Building Kadensy: Why We Chose Credits Over Subscriptions

#TL;DR

When we built Kadensy across 2024 and 2025, we had to pick a pricing model. The default for English-tutoring marketplaces is subscriptions, sometimes paired with package billing. We chose a credit wallet with no expiry instead. This article is the Kadensy team’s reflection on why we made that call, what we gave up, and what we would build differently if we started Kadensy v2 from a blank repo today.

This is a product-philosophy note, not a feature announcement. If you want the product tour, read What is Kadensy first.

#The subscription model, what it does well, and where it breaks

Subscriptions are an excellent fit for products with predictable, steady consumption. Netflix, Spotify, and gym memberships work because most users want roughly the same thing every month. The billing cadence and the consumption cadence line up, so the customer feels they are getting value each cycle.

Language learning is not that kind of product. Adult learners study in bursts: before a job interview, before a relocation, before a presentation in English, before a holiday abroad. Most weeks are normal, and some weeks are simply impossible. A subscription does not care which week is which. It bills on schedule whether you took four lessons or zero.

We saw three specific failure modes that pushed us away from a subscription default for the student side of Kadensy.

The first failure is variable consumption. If you have a brutal work week, a sick child, or a travel day, the four lessons your monthly tier paid for will not all happen. The unused credit usually rolls into a tightening reset window, then expires. The customer ends up paying for inventory they never consumed.

The second failure is internal motivation. Adult language learners are mostly intrinsically motivated. They are not trying to win a streak; they are trying to pass IELTS, get a promotion, or feel less anxious on a sales call. A monthly reset does not add motivation for that profile; it adds guilt.

The third failure is cancellation friction. People stay subscribed past the point of value because cancelling feels like admitting defeat on a personal goal. That is a dark pattern even when it is not designed to be one. We did not want to ship it.

#The credit wallet model and what makes it different

The credit wallet decouples payment from consumption. You buy credits when you have decided to commit. You spend them whenever you actually have a free hour. The two timelines do not have to match.

In the Kadensy database, the wallets table has no expires_at column. We mention this often because it is the most concrete way to prove a marketing claim. The “Credits never expire” line on our landing page is not aspirational copy; it is the literal shape of the schema. A six-week travel break does not penalize you. A three-month parental-leave pause does not penalize you. Your balance is waiting.

The bet behind that choice is simple. Adults learning English do not need a monthly invoice to keep them in. They need a platform that respects their rhythm and gets out of the way when life intervenes. If we were right about the audience, the credit wallet wins.

The tradeoff is honest: without a billing cadence pushing you each month, you need self-discipline to maintain a learning rhythm. We accept that. Some learners will do better with a subscription product that forces them to log in. That is fine; the market is big enough for both models.

#The 3 product principles behind the choice

The three principles below are the lens we used whenever a pricing decision came up in code review or in design.

• Honor the user’s actual rhythm. Most adult learners study in bursts. Credits accommodate bursts; subscriptions punish them. If a student wants to do ten lessons in two weeks before a job interview, then disappear for a month, our system should treat that as normal, not as a billing edge case.
• Make cancellation a non-event. There is nothing to cancel on Kadensy. If you stop using the platform for six months, your credits wait. If you return, they are still there. We never want a customer to feel trapped by a billing relationship they cannot exit cleanly.
• Let tutors set their own rates. Each tutor sets their hourly credit rate through their profile. Subscriptions force a single platform-set rate, which compresses tutor differentiation and turns the marketplace into a race to the bottom. Credits let an experienced specialist price their expertise, and let a newer tutor price for volume.

The third principle has the biggest second-order effect. When tutors set their own rates, the marketplace becomes a real two-sided market where supply and demand actually meet on price. Subscription platforms hide that price discovery from both sides.

#What we gave up by not running subscriptions

Three real costs we accepted.

The first is predictable revenue. Subscription MRR is much easier to forecast than credit-pack purchases. Investors prefer the smoother curve. We chose to live with the volatility because we did not want to optimize for a chart at the cost of the customer experience.

The second is easy upsell paths. Subscription tiers are natural upgrade ladders: Basic, Pro, Premium. Credit packs require us to make each pack feel like a good deal at its size, and to let the customer self-select. That is harder to design and harder to merchandise, but it puts the choice in the buyer’s hands.

The third is behavior nudging. “Your lessons reset next Monday” is a powerful push notification. We do not have that lever. Our retention nudges have to come from tutor relationships, lesson reminders, and the learner’s own goals. We cannot scare anyone into a session.

#The EUR-native angle

A subtler choice sits underneath the credit model. We ship dual EUR and USD pricing from day one, and our Cashier-Stripe configuration defaults to EUR. The Laravel config file actually carries the comment “Cadence is EUR-first” (Cadence being the internal codename of our app server, the user-facing brand is Kadensy).

Most English-tutoring platforms are USD-only on the rails. Per Preply’s own published tutor commission disclosure, their tier sits at 18 to 33 percent depending on lesson volume (Preply, 2026, source: their “best online English teaching jobs” article). Their payouts run in USD via PayPal or Payoneer. European learners and tutors absorb the FX cost on every transaction.

EUR-native pricing on Kadensy means EU expats and Eurozone-based learners see honest prices without bank-conversion fees stacking on top. On the tutor side, Stripe Connect Express derives each tutor’s payout currency from their bank country across 40 supported regions. A tutor in France with a French bank gets paid in EUR. A tutor in Texas with a US bank gets paid in USD. We do not impose a single currency on a global marketplace; we honor whichever currency the tutor’s bank actually uses.

#What we would change if we built Kadensy v2 from scratch

Three honest things we would revisit if we started over.

We would test a hybrid subscription tier. The market signal is mixed: there is a real segment of learners who DO want the monthly cadence as a forcing function. A hypothetical “Pro” tier at a fixed monthly price could grant a base allotment of credits plus a top-up bonus. We resisted this in v1 to keep the brand simple and the message crisp, but a clean opt-in subscription tier would not contradict the no-expiry promise on the wallet.

We would compress tutor onboarding. The 40-country Stripe Connect Express setup is robust, but the KYC and identity flow is more verbose than it needs to be for tutors in low-risk jurisdictions. We would invest more in a progressive flow that lets tutors create a profile, then complete payout setup just before their first payout.

We would ship a native mobile app earlier. We built web-first to control the surface area while the product matured. Per our public roadmap, native mobile sits in Phase 2. The cost of not having it today is real for drop-in conversion, because drop-in lessons benefit from a push notification a tab-switch cannot match.

We are not going to claim the v1 choices were perfect. They were deliberate, and we know which ones we would revisit. That feels more useful to share than a tidy origin story.

#Why this matters for you

If you are a learner, the credit model means you are not paying for our financial smoothing; you are paying for actual lessons. Your balance is yours, your rhythm is yours, and your cancellation cost is zero.

If you are a tutor, you set your own rate without platform pressure to discount, and you get paid in your bank’s currency.

If you are building a marketplace yourself, the lesson is simpler than it looks: the pricing model is a product choice, not a billing detail. It shapes which audience you attract, how they behave, and how your supply side competes. Pick it on purpose.

#Try Kadensy

If this philosophy fits how you actually learn, the next step is short. Browse vetted tutors at /tutors, read What is Kadensy for the full product tour, or check Who Kadensy is for to see which learner archetype matches yours. If you are ready, create an account at /register. No subscription, no commitment, no expiry on what you buy.