Reading time: 7–8 min
Every founder has heard the advice: talk to your customers. Almost none of them do it well. The difference between an interview that changes your direction and one that produces comfortable noise is almost entirely about technique — and technique can be learned.
Why Most Interviews Fail
I have watched hundreds of founders conduct customer interviews. The pattern is almost always the same. They schedule a 20-minute call. They open with their idea. They ask the customer what they think. The customer says it sounds interesting. The founder leaves feeling validated.
Nothing useful has been learned.
The problem is structural. The founder led with the solution, which primed the customer to evaluate it rather than describe their real experience. The interview was too short to build the trust needed for honest answers. The questions were closed — 'would you use this?' — rather than open. And the founder was listening for confirmation, not for contradiction.
Real customer interviews require a different approach from the start.
The goal of an interview is not to pitch your solution. It is to understand a problem.
The Setup — Before You Even Ask a Question
Good interviews start before the first question. Three things need to be in place.
Time
Book 45 minutes minimum. Serious conversations need time to warm up. The first 10 minutes are almost always surface-level — the customer is still figuring out who you are and what you want. The real insight usually emerges in the 20-to-40-minute window, after trust has been built and the customer has started thinking out loud. Twenty-minute interviews almost never surface anything useful.
Context
Explain clearly that you are doing research, not selling anything. Say it explicitly: 'I am not trying to sell you anything today. I am trying to understand your experience better.' This removes the defensive posture that comes from feeling like a sales call. Customers who believe they are helping a researcher, not evaluating a pitch, speak more honestly.
Permission to Go Deep
At the start, give the customer permission to take you wherever the conversation needs to go. Tell them: 'There are no wrong answers. If something I ask reminds you of a related experience, go there.' You want to follow threads, not run through a script.
The Four Question Types
Well-structured customer interviews use four types of questions, roughly in sequence:
1. Context Questions
Open the interview by understanding who this person is and how they relate to the problem space. These questions are not about your product at all — they are about understanding the customer's world.
Examples: 'Walk me through your typical week in this area.' / 'How long have you been dealing with this?' / 'Who else in your organisation is involved in this?'
2. Experience Questions
Ask them to describe specific past experiences, not hypothetical opinions. Past behavior is a far more reliable predictor of future behavior than stated preferences. 'Would you buy this?' is almost useless. 'Tell me about the last time you tried to solve this problem' is valuable.
Examples: 'Tell me about the last time this came up for you.' / 'What happened?' / 'How did you handle it?' / 'What did you try first?'
3. Emotion Questions
Go below the surface to understand how the problem feels. This is where willingness-to-pay is born — not in features, but in frustration, embarrassment, anxiety, or wasted time. Customers do not pay for products; they pay to relieve feelings.
Examples: 'How did that make you feel?' / 'What was the most frustrating part?' / 'If this keeps happening, what is the cost to you — in money, time, or something else?'
4. Ideal State Questions
Ask them to describe what a perfect solution would look like — but do this after they have described the problem in depth, not before. This sequencing matters enormously. If you ask for the ideal state before they have articulated the problem, you get wish-lists. If you ask after they have relived the frustration, you get genuine insight into what they actually need.
Examples: 'If this was solved perfectly, what would that look like?' / 'What would make this problem completely disappear for you?' / 'Have you ever experienced a moment — in any part of your life — where something similar worked exactly the way you wished this did?'
The Interview Framework
The Techniques That Make Interviews Work
Shut Up and Listen
The most important technique in interviewing is also the hardest: stay silent. After the customer stops speaking, count to three before you respond. You will be amazed how often they fill the silence with something more honest than what they said first. People rush to fill silence, and what comes out in those moments is usually unfiltered.
The Five Whys
When a customer gives you an answer, ask why. Then ask why again. Do this five times and you will move from surface behavior to underlying motivation. 'I use spreadsheets for this' → why? → 'because the current tools are too complicated' → why? → 'because they were designed for enterprises, not our team size' → why does that matter? → 'because I am the only person who knows how to use them, and that creates a dependency problem.' The fifth why is almost always where the real insight lives.
Silence as a Tool
Professional interviewers and therapists know that silence is one of the most powerful techniques available. When someone has just said something emotionally loaded — about frustration, fear, cost, or failure — resist the impulse to respond or move on. Hold the silence. They will often go deeper on their own.
Mirror and Reflect
When the customer says something interesting, repeat the last few words back to them as a question. 'You said it felt like a waste of time?' This simple technique — borrowed from crisis negotiation — signals that you heard them and invites them to elaborate without steering the conversation anywhere in particular.
Never Pitch Mid-Interview
If the customer asks about your product or idea mid-interview, deflect. Say: 'I would love to tell you more about that — can I finish understanding your experience first? I want to make sure what we are building actually addresses the right problem.' This keeps the conversation in discovery mode and also signals that you are serious about understanding before building.
Remote vs. In-Person Interviews
In-person interviews almost always produce better data than remote ones. The physical environment gives you context. You can observe body language. The conversation feels more natural. If you can get to the customer — their home, their office, their place of work — do it.
That said, remote interviews are better than no interviews. If geography or schedule makes in-person impossible, use video (not phone). Seeing someone's face adds meaningful signal. Watch for the micro-expressions when they describe a problem: the slight wince, the long pause, the hand gesture. These are data.
How Many Interviews Is Enough?
Jakob Nielsen's research at the Nielsen Norman Group found that five well-conducted user interviews typically surface 85% of usability issues in a product — and that the marginal return from additional interviews drops steeply beyond that. This finding has been applied broadly to customer discovery research.
In practice, for early-stage venture research: conduct interviews until you start hearing the same things. When three or four consecutive interviews are producing no new insights, you have reached saturation in that segment. This typically happens between five and twelve interviews for a well-defined target segment.
GVP requires five customer interviews minimum, across at least three different types of people in your target market. That is not an arbitrary number — it is the threshold at which patterns begin to emerge.
You are not looking for majority opinion. You are looking for pattern. Five good interviews will show you the pattern.
What to Do With What You Hear
After each interview, write down three things immediately: the moment that surprised you most, the exact phrase the customer used that you want to remember, and the one thing you would do differently in your next interview. Do this before you check your phone or talk to anyone else. Memory is unreliable and social influence is fast.
After five interviews, lay out your notes and look for patterns. What themes appear more than once? What surprised you every time? What did you expect to hear that you never heard? The gap between your expectations and reality is where the insight lives.
✎ A Note for GVP Students
.png)
