Too many products are built based on assumptions. The team thinks they know what works, but they've never actually tested it. Then they build the full thing, launch it, and discover that people don't use it the way they expected.

Testing and validation should happen early and often. You don't need a finished product to test. You need a prototype that's good enough to show what you're thinking, and real people to try it.

Start with the riskiest assumption

Every product design has assumptions. Some are safe bets. Others could break everything if they're wrong. Start by testing the riskiest assumption first.

If you're building a tool that claims to save time, test whether people actually save time using it. If you're building something that's supposed to be easier than the alternative, test whether people find it easier.

Don't test everything at once. Pick the one thing that, if it's wrong, makes the whole product pointless. Test that first.

Build a testable prototype

You don't need a fully functional product to test. You need something that's good enough to show the core idea and let people interact with it.

This could be a clickable prototype in Figma. It could be a simple HTML page with fake data. It could be a working prototype built in a few days. The key is that it's realistic enough that people can actually try to use it for the thing you're testing.

If you're testing whether people understand the value, a simple landing page might be enough. If you're testing whether a flow works, you need something they can click through. If you're testing whether something feels fast, you need something that actually responds.

Test with the right people

Testing with the wrong people gives you the wrong answers. If you're building for finance teams, test with finance teams. If you're building for solo consultants, test with solo consultants.

You don't need hundreds of people. You need a handful of the right people. Five to ten people who match your target customer is usually enough to spot major problems.

Find people who have the problem you're solving. They're the ones who can tell you if your solution actually helps.

Watch what they do, not just what they say

People say things they think you want to hear. They're polite. They don't want to hurt your feelings. But their actions tell the truth.

Watch where they hesitate. Watch where they get confused. Watch where they try to do something and can't figure out how. Watch where they give up.

These are the real problems. The things people do (or don't do) are more valuable than the things they say.

Ask the right questions

Don't ask "Do you like this?" That's a useless question. Ask "What would you do with this?" or "How would this fit into your day?" or "What would make you stop using this?"

Ask about specific moments. "When you first saw this, what did you think it was?" "When you clicked here, what did you expect to happen?" "If you were going to use this tomorrow, what would you do first?"

These questions reveal whether people understand what you built and whether it solves their actual problem.

Test one thing at a time

Don't try to test everything in one session. Focus. Test whether people understand the value proposition. Or test whether a specific flow works. Or test whether the pricing makes sense.

One focused test gives you clear answers. A test that tries to cover everything gives you confusing results.

Iterate based on what you learn

Testing is useless if you don't change anything based on what you learn. If people don't understand something, fix it. If a flow doesn't work, change it. If people get confused, simplify it.

Then test again. See if the fix worked. Keep iterating until people can actually use what you built to solve their problem.

Simple test checklist

Before you test, make sure you know:

  • What specific thing are you testing?
  • What would success look like?
  • What would failure look like?
  • Who are you testing with?
  • What will you do with what you learn?

If you can't answer these, you're not ready to test. Figure out what you're actually trying to learn, then test that.

The goal is confidence, not perfection

You'll never be 100% sure something will work. But you can get confident enough to build it. Testing helps you get that confidence.

Test early. Test often. Test with real people. Then build based on what you learned, not what you assumed.