The education onboarding teardown.
20 education app onboarding flows, every screen documented.
Enter your email to get the teardown.
You'll be subscribed. You'll get occasional growth notes in your inbox.
What you get
All 20 flows screen by screen, the 7 rules, and a recommended flow for each of the three flow segments.
What we ran the teardown with. Point them at your own vertical.
The screen-by-screen data the findings were read from.
What the flows do
Education onboarding is confirming a decision the user already made, not discovering one.
- 1.What did you come here to learn?
- 2.Where are you starting from?
- 3.How much will you realistically do?
Someone who downloads a language or coding app has already decided. Asking them to justify it spends goodwill on a question you know the answer to. The flow works when it takes the decision as given and spends its questions on sequencing the curriculum instead.
Start with what the user has already decided
By downloading a language or coding app they have shown intent. Asking why they are learning makes them justify a decision they already made. Babbel opens with language selection and treats it as a product decision rather than something to defend.
Do not ask proficiency unless a lesson changes
A proficiency question sets the expectation that the experience will adapt. If nothing downstream does, the user feels the broken promise on lesson one. Duolingo asks how much you already know, then personalizes accordingly.
Never ask for commitment before anything is invested
A streak pledge only works once there is something to protect. Mimo waits until near the end of its flow, after you have chosen a coding path, to set the daily reminder.
Put attribution last, or cut it
Asking how someone heard about you serves your analytics, not the user. Asked early, it spends goodwill before you have earned any. Babbel waits until the very end.
Hold the signup wall until the payoff is visible
Let the user pick topics and goals first, so the account feels like the key to something already built for them rather than a toll gate. Headway collects everything, shows the tailored plan, then asks you to sign up.
Keep questions near zero when the job is transactional
If someone opened the app to get one answer now, every question before that answer taxes the reason they came. Photomath asks nothing and goes straight to solving.
Do not split one decision across three screens
Padding the step count inflates the progress bar until the user notices the flow is busywork. Length should track real personalization and nothing else. Duolingo keeps each screen to a single real decision.
I've already done some of this work. I do it for my consulting business, I have all the teardowns, and I've done the work that other people don't have to do.
Want this run on your app?
The teardown is the category. An onboarding audit reads your flow, screen by screen.
See the onboarding audit