You describe your idea, they nod, and they say it sounds interesting. You leave feeling like you're onto something. That's politeness, and it tells you little. The conversations that validate a business idea sound different. You spend the whole time learning about the other person's problem, and you never mention your idea. Ten of them will show you whether the problem is worth building around.
These conversations are called customer discovery interviews. Here's how to find the right people, what to ask, and what to do with what you hear.
Who should you interview to validate a business idea?
Interview people who have the problem you want to solve right now, or who had it recently. Someone who dealt with it last week remembers the details. Someone who faced something similar two years ago will give you a fuzzy version.
Friends, family, and supportive LinkedIn connections are the hardest people to learn from. They like you, so they'll tell you what you want to hear. Look for people who have the problem and no stake in your success.
How do you ask a stranger for a customer discovery interview?
Keep the ask short and make it about them. Most people will give you twenty minutes to talk about a problem they've lived with. Far fewer will sit through a pitch for someone's business idea.
Here's a message you can adapt:
> "I'm researching [the problem area], and I'd love to hear about your experience with it. It would take about twenty minutes, and I won't be pitching anything. Would you be open to a quick call?"
Leave out what you're building. If someone asks, be honest: "I'm exploring whether there's an opportunity here. I'm early, and right now I'm trying to understand the problem." That answer tends to make people more willing to talk.
What questions should you ask in a customer discovery interview?
Three questions, asked in roughly this order, get you to the truth. Each one keeps the focus on what the person has experienced, which is where honest answers come from.
1. "Tell me about your experience with this problem." Let them describe it in their own words. Pay attention to the language they use. If they describe it the way you would, that's useful. If they describe it differently, that's more useful. Don't correct them or steer them toward your version.
2. "What have you done about it so far?" Their past attempts show you how serious the problem is. Someone who has paid for three solutions and still has the problem is ready to act. Someone who plans to look into it someday isn't there yet. You'll also learn who your competition is without asking.
3. "Have you ever paid for something to fix it?" This is the most important question, and the one founders are most nervous to ask. People will say they'd pay for something and never follow through. What they've already paid for is a record of what mattered enough to act on.
If they've never paid for a fix, ask, "What stopped you from looking for something?" The answer is often worth the whole conversation.
This approach matches the core idea in Rob Fitzpatrick's book The Mom Test. Ask about people's past behavior, and even someone who loves you can't give you a polite, useless answer.
What should you avoid in a customer discovery interview?
Pitching. The moment you describe your solution, the conversation changes. The other person starts being polite about your idea and filtering what they say to protect your feelings. You lose the honest version of their thinking.
Hypothetical questions. "Would you use something that..." and "Would you pay for..." get hypothetical answers. People picture their best selves, the version who would use the app every day, and answer from there. Keep your questions about their past and present.
Leading questions. If you have a theory about the problem, keep it out of how you ask. "I imagine that's really frustrating" pushes them toward agreeing. "What has that been like for you?" lets them tell you.
What signals should you listen for in a customer interview?
Some of the most useful information comes from how people answer.
Specific answers mean the problem is urgent. People with a painful problem have stories ready. They remember moments and numbers: the hours lost, the money spent, the time they almost quit over it. "Yeah, it can be kind of annoying" is what a mild annoyance sounds like.
Pauses can be a good sign. A pause often means someone is thinking it through. What comes after a pause tends to be more honest than a quick reply.
Early enthusiasm is weak evidence. Polite people get excited about ideas. Listen for the moment someone gets specific. "This sounds great" turns into "This would have helped when I dealt with that last year." I've watched founders miss that shift. They were so relieved to hear enthusiasm that they stopped listening to what came next.
What do you do after your customer interviews?
Look for patterns across all ten conversations. Which problem descriptions kept coming up? What had people tried? Who had paid for a fix, and what did they pay for? Where did the people with a painful version of the problem differ from the people who mentioned it in passing?
Ten honest conversations tend to point you toward one of three outcomes:
- Confirmation. The problem is painful, people have spent money trying to fix it, and the gap you saw is there. Keep going.
- Recalibration. The problem exists, but it's milder than you thought or affects a narrower group. Your idea needs more definition.
- Redirection. What you heard doesn't match the problem you set out to solve. This outcome saves you the most, because you learned it before you built anything.
Hold what you heard up against your assumptions. Where they match, you have something to build on. Where they don't, you know what to rework. Either way, you're working from what people told you, and that's a much better place to decide from.
Once your interviews are done, run your idea through the Founder Filter and see how it holds up against what you heard.
Frequently asked questions
Ten is a good target. It's enough to see patterns without turning into a full-time research project. By conversation six or seven, you'll start hearing the same things, and that repetition is the signal.
