How to ask customer interview questions that move beyond surface answers

Most customer interviews do not fail because the opening question was terrible. They fail because the first vague answer is accepted as if it were evidence. “It was confusing,” “we need better reporting,” and “I would probably pay for that” sound useful in the moment. On their own, they do not tell a team what happened, where the friction lived, or which decision deserves to change.

The interviewer’s real craft begins after the first answer. The job is to move from a polished summary toward a specific event without putting an explanation in the participant’s mouth.

The GOV.UK Service Manual recommends learning what users are trying to do, how they currently do it, and which problems they experience user-needs guidance.

That learning goal has to become answerable prompts. Interaction Design Foundation's interview guidance recommends translating the overall research question into concrete questions and asking about recent, specific experiences interview-question guidance.

Use a five-step follow-up ladder

Move down one level at a time. Each step earns a different kind of detail:

1. Claim: “Onboarding took too long.”
2. Example: “Tell me about the last time that happened.”
3. Sequence: “What did you do first, and what happened next?”
4. Evidence: “What did you open, send, or ask for?”
5. Meaning: “Why did that part matter?”

Do not force every answer through all five steps. Stop when you can reconstruct the event, understand the constraint, and explain the consequence. A clean three-minute account of one real attempt is often more useful than ten quick opinions about the product category.

See the difference in a worked example

Imagine you are studying why new team leads struggle to share a weekly customer update.

The participant begins with: “The reporting is pretty manual.” A leading interviewer might ask, “Would an automated dashboard solve that?” That question introduces both the diagnosis and the solution. Agreement would be easy, but it would reveal very little.

A neutral sequence sounds different:

• “Tell me about the most recent update you prepared.”
• “What did you have in front of you when you started?”
• “What did you do after opening the spreadsheet?”
• “Where did you have to stop or ask someone for help?”
• “What happened because the update went out late?”

Now the phrase “manual reporting” can unfold into a mechanism: the team lead exports three files, copies account names into a private spreadsheet, messages an analyst to explain two mismatched totals, and delays the update until the numbers reconcile. The opportunity might be better source traceability or a clearer handoff, not another dashboard.

That example is illustrative, not a customer finding. Its purpose is to show what a useful interview note contains: an event, actions, artifacts, constraints, and consequences.

Choose neutral probes deliberately

• Ask “What happened next?” before suggesting an explanation.
• Ask for the most recent example rather than a general average.
• Separate what the participant saw from what they inferred.
• Let a pause do some work; do not answer your own question.
• Repeat the participant’s own noun when you need clarification: “You called it a handoff. What was handed over?”
• Ask about a contrast: “Was the previous time different?”
• Ask for the decision point: “What made you choose that route?”

Neutral does not mean robotic. A short acknowledgement such as “That sounds frustrating” can be human without claiming to know the cause. The danger begins when empathy becomes an explanation: “So the real problem was that the dashboard was unreliable?”

Match the probe to the missing evidence

Use this quick diagnostic while listening:

The table is not a script to recite. It is a listening aid. Ask only the question that fills the most important gap.

Avoid four common interview traps

Treating hypotheticals as behaviour

“Would you use this?” invites optimism, politeness, and imagination. If the decision depends on likely adoption, first understand what the participant does today, what triggers the need, and what they already pay in time, effort, or money.

Collecting quotes instead of mechanisms

A vivid sentence is easy to remember. Before elevating it, ask what produced the feeling and what the participant did next. The mechanism makes the quote interpretable.

Turning the guide into a checklist

If the interviewer is racing to ask every prepared question, unexpected evidence has nowhere to go. Protect the learning goal, but spend time on the event that can actually change the team’s understanding.

Asking “why” too early

“Why?” can produce a neat explanation before the event is clear. Sequence questions are often easier to answer and less likely to invite rationalisation. Once the sequence is grounded, ask what the decision meant.

Prepare a guide that leaves room to listen

For a 20-minute conversation, a useful guide can be compact:

1. Context: What was happening when the need appeared?
2. Recent event: Walk me through the last attempt.
3. Friction: Where did the process slow, change, or require help?
4. Decision: What alternatives did you consider, and what shaped the choice?
5. Consequence: What did the outcome affect?
6. Reflection: What would a meaningfully better result look like?

Write two or three optional probes under each section, not ten mandatory questions. That preserves consistency across interviews while giving the interviewer permission to follow the evidence.

End with a usable note

Before ending, silently test whether you could explain the episode to a colleague who was not on the call. Capture:

• the situation and trigger;
• the participant’s goal;
• the actions in order;
• the tools, documents, or people involved;
• the obstacle or trade-off;
• the consequence;
• the participant’s own interpretation;
• what remains uncertain.

Then write one observation before writing an insight. “The participant exported three files and asked an analyst to reconcile totals” is an observation. “Team leads lack confidence in source consistency” is an interpretation that still needs comparison with other evidence.

The standard is not depth for its own sake. It is enough concrete detail to distinguish what happened from what the research team thinks it means.

Next, choose the right study shape in Which customer research study should you run?.