Good customer discovery interview questions help you understand what someone actually does when a problem occurs. Ask about a recent event, the workaround they used, the people involved and the consequences. Save your product pitch until you have learned enough to decide whether the problem is worth pursuing.
The practical output is not a collection of compliments. It is a short evidence record that can change your next experiment. The question bank and example below are original editorial worksheets, not a validated survey instrument.
Choose one learning decision
Write the decision before recruiting participants. For example: “Should we prototype a way for operations teams to coordinate unresolved supplier requests?” That is more useful than “Would people buy our platform?” Identify whose work you need to understand and recruit people who have recently performed it.
Make room for disagreement. Interview someone who solved the problem comfortably, not only people who already like your idea. Tell participants what you are researching and how you will use their information. Ask permission before recording or retaining private materials.
Copy this customer discovery question bank
| Purpose | Question | Useful follow-up |
|---|---|---|
| Find a real event | Tell me about the last time you handled this situation. | What happened immediately before it? |
| Understand the sequence | Walk me through what you did. | Which step took the most attention? |
| Find the workaround | What tools, documents or people did you use? | What did you still have to do manually? |
| Learn the consequence | What happened when the work was delayed or incorrect? | How do you know that was the consequence? |
| Identify other participants | Who else needed to be involved? | Who made the final decision? |
| Test competing explanations | Was this unusual or part of a recurring pattern? | Can you describe another example? |
| Understand alternatives | What have you already tried to change? | Why did you continue or stop? |
Ask follow-ups in the participant's language. If they say the problem was “annoying,” learn what happened rather than translating annoyance into a willingness-to-pay claim. A hypothetical purchase answer is not a completed sale.
A worked fictional interview
A founder asks an operations manager about supplier-request coordination. The manager describes forwarding the same email to three colleagues, then maintaining a spreadsheet because nobody knew who owned the reply.
The founder's first interpretation is “They need AI email automation.” The evidence supports a narrower statement: ownership was unclear during a recent request. It does not yet establish that automation is the right solution.
| Evidence record | Fictional entry |
|---|---|
| Recent event | A supplier request required three colleagues |
| Existing workaround | Forwarded email plus a manually maintained spreadsheet |
| Observation | Reply ownership remained unclear |
| Interpretation | A visible owner may reduce coordination effort |
| Alternative explanation | The team may simply lack a shared process |
| Next test | Ask the team to try a named-owner request board |
Questions to rewrite before the interview
“Would you use our AI tool?” introduces your solution too early. Replace it with a question about the work. “Would you pay $50?” asks for a prediction. Ask what they currently buy, who approves that spending and what happened the last time they changed tools.
Do not turn discovery into a covert sales call. If you offer a demo afterward, make the transition clear and optional. Keep a polite response separate from a requested next conversation.
For the underlying method, explore Steve Blank's current customer-development teaching and UC Davis's customer-discovery training. Both emphasize learning from customers. The worksheet here applies that approach to your next conversation; no fixed interview count proves a market.
Give Aurora a synthesis job
Organize these permitted interview notes into recent events, workarounds, consequences, contradictions and unanswered questions. Link each observation to its source. Keep participant statements separate from my interpretations. Suggest what to ask next without inventing quotes, willingness to pay or market validation.
Aurora can prepare the synthesis from available context. Use Documents for approved source notes and Finta CRM for permissioned participant relationships and follow-up. The value is preserving what you learned, not converting every conversation into a positive signal.
Continue with our product-market-fit evidence checklist and go-to-market worksheet when the interviews have produced a concrete next decision.