This article outlines a 14-day validation process: first, talk to ten real users about their problems; then test willingness to pay with a landing page or manual service; finally, decide whether to develop.
01 Start by writing hypotheses as falsifiable statements
A good hypothesis includes three things: who, what problem they face, and what action you expect them to take.
Example: 'Independent e-commerce sellers spend 3 hours weekly organizing ad data; they would pay $29 monthly for an automated report.' This has a target, a pain point, and observable behavior, making it testable.
02 Spend 3 days finding 10 people to talk to
Finding people is harder than coding, but it determines validation quality. Common channels include:
- Communities and groups where your target audience spends time
- Comments from peers, especially those complaining about the same issue
- Your own network, asking who they know doing this
Keep each conversation under 20 minutes, focusing on when they last faced the problem and how they solved it. Avoid asking 'Would you use this?'
03 Use manual service instead of development
From days 4 to 10, deliver results manually, e.g., using spreadsheets and scripts instead of a system, manually compiling reports for clients.
Manual service serves two purposes: testing if clients will pay for the outcome and revealing real workflows to identify steps worth automating.
04 Use willingness to pay as the final judgment
'Interested' and 'willing to pay' are different conclusions. Set up a real payment step, like prepaid slots or early purchases.
By day 14, focus on three numbers: users who started, actively continued, and were willing to pay. If all are low, pivot instead of adding features.