Customer Interview Questions for Startup Validation
Good customer discovery questions reconstruct a real experience. Ask what happened, what the person did and what consequence followed. Avoid opening with your solution or asking people to predict whether they would buy an imagined product.
Interviews explain context and motivations. They do not, by themselves, establish willingness to pay or the prevalence of a problem across a market. Use what you learn to design the next appropriate test.
Recruit for relevant experience
Define participants by a recent task or situation, rather than a broad label such as “business owner.” If you are exploring employee onboarding, recruit people who recently managed it and include different outcomes: smooth transitions, difficult ones and abandoned purchases.
Separate users, buyers and approvers. A person who enjoys a prototype may have no control over purchasing. Avoid relying entirely on friends or people already enthusiastic about your idea.
Explain the session’s purpose, approximate length and how information will be used. Obtain permission before recording. Offer an accessible way to participate and allow people to skip questions. Do not ask participants to reveal confidential customer or employer information.
Screen participants without leading them
Use a short recruitment screen focused on relevant experience. Ask when the person last completed the task, their role and whether someone else controlled the decision. Do not reveal the answer you hope to hear through a question such as “Do you struggle with slow reporting?”
For a handover-report service, a suitable screen might ask whether an agency completed a client project recently and who prepared the final materials. A person who has never handled a handover can still offer an opinion, but that opinion should not be presented as evidence from the target workflow.
Record the recruitment source and important exclusions. If every participant comes from one professional community, the pattern may reflect that community’s tools or practices. Recruit near the segment boundaries when you need to understand whether a proposed criterion actually predicts a different need.
Use a discussion guide, not an interrogation script
Begin with a neutral introduction and a few context questions. Then follow a specific recent event. The prompts below are original examples to adapt to your research question.
| Purpose | Useful prompt | Follow-up |
|---|---|---|
| Establish context | What is your role in this task? | Who else is involved? |
| Anchor in an event | Tell me about the last time you did it. | When was that? |
| Reconstruct the process | What happened first? | What happened after that? |
| Find friction | Where did the process become difficult? | What did you do to get past it? |
| Understand consequences | What happened because of the delay or error? | Who noticed or had to respond? |
| Explore alternatives | What else did you consider or try? | Why did you keep or reject it? |
| Understand buying | How was the current approach approved? | Who controlled the budget? |
Ask for concrete examples when an answer is general. “It is always frustrating” can lead to “Can you walk me through the most recent time?” Silence and patient follow-up often provide more detail than suggesting an answer.
Replace leading questions
“Would automated reminders save you time?” suggests both the solution and the desired outcome. Ask how reminders are handled today, what takes effort and what happens when one is missed.
“Would you pay $20?” invites a hypothetical answer without a complete offer or buying context. Investigate previous spending and approval first. Test an actual, accurately described offer separately when it is ready.
Do not debate participants who disagree. Contradictory evidence may expose a segment difference or a mistaken assumption. If you want feedback on a prototype, clearly move to a separate part of the session after discovery.
Follow an answer without inserting your theory
Illustrative interview exchange, not a real participant transcript:
Interviewer: Tell me about the last client handover you prepared.
Participant: It took longer than it should have.
Interviewer: What happened that made it take longer?
Participant: We were waiting for two files from a contractor.
Interviewer: What did you do while you were waiting?
The follow-up explores the sequence. Jumping to “Would reminders fix that?” would introduce a solution before understanding whether the difficulty was communication, permissions, an unclear agreement or missing work.
If participants say they “always” experience a problem, ask for the last example and a time it went smoothly. The contrast can reveal the conditions behind the difficulty. If they cannot recall an incident, record that uncertainty rather than supplying one for them.
When the story involves several people, clarify who knew what and when. A manager’s account of a colleague’s behavior is second-hand evidence. If that role is central to the decision, consider interviewing someone who performs it directly.
Investigate money without a hypothetical sales pitch
Ask what the current approach costs, who approves spending and how similar purchases happened. Participants may not know exact amounts or may be unable to share them. Accept that boundary and note what remains unknown.
Useful prompts include “What did you try before this?”, “Was there a budget already allocated?” and “What would have needed approval?” These questions investigate the buying process. They do not establish that the participant will purchase your proposed offer.
If you later introduce a pilot, clearly separate that part of the conversation. Explain the scope, terms and what exists today. Record the actual requested action and response separately from discovery notes. Otherwise polite feedback about a concept can be mistaken for a commitment.
Keep evidence separate from interpretation
Illustrative example: a manager describes exporting a spreadsheet every Friday and correcting employee names manually. The observation is a repeated correction step in that workflow. “Every manager needs a new HR platform” is an unsupported interpretation.
Use notes with participant codes, context, observed or reported behavior, and tentative implications. If quoting publicly, obtain appropriate permission and preserve the meaning. Limit identifiable details to what the research requires.
After each session, record what changed in your understanding. Across sessions, group patterns by person and situation. Repeated mentions by one participant should not be mistaken for agreement among many participants. Keep exceptions visible.
Synthesize by participant and situation
Use the interview notes template to keep observations traceable. Create a table with participant code, situation, current alternative, consequence, buying role and unresolved questions. Do not turn every sentence into an equally weighted vote.
Compare similar situations first. If agencies preparing one-off handovers describe different needs from agencies producing monthly reports, preserve that distinction. A combined average persona could obscure the two actual workflows.
Mark the strength of each statement. “Participant described a missed deadline” differs from “we observed the reporting task” and from “the team believes deadlines are the main buying trigger.” Store interpretations separately so later evidence can revise them without rewriting the original account.
Avoid common analysis shortcuts
Do not conclude that a topic is unimportant because it did not arise in an interview that never explored the relevant situation. Likewise, a participant agreeing after a leading question is weaker evidence than a spontaneously described experience. Keep enough context to judge how the answer was elicited.
An AI summary can assist organization only if you verify it against authorized source material. Check participant counts, omitted contradictions and whether the model invents motivations. Do not upload sensitive recordings or notes without assessing the intended data use and controls. The researcher remains responsible for the interpretation and the next decision.
Decide what to test next
Small qualitative rounds can identify useful problems and language; they cannot establish reliable population percentages. Look for a coherent problem within a reachable group, then choose a test of behavior or commitment.
Use the ideal customer profile guide to refine recruitment, market research to add external context and idea validation to select the next experiment. Start by piloting your discussion guide with one person and removing questions that invite the answer you hoped to hear.