Step 11: Build your interview guide
Phase 1: Preliminary Research · Step 11 of 13
Your hypotheses are ready to test. Before you talk to a single customer you need questions designed to surface whether your beliefs are right or wrong.
Article
Interview guides are not questionnaires. A questionnaire collects answers. An interview guide creates a conversation that surfaces truth. The difference is in the design. Questionnaires ask direct questions and get direct answers, which are often wrong because people tell you what they think you want to hear. Interview guides ask open-ended questions that invite stories, and stories contain the real data.
The structure of a good interview follows a deliberate arc. You start with broad context questions that establish who this person is and how they relate to the problem space. You move into specific experience questions that surface what they actually do today, not what they say they do. You then probe the pain points your hypotheses predict, looking for confirmation or disconfirmation. You close by inviting them to tell you what you missed.
The hardest discipline in interview design is avoiding leading questions. Every founder wants to ask questions that validate their idea. Questions like "Would you use a product that does X?" or "How much would you pay for Y?" feel productive but produce garbage data. People are polite. They will tell you your idea is great whether they believe it or not. Your questions need to surface behavior and experience, not opinions about hypothetical products.
Follow-up probes are where the real insights live. When a customer mentions a frustration, you need to be ready to go deeper. How often does this happen? What do you do when it happens? Have you tried to solve it? What happened? These follow-ups turn surface-level comments into rich data about the intensity and frequency of the problem.
Your interview guide should produce a thirty to forty-five minute conversation. Shorter than that and you will not get past surface-level responses. Longer than that and you are asking too much of someone who is giving you their time for free. Respect their time by being prepared.
Key Takeaways
- Why: Good questions make good interviews. Weak questions produce confirmation not insight.
- What: Use AI to convert your hypotheses into open-ended interview questions.
- Output: A draft interview guide with five to eight questions organized by topic.
AI Prompt
Act as a qualitative research specialist. Based on this hypothesis document: [paste your step 10 output here].
Build a draft interview guide for exploratory customer interviews. The guide should:
- Open with two to three rapport and context questions that establish the customer's situation without revealing what I am building
- Include one to two questions for each critical hypothesis that would surface whether the hypothesis is true or false
- Use exclusively open-ended questions that invite stories and examples rather than yes or no answers
- Avoid any question that leads the respondent toward confirming my hypothesis
- Include follow-up probes for the two or three most important questions
- Close with one question that invites the customer to share anything I have not asked about
The total interview should take 30 to 45 minutes. Do not include any questions that pitch my solution or ask whether the customer would buy something. Every question should surface what is true about their experience not what I want to hear.