How to Interview Customers for Startup Validation

Customer interviews are one of the most reliable ways to validate whether a problem is real, painful, and worth paying to solve. But most founders interview poorly. They describe their solution and ask whether the customer would use it. This approach reliably produces polite, encouraging, and almost entirely useless feedback. This guide is a practical, question-by-question reference for interviewing well — a companion to the broader customer discovery guide, focused specifically on what to say in the room.

The Mom Test Principles

The Mom Test, by Rob Fitzpatrick, is the definitive framework for customer interviewing, and its core insight is simple: people will lie to be polite, especially about hypotheticals. Your mom will tell you your idea is great. Strangers will tell you your idea is interesting. Neither response is reliable, because neither response is about the actual problem — it's about being nice to you.

Good Questions vs. Bad Questions

Bad questions describe your solution and ask for an opinion: "Would you use a tool that automates invoice tracking?" This kind of question invites polite agreement regardless of whether the person would actually change their behavior. Good questions explore the problem without mentioning your solution at all: "Tell me about the last time you had trouble getting paid on time."

A practical question bank

• "Walk me through what you did the last time this happened."
• "What are you using today to deal with this?"
• "What don't you like about your current approach?"
• "How much time or money would you say this costs you in a typical month?"
• "Have you looked for a better way to handle this? What did you find?"
• "What would need to be true for you to switch away from what you use now?"

Questions to Avoid

• "Would you use [my product]?" — hypothetical, invites polite agreement
• "Do you think this is a good idea?" — asks for an opinion about you, not evidence about the problem
• "How much would you pay for this?" — abstract pricing questions produce unreliable numbers; anchor to a specific price instead
• Any question that describes your solution before the person has described their own current behavior

How to Handle Common Interview Moments

When someone says "that sounds cool," don't take it as validation — redirect back to their behavior: "Have you actually looked for something like that before?" When someone starts pitching you feature ideas, let them, but treat it as information about their frustration, not a product roadmap. When an interview produces nothing useful, that's data too — note it and move to the next conversation rather than forcing a signal that isn't there.

How Many Interviews Do You Need

Most founders need 10 to 20 interviews before real patterns emerge. Around the 8-10 interview mark, you should start hearing the same themes repeated in similar language. When multiple, independent strangers describe the same frustration using similar words without prompting, that's a strong signal you've found a real problem worth pursuing further.

After the Interview: What to Do With Your Notes

Write down the person's exact phrasing, not your summary of it — exact language becomes useful later for marketing copy and for spotting patterns across interviews. After each batch of 5-6 interviews, review your notes together and look specifically for repeated words, repeated workarounds, and repeated complaints about the same aspect of the problem.

How DemandProofHQ Prepares You for Better Interviews

Running a DemandProofHQ check before your interviews gives you market context that sharpens your questions from the very first conversation. You'll already know what signals exist publicly, what language potential customers use to describe the problem, and which communities to find interview subjects in. Start at /idea-check.

DemandProofHQ helps review public demand signals, but it does not guarantee product-market fit or replace direct customer conversations.