List what must be true for your idea to work. Choose a risky assumption you can investigate cheaply, then design a test that might change your decision. Decide in advance what observation would encourage you, worry you, or leave you uncertain. A small experiment should reduce uncertainty; being easy to build is not enough.
Test the assumption that could stop the idea
The smallest useful experiment is the one that answers an important question.
A small example
Before building a supplier portal, a founder manually prepares a sample comparison for two buyers. She watches whether it helps them make a real purchasing decision, rather than only asking if they like the design.
Keep in mind
Small samples can reveal problems but may not estimate market demand reliably. Match the conclusion to what the test can actually show.
A note to yourself
Create an account or sign in to keep private notes and saved takas across devices.