How to validate a startup idea before you build it

Most ideas die not because they were bad, but because nobody checked whether anyone wanted them. Validation is the cheap insurance you buy before spending months building. Here is the same six-step process behind every idea in this library.

1. 01

   Find the existing pain, not the idea

   Start from a problem people already pay to avoid or work around. If they are using spreadsheets, manual labour, or a tool they complain about, the demand is proven before you build.

2. 02

   Talk to ten real buyers

   Have ten conversations with people in the target group before writing code. You are listening for the words they use, what they have already tried, and whether they would pay, not pitching.

3. 03

   Look for where money already moves

   Validated demand usually has a budget attached. Check job posts, competitor pricing, agency rates, and what adjacent tools charge. Silence on price is a red flag.

4. 04

   Build the smallest testable slice

   Ship the one workflow that delivers the core value and nothing else. A landing page with a waitlist, a manual "concierge" version, or a single working feature all count.

5. 05

   Charge before you scale

   Ask for money as early as is honest. A paid pilot, even a small one, is the only validation that survives contact with reality. Free interest is not demand.

6. 06

   Kill or commit on evidence

   Set a clear signal in advance (e.g. "5 paying pilots in 30 days"). If you hit it, commit. If you miss it, change the idea or the buyer, do not keep polishing.

Want ideas that already have the demand signal mapped? Browse the validated idea library →