Skip to content

How to interview a customer after a feature request

Use a short follow-up interview to understand the task behind a feature request, avoid leading questions and turn the conversation into a clear next step.

Two people having a conversation across an office desk
Photo: Mina Rad / Unsplash · Unsplash License

When a customer asks for a feature, a follow-up interview can reveal the task they are trying to complete, what currently gets in the way, and how much the problem matters. Start with a recent example. Leave the proposed solution open until you understand the situation.

A request such as “add a PDF export” might come from someone who needs a printable document, someone who sends evidence to an auditor, or someone whose client cannot open a shared link. Those situations may need different solutions. The examples in this guide are fictional interview scenarios, not customer research results.

Choose the decision the interview should inform

Before inviting the customer, write down what you need to learn. For the PDF request, that could be: “Understand how customers hand reports to people outside their organization.” That question leaves room to discover access restrictions, formatting requirements, or an existing workaround.

Avoid defining the research question as “confirm that PDF export is useful.” You would be asking the customer to validate a decision already made. If the team has committed to PDF export and only needs to test its design, say so and plan an appropriate usability session instead.

GOV.UK's guidance on in-depth interviews recommends a discussion guide with topics and follow-up questions, and a focus on real examples. The structure below is one way to apply that approach to a product request.

Send an invitation that sets expectations

Explain why you are asking to talk, what the conversation will cover, and how much time you propose. Make clear that discussing the request does not mean the feature has been approved.

For example: “You mentioned needing to send a report as a PDF. Could we spend 20 minutes walking through the last time you shared a report? We are trying to understand the workflow before deciding what to change.” The duration is a suggested format for this narrow question, not a rule for every research interview.

If you want to record the call, ask for consent and explain how the recording will be used. Give the person a way to participate without a recording. Do not ask them to expose a real client's personal or confidential information during a screen share.

Walk through the last real attempt

Start with the event that led to the request. Ask the customer to describe what they were doing and what happened next. Useful prompts include:

  • “When did you last need to send a report outside your team?”
  • “Who needed it, and what were they going to do with it?”
  • “Can you walk me through how you shared it?”
  • “Where did you have to stop or do extra work?”
  • “What did you do instead?”

Let the answer determine your next question. If the person says the recipient could not open the link, find out what happened when they tried. A sign-in requirement and an expired link are different obstacles.

Ask about consequences without supplying the answer. “What happened because of that?” leaves more room than “Did that waste a lot of time?” If the customer mentions a delay, ask how they handled it and whether the same situation has occurred again.

Separate evidence from a plausible story

Keep notes in two parts. Record what the person described or demonstrated, then record your interpretation separately. This prevents an assumption from turning into a customer quote as the request passes between teams.

For example:

Observation from the fictional interviewInterpretation to investigate
The recipient needed a document to attach to a procurement recordA downloadable file may be a requirement of the recipient's process
The customer copied the report into a document editorThe current export or sharing flow did not meet this particular need
The customer sends one report to this recipient each quarterThis case alone does not establish broad demand for weekly PDF exports

Keep the connection to the original request so a teammate can ask a follow-up question. Remove identifying details before sharing a summary more widely. A public feedback board is not an appropriate place for private research recordings or customer documents.

Ask about the proposed feature after the workflow

Once the situation is clear, return to the request. “You suggested PDF export. What would the file need to include?” can expose requirements that the original message left out.

Then ask what would make the solution unusable. The recipient might require selectable text, a particular page layout, or a record of the reporting period. Avoid promising these details on the call. Write them down for the team to assess.

If you show a prototype, describe it as a concept and give the customer a realistic task. Their ability to complete that task gives you different evidence from their opinion of a screenshot. For guidance on that method, see GOV.UK's moderated usability testing guide.

Finish with a clear next step

Summarize your understanding in plain language and invite a correction: “The main difficulty is getting a readable file into your client's procurement system. Have I understood that correctly?” This is a check of what you heard, not a request for the customer to endorse your proposed implementation.

Explain what happens next. You may need to compare the request with other reports, test an existing workaround, or discuss a technical constraint with engineering. Give an update date only if someone is responsible for meeting it.

After the call, attach a concise summary to the appropriate internal record and keep the public request's status accurate. Our feature request template can help capture the task, obstacle and expected outcome. Productroad's feedback boards keep the request and discussion together while the team decides what to do.

Use the interview alongside other evidence

One detailed conversation can explain a problem well. It cannot establish how often that problem occurs across your customer base. Look for other affected users, relevant support requests and observable product behavior before estimating reach.

Include people whose experience may differ from the original requester. An administrator, a report author and an external recipient can encounter different parts of the same workflow. People who stopped using the feature may also be missing from a voting board.

When you have enough evidence to compare options, bring the findings into prioritization. After a decision or release, return to the people who helped you understand the problem. A useful feedback loop includes telling them what changed and checking whether it helped.

Productroad helps you collect user feedback, prioritize it, make a roadmap of a project and inform customers via changelog about new cool features.

Free 14 day trial. No credit 💳 card required.