Why Language Tutors Deserve a Fair Go: The Hidden Economics of Online Tutoring (And What Needs to Change)

The uncomfortable truth about online language tutoring in 2026

If you're a language tutor reading this, you already know how the maths works. You spend an hour with a student and the platform takes a slice of your earnings. That's the deal you signed up for, and on its face it sounds reasonable — the platform finds you students, runs the booking system, handles the payments, and takes a fee. Fair enough.

Except the fee isn't always reasonable. And the rest of the deal — the parts nobody mentioned at sign-up — often isn't either.

Let's start with what the major platforms actually charge in 2026, because the numbers tell a story most marketing pages would rather you didn't see.

Preply's commission starts at 33% for new tutors. That's a third of your income, gone, on every lesson. The rate steps down as you teach more hours: 28% after 20 hours, 25% after 50, 22% after 100, 20% after 200, and finally 18% after 400 hours. To put that in human terms, you have to teach the equivalent of ten full-time weeks before the platform stops taking nearly a quarter of your money. Most tutors never reach the lowest tier. The ones who do have spent months working at rates that, after commission, often barely clear minimum wage in the country they live in.

iTalki and Verbling charge a flat 15%. That's better. It's predictable, it doesn't punish new tutors, and it's the rate most experienced tutors will tell you they consider the absolute ceiling of "fair" for what these platforms actually deliver.

Some platforms charge 100% on trial lessons. That's not a typo. On Preply, when a new student books their first 50-minute lesson with you, you receive nothing. Zero. The platform takes the entire fee on the basis that the trial is a customer acquisition cost — and they've decided you, the tutor, should pay it. The tutor is doing the work, taking the time, preparing the materials, and showing up to teach a real lesson. The platform keeps every cent.

If you're wondering whether tutors find this acceptable, the answer is very clearly no. Tutor forums, Reddit threads, Glassdoor reviews, and every honest comparison article published in the last three years cite high commissions and unpaid trial lessons as the single biggest reason experienced teachers either leave these platforms or refuse to join in the first place. One veteran IELTS coach with fifteen years of teaching experience put it bluntly in a 2025 review: she described Preply as taking "probably the largest commission of any platform" and noted the irony that high-quality teachers are routinely rejected from their application process while less qualified ones get through.

This isn't a few angry tutors having a bad week. It's a structural pattern that has held for years and shows no sign of changing.

The unpaid hours nobody talks about

Commission is the visible problem. The invisible problem is everything else.

Every hour of one-on-one language tutoring requires somewhere between thirty minutes and two hours of preparation, depending on the student's level, the lesson focus, and how customised the materials need to be. Beginner conversation classes can be lighter. Exam prep sessions for IELTS, TOEFL, or DELE are heavier — often heavier than the lesson itself. A tutor preparing a student for IELTS speaking might spend an hour finding appropriate practice prompts, building a custom vocabulary list, recording feedback notes from the previous lesson, and reviewing the student's progress before the lesson even starts. Then they teach for an hour. Then they spend another twenty minutes writing up notes for the student to take away.

That's two and a half hours of work for one hour of paid lesson time. The tutor is paid for one hour. The platform takes its commission on that one hour. The other ninety minutes of work — the part that actually makes the lesson valuable — is invisible to the platform's accounting and invisible to the student's invoice.

Multiply this across a working week. A tutor who teaches twenty paid hours is realistically working thirty to forty unpaid hours alongside them. At a notional $25 per paid hour, after a 20% commission, that's $400 a week in pay for forty to sixty hours of actual work. The effective hourly rate, once you count the prep, is between $7 and $10 per hour — well below the minimum wage in most developed countries, and often well below what the same tutor could earn doing almost anything else.

The platforms know this. The platforms have always known this. The reason they don't talk about it is that the unpaid work is what makes their margins look the way they do. If tutors were paid for prep time, lessons would cost more, fewer students would book, and the platforms would have to either raise their commission to maintain margins or accept lower margins. None of those options are appealing to a venture-backed marketplace optimising for growth metrics.

The lesson materials economy that doesn't exist (yet)

Now let's talk about the part that's even more invisible than prep time: the materials.

Every working language tutor builds a library. Over years of teaching, they accumulate hundreds of worksheets, vocabulary lists, conversation prompts, grammar drills, lesson plans, reading comprehension exercises, listening tasks, and exam preparation materials. Some are adaptations of textbook content. Some are original. The best tutors have entire pedagogical systems built up over a decade or more, refined lesson by lesson with real students.

Where does this work go on a traditional tutoring platform?

It goes onto the tutor's own laptop. It sits in folders organised by language, level, and topic. It gets used in lessons, often shared with students as a PDF after the class, and then it sits there. The tutor cannot sell it. The tutor cannot publish it on the platform for other students to discover. The tutor cannot earn from it after the lesson it was used in is over. The platform doesn't have a mechanism for any of this, because the platform is designed around the unit of "one lesson" as the transactional event, and once the lesson is done, the value the tutor created during prep is invisible again.

This is a strange situation when you think about it for more than ten seconds. We are in the middle of the largest creator economy expansion in human history. Substack pays writers. YouTube pays video creators. Spotify pays musicians. Patreon pays everyone. In every other knowledge industry, the people who make the materials get paid for the materials. In online language tutoring, the people who make the materials get paid only for the moments they personally spend in a video call. The materials themselves are treated as overhead — a personal cost that the tutor absorbs, like buying a webcam or a notebook.

This is not because tutoring materials lack value. They obviously have value. A great IELTS speaking practice pack would help thousands of students if they could find it. A well-built A2 Spanish vocabulary worksheet would save hundreds of tutors hours of prep time if they could buy it or licence it. The value exists. The infrastructure to capture it doesn't.

The platforms haven't built it because building it would mean acknowledging that tutors are creators, not just contractors. And acknowledging tutors as creators would mean rethinking the entire revenue split — because creators, in every other industry, get paid royalties on what they create, not just hourly rates on what they perform.

What "fair" actually looks like

So what would a fair-go tutoring platform look like in practice? Let's get specific, because this is the part where vague manifestos usually fall apart.

A fair platform charges a reasonable commission on lessons. Not 33%. Not 25%. Not 18% after you've worked yourself to exhaustion. Something in the 10% to 15% range, applied flatly from day one, with no tiered scheme that punishes new tutors. The platform's job is to find students, run the booking system, handle payments, and provide the technology — not to take a quarter of the tutor's income for doing it.

A fair platform pays tutors for trial lessons. A trial is a real lesson with real teaching. The platform's customer acquisition costs are not the tutor's problem to subsidise. If a platform wants to offer free or discounted trials to attract students, the platform should pay for them out of its marketing budget, not out of the tutor's earnings.

A fair platform recognises that tutors are creators, not just contractors. The materials a tutor builds over years of teaching are valuable intellectual property, and they should generate ongoing income for the tutor whenever they're used by other learners. This means a content library where tutors can publish their lesson materials and earn from views, downloads, or licences. It means clear attribution. It means royalties that flow even when the tutor isn't personally in a video call.

A fair platform is transparent about its rules and its money. Commission rates should be on the homepage, not buried in a help article. Payout rates should be predictable. Changes to rates should require advance notice — measured in months, not days. The relationship between the platform and the tutor should look more like a partnership and less like a gig-economy app squeezing the last cent of margin out of a labour pool.

A fair platform protects tutors' time. That means honest cancellation policies (a student who cancels with five minutes' notice should not result in the tutor receiving 50% of the lesson fee, as happens on some platforms — that's an insult dressed up as a policy). It means respecting that prep work is work. It means building product features that reduce prep burden — AI lesson planning tools that draw on existing materials, shared content libraries, ready-to-use templates — rather than treating prep as an externality the tutor handles in their own unpaid time.

A fair platform treats group sessions like the volume product they are. Conversation groups, mixed-level chats, and peer practice sessions are how a lot of tutors actually want to teach — they're more sustainable, less draining, and better for students who learn faster in social settings. But many platforms either don't support group sessions at all, or apply punitive commission rates to them on the basis that the per-student price is lower. A fair platform supports group teaching as a first-class option with commercial terms that make it worth a tutor's time.

What Parlora is doing differently

This is the part where every marketing post pivots into a sales pitch, and we want to be honest about what we're actually offering rather than overselling it. Parlora is in pre-beta. We have not solved every problem in language learning. We have not built a perfect platform. We are a small, founder-led product, and we are going to make mistakes.

Here is what we've decided, on the record, about how Parlora treats tutors.

Tutors keep 87% of every lesson fee. Not tiered, not conditional, not subject to trial-period withholding. Eighty-seven percent, from the first lesson, forever. If a student pays $30 for an hour with you, you get $26.10. Stripe processing fees and Parlora's share come from the remaining 13%. That is the rate, and it is on our home page rather than buried in a footnote.

We do not take 100% of trial lessons. We do not take any of them, because we don't think trial lessons should be unpaid. If a tutor offers a discounted trial as their own marketing decision, that's their call, but the platform doesn't pocket the proceeds.

Parlora Library is a content library where tutors earn from their teaching materials. This is the part we're most proud of. Tutors can upload their lesson plans, worksheets, exercises, and reading materials into Parlora Library, and earn payouts based on views from other learners on the platform. Our payout rate is A$2.00 per 200 unique views, paid monthly, with no minimum threshold tied to your subscription tier. It's not going to make anyone rich, but it's the first time most tutors have ever been paid for the materials they've been making for free for years. We think that matters even if the early numbers are modest, because the principle of paying creators for what they create is the principle the rest of the industry has been ignoring.

Conversation groups are a first-class product on Parlora, designed for tutors who want to build a sustainable teaching practice that doesn't depend on being personally available for every lesson. Group sessions are fixed at A$9 per person, and you keep 87% of every group fee — the same rate as one-on-ones. We started at 50% during early beta and got direct feedback from tutors that it felt high relative to the per-seat value. We listened, and 13% is where we landed — the same rate as one-on-ones.

We use Didit for identity verification, and the documents never touch our database. This isn't directly about pay, but it's part of how we think about respecting tutors. Your government ID is yours. Our verification provider checks it and tells us yes or no. We never see it, store it, or process it. We think that's how it should work everywhere.

Our cancellation policy is fair to both sides. Students who cancel with reasonable notice get a full refund. Students who cancel last-minute pay the tutor for their time. We're not going to hide behind a policy that lets students stiff tutors at five minutes' notice and call it customer service.

We are not perfect. Our group session commission needs to come down. Our content payout rates need to grow as the platform grows. There are corners of the product that aren't built yet. But the values we are starting from are different from the values the rest of the industry has been operating on, and we think those values matter.

Where this leaves you

If you're a tutor reading this, here's the honest takeaway. The big platforms are not going to change. The 33% commission, the unpaid trial lessons, the invisible prep, the materials economy that doesn't exist — these are not bugs that the existing players will fix. They are features of a business model that was designed around growth metrics for investors, not sustainability for the people doing the teaching.

The only way the model changes is if there are alternatives, and the only way there are alternatives is if tutors are willing to try them when they appear. We know that's a big ask. The big platforms have inertia, student bases, and brand recognition that a beta-stage product like Parlora cannot match. We are not asking you to leave them. We are asking you to consider, when you have a spare evening, whether there's room in your teaching life for a platform that treats you the way you should have been treated all along.

If you're a learner reading this, the takeaway is different. You are not the villain in this story. The platforms have made choices about how to treat tutors, and those choices are not your fault. But the next time you book a trial lesson on a platform that pays the tutor nothing for it, you might think about what that means for the person on the other end of the video call. Tutors who are paid fairly teach better. They burn out less. They build better relationships with their students. They stick around long enough to actually become great teachers. The economics matter for everyone, not just the people on the supply side.

And if you are a fellow founder in the language learning space, consider this an open invitation. The status quo serves nobody well in the long term. Investors want returns and the current model is producing churn instead. Students want quality and the current model is producing burnout. Tutors want sustainability and the current model is producing exits. There is room for more than one platform that builds differently, and the more of us doing the work, the faster the whole industry moves.

A few final words on what we mean by "a fair go"

The phrase "a fair go" is a deeply Australian one. It doesn't translate perfectly into other languages or other cultures, but the spirit is something like this: everyone deserves the chance to do their work, be recognised for it, and earn a reasonable living from it without being squeezed by the people in charge of the system. It's the principle that the person doing the labour should have a meaningful say in how the labour is valued. It's the principle that hidden costs and unpaid hours and silent extractions are not the price of doing business — they're a choice, and a different choice is possible.

We named Parlora to mean "place of speaking" — the conversation hub where tutors and learners come together. But the name also carries something about voice. About being heard. About the people who do the actual teaching having a seat at the table when decisions are made about how the teaching is valued.

If you want to tutor on a platform that takes that seriously, we'd love to hear from you. You can join our pre-beta waitlist at parlora.app, and we'll be in touch as we open up Founding Tutor spots in the coming weeks. There are no fake numbers, no inflated promises, and no marketing tricks. Just a small platform built by people who think the existing options can be done better, and who are willing to back that belief with the way we run the business.

A fair go. That's the whole pitch.