I was proud of the first version of my app's UI. It made sense to me. The layout felt logical. The colors looked good on my phone. I was ready to submit.
Then I watched three people use it for the first time.
None of them found the core feature. They tapped on things I had never expected anyone to tap on. They missed buttons that felt obvious to me. One person spent two minutes looking for a feature that was right in front of them the whole time.
This is one of the most important things you will experience as someone building your first app. The UI that makes sense to you — the person who built it, who knows what every button does — is completely different from the UI that makes sense to a first-time user.
I took notes on everything I saw and asked Claude to help me redesign the three screens that caused the most confusion. I gave it specific observations: "Users expect this button to be at the bottom of the screen, not the top" and "Nobody noticed the navigation icons without labels." I did not ask for a full redesign. I asked for surgical changes.
The second version took about two days to build. Not because the changes were large — they were not. But because I wanted to be careful. For each change, I committed to GitHub before I started, asked for one change at a time, and tested each one before moving to the next.
The third person I tested the revised version with found the core feature in under thirty seconds. That is the difference between a UI built for yourself and a UI built for a user you have actually watched.
I now test with real people before every submission. Even if it is just two or three people. It is the most valuable thing I do.