Building
How to validate an idea
Validation is not people telling you it’s a good idea. It’s you trying to kill the thing and failing.
-
Ask about the past, not the future.
“Would you use this?” measures politeness. “Tell me the last time you had this problem” measures reality. People predict their own future behaviour badly and describe their past accurately, which is the whole insight behind The Mom Test.
-
Count the workarounds.
A real problem already has a bad solution attached to it, a spreadsheet, a group chat, a note on someone’s phone. If nobody has bothered to hack around it, it probably doesn’t hurt enough to pay for.
-
Build the smallest thing that can fail.
Dropbox’s first product was a three-minute video. It went up on Digg, the waitlist went from 5,000 to 75,000 overnight, and the sync engine still barely existed. The demo tested the demand, not the code.
-
Ten conversations that don’t change the idea were ten bad conversations.
If nothing moved after talking to ten people, you asked leading questions and they were kind to you. Go back and ask worse-mannered ones.
-
Solve a problem you have.
You are the only user you can interview at midnight, for free, honestly. Building for a hypothetical person is how six months disappear into a product nobody asked for.
-
Borrow a habit instead of create one.
New habits are enormously expensive. Existing ones are free. A daily maths puzzle works because the Wordle-shaped slot in someone’s morning already exists. Pi-Squared moved into it not digging a new one.
-
Ship it to twelve people before you ship it to everyone.
Not a launch. A test with names attached, where you can ring each person and ask why they stopped.
-
Be suspicious of enthusiasm.
Friends will lie kindly and strangers will lie politely. The only unambiguous signals are money, repeated use, and someone getting annoyed when it breaks.
-
Give it a deadline, not a budget.
Braindex went from nothing to App Store submission in seventeen days. A deadline that short forces the idea to prove it’s simple enough to exist at all.
-
Kill it early, cheaply, and without ceremony.
MovieToons taught me more about validating an idea than anything else I have built. A wrong idea is only expensive if you keep paying for it after you know.