Build the app that carries your language school into the six days between lessons

The real work starts once enrolment season ends: of ten students who begin at A1, how many reach B1? Absences begin in the third week, and when a student paying in instalments drops out, both class sizes and collections suffer. Vocabulary review, pronunciation practice and homework tracking fill the six days between lessons. An app works because it takes the lesson outside the classroom and makes retention measurable.

How many separate apps does this need?

  • 01

    Student App (iOS + Android)

    Class schedule, live lesson access, offline listening, flashcard review, pronunciation recording, homework submission, attendance and course-fee tracking in one place.

  • 02

    Teacher App

    Take attendance, grade homework, listen to a student's recording and send back a spoken correction, share lesson notes, and flag struggling students early.

  • 03

    Course Management Panel (Web)

    Open levels and classes, place students by placement-test results, manage the content library, handle instalment collection and invoicing, and report on attendance and renewals.

What the app includes

  • Automatic class placement from level-test results
  • Teacher attaches a spoken correction to the learner's recording
  • Phoneme-level pronunciation feedback with instant retry
  • Missed words drop automatically into the review deck
  • Offline listening pack whose access expires when the term ends
  • Lesson recording released only to that class for a limited window
  • Learners crossing the absence threshold flagged to the teacher
  • Overdue instalment alerts with a payment link for term fees
  • Trial-lesson request that flows straight into the placement test
  • Peer-pairing speaking assignment for practice between lessons
  • Input support for umlauts, accents, Cyrillic and right-to-left script
  • Parent view for attendance, homework and payments, with separate messaging consent

State of the sector

Competition no longer comes from the school down the street. Students have global language apps on their phones that they open every day on a free tier; four hours a week are spent in your classroom and all the rest inside someone else's product. Online tutoring platforms, meanwhile, are undercutting prices on one-to-one speaking lessons. A school without an app has no contact with students between lessons and cannot see who is studying and who is drifting away. When renewal season arrives, all it has is the attendance register. For a local school, the differentiator is being able to deliver its own teacher's spoken correction straight to the student's pocket.

Metrics that matter here

  • Term renewal rate — Students who practise between lessons enrol in the next term; the app makes that contact measurable.
  • Active days between lessons — Students who review vocabulary and practise pronunciation on the days between their two weekly lessons move up a level; the panel shows this class by class.
  • Absences in the first three weeks — Dropping out starts early; attendance and homework data show the teacher which students are at risk before they are lost.
  • Trial lesson to paid enrolment conversion — Follow-up from the app after a placement test and trial lesson speeds up the enrolment decision and can be measured by source.

Common mistakes in this sector

  • Dumping every PDF and video into the app Hundreds of files uploaded without a content library management screen become a mess in no time. Students cannot find what they are looking for and teachers cannot update anything. The right approach is to tie content to a term, unit and skill, and make it manageable from the panel.
  • Setting the pronunciation threshold too harshly An app that shows the raw score gives a student low marks over and over on day one, and it is never opened again. The threshold should be softened according to level, and feedback should say which sound needs fixing.
  • Trying to sell digital packages through your own payment page Routing a lesson package consumed inside the app to an out-of-store payment ends in rejection and costs weeks. Physical course enrolment and digital content sales must be separated from the very beginning.
  • Sending everyone the same bulk notification at the same time The same reminder going to every student at nine in the morning gets notifications switched off. Notifications should be built around lesson days, absences and each student's study hours, reminding them where they left off rather than pushing a streak.

Regulation and compliance

Language schools in Turkey fall under MEB (Turkish Ministry of National Education) private-education regulations: the student contract, lesson hours and refund terms must match the purchase flow in the app. For digital lesson packages, the distance-selling agreement and the exemption from the right of withdrawal are shown explicitly. Lesson and pronunciation recordings are personal data; a retention period, an explicit-consent text and a privacy-notice screen are required under KVKK (Turkey's data protection law, aligned with GDPR), and parental approval is collected for students under 18. Course fee invoices are issued through GİB (Turkish Revenue Administration) e-Arşiv, part of Turkey's mandatory e-invoicing system.

The real challenges in this sector

  • Setting a pronunciation-scoring threshold that fits Turkish speakers

    Off-the-shelf speech recognition services are calibrated for native English speakers; “th”, the open “æ” and final consonant clusters all produce low scores for Turkish speakers. Showing the raw score as-is loses the student in week one. We build a layer that takes phoneme-level scores, applies a threshold based on the learner's level, and instead of saying “wrong” tells them which sound to fix and how; the recording is trimmed on the device and feedback is kept under two seconds.

  • Balancing offline listening against content protection

    Students do their listening practice on the bus or when their data runs out; if the audio cannot be downloaded, the feature simply goes unused. Yet those audio files are the school's core asset. Files are stored encrypted on the device, the playback key is tied to the session, and downloaded content expires automatically at the end of the term. For video lessons, the platform's own protection layer (FairPlay/Widevine) is enabled against screen recording.

  • Keeping spaced-repetition data consistent across devices

    When flashcard progress is updated offline on a phone and online on a tablet, it becomes unclear which review date is valid; the crude fix wipes out a student's week of work. Each review is written as a separate event with its own timestamp, and the card's state is recomputed from those events. So even if the same card is studied on two devices, no data is lost, and the teacher panel shows the real study history.

  • Keeping live lessons alive on a mobile connection

    In one-to-one speaking lessons the student is usually on mobile data, often on the move; if audio is not prioritised, speech drops out and the lesson is wasted. The audio channel is kept separate from video, video switches off automatically when bandwidth falls, and the audio session continues when the app goes to the background. In-lesson chat and the sentences the teacher corrects land in the student's notebook the moment the lesson ends.

  • Handling input and answer matching across different alphabets

    German umlauts, French accents, Russian Cyrillic, right-to-left Arabic and Japanese kana input can all live in the same app. Keyboard direction and text alignment change by language. If answer checking rejects unaccented spelling without a warning, a correct answer is marked wrong. When Arabic is selected, not just the text but the progress bar and button layout flip direction along with it.

Required integrations

  • Instalment-based course fee collection and recurring payments with iyzico or PayTR (Turkish payment providers)
  • App Store / Google Play in-app purchase (the mandatory flow for digital lesson packages)
  • GİB (Turkish Revenue Administration) e-Arşiv invoicing integration for student invoices via Paraşüt or Logo
  • Speech recognition and pronunciation scoring service (Azure Speech / Google Speech)
  • Live lesson infrastructure: Zoom API or a WebRTC-based service (Agora, Daily, 100ms)
  • Lesson reminders, absence alerts and payment notices by SMS via Netgsm or İletimerkezi (Turkish SMS providers)

Who this page is for

  • Single-branch regional language school

    The priority is contact between lessons and term renewal. A core first release with timetable, attendance, homework and instalment tracking is enough; pronunciation and live lessons come in a later phase.

  • Newly founded online one-to-one tutoring team

    Not class management but lesson booking, tutor calendars, post-lesson notes and payment flow. Live sessions are bridged to an existing service at first, and digital packages are structured around app-store rules.

  • School working with children and teenagers

    Parent accounts, guardian consent and separated permissions belong in the data model from day one. Attendance, homework and payment sit with the parent, while recordings and teacher messaging need their own consent, retention period and automatic deletion.

A phased plan that splits the budget

  1. 1

    Phase 1 — Enrolment and lesson core for one branch 5-7 hafta

    Learner app plus web panel: opening terms and classes, timetable, attendance, homework submission, announcements and instalment tracking. It stands on its own and replaces the paper register. This comes first because practice modules have nothing to attach to without class, term and learner data.

  2. 2

    Phase 2 — Between-lesson practice engine 4-6 hafta

    Flashcards with spaced repetition, offline listening packs, pronunciation recording and the teacher's spoken corrections. The teacher app arrives here. It follows Phase 1 because review schedules can only be personalised once each learner's term and level are known.

  3. 3

    Phase 3 — Live lessons, parent access and digital package sales 4-6 hafta

    Live lessons brought inside the app, parent accounts with separate consents, digital lesson packages sold through in-app purchase, plus renewal and absence reporting. It comes last because it is both the costliest part and the one most easily built wrong without real usage data.

Typical scope and timeline

A typical first release: student and teacher apps plus a web management panel. Placement testing and class assignment, class schedule and attendance, homework submission and spoken feedback, flashcard review, offline listening, live lesson connection, instalment collection and the e-Arşiv invoicing flow. Pronunciation scoring and the parent screen are usually added in a second phase.

Estimated timeline: 12-18 weeks

Off-the-shelf or custom build?

Topic Off-the-shelf Custom build
Time to start and cost An off-the-shelf LMS or classroom platform can be running the same week, and for a small school it genuinely covers sharing material and collecting homework. Custom development takes weeks to reach a working version, but terms, instalments and attendance are shaped around how the school actually runs.
Fit to the course workflow Content, quiz and homework modules are mature, but placement testing, class assignment, term renewal and absence thresholds usually sit outside the product. One chain runs from placement test to class, attendance and instalment, so a teacher sees an at-risk learner before the term ends.
Ownership of content and learner data Your audio and video library lives in the platform's storage; export is usually limited to downloading files, while review history and pronunciation recordings cannot move with you. Content, review history and recordings sit in your own database, and retention or deletion rules follow the policy you define under Turkish data-protection law (KVKK).
Monthly subscription versus one-off investment A per-learner monthly fee is cheap at low volumes and needs no maintenance effort, though it keeps billing outside the busy season too. A one-off build leaves only hosting, store accounts and maintenance behind, so unit cost falls as learner numbers grow.

Sector glossary

CEFR level (A1–C2)
The shared scale that course plans, class names and price lists are built on. In the app, content, homework and tests are tied to these levels; “moving up a term” means stepping one rung higher on this scale.
Placement test
A short pre-enrolment assessment used to put a learner in the right class. Wrong placement is the most common cause of dropping out in the first three weeks, so results are read alongside class capacity.
Spaced repetition (SRS)
Showing a word again just before it would be forgotten. In a school setting it means the system, not the learner, decides the daily card load, and weak words are pulled forward before an exam.
Phoneme-level score
Scoring individual sounds rather than giving one mark for the whole word. This is what lets the app tell a learner which sound to fix and how, instead of simply saying “wrong”.
Term renewal (re-enrolment)
A learner who finishes one term signing up for the next. Most of a school's revenue comes from here rather than from new sign-ups, and attendance, homework and practice data set the ground for that conversation.

Frequently asked questions

Let us talk about your project

A free 30-minute call to scope it out, then an itemised quote within 48 hours.

Get a Free Quote

This page explains the problems an app must solve in this sector. We share references and comparable projects during the call.

Other sectors

All sectors →