conduct-user-interviews

# conduct-user-interviews Learn from specific past behavior, not predictions or praise. Keep recruitment, consent, notes, and synthesis transparent enough that another person can distinguish participant evidence from interpretation. ## Inputs - Write the research question, target participants, decision this research informs, and research owner. - Record incentives, accessibility needs, recording policy, retention period, and who may see raw notes. - Decide how findings will be reported without exposing participant identity unnecessarily. ## Procedure 1. Turn the product question into two or three learning goals about behavior, context, constraints, or decision criteria. 2. Define participant criteria and exclusions based on the real research question. Avoid using friends or convenient contacts as though they represent the target group. 3. Write a neutral guide that starts with a recent concrete event: "Tell me about the last time..." Follow with sequence, tools, workarounds, stakes, and what happened next. 4. Remove questions that pitch the solution, ask participants to predict purchases, or imply a preferred answer. 5. Recruit with an accurate invitation covering purpose, duration, incentive, recording, privacy, and the voluntary nature of participation. 6. Obtain consent before notes or recording. Let participants skip questions or stop without penalty. 7. During each session, listen for examples and ask for artifacts only with permission. Mark quotes, observations, and interviewer interpretations distinctly. 8. Debrief immediately. Capture surprising evidence, contradictions, unanswered questions, and potential influence from the interviewer. 9. Synthesize across participants into patterns with counts and counterexamples. Do not convert a few interviews into population percentages. 10. Connect each finding to the research decision, confidence, evidence excerpts, and an appropriate next test. 11. Return or delete raw data according to the consent and retention plan. ## Boundaries Do not interview minors or vulnerable groups without appropriate approval and safeguards. Never collect credentials, unnecessary sensitive details, or secretly recorded material. Do not quote identifiable participants publicly without explicit permission. ## Done - A research plan records learning goals, participant criteria, consent, incentives, and data handling - Session notes separate participant evidence, direct quotes, and researcher interpretation - A synthesis report includes patterns, counterexamples, confidence, and implications for a named decision - Raw recordings and notes are stored, shared, retained, or deleted exactly as participants were told Then use validate-a-product-idea to test the riskiest finding or run-a-survey to measure a pattern at larger scale.