What counts as real validation

Behavior is the only signal that counts. Everything else is noise.

Someone signing up is a weak signal. Someone paying is stronger. Someone paying, coming back, and telling someone else, that's validation. The bar is a change in behavior, not a change in opinion.

What does a behavioural signal look like in practice? Someone pays before the product exists. Someone switches from a tool they've used for years. Someone introduces you to three people with the same problem. Someone uses your product every day without being asked.

None of these require a finished product. All require you to be in front of real people with a real ask — not a survey, not a landing page, not a waitlist. A conversation with a specific question at the end of it.

The most common validation mistakes

Asking "would you use this?" — people say yes to avoid awkwardness. Ask "how are you solving this today?" instead. The answer tells you more than any hypothetical.

Building before testing. You don't need a product to validate demand. You need a problem worth solving and someone willing to pay to solve it. A deck, a prototype, or a manual version of the thing is enough to get a real signal.

Validating with the wrong people. Friends and family are not your market. Early adopters are a specific type of person — they feel the problem acutely enough to try an unfinished solution. Find those people first.

How much validation is enough

There's no universal number. But a useful rule: if you can't find ten people who feel this problem so deeply they'd pay to fix it right now, you either haven't found your customer yet or the problem isn't painful enough to build a business on.

Ten isn't a big number. If you can't find ten, that's the signal.