How to Run Customer Interviews That Tell You the Truth

Because founders ask people to predict the future. “Would you use this?” “Would you pay for that?” People are terrible at predicting their own behavior, and they're kind enough to guess in your favor.

Founders also pitch instead of listen. Once you describe your idea, the conversation becomes about being polite to you. You stop learning about their problem and start collecting compliments.

The result is a stack of encouraging notes that don't match reality. The fix is simple but uncomfortable: stop talking about your idea at all.

Ask about the last time the problem happened. What were they doing? What did they try? How long did it take? What did it cost? Specific stories reveal how painful the problem really is.

Ask what they currently use and what they dislike about it. If they've cobbled together a workaround, the pain is real. If they haven't done anything, the pain may be smaller than you think.

Ask what they've spent. Money, hours, hiring help. Past spending is the most reliable predictor of future spending. People who have paid before will likely pay again for something better.

Go where your target customer already spends time. Online communities, industry groups, local events, LinkedIn. Look for people actively discussing the problem you want to solve.

Reach out with a short, honest message. Explain that you're researching a problem, not selling, and ask for fifteen minutes. Be specific about why you chose them.

Avoid friends and family unless they truly fit your customer profile. Even then, their desire to support you can color their answers. Strangers give cleaner data.

Talk less than a third of the time. Ask open questions and let silence do the work. People often share the most important detail after a pause.

Don't correct or defend. If someone says they don't care about a problem you think is huge, don't argue. Ask why. Their reasoning is the data you came for.

Follow the emotion. When someone's voice changes, when they sigh or get animated, dig deeper. That's where the real pain lives.

After each conversation, write down the key facts: the problem, what they've tried, what it cost, and any exact phrases that stood out. Do it immediately, before memory smooths it over.

After five to ten interviews, look for patterns. Do most people describe the same pain? Do they spend money on it? Do they use similar words? Consistent patterns are a strong signal.

If the patterns are weak or scattered, your customer definition may be too broad, or the problem may not be as painful as you assumed. Both are worth knowing before you build.

At the very end, if at all. Once you've learned about their problem, you can briefly describe what you're considering and watch their reaction. But don't treat enthusiasm as a yes.

Instead, make a small ask. Would they join a waitlist with the price listed? Would they pay a deposit? Would they introduce you to someone else with the problem? Commitments mean something. Compliments don't.

If they decline every ask, that's useful too. Thank them warmly and take the lesson seriously.