Your prototype is not the whole product. It is the single moment your user gets value — what changes for them in the first 30 seconds. Build that loop. Skip everything else. No navigation. No settings. No profile pages. Not yet.
If a stranger can complete the core task without you whispering hints over their shoulder, you are done.
Pull the one-sentence V1 from waypoint 14. The verb in that sentence is the action you are prototyping right now.
Map the smallest possible path:
Three or four screens, max. If you are no-code, build it live in Bubble or FlutterFlow. If you are designing first, make it clickable in Figma. If you are coding, hardcode the data — do not build a database for a demo.
Build it. Click through it yourself end to end. Fix anything that throws an error or stalls.
Then run the silent test:
- Hand your phone or laptop to a friend.
- Say "try this" and stop talking.
- Watch where they pause, tap the wrong thing, or ask a question.
The pauses are your bugs. Fix the worst one. Repeat with another friend if you can.
Bad: A polished sign-up screen, a settings page, and no working core action.
Good: An ugly button that, when tapped, actually performs the thing your product promises.
Bad: "It works if I explain it first." That is not working.
Good: A friend completes the task on the first try, in silence.
If your friend gets stuck, the fix is almost always to remove a step, not to add a tooltip.
- Building the whole app at once. One screen, one flow, one button.
- Hiding the core action behind sign-up before the user knows what they are signing up for.
- Adding a tutorial because the flow is confusing. Simplify the flow instead.
- Testing only on yourself. You wrote it — of course you understand it.
Open your build tool. Construct three to four screens that take a user from open to first win. Test it silently on at least one friend. Fix the worst friction point.
A working prototype of the core interaction. At least one stranger or friend has completed it without help, in under two minutes, and confirmed they understood what just happened.
Your prototype is not the whole product. It is the single moment your user gets value — what changes for them in the first 30 seconds. Build that loop. Skip everything else. No navigation. No settings. No profile pages. Not yet.
If a stranger can complete the core task without you whispering hints over their shoulder, you are done.
Pull the one-sentence V1 from waypoint 14. The verb in that sentence is the action you are prototyping right now.
Map the smallest possible path:
Three or four screens, max. If you are no-code, build it live in Bubble or FlutterFlow. If you are designing first, make it clickable in Figma. If you are coding, hardcode the data — do not build a database for a demo.
Build it. Click through it yourself end to end. Fix anything that throws an error or stalls.
Then run the silent test:
- Hand your phone or laptop to a friend.
- Say "try this" and stop talking.
- Watch where they pause, tap the wrong thing, or ask a question.
The pauses are your bugs. Fix the worst one. Repeat with another friend if you can.
Bad: A polished sign-up screen, a settings page, and no working core action.
Good: An ugly button that, when tapped, actually performs the thing your product promises.
Bad: "It works if I explain it first." That is not working.
Good: A friend completes the task on the first try, in silence.
If your friend gets stuck, the fix is almost always to remove a step, not to add a tooltip.
- Building the whole app at once. One screen, one flow, one button.
- Hiding the core action behind sign-up before the user knows what they are signing up for.
- Adding a tutorial because the flow is confusing. Simplify the flow instead.
- Testing only on yourself. You wrote it — of course you understand it.
Open your build tool. Construct three to four screens that take a user from open to first win. Test it silently on at least one friend. Fix the worst friction point.
A working prototype of the core interaction. At least one stranger or friend has completed it without help, in under two minutes, and confirmed they understood what just happened.