Customer interviews

People are polite: ask them whether they like your idea and most will say yes. This skill designs interviews about their life and their past behavior, so that the answers stay useful even when the idea is wrong.

When to use

• Before building, to check that the problem exists and matters to the target.
• After a launch, to understand why users stay or leave.
• Not to measure how common something is: use survey-design once the interviews have shown what to measure.

Inputs

1. The target customer and the problem to explore.
2. The three things the founder most needs to learn.
3. How to reach interviewees: network, communities, professional networks, shops, events.

Method

1. Set the learning goal: the three questions whose answers would change the plan. Keep the idea itself out of the first part of the interview.
2. Recruit the right people: real members of the target, not friends who want to help. Look for variety (sizes, contexts, regions) and note who refused: refusals are data too.
3. Write the guide around past, specific behavior, following Fitzpatrick's rules: talk about their life rather than the idea, ask about specific past events rather than opinions or the future, and listen more than you talk.
   • Useful: "Tell me about the last time you had to… What did you do? What did it cost you?"
   • Useless: "Would you use an app that…?", "How much would you pay for…?", "Do you think it's a good idea?"
4. Run it: 20 to 40 minutes, ideally one person asking and one taking notes, with consent before taking notes or recording. Dig where the person shows emotion or has already spent money or time.
5. Filter the noise: compliments, promises about the future ("I would definitely buy it") and generic statements ("I always…") are not evidence. Bring the person back to a concrete past example.
6. Ask for a commitment when it fits: an introduction to someone else, a follow-up call, a pilot, a pre-order. A commitment is the strongest signal an interview can give.
7. Debrief within 24 hours: facts, quotes, problems ranked by intensity, current solutions and what they cost, surprises, commitments. After every five interviews, look for patterns and update the assumptions (idea-validation).

Output

### Interview guide: <target>
Learning goals: 1. … 2. … 3. …
Warm-up (2 min): …
Their world (15 min): "Tell me about the last time…"
Current solutions (10 min): what they use, what it costs them
Wrap-up: "Who else should I talk to?" and a commitment ask

### Debrief: interview <n> (<role, context>)
| Fact (what they did) | Quote | Pain intensity (1 to 3) |
|---|---|---|
Current solution and its cost: …
Commitment obtained: …
Assumptions to update: …

Quality bar

• No question asks the person to predict their future behavior or to rate the idea.
• Every insight in a summary is backed by facts or quotes, with the number of people who said it.
• What many people said is kept apart from what one person said.
• Personal data is kept to a minimum and stored with consent (GDPR in the European Union).
• Written in the user's language, with the terms below.

Terminology

Sources

• Fitzpatrick, R. (2013). The Mom Test: How to Talk to Customers and Learn If Your Business Is a Good Idea When Everyone Is Lying to You. CreateSpace.
• Portigal, S. (2013). Interviewing Users: How to Uncover Compelling Insights. Rosenfeld Media.