How to Do Customer Interviews as a Beginner
To do a good customer interview, ask people about what they already do, not what they would hypothetically buy. Your goal is truth, and truth lives in the past: the last time they hit the problem, what they tried, what it cost them. The moment you describe your idea and ask “would you use this?”, the interview is dead, because everyone is nice and nobody wants to hurt your feelings.
That’s the whole trick. Most beginners think an interview is a chance to sell. It’s the opposite. You are a detective, not a salesperson, and you’re trying to figure out whether the problem you care about is a real bleeding problem in someone’s actual life.
Why most people run interviews wrong
The default instinct is to walk in and pitch. You say “I’m building an app that does X, would you use it?” and the person says “yeah, sounds cool.” You leave feeling great and learn nothing.
Here’s why that answer is worthless. It’s about the future, which people are bad at predicting. It’s about your idea, which they’ll compliment to be polite. And it’s a yes/no question, which invites the easiest possible answer.
Compliments, hypotheticals, and vague futures are the three ways an interview lies to you. The fix is to stop talking about your idea at all and start asking about their life.
What is a customer interview?
A customer interview is a short, structured conversation where you learn how someone currently deals with a problem, without pitching them a solution. You’re gathering evidence, not feedback.
The distinction matters. Feedback is what someone thinks of your idea, and it’s usually inflated. Evidence is what someone actually did last week, and it’s hard to fake.
This whole approach comes from a short book called The Mom Test by Rob Fitzpatrick. The core rule: ask questions so grounded in the person’s real life that even your mom couldn’t lie to you about the answer. You don’t ask “do you think my idea is good?” (Mom says yes.) You ask “when did you last spend money trying to fix this?” (Now she has to tell the truth.)
Where do I find people to interview?
You need people who plausibly have the problem. Start closer than you think.
- Your own network, one degree out. Not your best friend who will lie to be kind. Their older sibling, a teammate’s parent, the coach, the teacher who runs the club.
- Communities where the problem lives. Subreddits, Discord servers, Facebook groups, Slack communities built around the topic you care about. Read first, ask second.
- The place the behavior happens. If you’re solving something for student athletes, go to a practice. For small local businesses, walk in during a slow hour.
- A short post asking for 15 minutes. “I’m a high school student researching how people handle [problem]. Not selling anything. Can I ask you a few questions?” People say yes to helping a student more than you’d guess.
You don’t need 100 interviews. Five to ten focused conversations with the right people will teach you more than a 500-person survey, because you can follow up and dig. If you’re still figuring out which problem is even worth this effort, start with how to find a startup problem worth solving before you book a single call.
A question script you can actually use
Steal this. Adapt the topic to your idea, but keep the shape: past tense, specific, no pitching. Aim for a 15-20 minute call.
- “Walk me through the last time you dealt with [problem]. What happened?” This is your best question. It forces a real story instead of a generalization.
- “Why was that annoying?” or “What made that hard?” Let them explain the pain in their own words. Don’t put words in their mouth.
- “What did you do about it?” If the answer is “nothing,” that’s a huge signal — the problem may not hurt enough to act on.
- “What tools or workarounds do you use now?” People duct-taping together spreadsheets and group chats are telling you where a product could go.
- “How often does this come up?” Once a year is a vitamin. Every day is a painkiller.
- “Have you ever paid for anything to solve this? What?” Money already spent is the strongest evidence a problem is real.
- “What else have you tried that didn’t work?” This maps the graveyard of solutions and shows you why the problem is still open.
- “Who else deals with this? Can you intro me?” Ends the call and grows your list at the same time.
Notice what’s missing: you never describe your product. If they ask what you’re building, say “I’ll show you in a minute, but I want to hear your experience first,” and get back to their story.
Questions to ask vs. questions to avoid
The difference between a useful and a useless interview usually comes down to a handful of word choices.
| Avoid this (invites lies) | Ask this instead (surfaces truth) |
|---|---|
| “Would you use an app that…?" | "How do you handle this today?" |
| "Do you think this is a good idea?" | "Walk me through the last time it happened." |
| "Would you pay $10 for this?" | "What have you already paid to solve this?" |
| "Don’t you hate when…?" | "Tell me about the last time that came up." |
| "Would this feature be helpful?" | "What did you do when you hit that wall?” |
The left column asks about the future and about your idea. The right column asks about the past and about their life. That’s the only pattern you have to remember.
What to avoid: pitching, leading, and yes-bias
Three traps kill beginner interviews.
Pitching. The second you describe your solution, the person switches from “reporting my life” to “reacting to your thing,” and reactions are polite noise. Save the pitch for last, or skip it entirely.
Leading questions. “Isn’t it frustrating when the app crashes?” tells them the answer you want. Ask open questions that start with what, how, when, or why, and then shut up.
Yes-bias. Humans are wired to agree and encourage, especially with a kid asking earnest questions. Assume every “that’s a great idea” is worth zero. Only weigh what they’ve actually done. For more on why enthusiasm is the least reliable signal, see how to validate a startup idea in high school.
One more: talking too much. If you’re speaking more than a third of the time, you’re doing it wrong. Ask, then let the silence sit. People fill silence with the good stuff.
How to take notes and spot real signal
Record the call if they consent, but also take live notes, because the act of writing forces you to catch specifics. Capture exact quotes, not your summary. “I spent an hour every Sunday copying numbers into a spreadsheet” is gold. “User frustrated with process” is a paraphrase that loses the evidence.
After each call, mark three things:
- The specific pain moment — the actual event they described.
- What it cost them — time, money, stress, a missed outcome.
- What they already do about it — the current workaround.
Then sort your signal by strength. Here’s the ladder, weakest to strongest:
| Signal | What it means | How much to trust it |
|---|---|---|
| ”That’s a cool idea” | Politeness | Ignore |
| ”I’d definitely use that” | A future promise | Ignore |
| ”This happens to me all the time” | A real, frequent pain | Note it |
| ”I currently do [messy workaround]“ | Active behavior | Trust it |
| ”I already pay for [X]” or “Can I try it now?” | Money or urgency | Trust it a lot |
The bottom two rows are what you’re mining for. Anything a person already spends time or money on is a problem they’ve voted for with their behavior. Anything they only promise about the future is a wish.
If you run ten interviews and nobody has a workaround, nobody has paid for anything, and nobody can name the last time it hurt — that’s not a failed study. That’s a successful one. You just saved yourself months building something nobody needs. At batch0, the first of our four build sprints is entirely about this kind of validation, because founders who skip it tend to build beautiful things for problems that don’t exist.
What customer interviews are not
- Not a survey. Surveys give you shallow answers at scale. Interviews give you depth and the ability to ask “why” five times. Once your interviews reveal a pattern worth measuring across more people, learn how to run a survey that actually tells you something.
- Not a focus group. You want one person’s real story, not a room agreeing with the loudest voice.
- Not a sales call. No pitching, no closing, no “so, interested?”
- Not proof by itself. Interviews tell you a problem is real. They don’t prove people will pay. For that you need a test with skin in the game — a pre-order, a signup, a deposit. That’s the next step, and it’s covered in how to test a business idea before you build it.
How long does this take, and what’s next?
Plan for a week. You can book five to ten conversations in a few days if you ask enough people, and each call is 15-20 minutes plus a few minutes to write up. The bottleneck is usually reaching out, not the interviews themselves, so send more requests than you think you need.
When patterns start repeating — the same pain, the same workaround, the same words — you’ve heard enough to move. Don’t interview forever to avoid building. The point of talking to people is to earn the right to make something, then put that something in front of them and watch what they actually do.
The people who gave you the sharpest answers are also your warmest early leads, so keep their names. Many of them become your first users once you have something to show, which is exactly the moment to read how to get your first 10 customers as a student founder.