Problem Interviews

What is a problem interview?

A problem interview is a 30 to 45 minute conversation with someone who seems to have the problem you want to solve, where you ask about their real past behaviour and say almost nothing about your idea. You want three facts: does the problem happen, how do they deal with it today, and what does it cost them in money, time or nerves.

The best-known guide is The Mom Test by Rob Fitzpatrick (momtestbook.com). The name comes from his point that your mum loves you and will lie to protect your feelings, so you need questions even she can’t answer with a polite lie. His three rules, in short: talk about their life instead of your idea, ask about specifics in the past instead of opinions about the future, and talk less while you listen more.

When should you use it?

Use problem interviews when:

• You have a problem you suspect is real, say from Reddit problem research or a problem journal, and a rough guess of who has it.
• You have NOT built anything yet, including a landing page.
• You sell to businesses or consumers, online or offline. Apps, services and physical products all fit, because the questions are about the customer’s week.

Skip it, or wait, when:

• You can’t name a group of people who have the problem. Go back to find a problem worth solving first.
• The question you need answered is “will they pay €29 for this?” Price needs a pre-sale or a payment link.

How do you do it, step by step?

1. Write one sentence you want to test. Keep it about people and a problem: “Independent dog groomers lose income every week to clients who don’t show up.” No solution in it. Every question you ask should help prove or break that sentence.

2. Prepare 5 to 7 questions about the past. These work for almost any problem:

   • “Walk me through the last time [the problem] happened. When was it?”
   • “What was the hardest part of that?”
   • “What did you do about it?”
   • “What else have you tried? What didn’t you like about it?”
   • “What does this cost you now, in money or hours a week?”
   • “How do you handle it today, step by step?”
   • “Who else should I talk to about this?”

3. Book 10 to 15 people who had the problem in the last 3 months. Recent memories are sharp. See finding people to interview for message templates. Friends and family do not count unless they truly live the problem.

4. Open with a low-pressure frame. Say something like: “I’m researching how independent groomers handle bookings. I’m not selling anything today, I just want to learn how your week works. Is it OK if I record this for my notes?” Then ask your first question and stop talking.

5. Dig three levels down on every answer. When someone says “no-shows are annoying”, ask when it last happened, what they did next and what it cost. When they drift into the future (“I would totally use an app for that”), pull them back: “Have you looked for one? What happened?” When they compliment your idea, thank them and ask about their week again. Compliments are noise.

6. Take notes in their words, the same day. Use one Google Doc per interview with four headings: exact quotes, what they DID (actions, workarounds), what it costs (money, hours, lost clients), and surprises. Fill it within 15 minutes of hanging up. Add a line on tone and hesitation, which no transcript catches.

7. End with a small ask. “Who else has this problem that I should talk to?” and “Could I show you a rough version in 3 weeks?” A yes with a date, an intro or a calendar slot is a real commitment. “Sounds cool, keep me posted” is a polite goodbye.

What does a worked example look like?

Say Lena, a hypothetical founder, wants to build booking software for independent dog groomers. Her test sentence is the one from step 1.

Through 3 local pet-business Facebook groups and 4 salon visits, she books 12 interviews in 3 weeks.

Her opening question is “Walk me through the last client who didn’t show up.” 9 of the 12 describe a no-show from the past month without hesitation. 7 already ask new clients for a deposit by bank transfer and chase it by hand. 3 send reminder texts copied from a note on their phone, one by one, every evening. One groomer estimates she loses 2 slots a week.

The surprise sits in her “surprises” heading: 5 groomers say they HATE asking regulars for deposits because it feels rude. So the problem is partly social. Lena rewrites her sentence: “Groomers want to protect themselves from new clients who don’t show up, without upsetting loyal ones.” She ends the round with 5 groomers who agreed to try a rough version within a month.

Which free tools help?

What mistakes should you avoid?

• Pitching in the first 10 minutes. Once you describe your idea, people switch to being nice and you lose the truth.
• Asking “would you use” or “would you pay”. Opinions about the future are cheap. Ask what they did last time.
• Leading questions. “Don’t you hate it when clients cancel?” tells them the answer you want. Ask “How do cancellations go for you?”
• Counting enthusiasm as evidence. Excitement fades by Monday. Workarounds, money spent and hours lost are the evidence.
• Stopping at 3 interviews. Three people can all be outliers. Patterns start to show around 8 to 10.

How do you know it worked?

Our rule of thumb for a round of 10 to 15 interviews:

• At least 6 in 10 people describe a SPECIFIC recent time the problem happened, without you prompting the details.
• At least 3 have already spent money, built a workaround (a spreadsheet, a WhatsApp routine, a hired helper) or switched tools because of it.
• At least 2 give you a commitment: an intro, a follow-up date, a request to try something.

Go if you hit all three. If most people only agree the problem exists once you describe it, and nobody has tried to fix it, the problem is mild. Change the segment or the problem and run another round before you build.

What should you do next?

Turn your notes into patterns with interview synthesis, then go deeper on why people switch solutions with jobs-to-be-done interviews. If the problem holds up, test the promise with a landing page test. For feedback on your question list before you start, the F/MS startup game walks you through validation with an AI co-founder.

FAQ

How many problem interviews do I need?

Plan for 10 to 15 per customer segment. Our rule of thumb is to stop when 3 interviews in a row teach you nothing new. If you are still hearing new problems after 15, your segment is too broad.

Can I do problem interviews by email or chat?

You can, but you lose most of the signal. Hesitation, tone and the stories people tell once they relax only show up live, so a 30 minute call beats 10 written answers.

What if people ask about my idea during the interview?

Promise to share it at the end and keep asking about their experience. In the last 5 minutes, describe it in 2 sentences and watch whether they ask how to get it or just say “cool”.

Updated 26 October 2026 · Written by the Fe/male Switch team

Mark this method as done