Edtech isn't one category, it's two, and they need different surveys. A consumer language app or test-prep tool sells directly to the person who uses it: the learner decides to pay, decides to keep going, and would know exactly how they'd feel losing access. A school-sold platform is the opposite: a district or school buys it, a teacher is handed it as a requirement, and the students who spend the most time in it never signed up for anything. Treat both the same way and you'll survey the wrong crowd in at least one of them.
Is the PMF benchmark different for edtech? No, it's still 40%. What changes is who counts as engaged, and it depends entirely on whether your product is consumer-bought or school-bought.
Consumer and self-directed edtech: survey the learner
When the learner chose the product themselves, a language app, a coding bootcamp, a test-prep subscription, they are both the user and the buyer, so the standard Sean Ellis approach applies directly. An engaged user is someone who has completed a lesson, session, or module recently, not someone who signed up and never returned. "Very disappointed" here usually means the product replaced a habit, a daily streak, a study routine, not just that it was fun to try once.
School-sold edtech: survey the teacher and the student separately
When a school or district is the buyer, you have two distinct engaged-user populations, and their experiences can diverge sharply. A teacher might love a platform for cutting grading time in half while their students find the interface confusing or the content mismatched to the lesson. Run the Sean Ellis survey on each group on its own, the same way a marketplace scores supply and demand separately, rather than blending an administrator's satisfaction with a teacher's and calling it one number.
Engagement means actual use, not a seat license
A school buying a site license for every student in a grade doesn't mean every student is an engaged user, most site licenses go mostly unused by design, teachers assign what fits their curriculum, and plenty of licensed seats never open the product. Define engagement the same way you would anywhere else: someone who has actually used the core product, a completed lesson, an assignment submitted through it, a session logged, in the recent past. Surveying every licensed seat will bury your real signal under students who were never meaningfully exposed to the product.
Time it to the school calendar
Usage in school-sold edtech follows the academic year, not the calendar quarter. A survey sent during summer break, winter break, or exam week will catch a naturally quiet period and read as disengagement that has nothing to do with product-market fit. Run the survey mid-semester, on users who've had a few consecutive weeks of normal access, and note the academic-calendar context if you're comparing scores across surveys run at different points in the year.
How to measure PMF in edtech
- Identify your model first. Consumer-bought means one user population; school-bought means at least two.
- Define engagement as real use. A completed lesson or submitted assignment, not a licensed seat.
- Score teacher and student separately when both exist. Don't blend an administrator's or teacher's view with a student's.
- Survey mid-semester. Avoid breaks and exam periods, which distort engagement for reasons unrelated to fit.
The 40% line doesn't change for edtech. What changes is recognizing which of the two edtech businesses you're actually running, and surveying accordingly.
Measure your fit, find your ICP, track the trend
PMFtracker runs the Sean Ellis survey on your engaged users, scores you against the 40% benchmark, surfaces your ICP from the open-ended answers, and tracks the trend over time.
Start measuring free → Set up in 5 minutes · No credit card required
