Customer interviews: eight neutral questions on real productivity workflows

Good interview questions ask for a recent example, not a prediction about whether someone would buy an imagined tool. The eight prompts below help a researcher reconstruct a workflow and the decision around changing it.

The list was compiled on 5 September 2026 for research with actual or likely business-software users in England. A question was included when it could elicit observable behaviour, a constraint or buying responsibility. It excludes satisfaction scores, pricing research and usability tasks. The order follows a conversation rather than ranking the questions.

What to take away

1. What happened the last time you did this task?

Ask for a specific occasion and let the participant choose where to begin. Follow the event from trigger to completion. The GOV.UK guide to in-depth interviews recommends open, neutral questions and real stories instead of general accounts of how work should happen.

2. What information did you need at the start?

Find the source, format and person responsible for each input. Ask what happens when a record is late, incomplete or inconsistent. Do not request confidential customer or employee data; a blank example or verbal description may answer the research question.

3. Who touched the work before it was finished?

Map operators, reviewers, approvers and recipients. A hand-off may occur by message, spreadsheet, meeting or another system. Note who is accountable when the next person does not respond.

4. Where did you leave the main system?

This prompt can uncover paper notes, private reminders, duplicate entry and offline steps. Ask why the detour exists before treating it as a defect. It may be a workaround for access, speed or a client requirement.

5. How did you know the result was correct?

Look for checks, reconciliations, approvals and later corrections. If a participant says the process saves time or causes errors, ask how that is recorded. Their explanation is evidence from one account, not a measured market result.

6. What changed during the most recent exception?

Routine demonstrations hide the awkward cases. An urgent request, staff absence, failed integration or unusual customer condition can expose the permissions and recovery behaviour the product must support.

7. Who could approve a different way of working?

Separate interest from authority. Ask who controls the budget, security review, contract, implementation time and final acceptance. The answer may identify other roles for a later round rather than a sales lead.

8. What would you need to see before trying a change?

This closes on evidence rather than a purchase promise. A participant may need an export test, security information, colleague involvement or proof that an old process can still run during transition.

Run the session responsibly

Use the same discussion topics across participants but follow relevant details. The Service Manual's round-planning guidance asks teams to define objectives, arrange notes or recordings, collect informed consent and allow time for analysis.

Research records can contain personal information. GOV.UK's participant privacy guidance recommends explaining collection and use, holding only what is needed, restricting access and setting deletion arrangements. A private company should obtain current qualified advice for its own lawful basis, notices and suppliers.

After each interview, label direct observation, participant recollection and researcher interpretation separately. Compare contrasting cases before changing the product. Never invent a quotation or report a frequency that the recruitment and method cannot support.

If two colleagues share the task, consider interviewing them together so their hand-offs can be examined rather than reconstructed separately.

Before you act

Common questions

What kind of questions does the article recommend asking in an interview?

It recommends open, neutral questions that ask for a recent example rather than a prediction about whether someone would buy an imagined tool. The GOV.UK guide to in-depth interviews recommends real stories instead of general accounts of how work should happen.

How should a researcher handle personal information collected during interviews?

GOV.UK's participant privacy guidance recommends explaining collection and use, holding only what is needed, restricting access and setting deletion arrangements. A private company should obtain current qualified advice for its own lawful basis, notices and suppliers. Do not request confidential customer or employee data.

What should a researcher do after each interview?

Label direct observation, participant recollection and researcher interpretation separately. Compare contrasting cases before changing the product. Never invent a quotation or report a frequency that the recruitment and method cannot support. If two colleagues share the task, consider interviewing them together so their hand-offs can be examined.