Research Methods for EdTech Founders: A Practical Guide

EdTech founders don't need a PhD to do credible research. They need three lean methods used well: structured interviews to discover problems, surveys to size them, and small tests to check that the product actually helps. This guide covers how to use each — and the ethics that apply when your users are learners.

Start with a hypothesis, not a method

The most common research mistake founders make is starting with a tool ("let's send a survey") instead of a question. Write down the assumption your product depends on — "teachers will use this weekly", "students improve faster with feature X" — and pick the cheapest method that could prove it wrong. A research plan is just a list of your riskiest assumptions, each paired with a test.

Method 1: Structured interviews — for discovering problems

Interviews are the fastest way to understand how learners, teachers, and buyers actually experience the problem you're solving. To make them count:

• Ask about past behaviour ("walk me through the last time…"), not predictions ("would you use…")
• Use the same core questions across interviews so answers are comparable
• Interview each stakeholder group separately — in education, the buyer (school leader), deployer (teacher), and user (student) see different problems
• Record and summarise each interview the same day, tagging recurring themes

Method 2: Lean surveys — for sizing what you found

Once interviews surface a pattern, a short survey tells you how widespread it is. Keep surveys under ten questions, lead with the behavioural questions (not demographics), and pilot the survey on two or three people first to catch confusing wording. A survey that converts into answers is one where every question maps to a decision you'll make differently depending on the result.

Method 3: Small structured tests — for checking the product works

Usability tests, A/B comparisons, and short classroom pilots test whether the product delivers. The discipline that separates evidence from anecdote is deciding before the test what outcome you'll measure and what result would count as failure. For pilots with schools, see the dedicated guide to running EdTech pilots that generate real evidence.

Research ethics: non-negotiable in education

Research with learners — especially children — carries obligations that consumer startups can ignore but EdTech startups cannot:

• Informed consent — participants (and parents/guardians for minors) must understand what data you collect and why
• Data minimisation — collect only what your research question needs, and store it securely
• Transparency with schools — teachers and school leaders should see findings, including unflattering ones
• Fairness — check that your product works across learner groups, not just the average (this matters doubly for AI products; see responsible AI in education)

Handled well, ethics is a commercial asset: schools trust — and buy from — companies that treat their students' data with visible care.

How UCL EdTech Labs teaches this

Research methods are Stage 2 of the accelerator — founders learn practical methods without the jargon and build lean surveys, interviews, and tests that convert into answers, leaving with a research library and a live research plan. See how the accelerator programme works and the broader evidence-based methodology.