Building an EdTech Product in 2026: What to Build First, the Rules and the Cost

In 2022 this post named five edtech trends: upskilling, AI personalisation, immersive learning, subscriptions and gamification. In 2026 most of them are what learners expect by default, and a few are expensive detours. For a new learning product, our advice is to pick one learner and one result they want, and build four things well: lessons, practice, visible progress and an admin panel your content team can run. If your users include children, US public schools or EU institutions, check three rules before you design anything.

The five 2022 trends, checked in 2026

2022 trend	What it means in 2026	What we would do
Upskilling and reskilling	Employers pay, so they want proof that people finished and learned	Build completion reports for the manager, not only screens for the learner
AI personalisation	Useful for practice and feedback. In the EU, high-risk rules for AI that scores exams in schools apply from 2 December 2027	Use AI for practice in the first version; a person reviews any grade that counts
Immersive learning (VR, AR, 3D)	Costly to make and to update	Use video first; build 3D only when it is the product
Subscription learning	Normal, but app store fees cut into the price	Decide early who pays: the learner, a parent, an employer or a school
Gamification	Points, levels and streaks are expected	Tie every point to a skill the learner practised

Three rules to check before you build

This is general information, not legal advice. Check the rules for your market with the regulator or a lawyer.

1. Children under 13 in the US (COPPA): the FTC finalised changes to the COPPA Rule in January 2025, with one year to comply after publication. The rule was published in April 2025, so the deadline passed in April 2026. Two changes matter for product design: a separate parental consent before sharing a child’s data with third parties, such as for targeted ads, and no keeping children’s data forever.
2. US public schools and universities (ADA Title II): the Justice Department’s web rule sets WCAG 2.1 Level AA as the standard for the websites and mobile apps of state and local governments, which includes public schools and universities. An interim rule published on 20 April 2026 moved the deadlines to 26 April 2027 for entities serving 50,000 people or more, and 26 April 2028 for smaller ones. If you sell to them, expect accessibility questions when they buy.
3. AI in EU education (the AI Act): the European Commission’s AI Act page lists AI used in education institutions, with exam scoring as an example, among the high-risk uses. After the AI Omnibus entered into force on 27 July 2026, the rules for these uses apply from 2 December 2027.

None of these stops a small first version. Each one is easier to design in than to add later.

What a first version should include

Product	Build first	Leave for later
Language or skill app for adults	Lessons, short practice, daily progress, a subscription	Social features, an AI tutor
Upskilling platform sold to companies	Courses, assignments, completion reports for managers	A marketplace of outside courses
Tool for schools	Teacher dashboard, class lists, accessible screens	A parent app
Live lessons with experts	Booking, paid video calls, schedules across time zones	Group classes, recordings library
Quiz platform for a creator’s audience	Quizzes, creator tools, subscriptions; a no-code start can work	Custom code, until the limits show

Whatever the type, build the admin panel early. If every new lesson needs a developer, content becomes your slowest and most expensive step.

What we have built

Six of our case studies are edtech products. Our edtech app development page lists them all. Three examples:

• Akler is a language learning app for iPhone with its own LMS. Admins add languages, words and translations from the admin panel, with no new app release. It went from idea to a launch-ready MVP in 4 months.
• QuizWithIt runs on Bubble, a no-code tool. It reached 5K+ users in its first 3 months and broke even in 4 months. Its case study also lists the limits it hit, such as a points system that slowed the platform down.
• Kanku is a karate learning app built in React Native. It packs a 3D anatomy model and 4,500+ pages into a 9 MB Android app and had 5,000+ downloads at launch.

What it costs

A first version of an edtech app typically costs $5,000 to $15,000 with us and takes 2 to 4 weeks. Bigger products take longer: Akler took 4 months, and Everytalent had a working MVP in 2 months.

Across agencies with a published price, the median MVP starting price is $13,234. AI can change that a lot: among agencies that price both, an AI MVP starts at a median 4.8 times a regular MVP at the same agency (15 agencies), according to the Rocketech MVP Price Index 2026, H2 edition. Our MVP development page shows that we build AI MVPs for the same price as any other MVP.

How we would estimate it

1. Our project estimate turns your brief into user stories, such as “a teacher sees which students finished the module”. You confirm what is in scope, and every hour in the budget traces back to a story you kept.
2. We check the hours the model suggests against what our team actually books on finished projects, then multiply by our published rate of $39-55 an hour.
3. The deeper version suggests customer groups, for example adult learners, parents or employers, and gives each a TAM, SAM and SOM with a note on what the number covers. Treat them as hypotheses to test, not verified demand.
4. For the groups you keep, it shows profit a month, lifetime value, how many customers pay for the build and the month it is paid back. You can change any starting number.

It takes about four minutes to estimate your learning app’s cost and timeline. The result is a planning range, not a fixed quote.

What we would do

1. Pick one learner and one result you can measure, such as “reads 500 new words a month”.
2. Decide who pays: the learner, a parent, an employer or a school. It changes the whole design.
3. If children under 13 might sign up, design parental consent and data deletion first.
4. Build the admin panel in the first version so content doesn’t wait for developers.
5. Write WCAG 2.1 AA checks into every screen if you plan to sell to public schools or universities.
6. Keep AI to practice and feedback at first. A person checks any grade that counts.
7. Track completion every week. It tells you more than sign-ups do.

FAQ

How much does it cost to build an edtech app in 2026?

A first version typically costs $5,000 to $15,000 with us and takes 2 to 4 weeks. Larger products take longer: Akler went from idea to a launch-ready MVP in 4 months.

Which rules apply to an edtech app for children?

In the US, the FTC’s COPPA Rule covers children under 13. Its 2025 changes require separate parental consent before sharing a child’s data with third parties and limit how long data can be kept. Check with a lawyer for your markets.

Can an edtech product start on a no-code tool?

Yes. QuizWithIt runs on Bubble and reached 5K+ users in its first 3 months. Plan for the limits: its points system slowed the platform down as it grew.

Is AI grading allowed in EU schools?

The AI Act treats AI used to score exams in education institutions as high-risk. Those rules apply from 2 December 2027, so plan for human review and documentation if you sell AI grading in the EU.