How to run customer discovery interviews: questions that get honest answers

Most product teams think they’re doing customer discovery interviews, but they’re actually just collecting compliments. They ask leading questions, accept vague answers, and walk away believing customers love an idea that will flop the moment it ships. Real discovery is harder—and more valuable—than that.

The difference between teams that build products people actually want and teams that build expensive failures often comes down to how well they run these conversations. This guide will show you exactly how to conduct customer discovery interviews that reveal genuine insights, including the questions that actually work and the mistakes that sabotage most teams.

Discovery interviews vs. usability tests: know the difference

Before you schedule a single interview, get clear on what you’re trying to learn. Discovery interviews and usability tests answer fundamentally different questions:

Customer discovery interviews explore the problem space. You’re trying to understand how people currently behave, what frustrates them, and what they’ve already tried. You don’t show them anything—you just listen to their stories.

Usability tests evaluate a specific solution. You put a prototype or product in front of someone and watch them try to complete tasks. You’re testing whether your design works, not whether you’re solving the right problem.

The mistake many teams make is jumping to usability testing before validating that the problem exists and matters. As Teresa Torres puts it in her Continuous Discovery Habits framework [INTERNAL_LINK: continuous discovery habits], you need to understand the opportunity space before you can evaluate solutions effectively.

Run discovery interviews when you’re:

• Exploring a new problem area or market
• Validating that a problem is worth solving
• Understanding why users abandoned a feature or product
• Investigating unexpected user behavior in your data
• Building conviction before a major investment

The Mom Test: the only framework you need

Rob Fitzpatrick’s The Mom Test should be required reading for every PM. The core principle is simple: ask questions so good that even your mom couldn’t lie to you.

Most interview questions fail the Mom Test. “Would you use an app that helps you save money?” Your mom would say yes. So would everyone else—even people who would never actually use it. The question invites false positives.

Three rules make customer discovery interviews actually useful:

1. Talk about their life, not your idea

Instead of pitching your solution and asking for opinions, ask about their past behavior and current struggles. People are bad at predicting what they’ll do in the future, but they can accurately describe what they’ve already done.

Bad: “Would you pay for a tool that organizes your receipts?”
Good: “Walk me through what happens when you need to find a receipt from a few months ago.”

2. Ask about specifics, not hypotheticals

Generic questions get generic answers. Anchor every question to a specific, real situation. “Last time this happened” is your best friend.

Bad: “How often do you have this problem?”
Good: “When was the last time this happened? Tell me about that.”

3. Listen more than you talk

If you’re talking more than 20% of the interview, you’re doing it wrong. Your job is to create space for them to share stories, then dig deeper into the interesting parts.

Recruiting participants who actually matter

Talking to the wrong people is worse than talking to no one—it gives you false confidence. Here’s how to find the right participants:

Define your target precisely

Don’t recruit “small business owners.” Recruit “founders of B2B SaaS companies with 10-50 employees who are actively hiring.” The more specific your criteria, the more useful your insights.

Find people with recent, relevant experience

You want participants who have encountered the problem recently—ideally within the last few weeks. Their memories are fresh, and they can give you specific details instead of hazy generalizations.

Recruitment channels that work

• Your existing users: Filter for specific behaviors in your analytics, then reach out directly
• LinkedIn: Search for job titles and industries, then send personalized messages
• Respondent.io or User Interviews: Paid panels with detailed screeners
• Industry communities: Slack groups, Reddit, Discord servers where your target users hang out
• Customer-facing teams: Ask sales, support, and success who they’ve talked to recently with relevant problems

How many interviews do you need?

The classic answer is “until you stop hearing new things.” In practice, 8-12 interviews usually surface the major patterns for a specific problem area. But quality matters more than quantity—five interviews with precisely the right people beats twenty with loosely-qualified participants.

20+ interview questions that actually work

These questions follow the Mom Test principles. They focus on past behavior, specific situations, and observable actions rather than opinions and predictions.

Opening questions to understand context

1. “Tell me about your role. What does a typical week look like?”
2. “What are the biggest challenges you’re dealing with right now in [specific area]?”
3. “Walk me through how you currently handle [process you’re investigating].”

Problem exploration questions

4. “When was the last time [problem] happened? Tell me about that situation.”
5. “What were you trying to accomplish when that happened?”
6. “How did that make you feel? What did you do next?”
7. “What’s the hardest part about [task or process]?”
8. “Why is that hard?” (Then ask “why” again—at least 3-5 times)
9. “How often does this come up? Can you give me another example?”
10. “What happens if you don’t solve this problem? What’s the consequence?”

Current solution questions

11. “What have you already tried to solve this?”
12. “What tools or workarounds are you currently using?”
13. “How much time/money are you spending on this now?”
14. “What do you like about your current approach? What’s frustrating about it?”
15. “Have you looked for a better solution? What did you find?”
16. “Why haven’t you switched to [competitor or alternative approach]?”

Value and priority questions

17. “Where does this rank among your priorities right now? What’s above it?”
18. “If you could wave a magic wand and fix one thing about this, what would it be?”
19. “Who else cares about this problem? Who would need to approve a solution?”
20. “What would a successful outcome look like for you?”

Follow-up questions to go deeper

21. “Can you show me how you do that?” (screen share or demonstration)
22. “You mentioned [interesting thing]. Tell me more about that.”
23. “What do you mean when you say [vague term they used]?”
24. “That’s interesting—why do you think that is?”
25. “Is there anything else about this I should have asked?”

What to listen for during interviews

The words people say matter less than the signals underneath them. Train yourself to notice:

Emotional intensity

When someone’s voice changes, when they lean in, when they use words like “hate,” “nightmare,” or “constantly”—you’ve found something that actually matters. Mild frustration won’t drive purchase decisions. Strong emotion does.

Specific workarounds

If someone has built a spreadsheet, written a script, or developed a manual process to solve this problem, they’ve already proven they care enough to act. This is stronger evidence than any verbal claim about willingness to pay.

Money or time already spent

Past investment predicts future investment. If they’ve tried paid solutions, hired someone, or dedicated significant time—the problem is real to them. If they’ve done nothing despite supposedly caring, be skeptical.

Contradictions between words and actions

“This is really important to me, but I haven’t tried anything to fix it” is a red flag. Dig into why. Sometimes there’s a valid blocker. Often, it’s just not that important.

Who else is involved

B2B buying decisions involve multiple stakeholders. Understanding who else cares—and who might block a purchase—helps you map the real decision process.

Synthesizing findings across multiple interviews

Individual interviews generate stories. Synthesis generates insights. Here’s how to move from raw notes to actionable patterns:

Capture notes in a consistent format

During or immediately after each interview, document:

• Key quotes (verbatim when possible)
• Observed behaviors and workarounds
• Emotions expressed
• Problems mentioned and their severity
• Solutions they’ve tried
• Surprising moments or contradictions

Look for patterns, not just frequency

A problem mentioned by 8 of 10 participants is significant. But a problem mentioned by only 2 participants with extreme emotional intensity—who’ve spent thousands trying to solve it—might be more important. Weight by severity, not just occurrence.

Create an insight repository

Many teams use tools like Dovetail, Notion, or Airtable to tag and organize findings. The specific tool matters less than having a searchable system your team actually uses. Intercom famously uses a shared research repository that any team member can access and contribute to.

Share findings visually

A one-page summary with the top 3-5 patterns, supported by specific quotes, travels further than a 20-page report. Include the voice of the customer directly—let stakeholders hear what you heard.

Common mistakes that ruin discovery interviews

Even experienced PMs fall into these traps:

• Pitching instead of listening: If you’re explaining your idea, you’re not learning.
• Accepting first answers: The interesting stuff comes when you ask “why” and “tell me more.”
• Talking to convenient participants: Your friend’s opinion doesn’t represent your target market.
• Leading questions: “Don’t you think it would be helpful if…” guarantees useless data.
• Stopping at surface problems: The first problem they mention is rarely the root cause.
• Conducting discovery once: Customer understanding isn’t a phase—it’s an ongoing practice. [INTERNAL_LINK: continuous product discovery]

Your next step

Schedule three customer discovery interviews this week. Use the questions in this guide. After each one, write down the single most surprising thing you learned. If nothing surprises you, you’re either asking the wrong questions or talking to people you already understand too well.

The best product managers treat discovery as a habit, not a project phase. Start building that habit today.

Frequently asked questions