TL;DR: Language learners are saying two things out loud in app-store reviews: their paid app measures streaks instead of fluency, and a free LLM out-teaches it. PainHunt's EdTech cluster describes energy-gated lessons, bait-and-switch paywalls, and two years of use that underperformed two weeks of immersion. The opportunity is outcome-measured, LLM-native practice — not another freemium tier.
The evidence
Language Learning surfaces in PainHunt with average score 11.2/15 and intensity 8.2/10 across 771 high-signal posts, arriving overwhelmingly through GooglePlay (42) and AppStore (18) — paying mobile learners rating the apps they actually use.
The complaints cluster into two layers. The first is monetization friction: an app becomes inaccessible after a 3-day trial, an energy system blocks lessons until you subscribe, and basic functionality is locked behind an annual commitment learners call "way too expensive" — a sudden paywall after a good first impression that reads as bait-and-switch. The second, and more damaging, is an efficacy verdict: one learner reports two years in an app yielding less than two weeks of travel immersion, and the recurring, quietly devastating observation that a free LLM now provides better practice than the paid subscription. The apps are tuned to maximize daily engagement — streaks, energy, notifications — and learners have noticed that engagement is not the same thing as learning.
The requested features name the gap: a genuinely useful free tier with real lessons, a methodology that produces measurable results, and pricing that does not feel like it is optimized against the learner. Underneath all of them sits one demand — practice that is measured by whether you can use the language, not by how many days in a row you opened the app.
Why now
Conversation practice went from scarce to nearly free. The single most expensive part of language learning used to be a patient partner who would talk with you at your level and correct you; an LLM does that on demand, in any language, at near-zero marginal cost. The thing the incumbents' gamification was a substitute for is now directly available.
At the same time, freemium fatigue peaked. Learners are no longer comparing paid apps to each other — they are comparing them to a free chat window that, for pure practice, many now prefer. When your paying users publicly recommend a free alternative, the pricing model is not the problem; the value model is. The opening is for a product that treats the LLM as the practice engine and competes on outcomes the incumbents never measured.
The wedge
Do not build another gamified freemium app. Build the outcome-measured practice layer that a blank chat box lacks.
- Structured, level-appropriate conversation built on an LLM, with a real curriculum and spaced review — the scaffolding a raw chat window does not give you.
- Error tracking that remembers what you get wrong and brings it back, so practice compounds instead of resetting each session.
- Progress measured against usable outcomes — can you order a meal, hold a five-minute conversation, read a menu — reported honestly instead of a streak counter.
Start with a single language pair and a single measurable milestone — "hold a real conversation in eight weeks" — and let the honest measurement be the differentiator the streak-driven incumbents cannot copy without abandoning their retention model.
Risks and honest caveats
- Outcomes are hard to measure objectively. Fluency is fuzzy, and self-reported milestones can drift into the same engagement theater you are trying to escape. The measurement has to be real enough to trust, which is a genuine product problem, not a footnote.
- Distribution is brutal. Duolingo and its peers own the category's mindshare and app-store gravity. A better methodology does not distribute itself; the wedge has to be narrow and the outcome claim has to be verifiable enough to earn word of mouth.
- The free LLM is both your engine and your competitor. If the value is only "talk to an AI," learners will do that for free. The durable moat is the curriculum, the review system, and the honest progress measurement around the model — not the model access itself.
- Willingness to pay is seasonal and shallow. Consumer language learning attracts New Year's-resolution spending that churns fast. Price and retention have to assume the motivation curve, not fight it.
How to validate this further
Read the EdTech and language-learning threads in the Pain Point Browser and pressure-test the outcome-measured framing with the Idea Validator. Related: language learning without energy paywalls.