Founder guide · 7 minute read
How to validate a business idea before you build
Idea validation is not proving your idea will succeed. It is replacing your most dangerous guesses with enough real-world evidence to decide what to test, change, or stop.
The short answer
How do I validate a startup idea?
Define a narrow customer and painful problem, list the assumptions that must be true, test the riskiest assumption with real buyers, ask about past behavior instead of future opinions, and use a pre-set decision rule to continue, revise, or stop.
A practical five-step process
Define one customer and one painful problem
Describe a specific person, the situation they are in, and the cost of leaving the problem unsolved. Avoid broad labels such as ‘small businesses.’
List the assumptions that must be true
Write down what you believe about urgency, buyer access, willingness to pay, market size, and your ability to deliver the outcome.
Test the riskiest assumption first
Choose the assumption that would kill the idea if false. Run the cheapest credible test—usually interviews, an offer, or a price conversation.
Ask about behavior, not opinions
Discuss the last time the problem happened, what the customer tried, what it cost, and who approved the spend. Past action is stronger evidence than praise.
Set a decision rule before the test
Decide what evidence means continue, revise, or stop. A clear threshold prevents you from explaining away weak results after the fact.
What counts as strong validation evidence?
Strong evidence requires meaningful action: customers paying for an alternative, agreeing to a priced pilot, introducing you to the buyer, pre-ordering, or repeatedly trying to solve the problem themselves. Survey interest, compliments, and hypothetical willingness to pay are useful clues, but they are not demand.
When should you start building?
Build the smallest version only after you can identify a specific buyer, show that the problem matters, reach enough people like them, and get credible evidence that the outcome is worth paying for. Even then, build only enough to test the next uncertainty.