I built a learning app for my own kid. Here's what actually worked.

I work in edtech. I have twins. I have, over the last three years, downloaded approximately 47 "educational" apps for them. Four of them survived past week one.

The four that survived all had one thing in common, and I'll get to it.

What the research says about toddler apps

A landmark 2015 paper by Hirsh-Pasek, Zosh, Golinkoff and colleagues in *Psychological Science in the Public Interest* set out four criteria for an app to be genuinely educational for young children. It should be:

1. **Minds-on**, not just hands-on (engages thought, not just tapping) 2. **Engaging**, but not distracting (no flashy rewards that derail learning) 3. **Meaningful**, connected to what the child already knows 4. **Socially interactive**, ideally with a caregiver

Most apps in the "Kids" category of any app store fail on at least two of these.

The one thing the good ones had in common

They were designed to be done *with* an adult, not instead of one. The app was a prompt for a conversation, not a babysitter.

This matches the broader research: a 2020 review in *Acta Paediatrica* found that "joint media engagement" — parent and child using a screen together — produced learning outcomes comparable to live, in-person teaching. Solo screen use, for the same content, produced almost none.

What I tell other parents now

If a toddler app does not invite you to sit down with your child, it is, at best, entertainment. That's fine — sometimes you need entertainment. Just don't call it learning.

The best digital tool I've ever seen for a toddler is one that turns into a moment with a parent. Everything else is wallpaper.

— Aarav