Customer interview

Ask about their life, not your idea, and let them tell you the problem.

The trap in every early conversation is that people are kind. Ask "would you use this?" and most will say yes to be nice. Friends and your existing network are the worst for this: their encouragement feels like validation and isn't. Useful conversations ask about what someone already did, what it cost them, and what they tried, because the past doesn't lie to be polite.

This skill follows the principles Rob Fitzpatrick set out in The Mom Test, and a second idea: the best sales conversation is one where the customer reaches the conclusion themselves. You ask, they explain the problem in their own words, and the product becomes the obvious answer without you pitching it.

What you need

1. Who you're talking to, and why you think they fit.
2. What you most need to learn. One or two unknowns, not ten.
3. Whether this is discovery (before or around building) or a buying conversation (they may pay soon).

Steps

1. Pick the one thing you must learn. Write it as a belief that could be wrong. "Solo founders spend at least two hours a week hunting for customers by hand." Every question should test it.

2. Write questions about the past. Use this shape:

   • "Tell me about the last time you..."
   • "What did you do about it?"
   • "What else have you tried? What happened?"
   • "What did that cost you, in money or time?"
   • "Who else was involved in deciding?"
   • "What are you using for it right now?"

3. Strike the bad questions. Delete anything that asks for an opinion about your idea ("would you", "do you think", "how much would you pay") or describes your product. Those come at the end, if at all.

4. Plan the path, not a script. Order the questions so each answer opens the next. Start broad on their situation, narrow to the moment the problem bites, then to what they've done about it. If the problem is real, they'll describe it before you mention it. When they do, ask "why was that hard?" and listen.

5. Agree how you'll know. Before the conversation, write what answer would prove you wrong. If you can't be wrong, you're not learning anything.

6. After the conversation, write it up the same day. Separate what they did (facts) from what they said they'd do (opinions) and what you hoped (bias).

7. Look for the commitment. A useful conversation ends with them giving something: time for a second call, an introduction to someone else with the problem, early access in exchange for feedback, or money. Compliments are not commitments.

Output

Before the conversation:

What I must learn: <belief that could be wrong>
It's wrong if: <answer that would change my mind>

Questions (in order):
1. ...
Follow-ups when the problem appears:
- "Why was that hard?"
- "What did you do next?"
Questions to avoid: <the opinion questions I'm tempted to ask>

After the conversation:

Facts (what they did):
Their words for the problem (quote exactly):
Opinions (weight lightly):
Commitment offered: <none / time / introduction / money>
What this changes: <decision>

Rules

• Talk less than a third of the time.
• Never pitch in a discovery conversation. If they ask what you're building, tell them briefly at the end.
• Write down their exact words. Those phrases become your homepage, your pitch and the words you search for when you look for more people like them.
• A paying stranger is the strongest signal you can get. Free users and friends tell you much less.
• If you hear the same problem described the same way from three unrelated people, that's a pattern worth acting on. One enthusiastic person is not.

Questions that work for founders looking for their first customers

• "Walk me through how you found your last customer or user."
• "What have you tried to get people to notice what you built? Which one did anything?"
• "When did you last talk to someone who might pay you? How did that start?"
• "What's the part of this you keep putting off?"

With MyPip (coming)

When the MyPip MCP is live, this skill will be able to pull the people MyPip has introduced you to, with what they said publicly about the problem, so each conversation starts from their own words.