Start with a learning goal, not a sales goal
Customer discovery interviews help you understand how people handle a task, experience a problem, and make decisions. They are not demonstrations or opportunities to collect compliments. A useful conversation may reveal that your proposed problem is uncommon, already solved, or not important enough to justify a change.
Before recruiting anyone, write down the decision the interviews should inform. For example: should your team investigate expense reporting for small agencies? Then identify the assumptions behind that decision, such as who prepares reports, where delays occur, and whether those delays have meaningful consequences. Keep each assumption separate so one memorable story does not appear to validate the entire business idea.
Turn those assumptions into topics rather than statements to confirm. Instead of asking, “Do manual expense reports waste your time?” plan to explore the participant’s latest reporting cycle, tools, handoffs, and outcomes. This leaves room for the process to work well.
Recruit people with relevant experience
Interview people who have recently encountered the task or situation you want to understand. Someone who sounds like your target customer but has never performed the relevant work may offer thoughtful opinions without providing evidence about actual behavior. Use simple screening questions about responsibilities and recent experience, not agreement with your problem statement.
Avoid recruiting only friends, enthusiastic followers, or people already interested in your solution. Their experiences can still be useful, but their relationship with you may encourage supportive answers. Where relevant, include people who use competing tools, rely on manual workarounds, or have decided the task does not deserve additional investment.
At the start, explain that you are learning about their experience, not evaluating their performance. Ask permission before recording and make declining easy. Introduce the subject without presenting your preferred conclusion: “I’m learning how small teams handle expenses,” rather than “I’m researching why expense software is frustrating.”
Ask about a specific past event
Begin with a concrete invitation: “Walk me through the last time you submitted an expense report.” Follow the sequence from the initial trigger to completion. Ask what the participant did, which tools they used, who else was involved, and what happened next. Specific episodes provide more useful detail than broad descriptions of what someone usually does.
Replace leading questions with neutral alternatives. Instead of “Was the approval process confusing?” ask, “How did approval work?” Instead of “Would automatic reminders help?” ask, “How did you keep track of outstanding approvals?” Instead of “How much time did you waste?” ask, “About how long did it take, and what affected that?”
Avoid assuming that a problem exists. “What problems did you encounter?” can nudge someone to find a complaint. A more balanced prompt is, “What went smoothly, and what, if anything, was difficult?” If an answer is vague, ask for an example rather than suggesting one yourself.
Follow up without supplying the answer
Neutral follow-ups include “What happened next?”, “Can you tell me more about that?”, and “What do you mean by complicated?” Use the participant’s language, but do not intensify it. If they describe a process as inconvenient, do not summarize it as a major operational failure.
Allow brief silences. Filling every pause with possible answers can steer the conversation toward your assumptions. Ask one question at a time, and avoid rewarding particular responses with enthusiasm. “That’s exactly what we thought” signals which answers you want to hear.
Explore consequences without exaggerating them. Ask, “What happened because of that delay?” or “How did you handle it?” When appropriate, invite participants to describe a document or workflow, without requesting confidential information. Concrete examples can clarify a story, but participants should control what they share.
Separate evidence of demand from polite interest
Questions about future behavior, such as “Would you use this?” or “Would you pay for that?”, invite speculation. A positive answer may reflect courtesy or curiosity rather than a genuine intention to change. First investigate what the person has already tried, purchased, abandoned, or worked around.
Ask, “Have you looked for another way to do this?” If yes, explore what triggered the search, what alternatives they considered, and why they chose or rejected them. If not, investigate that too. An annoying task may still be too infrequent or low-priority to motivate action.
If you need feedback on a concept, place it after the discovery discussion and clearly treat it as a separate activity. Reactions to your proposal should not be mixed with evidence gathered before participants knew what you wanted to build.
Review patterns and preserve contradictions
After each interview, separate observations from interpretations. “The participant copied receipts into a spreadsheet” is an observation. “They need our automation product” is an interpretation. Record the context, consequences, workarounds, and unanswered questions while the conversation is fresh.
Compare interviews by relevant characteristics, such as role, workload, and purchasing responsibility. Preserve contradictory accounts rather than averaging them away; they may reveal distinct customer groups. Interviews can uncover patterns and explanations, but they do not establish how common a need is across an entire market.
End the review with a decision and a next test. You might narrow the target audience, investigate an alternative explanation, or test a prototype against an existing workflow. The goal is not unanimous approval. It is a clearer understanding of which assumptions remain credible and what evidence you still need.
