Customer Discovery Questions That Reveal Behaviour, Not Opinions

The best customer discovery questions make it easier for a person to describe what actually happened. Weak questions invite encouragement, guesses and feature requests.

Your objective is not to persuade the interviewee that the idea is good. It is to understand the situation well enough to decide what deserves testing.

Begin with context

Start by learning how the person works:

• What is your role in this process?
• When does this task occur?
• Who else is involved?
• What outcome are you responsible for?
• Which tools or services do you use today?

Context helps you distinguish the user, buyer, approver and person affected by failure.

Ask about the last real event

Specific events produce better evidence than general opinions:

• Tell me about the last time this problem occurred.
• What triggered it?
• What did you do first?
• What happened next?
• How long did the process take?
• What was the result?
• What would have happened if you had done nothing?

Follow the timeline. If the answer becomes abstract, return to the event: “What happened on that occasion?”

Investigate frequency and severity

A vivid problem may still be rare. Ask:

• How often has this happened in the last three months?
• Which cases were most expensive or disruptive?
• Who notices when it goes wrong?
• What gets delayed?
• Is there a deadline, risk or penalty?
• How do you measure the impact?

Avoid asking the interviewee to score pain from one to ten without context. Observable consequences are more useful.

Understand the workaround

Workarounds reveal both demand and competition:

• How do you handle this today?
• What have you built or adapted?
• Which parts require manual effort?
• What is frustrating about the current approach?
• What works well enough that you would not want to lose it?
• Have you tried to replace it?
• Why did previous attempts fail?

Respect the current solution. If it survives despite obvious inefficiency, it probably has advantages your proposed product must match.

Explore spending and authority

Do not begin with “Would you pay?” Ask about real decisions:

• What does the current approach cost?
• Which budget pays for it?
• Who approves a change?
• What information does that person require?
• When is the budget reviewed?
• Have you purchased something similar?
• What caused that purchase to happen?

In B2B markets, the enthusiastic user may not control the money, data or implementation.

Ask for evidence and introductions

Where appropriate:

• Could you show me the template or tool you use?
• Is there a process document, ticket or report that captures this?
• Who else experiences the problem differently?
• Could you introduce me to the person who approves the current solution?
• Would you be willing to test a manual version?

An artefact or introduction is a stronger signal than praise because it requires some effort.

Questions to avoid

Avoid leading questions:

• “Don’t you find this frustrating?”
• “Wouldn’t an automated tool be better?”
• “Would you pay £20 for this?”
• “Do you think this is a good idea?”
• “Which features should I build?”

These questions reveal how the interviewee responds to your framing. They do not reliably reveal market behaviour.

Replace them with neutral prompts: “How do you handle it now?”, “What happened the last time?” and “What would need to be true for you to change?”

Separate problem and solution interviews

Early conversations should establish the problem, segment and buying context. Once that evidence converges, show a solution and observe whether people can use it, whether it fits the workflow and whether they commit to a next step.

Read problem validation versus solution validation for the distinction.

Debrief immediately

After each interview, record:

• facts and observed behaviour;
• direct quotations that preserve customer language;
• your interpretation;
• contradictions;
• unanswered questions;
• promised follow-ups;
• evidence strength.

Keep facts separate from conclusions. “The customer exports three spreadsheets weekly” is evidence. “The customer needs our dashboard” is an interpretation.

Use interviews inside a wider research process

Interviews reveal context that public research misses, but a few conversations can also mislead. Compare what you hear with reviews, competitor adoption, market structure and real commitments.

The startup market research guide shows how interviews fit with other sources. When you are ready to organise public evidence before interviewing, explore the SignalCraft methodology.