7 Mistakes Founders Make in Customer Interviews

Most founders who fail to find product-market fit do not fail because their idea was bad. They fail because the customer interviews they relied on quietly told them the wrong thing. Having shadowed dozens of founders running their own interviews, the same handful of mistakes come up over and over, and each one sends you off in a confident, expensive, wrong direction.

The good news is that these mistakes are easy to recognise once you know them. Here are the seven that do the most damage, why each one costs you, and how to fix it.

1. Getting overexcited by positive feedback

A participant says your idea sounds "interesting." You nod happily and move on, quietly filing it as a win. This is the most common and most expensive mistake of all. Interest is not a need. People are interested in all sorts of things they will never pay for or use. Treating a warm reaction as validation is how founders convince themselves they have demand when they have only politeness.

Fix: when someone shows interest, dig deeper. Ask what they do about this problem today and what it costs them. Real demand shows up in behaviour, not enthusiasm.

2. Trying to sell the product

The moment you start pitching, the interview is over, even if the conversation keeps going. Selling instead of listening teaches you nothing, because the person stops telling you about their world and starts reacting to yours. Your goal in an interview is insight, not a demo.

Fix: keep your idea out of the room for as long as possible. Let them describe how they handle the problem now, before they know what you are building.

3. Asking leading questions

"Don't you think this would be useful?" feels like a reasonable question. It is not. It signals the answer you want, and most people will agree to be polite rather than risk seeming difficult. A leading question contaminates your data and hands you the conclusion you already hoped for.

Fix: ask open, neutral questions that do not hint at a right answer. "How do you handle this now?" beats "Wouldn't this make it easier?"

4. Talking more than listening

If you are talking for 70% of the interview, you are not running research, you are running a monologue. Every minute you spend explaining is a minute the participant is not revealing the signal you came for. The most useful thing you can do in an interview is stay quiet and let the silence do the work.

Fix: aim to talk far less than the participant. When an answer feels incomplete, wait. The first answer is usually the tidy version; the real one arrives in the pause.

5. Interviewing the wrong people

Friends, family, and your own network are the easiest people to recruit and the worst people to learn from. They want to support you, so they flatter the idea. Friends are not users. The only feedback that matters comes from people who actually feel the pain you are trying to solve.

Fix: define who genuinely has the problem and recruit strangers who match that profile, even though it is harder. Reaching the right people is the single most important factor in useful research.

6. Asking hypothetical questions

"If we built X, would you use it?" is the question that has misled more founders than any other. Faced with a hypothetical, almost everyone says yes, because saying yes is easy and costs nothing. Guesses about the future are not evidence. Past behaviour is.

Fix: ask what they have actually done. "When did you last face this? What did you do? What did it cost you?" tells you far more than any hypothetical ever will.

7. Ignoring emotional cues

The words a participant uses are only part of the data. Tone, hesitation, frustration, and the things they avoid saying are often the richest signal in the entire conversation. If you only capture the words, you miss the meaning behind them.

Fix: listen beyond the words. When you hear hesitation or frustration, slow down and follow it. That is usually where the real problem lives.

How many interviews do you actually need?

There is no magic number, but for early discovery, five to eight interviews per customer segment is usually enough to start hearing the same patterns repeat. When several people in a row tell you nothing new, you have reached the point of diminishing returns. The quality of who you talk to matters far more than the quantity.