies.
Number line (5 min): Place a 0-20 line. "Can you hop from 0 to 12?" Then: "Now hop back from 12 to 4." Watching for: one-to-one correspondence on hopping, whether they need to count each tick or can jump multiple.
Written notation (5 min): "Can you write 5 + 3 = 8 in a box for me?" Watching for: number formation, confident use of the + and = signs, whether the equation reads left-to-right cleanly.
Reward and close (5 min): A short Playground visit or a Gems handout — see the gamified virtual classroom answer for the engagement layer. For a first-lesson K-2 student, this closing engagement moment is what makes them want to come back.

Grades 3–5 (ages 8–10): fluency + fractions + strategies

Warm-up (3 min): "Tell me about a math thing you're good at and a math thing that's tricky." Cheap information from the student's own account; parents don't always relay this accurately.
Times-table fluency (7 min): Drop a text box: "6 × 7 = ?" Ask them to type the answer. Then "8 × 4?" "9 × 6?" "7 × 8?" You're watching timing (do they answer inside two seconds, or is there a long pause?) and error pattern (near-misses vs. wild guesses). If you have Platinum, place a Number Roller set to 1-12 and roll five random pairs — the roller removes tutor-invented bias in the fluency check.
Fractions conceptual check (7 min): If Platinum, place a Fraction Bars set and ask "Can you build 3/4? Now build 1/2. Which is bigger?" On Pro, ask them to draw the fractions with the pen and shade — a rougher version of the same diagnostic. Then: "Can you build 1/2 in a different way?" Watching for: equivalent-fraction concept, comfort with denominators up to 8, whether they can see a half as 2/4 or 3/6 or 4/8.
Word problem (5 min): Drop a text box: "There are 24 pencils in a box. If you share them equally among 6 kids, how many does each kid get?" Hand the student the pen. Watching for: strategy choice (skip-count, division, drawing), setup vs. computation, whether they check their answer.
Notation snapshot (3 min): "Can you write 3 × 4 = 12 in a box, using the math symbols palette for the signs?" See whether they use × (proper multiplication sign) or "*" or "x" from the keyboard.
Reward and close (5 min): Gems for effort on the harder parts; a short Playground moment if age-appropriate. Homework ask: one specific practice thing that came out of the diagnostic (see the homework answer for the between-lesson practice framing).

Middle school (ages 11–13): notation, fractions, integers, and a word problem

Warm-up (3 min): Student's own account of what's going well and what isn't. This age band self-reports more accurately than younger.
Fractions and decimals (7 min): Drop "3/4 + 1/2 = ?" as a text box. Then "0.3 × 0.2 = ?" Then "What's 40% of 60?" You're diagnosing the fluency of the rational-number arithmetic that pre-algebra depends on. Errors here explain a great deal of the trouble that surfaces in algebra later.
Integer arithmetic (5 min): Place the −10 to 10 number line. "Can you hop from -3 to 5?" (Eight steps.) "From 4 to -2?" (Six steps.) Then in a text box: "−7 + 4 = ?" and "−3 × −5 = ?". Signed-number arithmetic is a common gap in this age band and a common cause of "my kid understands algebra when we walk through it, but can't do it alone" — because the arithmetic underneath breaks.
A pre-algebra word problem (7 min): "A pencil costs $2. A notebook costs $5. Sam buys some pencils and 3 notebooks. He spends $23 in total. How many pencils did he buy?" Hand the student the pen. Watching for: whether they write an equation, whether they solve it or guess-and-check, whether they name the unknown as "x" or as "the number of pencils". This is the pre-algebra transition — many middle-schoolers can solve it by guess-and-check but not by algebra.
Notation snapshot (3 min): "Can you type x² + 3x − 4 = 0, using the palette for the symbols?" Watching for: comfort with the superscript (Platinum), whether they know the − is not the same as the hyphen, whether they use the palette's × or "x" for multiplication.
Debrief (5 min): Summarise what you saw, ask them what they thought of the session, and outline a week 1 plan. Send a short note to the parent inside 24 hours (parent-comms cadence in the progress-updates answer).

High school (ages 14–17): past paper first, algebra depth, cobrowser to Desmos

Warm-up (3 min): The high-school student's own read on what they need help with. At this age band they've usually named the specific exam, unit, or topic accurately.
Past paper or worksheet (15 min): Upload the PDF of the most recent piece of graded work the family shared, or a past-paper question at grade level. Hand the student the pen. Watching for: setup, algebraic manipulation, calculator use, notation habit, checking. Fifteen minutes of live-annotated past-paper work produces more diagnostic signal than any other move at this age.
Graphing diagnostic (5 min): Open the cobrowser to Desmos, plot y = 2x + 3. Ask the student to change the slope, then the intercept, then to make a line perpendicular to it. Watching for: comfort with slope-intercept form, whether they see the geometric meaning of slope, whether they can predict the effect before pressing enter. This is the diagnostic move that pairs well with the algebra spoke walkthrough at how do I teach algebra online?
Exam-technique check (5 min): If the student is prepping for a specific exam, ask them to answer one exam-style question with the palette's proper notation (√ and ° are on Pro; ∠, ≤, and π need Platinum). Watching for: mark-earning notation habits, whether they're leaving working out visible, whether they know the difference between an approximate and an exact answer. Exam-focused workflow at depth in how do I teach exam prep online?
Debrief (5 min): With this age band the student is usually in the room for the parent debrief too, or is themselves the primary decision-maker. Give an honest read of what you saw and what you'd suggest as a week 1 focus.

Reading the results: which spoke to route the student into

A diagnostic without a downstream plan is a diagnostic that wastes the family's money. Once you've run the session and know where the gaps live, the math cluster is set up so each common gap has a specific follow-up spoke you can teach from. The rough mapping:

K-2 number-sense gap (student can't reliably compose to ten, can't skip-count on a line, isn't yet subitising): route into the workflow described in how do I teach number sense online with virtual manipulatives? The rods and lines are your primary tools for the coming weeks.
Times-tables fluency gap (grade 3+ student stalling on any single-digit multiplication, can't handle skip-counting on the 6s / 7s / 8s / 9s): route into how do I teach multiplication and times tables online? — arrays, skip-counting, derived facts, and Number Roller drill on Platinum.
Fractions concept gap (student adds "across", can't produce equivalent fractions, can't compare 3/4 and 2/3): route into how do I teach fractions online? — the fraction bars and circles workflow.
Math notation habit gap (student typing "1/2" instead of using the slash-fraction, using "x" for multiplication, hyphens for minus signs): route through how do I write math equations and notation on an online whiteboard? — the palette entry points and six notation moves. This tends to be a quick fix; the notation vocabulary catches up fast once demonstrated.
Pre-algebra / algebra gap (middle- or high-school student stalling on solving equations, expanding, factoring, working with variables): route into how do I teach algebra online? — the balance-scale, area-model, and cobrowser-to-Desmos workflow.
Geometry gap (student can't identify shape properties, area/perimeter formulas don't stick, coordinate geometry is opaque): route into how do I teach geometry online? — the 2D/3D shape library, shape labels, and GeoGebra cobrowser pairing pattern.
SEN pattern surfaced (persistent pattern that suggests a math learning difference — magnitude confusion that doesn't respond to concrete work, fact retrieval that stays laboured after weeks of drill, sequencing difficulty that's out of proportion to age): the diagnostic is not a diagnosis, and it isn't our job to hand out one. Read how do I tutor a student with dyscalculia online? for how to adapt the workflow, and — separately — recommend the family talk to an educational psychologist or their school's SEN coordinator for actual assessment. A tutor's diagnostic surfaces patterns; a qualified specialist diagnoses.

What NOT to do in a first-session math diagnostic

Six specific mistakes we've seen new online math tutors make when the first-session diagnostic goes wrong:

Don't skip the diagnostic because "the parent already told me the gap". The parent's account of the gap is a starting hypothesis, not a placement. More often than not, the actual break is one to three levels below what the parent named. Run the diagnostic anyway.
Don't turn the session into a formal test. A young student who feels they're being tested will freeze, shut down, or perform below their real level. Frame it as "let's play with some math and see what you know" for younger students; frame it as "let me get a read on where you are so I can plan our sessions" for older students. Both framings are honest and both preserve engagement.
Don't collect too much data in one session. The temptation is to run every diagnostic on the list because the surfaces make it possible. Stop at "enough to know where to teach next week" — usually 20 to 30 minutes of diagnostic, then a short teaching move and a close. Save the deeper diagnostics for weeks 2 and 3 as new gaps become visible.
Don't run adaptive third-party diagnostics (IXL, ALEKS) as if they're definitive placements. They're calibrated for independent student work; your presence changes the result. Use them as one data point among several, and read the printable snapshot with the parent as a rough baseline, not a final ranking.
Don't over-share the diagnostic result with the parent as a list of deficits. A three-minute parent debrief is "here's what your child does well, here's the one thing we'll focus on first, and here's why". Not a comprehensive audit of every gap you found. The full picture goes into your own planning notes, not the parent update. Cadence in how do I share progress updates with parents?
Don't skip the "one small win" close. A diagnostic session that ends with the student having successfully done something — built a fraction with bars, hopped a number line correctly, factored one small expression — leaves them wanting to come back. A diagnostic that ends with the student stuck on the last problem leaves them thinking they're bad at math. The last problem you drop should be one they can do; save the harder unknowns for week 2.

Honest limits

Two limits worth naming plainly rather than papering over:

This is a tutoring diagnostic, not a formal educational assessment. Instruments like the Woodcock-Johnson, KeyMath, TEMA, or a curriculum-specific standardised placement test are administered by qualified educational psychologists or assessors, take longer, and produce results that stand up in school placement conversations. A tutor's first-session diagnostic is fit for purpose (planning the next few weeks of tutoring) but not a substitute for those instruments where they're needed. If the family needs a formal placement or the school is asking for one, point them to a qualified assessor.
Platinum unlocks help but Pro is enough for most first-session diagnostics. The Fraction Bars, Fraction Circles, Number Roller, "make your own" number line, and full Math Symbols set are Platinum-tier features (Koala Platinum is $49.99/mo monthly, $39.99/mo annual; Pro is $25.99/mo monthly, $21.99/mo annual). Pro (which covers Number Rods, all eight preset Number Lines, PDF upload, core shape library, and the base Math Symbols set — the four operations and equals, the basic comparisons < and >, /, :, %, √, °, and round brackets) is sufficient to run a good first-session diagnostic through most age bands; the Platinum tools sharpen the fractions and fluency-drill diagnostic specifically, and matter most if you have a fractions-heavy or drill-heavy caseload. Enterprise / B2B tutors are provisioned at the Pro feature level, and the Platinum tools are hidden from their panel rather than shown locked — so an enterprise tutor runs the diagnostic on the Pro set. On Free, the diagnostic runs on the whiteboard and the capped cobrowser (10 minutes per session, 20-minute cool-down) — enough to try a single 30-minute diagnostic, tight for anything longer.

Two implementation notes: (a) if you'd like to record the diagnostic session for your own notes or to share a short clip with the family (a parent seeing the child successfully build a fraction bar is worth a thousand words), see how do I record online tutoring lessons? — including the family-permission step; (b) if the student is in mainland China, the whiteboard and Math Tools all work through our Hong Kong proxy, but cobrowser reliability into US-hosted diagnostic sites can vary — see the China-specific answer for the workaround pattern.

Where the diagnostic fits inside the broader first-lesson workflow

The math diagnostic is one specific move inside a broader first-week onboarding. Three neighbouring pages that describe the rest of the shape:

How do I onboard a new online tutoring student? — the pre-lesson checklist, parent-invite mechanics, and first-week parent conversation. Subject-agnostic; this page is the math-specific diagnostic sub-step that lives inside it.
How do I run a free trial lesson that converts? — when the diagnostic doubles as a trial (typical for a new indie tutor still building their first ten paid students). A well-run diagnostic-trial answers two questions at once: is this tutor a good fit for my child, and what would we work on if we hired them.
How do I structure a 1-on-1 online lesson? — the per-lesson shape (opening, activities, close, homework) that applies to every ongoing lesson after the diagnostic session.

Why math word problems are the hardest thing to teach on a webcam

Every math tutor who's worked online with a student stuck on word problems knows the shape of the failure: the student can do the arithmetic when it's presented as "27 + 15 = ?" and freezes when the same numbers arrive inside a sentence — "Alex has 27 stickers. She gets 15 more for her birthday. How many stickers does she have now?" The instinct is that the child "isn't a word-problem person", or "just needs more practice", or "hasn't read carefully enough". Occasionally that's true. Much more often the block is in translation: the student can compute, but they can't figure out which operation to use because the problem's structure isn't visible to them. In an in-person lesson you'd reach for a pencil, sketch two boxes side by side, put "27" in one and "15" in the other, draw an arrow with a question mark, and the "oh — I add them" moment lands in ten seconds. On a video call with a screen share, that same sketch happens on your side of the screen and the student watches. They see the answer without doing the work to see the structure themselves.

Three problems compound to make this specifically hard remotely, and the online workflow has to answer all three:

Word problems are read differently from prose. A word problem is not a paragraph you read for meaning; it's a set of relationships you have to extract from a paragraph. Skilled solvers routinely re-read a word problem three times — once for what's happening, once for the numbers, once for the question — and mark it up as they go. On paper, the student annotates the sheet: underlining the numbers, circling the question, crossing out red herrings. On a webcam, if the problem is inside a PDF the student opens themselves and you can't see, none of that mark-up is happening on your surface. You're teaching to a black box.
The structure of a word problem is spatial. "Alex has twice as many stickers as Maya" is a comparison between two quantities of different sizes. That relationship is naturally visual — two rectangles side by side, one twice as long as the other. Text can't carry that meaning as directly as a picture can. If your whiteboard forces you to type the relationship as a sentence, you're translating twice — from the problem's English to your typed English, and from your typed English to the student's mental model. Bar models cut out the middle translation.
Word problems trigger the strongest math anxiety. More students freeze on word problems than on any other kind of math task. The freeze isn't about the math; it's about a paragraph of English that they know they're going to be judged on. Online, the freeze is worse — the student is in their bedroom with a webcam pointed at them and no easy way to hide "I don't know where to start". The workflow has to make it safe to begin before anyone gets to a correct answer. A bar model with the first bar wrong is still a start. A blank whiteboard is not.

An online word-problem workflow that works has an answer to each of the three: it puts the problem on a shared surface both people can annotate; it makes the structure visible through a bar model rather than through a sentence; and it gives the student a first move (place one bar, stretch it, label it) that isn't the same thing as committing to the final answer. The rest of this page is that workflow.

The move: bar models (tape diagrams) on the shared whiteboard

The bar model is the single technique that carries most word-problem teaching from about grade 2 through pre-algebra. It came out of Singapore's Primary Mathematics curriculum in the 1980s and now shows up under different names in Beast Academy (Art of Problem Solving), Eureka Math / EngageNY ("tape diagrams"), Maths — No Problem! (UK), Math in Focus (US), and most other mainstream problem-solving curricula. The technique is deceptively simple: for every quantity in the problem, draw a rectangle whose length is proportional to that quantity; for every relationship between quantities, put the rectangles next to each other so the relationship is visible; mark the unknown with a "?"; then read off the equation from the picture. It's the pictorial step of the concrete-pictorial-abstract sequence — students who can do the picture almost always can do the equation, and students who can't do the picture usually haven't understood the problem.

Koala Go's whiteboard ships the bar-model primitives inside the Whiteboard Library. What matters for word problems, before any workflow talk:

The Rectangle shape (Pro tier) is the bar-model workhorse. On the Pro plan (currently $25.99/month monthly, $21.99/month annual) you get a rectangle you can drop on the canvas and stretch to any width and any height independently — that's the "flexible" resize behaviour, contrasted with the square (which is proportional-only). So you can place one rectangle, drag its right edge until it looks the right length for "27 stickers", place a second rectangle underneath, stretch it to about 15/27 of the first's length, and the child can see the ratio in the picture before you write a number down. Because the rectangle is a first-class whiteboard object, both you AND the student can grab, resize, and move it. Which matters — the student doing the placement is where the understanding lives; a demonstration is not the same thing.
Number Rods (Pro tier) are the small-whole-number variant. When the quantities in the problem are small (typically ≤ 10) and the child is still building the concept, the 10 pre-sized coloured rods — values 1 through 10 — are often more useful than free rectangles, because the sizes are already exact and the student can count units directly. "Sam has 3 apples, Ana has twice as many" is one 3-rod above two 3-rods (or one 3-rod above a 6-rod), and the picture answers the question without any arithmetic. Rods are what you reach for in grades 2 through 4 with numbers that fit; rectangles are what you reach for beyond that.
Fraction Bars (Platinum tier) handle fraction word problems. "3/4 of the class went on the trip; 12 children stayed behind — how many went?" is a fraction-of-a-whole problem, and the model is the same shape as a whole-number bar model but the fractional pieces are placed one at a time. On Platinum ($49.99/month monthly, $39.99/month annual) you get individual 1/n bar segments for denominators 1/2 through 1/12 plus 1/16 — the same set the fractions spoke walks through in depth. See how do I teach fractions online? for the fraction-manipulative side; this spoke describes how those pieces show up specifically inside a bar-model word problem.
Shape Labels (Platinum tier) name each bar with a rescaling label. A "boys" label at the top of one bar, a "girls" label at the top of the second, and a "?" label above the difference — all four corner-anchored so they move and rescale with the bar when you stretch it. On Pro, the workaround is Sticky notes placed next to each bar (which drift if you resize the bar afterwards); most working tutors on Pro live with the sticky-note workaround for a while and upgrade to Platinum when they start doing multi-part bar models routinely.
Straight lines and arrows (Pro tier) are the "curly-brace" and "unknown" marks. A dashed line under a section of the bar with a "?" sticky above it marks the unknown quantity; an arrow from one bar to another with a "= 3 × ?" sticky above it marks a multiplicative relationship. Both are Pro-tier line primitives on the same canvas — the lines themselves don't take Shape Labels, so the annotation next to them is a sticky note or a text box. These small marks are what turn a bar-model diagram into a language you can read off — without them, two rectangles are just two rectangles.
The Math Symbols palette writes the equation on the same canvas. Once the picture is finished, the equation goes right next to it. The base Symbols set (Pro) covers the four operations, equals, and the basic "less than" / "greater than" comparisons — everything a pre-algebraic word problem needs. Advanced symbols (² ³ π ± infinity, ≤ ≥ ≠ ≈) sit on Platinum and are what you reach for once the natural next step becomes "let x = ...", which is where this page hands off to how do I teach algebra online?

Because all of these primitives live on the same shared whiteboard, the picture and the equation live in the same room — not in a slide-plus-Doc split that adds a cognitive tax every time the student's eyes move. That single property is most of the difference between "let me share my screen and we'll go through a problem" and "let's draw this together and figure out what it's asking".

The five bar-model shapes that carry most word problems

Working online math tutors typically converge on five bar-model shapes that between them handle most of the word problems in elementary and middle-school curricula. Learning to recognise which shape a problem needs is the highest-leverage move a tutor can teach — the shape names below are the ones used across Singapore's Model Method, Beast Academy, and most Eureka-adjacent curricula, so they'll be familiar if the student is already working in one of those systems.

Part-whole (single-quantity addition and subtraction). One bar divided into two labelled parts; the whole is either known or the unknown. "Alex has 27 stickers, then gets 15 more" is one bar of length 27, a second joined to it of length 15, and a "?" over the total. "Maya had 40 marbles, gave some away, and has 24 left" is one bar of length 40, split into a known part (24) and an unknown part ("?"). This is the earliest bar-model shape and the one to start with in grade 2 or 3 — students who own this can usually generalise the others in a few lessons.
Comparison (two quantities of different sizes). Two bars side by side; the difference between them is the answer, or the ratio between them is the answer. "Alex has 27 stickers. Maya has 12 more than Alex. How many does Maya have?" is one bar of length 27 with a second bar underneath that extends 12 beyond it, both labelled, the "?" over Maya's total. "Sam has twice as many pencils as Jamie" is one bar under a bar of twice the length. The comparison shape unlocks most of the elementary "how many more" / "how many fewer" / "how many times as many" problems.
Before-and-after (change over time). Two rows of bars, one showing the state before and one showing the state after. "Jamie had 60 baseball cards. She gave 1/3 of them to her sister. How many does she have left?" — top row: one bar of length 60, split into three equal parts (three thirds); bottom row: two parts remaining, one part labelled "?". The before-and-after shape is what makes "some were removed" and "then more arrived" problems traceable. Younger students often need to draw both rows before the shift makes sense.
Fraction-of-whole (fractional bar-model). One bar representing the whole, subdivided into fractional parts, with either the whole known and a fractional part unknown or vice versa. "3/4 of the class went on the trip; 12 stayed behind" is one bar of length "24 children" (student figures this out) split into four quarters, one quarter labelled 12 and three quarters labelled "?". The fraction bar-model is where the fraction spoke and this spoke intersect — the drag-and-drop 1/n fraction bars on Platinum are the tightest version, though a Pro tutor can achieve the same result by hand-drawing a rectangle and using sticky-note fractional labels. See how do I teach fractions online? for the fraction manipulative deep-dive.
Ratio (two related bars with proportional units). Two bars whose lengths are in a stated ratio. "In a bag of 24 sweets, the ratio of red to yellow is 3:5. How many are red?" is one bar of length 24 split into 8 equal units, 3 shaded as red, 5 shaded as yellow. Ratio is the last of the five and usually shows up in grades 5 through 8; it's the bar-model shape that most directly bridges into algebra.

The fastest way to build a student's bar-model fluency is not to teach all five at once — it's to run one shape for two or three weeks until the student can draw it without prompting, then introduce the next shape when the student meets a problem the previous shape doesn't fit. Part-whole first, then comparison, then before-and-after, then fraction-of-whole, then ratio, is the order most curricula use in some form.

A working lesson: from the word problem to the answer, on the shared canvas

What the workflow looks like in a real 30 to 45-minute session. This is not a script — adapt to the student and the problem — but the shape is broadly what most working online math tutors run once they've been doing bar-modelling with a student for a few weeks.

Put the problem where you can both mark it up (2 min). If the parent sent a worksheet, upload the PDF onto the whiteboard on Pro — it becomes a page both you and the student can annotate with the pen, sticky notes, and shapes. If the problem is from a textbook or a photo, drop it in as an image. If the problem is one you're inventing on the fly, type it in a sticky note at the top. What matters is that the problem lives on the same canvas as the bar model you're about to build — not in a link the student opens elsewhere.
Read the problem twice, out loud, together (2 min). First read, both of you read it silently at your own pace. Second read, the student reads it out loud. This is the reading-strategy step and it matters more than tutors typically credit — students who freeze on word problems often haven't finished reading them; forcing a slow spoken read anchors the language. As the student reads, underline the numbers with the pen tool and circle the question.
Name what's known and what's unknown (2 min). Ask the student: "What are the quantities in this problem?" and "What's the question asking us to find?" Take their answers verbatim onto a sticky note ("Alex: 27 stickers. Maya: ? stickers. Difference: 15 stickers."). This is the moment you catch a misread — if the student names quantities that aren't in the problem, or names the wrong unknown, stop and re-read. Fixing the misread here saves the next fifteen minutes.
Choose the bar-model shape (2 min). Ask: "Does this problem have one quantity, or two? If two, is one bigger and we care about the difference (comparison), or do we care about the total (part-whole)? Or is the same quantity changing over time (before-and-after)?" For the first few weeks of bar-modelling, walk the student through this choice; after a few weeks, they'll start naming the shape themselves. The bar-model shape is the abstraction the student is really learning; the arithmetic is downstream.
Draw the picture together (5 min). Place the rectangles or rods on the canvas. Stretch the first to roughly the right length; place the second and stretch it to a proportional length. Add the labels — Shape Labels on Platinum stick to the bar corners; sticky notes on Pro sit next to each bar. Add the "?" over the unknown and an arrow to the quantity that answers it. The student should be doing at least half of this — hand them the pen or the shape tool and let them place the second bar. If they place it wrong (too short, too long, in the wrong place), leave it and ask "does that look right compared to the first bar?" — the correction is the learning.
Read the equation off the picture (3 min). Once the diagram is complete, the equation is usually visible in the shape. Two bars totaling a "?" means addition — write "27 + 15 = ?" on the canvas using the Math Symbols palette. Two bars where the difference is the "?" means subtraction. Ask the student to write the equation next to the picture (not you). This is the pictorial-to-abstract step of the CPA sequence, and it's the step that most naturally transfers to problems the student will meet without you.
Solve the arithmetic (2 min). Almost always the shortest step — the student has already done the hard work of translating. If the arithmetic itself is a struggle (times-tables not automatic, addition-with-carrying wobbly), you've surfaced a separate gap; note it and route into the multiplication spoke or the number-sense spoke in a later lesson.
Check by reading back to the problem (2 min). "So the answer is 42. Does that make sense — Maya has 42 stickers, which is 15 more than Alex's 27?" Yes / no + why. Answer-checking against the story is often skipped and always worth doing.
Repeat with a variation (10-15 min). Same shape, different numbers. Same numbers, different question. Fresh problem of the same shape. The bar model is a template the student is grooving; three or four problems of the same shape in one session is what turns "we did that with the tutor" into "I know how to do these".

A single 30-minute session usually gets through 3-5 problems of the same bar-model shape at this pace. That's the target — not "how many problems can we get through" but "did the student's bar model for the third problem look better than the first". Fluency comes from repetition of the shape, not from a broad tour of shapes.

Where practice sites fit: the cobrowser for Beast Academy, Khan Academy, IXL, NRICH

Between lessons, the student needs word-problem practice they can actually complete without the tutor sitting next to them. Sending a worksheet home works, but the highest-leverage move is to open a curated problem-set inside the lesson using the cobrowser — a shared browser window where both you and the student can click, type, and interact with the same page — and finish the session with a few problems the student attempts semi-independently while you watch. Well-shaped problem sets for bar-model practice, in rough order of what most tutors reach for:

Beast Academy (Art of Problem Solving) — problem-solving-heavy curriculum from grades 2 through 5, with word problems in every chapter designed to be attacked with a bar model. Their problem sets are especially good for the "novel problem that doesn't fit the exact template" case.
Khan Academy — free, broad coverage, hints available. Their word-problem practice is calibrated for typical grade-band expectations and works well as spaced-practice homework. Search the site for the specific concept ("multi-step word problems 4th grade"; "fractions word problems").
IXL — subscription, but very deep problem-set library across every US state's curriculum. Good for parents who already have an IXL subscription. IXL's adaptive difficulty is a mixed blessing for bar-model practice — the student can slide into problem types the bar model doesn't fit; steer them back to a specific topic strand rather than free-choice work.
NRICH (Cambridge) — free, world-class problem collection. Older-student oriented (upper elementary and up). Good for the student who's fluent with the standard shapes and wants harder problems.
Bedtime Math and Illustrative Mathematics tasks — additional free sources for specific shapes.

Open the practice site in the cobrowser rather than as a link the student opens themselves. The cobrowser gives you actual shared control (not screen sharing) — you see exactly what the student clicks, why they got it wrong, and can jump in without them switching windows. On Koala Free the cobrowser is capped at 10 minutes per session with a 20-minute cool-down; on Pro it's unlimited. See what is a cobrowser? for the mechanic in depth and how do I give homework to online tutoring students? for the between-lesson practice cadence — the bar-model habit is one of the highest-ROI homework asks in elementary math.

Common failure modes and what to do about them

Six patterns that show up regularly with word-problem tutoring online, and the workflow adjustment for each.

The student draws the bar model correctly and still writes the wrong equation. The picture-to-equation gap. Usually the fix is one more round of "what does the picture say?" — literally reading each bar out loud and pointing at it: "This bar is 27. This bar is 15. The two together are what?" Sometimes the student can talk their way to the answer even though they can't write the equation; that's still progress. Write it for them once, ask them to write it themselves the next problem, and let the habit build.
The student can't decide which bar-model shape to use. Common in weeks 2-4. Fix: for a few sessions, tell them the shape ("This one's comparison — two bars of different lengths") rather than asking them to name it. The naming skill lags the drawing skill by a few weeks; that's fine. Once they've drawn twenty part-whole models and twenty comparison models, they'll start recognising the shape from the language.
The student stretches every bar to the same length regardless of the numbers. A common short-cut. Fix: place the "1" rod next to their bar and ask "how many of these fit inside your bar?" — the numerical mismatch becomes visible. For older students, add tick marks to the bar (labelled units) so the length constraint is explicit.
The student gets the bar-model but freezes when the problem has more than one step. Multi-step word problems are their own harder skill. Fix: draw the bar model in two stages ("first, let's answer the sub-question this problem needs before we can get to the real question"), and use the before-and-after shape (top row = state at step one; bottom row = state at step two). The two-row layout makes multi-step visible.
The student's block is in reading the problem, not in solving it. If the student reads "twice as many" as "two", or reads "combined" as an operation they don't recognise, or can't decode a word in the problem at all, you may have surfaced a reading rather than a math gap. Slow down; read the problem out loud with them; look up the confusing words together. If the pattern is persistent — the student reads the numbers fluently but can't parse the sentences — cross-link to how do I tutor a student with dyslexia online? for the reading-side workflow and consider suggesting the family talk to a reading specialist. Reading fluency is upstream of word-problem fluency.
The student's block is persistent and structural — even simple bar models don't stick after weeks of practice. If magnitude comparison stays unreliable, number-symbol mapping is confused, or the bar model keeps getting drawn without any grasp of what it represents, you may be seeing a pattern that could suggest a math learning difference. Read how do I tutor a student with dyscalculia online? for the SEN-specific workflow, and — separately — suggest the family talk to an educational psychologist or the school's SEN coordinator for actual assessment. A tutor's observation of a pattern is not a diagnosis; a qualified specialist diagnoses.

When to hand off to algebra: the "let x = " moment

Bar-modelling scales up remarkably well — Singapore's curriculum runs bar models through pre-algebra, and there are respectable versions of quadratic and simultaneous-equation bar models. But at some point in middle school, the natural next step becomes "let x = the number of pencils" and the picture stops being the primary tool. Signs the student is ready for the algebra hand-off:

The student is drawing the bar model in their head and getting to the equation without drawing.
Problems introduce quantities that need multiple operations combined ("If she buys 4 pencils and 3 notebooks and spends $23, and each notebook costs $2 more than a pencil…").
Problems introduce a variable in more than one place ("A rectangle has length twice its width; the perimeter is 30cm…").
The student is comfortable naming unknowns with letters rather than "?" boxes.

At that point, the workflow shifts: the bar model becomes a supporting scaffold ("let's draw the picture first, then write x = the unknown quantity"), and the algebra takes over as the primary technique. The algebra spoke covers this in depth — see how do I teach algebra online? for the balance-scale equation-solving workflow, the area-model for expanding products, and the Desmos cobrowser pairing for graphing. Bar-modelling is what gets a student ready for algebra; algebra is what they need once they arrive.

Honest limits

Four limits worth naming plainly rather than papering over.

Koala's whiteboard has no automatic bar-model generator. The tutor draws each rectangle by hand and stretches it to the right proportion. There are curriculum-specific tools (Thinking Blocks, some Beast Academy in-book apps) that let you type "27 and 15" and produce a bar model automatically; Koala doesn't. In practice most working tutors prefer hand-drawing because the manual step is where the student's participation lives — but if you want a click-to-generate diagram, the whiteboard isn't that.
Shape Labels are Platinum-only. If you're on Koala Pro ($25.99/month monthly, $21.99/month annual) and want your labels to rescale with the bar, you'll need to upgrade to Platinum ($49.99/month monthly, $39.99/month annual). On Pro, sticky notes next to each bar work as the label workaround — they don't drift much if you finalise the bar dimensions first and place the stickies after, and multi-part bar models can absolutely be run on Pro.
iPad is rougher than desktop for the drag-and-stretch gesture. The bar-model workflow depends on precise resizing of small rectangles, which is easier with a mouse than with a fingertip on an iPad. Students on iPads can still do bar models, but they often benefit from a stylus or from a slightly larger placement. On a desktop or laptop with a mouse, the gesture is smooth and quick.
The Playground and the 3D environment aren't the teaching surface here. Word problems live on the shared whiteboard; the Playground is where younger students take a 3-minute movement break between problems if focus is flagging. See what is a gamified virtual classroom? for the engagement layer around a math lesson without confusing it for the teaching surface.

Two implementation notes: (a) if you'd like to record a strong bar-model session for your own notes or to share a short clip with the family (a parent watching their child correctly draw a comparison bar for a problem they'd been stuck on is worth an hour of parent-conference explanation), see how do I record online tutoring lessons? — including the family-permission step; (b) if the student is in mainland China, the whiteboard, shape library, and Math Tools all work through our Hong Kong proxy, but cobrowser reliability into US-hosted practice sites can vary — see the China-specific answer for the workaround pattern.

Where word-problem teaching fits in the wider math cluster

Bar-modelling is one of several skills a working online math tutor teaches. The neighbouring pages that cover an adjacent question:

How do I teach math online? — the umbrella pillar. Start here if you're figuring out the whole math workflow, not just word problems.
How do I assess a new online math tutoring student? — the first-session diagnostic. A well-run diagnostic tells you whether word-problem work is where the student's gap sits before you build a bar-model curriculum around it.
How do I teach fractions online? — for fraction-of-whole word problems and the Fraction Bars manipulative deep-dive.
How do I teach number sense online with virtual manipulatives? — for the Number Rods bar-model variant with small whole numbers.
How do I teach multiplication and times tables online? — for equal-group word problems and the fluency-drill side that unblocks arithmetic once the bar model is right.
How do I teach geometry online? — for area, perimeter, and volume word problems that use the shape library alongside bar-model translation.
How do I teach algebra online? — for the "let x = " transition and algebraic word problems once bar-modelling has done its job.
How do I write math equations and notation on an online whiteboard? — for the Math Symbols palette that writes the equation next to the bar model.
How do I tutor a student with dyscalculia online? — the SEN spoke for persistent structural blocks in word-problem understanding.
How do I tutor a student with dyslexia online? — the SEN spoke when the block is reading the problem rather than doing the math.
How do I give homework to online tutoring students? — for the between-lesson bar-model practice workflow.
How do I share progress updates with parents? — for the parent debrief when word-problem fluency starts moving.
What is a cobrowser, and how is it different from screen sharing? — for the practice-site pairing mechanic (Beast Academy, Khan Academy, IXL, NRICH inside the lesson).
What is a virtual classroom, and how is it different from Zoom or Google Meet? — top-of-funnel for families still deciding whether they need a virtual classroom rather than a video call.
What are the best online whiteboards for tutoring? — whiteboard-tool comparison hub.
How much should I charge for online tutoring? — includes the math-tutoring premium band, which typically applies more strongly for tutors who demonstrably teach the bar-model workflow rather than only running problems.

If you'd like a walk-through of the bar-model surfaces for the specific age band, curriculum, or word-problem type you're teaching, open Koala Go at classroom.teachwithkoala.com and try one of the shape types above on the whiteboard, or write to koala@teachwithkoala.com with a sentence about the student and the problem set you're working with, and we'll give you our honest read on the fit — including the parts of your specific curriculum (Beast Academy, Singapore Math, Eureka Math, or something else) where the bar-model on a shared whiteboard is the right primary tool and the parts where a subject-specific practice site in the cobrowser is doing more of the work.

View full answer page →

Choosing an Online Tutoring Platform — A Practical Q&A

What an "online tutoring platform" actually is

How to actually choose one

Questions answered in this category

Choosing between platforms

Platform features that decide it

Running a tutoring business on a platform

Teaching specific students and subjects

Where Koala Go fits — and where it doesn't

The 6 best Zoom alternatives for online tutors

1. Koala Go — best for independent K-12 tutors and engagement-driven lessons

2. Lessonspace — best for STEM and subject-specific tutoring

3. LearnCube — best for ESL and language teaching with a ready curriculum

4. Vedamo — best for group classes and classroom-style live tutoring

5. Whereby — best for clean 1-on-1 video calls with adult learners

6. Google Meet — best free fallback when you don't need teaching tools

Feature comparison at a glance

How to choose between Zoom alternatives

Free Zoom alternatives for online teaching

Koala Go (Free) — best for free 1-on-1 and small-group K-12 tutoring

Whereby (Free) — best for short adult 1-on-1 calls with no install

Google Meet (Free) — best free fallback with maximum parent familiarity

What about Lessonspace, LearnCube, and Vedamo?

A note on free open-source video tools

Tradeoffs of going free for online teaching

When sticking with Zoom is still the right call

Questions to ask any platform before you commit

The 6 best online whiteboards for online tutors

1. BitPaper — best for general 1-on-1 tutoring with handwriting

2. Ziteboard — best for math tutors on a vector canvas

3. Lessonspace — best for STEM tutors who need equation, graph, and code editors

4. Koala Go — best whiteboard inside a full virtual classroom for kid-focused tutors

5. Miro — best for adult learners and design or business teaching

6. Microsoft Whiteboard — best free option for tutors already on Teams or Microsoft 365

Feature comparison at a glance

How to choose between online whiteboards for tutoring

A note on Google Jamboard

What makes a whiteboard good for tutoring, specifically?

Related answers

What a cobrowser actually is, under the hood

How screen sharing fails for tutoring (the do-the-thing-yourself problem)

What changes when you teach on a cobrowser instead

Where screen sharing is still the right tool

Which online tutoring platforms have a cobrowser, and how they differ

What a cobrowser is not

Real classroom workflows tutors use a cobrowser for

Cobrowser support in Koala Go specifically

How to decide whether you need a cobrowser

Related answers

What Google Meet actually is

What's missing for online tutoring

The typical "Google Meet + bolt-ons" tutoring stack

When Google Meet is a reasonable fit for tutoring

When it isn't

What about recording on Meet?

Purpose-built alternatives — what tutors typically switch to

Bottom line

Related answers

What actually happened to Skype

Why Skype worked for tutors (and what it always left you doing manually)

What Microsoft Teams Free actually gives you

The typical "Teams Free + bolt-ons" tutoring stack

When Teams Free is a reasonable fit for tutoring

When it isn't

Purpose-built alternatives — what tutors typically switch to

Bottom line

Related answers

Where WhatsApp actually fits in most tutors' workflows

What WhatsApp itself gives you

Where a WhatsApp-as-classroom stack breaks for tutoring

The typical "WhatsApp + bolt-ons" tutoring stack

When WhatsApp-as-classroom is a reasonable fit

When it isn't

Purpose-built alternatives — what tutors typically switch to

Keep WhatsApp for what it's good at

Bottom line

Related answers

Where Discord actually fits in tutoring today

What Discord actually is (as a category)

Where a Discord-as-classroom stack breaks for tutoring