Designing for users who can't read yet: what 3-year-olds changed in our UX rules
Our users are mostly 3–6 years old, and many can't read. Some rules we ended up with:
Every tappable thing is at least 88 pt on iPad. For a small child, a missed tap doesn't feel like a miss, it feels like the app is broken.
A kid holds the tablet with one hand, so a thumb is often resting on the glass. That was the #1 cause of "taps don't work" and "the picture zooms by itself" for us, so the canvas now tracks one finger and ignores the resting one instead of treating two fingers as a gesture.
If an instruction is only written, it doesn't exist. Everything is spoken.
No red X, and nobody ever ends with zero stars. A wrong answer just gets another try.
For anyone designing for users very different from yourself (kids, older adults, people with disabilities, non-native speakers): which of your own assumptions did your users break first, and how did you find out?


Replies
For us it was language. We build Victra, a job search site for people in Armenia, and we assumed English would be fine because most listings are in English. Then we saw how many people search in Armenian or Russian and read English only partly. Showing them jobs in a language they can't work in just felt like noise to them.
So now the assistant takes into account which languages you actually know and doesn't push jobs you can't apply to. Your "if it's only written, it doesn't exist" rule hits close. Same idea, different users.
@gevorgarakelyann the "reads English only partly" group is the tricky one, there's no setting for that. do people tell you which languages they're ok with, or do you guess it from what they search?
@digi_dinos they tell us. Languages with a level are part of the profile, and the assistant uses that. Guessing from searches felt too risky: someone who searches in Armenian may still be fine with English at work, and the other way around.
@gevorgarakelyann makes sense, self-declared beats inferred when the same person flips by context. kids have the same thing tbh: a 4yo can hear one language at home and another at daycare, so the parent's language setting is only half the picture
@digi_dinos ha, same problem, just 30 years later 😄 Do kids get to switch the language themselves mid-game, or is it locked to what the parent set?
the resting thumb one is such a good catch, would never have guessed thats why taps dont work. how did you find it, watching kids play live or from session recordings?
@bacelj in person, sitting next to a 3yo. a recording would've just shown taps doing nothing. you only notice the other thumb parked on the edge of the glass when you're right there holding your breath lol. do you get to sit in with your users or is it mostly analytics?