# Conducting User Interviews

> Help plan and run customer interviews — question design, interview structure, research planning. Use BEFORE interviews to prepare. For debriefing AFTER a conversation, use discovery-debrief instead.

- Skill: `layneformalized225/conducting-user-interviews` (Agent Skill, multi-file: 3 files)
- Install (CLI): `npx skillmds@latest add layneformalized225/conducting-user-interviews`
- Raw SKILL.md: https://api.skillmd.com/api/skills/layneformalized225/conducting-user-interviews/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Product & Planning
- Author: Layneformalized225 (https://skillmd.com/u/layneformalized225)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/layneformalized225/conducting-user-interviews

---

## Company Context

Before helping, read `MEMORY.md` for: current wedge, ICP, competitors, PMF stage, system constraint.
Apply all frameworks to the user's specific company and stage (read from MEMORY.md).
Follow output preferences from USER.md (language, format, platform constraints).

# Conducting User Interviews

Help the user run better discovery conversations and extract real insights using techniques from 43 product leaders.

## How to Help

When the user asks for help with user interviews:

1. **Understand their goal** - Ask what they're trying to learn (validating a problem, testing a solution, understanding behavior, pricing research)
2. **Help them prepare** - Suggest questions, warn against common mistakes, help them find the right participants
3. **Coach on technique** - Share principles for getting honest, useful answers rather than polite validation
4. **Help analyze findings** - Assist in synthesizing what they learned into actionable insights

## Core Principles

### Collect stories, not opinions
Teresa Torres: "Interviewing is a grossly underestimated skill. If you're not collecting rich stories, you won't identify opportunities." Don't ask "What do you like?" Ask "Tell me about the last time you..."

### Only interview people who've taken action
Bob Moesta: "I only talk to people who've already tried to make progress. What made them try? Ignore 'bitching' (complaining)—look for 'switching' (actual behavior change)."

### Watch, don't just ask
Gustaf Alstromer: "The best way to understand problem intensity isn't asking—it's watching. Have them screen share and walk through their daily workflow. Look for pain they've normalized."

### Avoid pitching
Jeff Weinstein: "Don't start with 'Hi, I'm the CEO of X, we do Y, let me show you a demo.' What a wasted opportunity. Listen first. Use silence to let them open up."

### Falsify, don't validate
Judd Antin: "We don't validate, we falsify. We look to be wrong. Many PMs want to be right—they do user-centered performance, not real research."

### Never ask what they want built
Judd Antin: "A researcher who asks customers what they want is a bad researcher. Focus on understanding behaviors and problems—not having users design your solution."

### Probe for the emotion
Nan Yu: "My goal is to feel bad the same way customers feel bad. Dig past the feature request to find the underlying negative emotion motivating it."

### Drop the discussion guide
Bob Moesta: "Not having a script drives people crazy, but rigid guides prevent you from following meaningful threads. Use the Four Forces (push, pull, anxiety, habit) as mental framework instead."

### Right-size your sample
Shaun Clowes: "Between 7-14 interviews, you stop learning new things. Less than 7, not enough data. More than 14, diminishing returns."

### Expect 90% rejection
Gustaf Alstromer: "90% of people aren't early adopters. You need to reach 10 to find 1. Rejection isn't failure—it's filtering for the right users."

### Get direct exposure
Marty Cagan: "I wasn't allowed to make product decisions until I'd visited 30 customers. Those visits changed my life—I thought I knew our customers and I really didn't."

### Respond with extreme speed
Jeff Weinstein: "When a customer goes out of their way to share a problem, that's a gift. I'll leave a meeting to reply. Be 'text message friendly' with 5-10 power users."

### Interview the non-users
Mihika Kapoor: "The most insightful conversations are with non-users. Ask why they're not using your product—you'll find perception gaps users can't see."

### Test willingness to pay
Jeff Weinstein: "Have them send you a $1 invoice right now. The gap between 'willingness to pay' and actually paying is massive. This tests real commitment."

### Co-create with lighthouse users
Tanguy Crusson: "Work with 10 'lighthouse' users over months. Put them in Slack with your team. Involve engineers directly so they build empathy."

## Questions to Help Users

- "What are you trying to learn from these interviews?"
- "Are you interviewing people who've already tried to solve this problem?"
- "How are you recruiting participants?"
- "What's your opening question? (Make sure it asks for a story, not an opinion)"
- "How will you avoid leading questions?"
- "What will you do with the findings?"

## Common Mistakes to Flag

- **Leading questions** - "Don't you think X would be better?" just gets agreement
- **Asking about hypotheticals** - "Would you use this?" is meaningless; behavior matters
- **Pitching during research** - You're there to learn, not sell
- **Too few interviews** - 2 isn't enough; aim for 7-14
- **Delegating observation** - PMs and designers must be in the room, not reading reports

## Deep Dive

For all 64 insights from 43 guests, see `references/guest-insights.md`

## Related Skills

- Analyzing User Feedback
- Defining Product Vision
- Measuring Product-Market Fit
- Designing Surveys