A useful customer interview reconstructs a recent real event. Ask what triggered the problem, what the person did, which alternatives were considered, who influenced the decision, and what happened. Avoid pitching, leading questions, and hypothetical purchase predictions. Interviews reveal context and language; they do not establish market prevalence or willingness to pay by themselves.
Who this is for: Founders and small product or service teams using interviews to understand a customer problem before making product and demand claims.
- Ask for a recent specific story before asking for opinions about a concept or category.
- Follow the sequence of actions, constraints, alternatives, and decisions while preserving the participant's own words.
- Analyze contradictions and selection limits instead of counting interview comments as population statistics.
Recruit for relevant experience
Participants should have recent experience with the situation under study rather than merely fitting a broad demographic profile. The decision is which problem assumptions deserve a behavioral test and which should be revised or dropped. State assumptions and separate direct observations from interpretation.
Screen for the event, role, timing, current responsibility, and conflicts, then record how recruiting may bias the sample. Look for specific recent stories, artifacts, actual workarounds, decision participants, sequence, and contradictory cases. Record support, contradictions, unknowns, and the next decision this evidence can change.
Ask for a real story
Memory of a concrete event gives more usable detail than a general opinion about what people usually do. The decision is which problem assumptions deserve a behavioral test and which should be revised or dropped. State assumptions and separate direct observations from interpretation.
Start with the last occurrence and follow the timeline through trigger, action, alternative, obstacle, decision, and outcome. Look for specific recent stories, artifacts, actual workarounds, decision participants, sequence, and contradictory cases. Record support, contradictions, unknowns, and the next decision this evidence can change.
Probe without selling
Explaining the idea too early encourages politeness and changes the evidence being collected. The decision is which problem assumptions deserve a behavioral test and which should be revised or dropped. State assumptions and separate direct observations from interpretation.
Use neutral follow-ups, ask for examples and artifacts, tolerate silence, and avoid correcting language or defending the proposed solution. Look for specific recent stories, artifacts, actual workarounds, decision participants, sequence, and contradictory cases. Record support, contradictions, unknowns, and the next decision this evidence can change.
Synthesize with limits
Interview patterns can generate hypotheses but do not measure how common a behavior is in a market. The decision is which problem assumptions deserve a behavioral test and which should be revised or dropped. State assumptions and separate direct observations from interpretation.
Tag observations, preserve contrary cases, compare participant contexts, and write what the interviews cannot establish before choosing the next test. Look for specific recent stories, artifacts, actual workarounds, decision participants, sequence, and contradictory cases. Record support, contradictions, unknowns, and the next decision this evidence can change.
Investigate appointment scheduling problems
A founder wants to understand how independent tutors schedule recurring sessions before designing a scheduling tool. Every quantity in this worked example is a hypothetical input used only to show the method. It is not a benchmark, forecast, validation rate, or claim about likely results.
- Recruit tutors who recently changed, canceled, or rescheduled a recurring appointment and record the recruiting channel.
- Ask each person to reconstruct the most recent event, including messages, tools, participants, and the consequence of the change.
- Probe current workarounds and actual costs without showing screens or asking whether they would buy the proposed tool.
- Group observations by context, preserve contradictory stories, and select a behavior-based test for the strongest unresolved assumption.
Customer interview record
Use the same evidence fields while leaving room for unexpected language and events.
- Participant context, qualifying event, recruiting source, and possible selection bias.
- Trigger, timeline, actions, tools, people, alternatives, and decision point.
- Exact phrases, observed artifact, emotional language, and factual follow-up needed.
- Contradiction, missing perspective, interviewer influence, and confidence limit.
- Supported hypothesis, weakened hypothesis, unknown, and next behavioral test.
Common mistakes
- Asking whether someone likes an idea instead of reconstructing what the person recently did.
- Pitching during the interview and interpreting politeness as customer evidence.
- Reporting themes from a recruited group as validation rates or market-wide findings.
Try one
A participant says they would definitely buy an automated reminder tool. Write a better follow-up and explain the limit of the answer.
Ask about the last missed reminder, what the participant did before and after it, which tool was used, whether another solution was considered, and what consequence followed. This evidence can clarify behavior and stakes, but even a confident future statement does not establish an actual purchase. A strong response names uncertainty and avoids predicting outcomes from assumptions. Seek legal or research guidance when interviews involve sensitive information, vulnerable participants, recording, incentives, privacy duties, or regulated subjects.
Sources
- SBA market research and competitive analysis guideOfficial guidance on using market research and competitive analysis to understand customers and competitors.
- SBA business plan guideOfficial guidance on traditional and lean startup business plans.