How to talk to users: a practical framework for founder interviews

Executive overview

Most founders pitch during user interviews instead of listening. The goal is to extract hard data about users' lives — not sell them on your product.

Three core traps: talking about your idea, asking hypotheticals, and talking too much. Avoiding all three is the foundation of a useful interview.

The best user interviews surface specific past experiences, not opinions about future features.

The Mom Test: three traps to avoid

• Pitching your idea instead of extracting information
• Asking hypothetical questions ("would you use this?") instead of probing past events
• Talking too much — listen, take notes, bring back hard data

Five questions for early customer interviews

1. What is the hardest part about doing [the thing you're trying to solve]?
2. Tell me about the last time you encountered this problem — extract the specific context
3. Why was that hard? Answers often reveal your future marketing copy
4. What, if anything, have you done to try to solve this problem? (No existing attempts = not a burning problem)
5. What don't you love about the solutions you've already tried? (This is your initial feature set)

Finding first users at the idea stage

• Start with yourself — test your interview strategy on your own experience
• Warm introductions: friends, coworkers, industry contacts
• Show up in person — a YC batch company cold-visited fire stations and got dozens of meetings
• Industry events work; guerrilla-style outreach costs nothing
• Take detailed notes, or bring a co-founder to do it
• Keep it casual; 10–15 minutes is enough for a first interview
• Start with one or three interviews — your process improves fast

Identifying your best first customer at the prototype stage

Use three numerical questions to rank potential customers:

1. How much does this problem cost them today? (revenue at stake or current spend)
2. How frequently do they encounter it? (more frequent = more receptive, more feedback loops)
3. Do they have the budget and authority to actually solve it?

Map answers across candidates — the best first customer scores highest on all three. A low-price-per-transaction customer with massive volume and a decision-maker's budget beats a prestigious but low-frequency one.

Iterating towards product market fit

• Ask users: "How would you feel if you could no longer use this product?" — very disappointed / somewhat disappointed / not disappointed
• If 40%+ say "very disappointed," the product is likely past the inflection point for organic growth
• Run this weekly to measure whether new features are adding or subtracting from fit
• Collect phone numbers at signup — when data raises a question, talk to one real person
• Don't design by committee: instead of asking what features users want, test willingness to pay before building
• Discard compliments (not specific) and hypotheticals (not grounded in past behavior) — both are bad data