Running customer interviews means having structured conversations with real or potential users to learn the truth about their problems—before you build anything. Done well, they reveal what people actually do, struggle with, and pay for; done badly, they produce flattering lies that send you confidently in the wrong direction. This guide covers who to interview, the questions that extract honest answers, how to run the conversation itself, and how to turn what you hear into decisions. The skill is mostly about listening, and it's learnable.
Why customer interviews beat surveys and guessing
Founders have three ways to figure out what customers want: guess, survey, or talk to them. The first two fail in predictable ways.
Guessing—building what you think is needed—is how most failed startups begin. Surveys feel more rigorous but share a fatal flaw: they ask people to predict their own future behavior, which humans do terribly. "Would you use this?" gets a cheerful yes that means nothing. Surveys also can't follow up, so they miss the why behind an answer and the messy workarounds people have rigged up.
Interviews fix this by digging into real, past behavior and the emotion around it. A good conversation surfaces what someone actually did the last time they hit the problem, how much it cost them, and what they tried—evidence, not speculation. This is why interviews are the engine of validating a startup idea properly: they're the cheapest way to replace your assumptions with facts. Where surveys and analytics tell you what and how many—the territory of sizing a market through research—interviews tell you why, and the why is what you can act on.
The mindset that matters most: an interview is for learning, not selling. The moment you start pitching, you've stopped gathering data.
Who to interview and how to find them
The best questions are worthless if you ask the wrong people. Interview people who have the problem right now—not your friends, family, or fellow founders, who want you to succeed and will tell you what you want to hear.
Find them where they already gather: niche subreddits and forums, industry Slack and Discord communities, LinkedIn, professional groups, or by asking early contacts for referrals. Cold outreach works better than expected when you're honest and ask for help rather than a sale: a short message saying "I'm researching how people handle [specific problem]—not selling anything—could I ask you a few questions for 20 minutes?" converts surprisingly well.
On volume, aim for roughly 15 to 30 interviews within a customer segment before drawing firm conclusions. You'll often hear the same patterns emerge by the fifth or sixth conversation, but a larger sample protects you from over-weighting one vivid opinion. Quality of fit beats quantity every time—ten conversations with people who genuinely live the problem teach you more than fifty random ones.
Asking questions that get the truth: the Mom Test
The definitive framework here is Rob Fitzpatrick's The Mom Test, built on a simple premise: ask questions so grounded in fact that even your mother couldn't lie to you to make you feel good. It comes down to three rules: talk about their life, not your idea; ask about specifics in the past, not opinions about the future; and talk less, listen more.
The difference shows up entirely in how you phrase questions:
| Bad question (invites lies) | Good question (extracts truth) |
|---|---|
| "Would you use an app that did X?" | "Tell me about the last time you faced this problem." |
| "Do you think this is a good idea?" | "What did you do to solve it? Walk me through it." |
| "Would you pay $20/month for this?" | "What are you using today, and what does it cost you?" |
| "Don't you hate it when…?" | "How often does this come up, and what happens when it does?" |
Every bad question on the left asks for a prediction, an opinion, or agreement—all cheap to give and worthless as data. Every good question on the right asks about concrete past behavior, which is real. Notice too that the good questions never mention your idea. The instant you describe your solution, the conversation turns into a sales pitch and the honesty evaporates. Keep your idea out of the room and investigate the problem like a journalist.
Running the interview and making sense of it
Knowing what to ask is half of it; the conversation itself has its own craft.
Here's a practical structure for each interview:
- Set the frame (1–2 min). Explain you're researching the problem, not selling, and ask permission to record. This relaxes people and gets you honest answers.
- Start broad, then narrow. Open with their context and their last encounter with the problem before zooming into specifics.
- Shut up and listen. Aim to talk less than 20–30% of the time. Silence is your friend—people fill it with the most revealing material. Resist the urge to jump in or defend.
- Follow the emotion and the "why." When someone gets animated or frustrated, dig there: "Why was that so annoying? What did you do next?" The strongest problems show up as workarounds, spreadsheets, and visible irritation.
- Chase specifics, not generalities. If they say "it's a hassle," ask them to walk you through the last time, step by step, with real numbers.
- Don't pitch, even at the end. If you must mention what you're building, do it after you've learned everything, and frame it as seeking honest reaction, not approval.
Keep interviews short—20 to 30 minutes respects people's time and keeps focus. Recording (with consent) frees you to listen instead of scribbling; if you can, bring a second person to take notes so the interviewer stays present.
Then comes synthesis, which is where the value is realized. After each conversation, capture verbatim quotes and the concrete facts—what they did, how often, what it cost. Across interviews, group these into themes (a simple affinity map of sticky notes or a spreadsheet works) and separate two things ruthlessly: facts (what people actually did) carry far more weight than opinions (what they say they'd do). Weight signals of real commitment—time spent, money already paid for a workaround, an offer to introduce you to a colleague—over compliments, which cost nothing. The patterns that emerge tell you whether the problem is real and acute, which in turn shapes what to build into your MVP and, eventually, your path toward genuine product-market fit.
Common mistakes to avoid
Pitching instead of listening. The single most common error. The interview is for learning; selling poisons the data. Keep your idea out of it.
Asking leading or hypothetical questions. "Would you use this?" and "Don't you think…?" beg for a polite yes. Ask what people did, not what they imagine they'd do.
Talking too much. If you're talking more than a third of the time, you're not interviewing—you're presenting. Embrace silence.
Interviewing the wrong people. Friends and supporters give encouragement, not evidence. Seek out strangers who have the problem.
Hunting for validation. Going in hoping to hear yes guarantees you'll find someone to say it. Go in trying to disprove your assumption instead.
Collecting compliments and calling it research. "That's a great idea" is not data. Commitment—time, money, referrals—is the only signal that counts.
Not capturing quotes. Relying on memory loses the specific language and details that matter most. Record or take careful notes, every time.
Frequently asked questions
How many customer interviews should I do? Aim for about 15 to 30 within a single customer segment. Clear patterns often appear by the fifth or sixth conversation, but a larger sample guards against being misled by one strong opinion. Talking to the right people—those who actually have the problem—matters more than the raw count.
What questions should I ask in a customer interview? Ask about specific past behavior, not hypotheticals: "Tell me about the last time you dealt with this," "What did you do about it," and "What are you using now and what does it cost you?" Avoid questions like "Would you use this?" that invite a polite, meaningless yes.
How do I find people to interview? Look where your target customers already gather—relevant online communities, forums, LinkedIn, and industry groups—and ask early contacts for referrals. Honest cold outreach that asks for help researching a problem, rather than pitching a product, gets a surprisingly good response.
Should I tell people about my product idea during the interview? Generally no. Describing your idea turns the conversation into a sales pitch and prompts polite, flattering answers. Investigate the problem first; if you mention your solution at all, do it at the end and frame it as seeking honest reactions, not approval.
What's the difference between customer interviews and surveys? Interviews are qualitative conversations that uncover why people behave as they do, with the ability to follow up and dig into specifics. Surveys gather quantitative data at scale but can't probe or interpret, and they rely on people predicting their own behavior, which is unreliable. The two complement each other.
The takeaway
Running customer interviews well comes down to one discipline: investigate the customer's real past behavior instead of pitching your idea or fishing for approval. Talk to people who actually have the problem, ask what they did rather than what they would do, listen far more than you speak, and weight commitment over compliments. Your next step is to line up five conversations this week with people who live the problem you're trying to solve—and go in trying to learn the truth, not to hear that you're right.