Customer discovery interview template
A customer discovery interview template with screeners, severity ratings, and AI-probed questions on workflows, workarounds, and switching triggers.
Questions and setup
Customize the placeholders and apply the setup notes in the study builder before launch.
- 01
Multiple Choice
In the past 30 days, which best describes what you personally did in [Specific process or task]?
- Completed the task myself
- Completed it jointly with someone else
- Reviewed or approved someone else's work
- Managed the work but did not take part directly
- Received the output but did not take part
- Did not encounter this task
Setup: Qualify the roles with firsthand experience that match the research scope; by default, use the first three.
- 02
Multiple Choice
In the past 30 days, on about how many days did you personally do this task?
- 0 days
- 1 day
- 2-3 days
- 4-7 days
- 8 or more days
Setup: Screen out 0 days; consider screening out 1 day when the study requires frequent experience.
- 03
AI Question
Think about the most recent time in the past 30 days when you did this task. What were you trying to accomplish, and what happened step by step?
Setup: Allow up to two follow-ups. Stay with the specific incident and establish actors, tools, handoffs, duration, and outcome.
- 04
AI Question
Which parts, if any, required the most effort, created uncertainty, or had to be repeated? What happened at those points?
Setup: Allow one follow-up and permit a genuine “none” answer.
- 05
Rating Scale
Over the past 30 days, how much did issues with this process interfere with getting the intended result?
Setup: Use a 1-5 scale from Not at all to Completely, with Moderately at the midpoint.
- 06
AI Question
Thinking about the past 30 days, when did the issue occur and when did it not? What was different?
Setup: Allow one follow-up focused on conditions, exceptions, or frequency.
- 07
AI Question
In the past six months, what, if anything, have you changed, bought, built, or asked for to improve this process? For the most recent attempt, what happened?
Setup: Allow up to two follow-ups to establish effort, outcome, and why the attempt continued or stopped.
- 08
Multiple Choice
Which of these are part of how you currently complete this task? Select all that apply.
- Spreadsheets, documents, email, or manual steps
- One or more third-party products
- An internal or homegrown tool
- An outside service or provider
- Help from another person or team
- We have no consistent approach
Setup: Enable multi-select and the built-in Other response. Do not combine “We have no consistent approach” with another choice.
- 09
AI Question
In the past six months, have you seriously considered changing how you handle this task? What prompted that, what did you evaluate or try, and why did you change or stay?
Setup: Allow up to two follow-ups; distinguish casual interest from demos, trials, budget requests, or implemented workarounds.
- 10
AI Question
The most recent time the issue occurred, what consequences followed, if any? What could you estimate in time, delay, rework, money, or risk?
Setup: Allow one follow-up and ask for observed consequences rather than hypothetical value.
- 11
Open Text
Is there anything else about how you currently handle this that would help us understand your day-to-day?
This customer discovery interview focuses on recent behavior: what someone did, where the process interfered with the outcome, and what they tried to change. It qualifies firsthand experience, reconstructs one recent event, and uses explicit recall windows to avoid broad opinions about a hypothetical problem.
What this template measures
The template covers the current workflow, interference severity, conditions that trigger the issue, attempted solutions, the mix of tools in use, actual switching behavior, and observed consequences. An anchored interference scale provides a comparable measure without asking participants to decide how “big” the problem should feel.
The AI Questions ask about a specific recent instance: the last time the issue arose, what happened, and what consequences followed. Specific accounts provide stronger evidence of switching triggers than opinions about a hypothetical solution.
Who it's for
This template is for product managers, founders, and product marketers evaluating a problem before committing a roadmap or building a solution. It fits early-stage discovery for a new feature or product, an investigation of an adjacent problem, or a periodic check on a known pain point. Once you have a specific concept for participants to review, use a concept test.
How the AI moderator probes
The opening questions route out people without recent firsthand experience. From there, the moderator stays with a specific event before generalizing. If someone says they “just use a spreadsheet,” it asks what happened in the most recent instance without assuming the spreadsheet is painful. AI follow-up questions explains adaptive probing.
Tips before you launch
Replace the bracketed placeholders with the process or problem area you are studying. A specific prompt such as "approving vendor invoices over 5,000 dollars" is more likely to elicit a specific account than "the process you are exploring." Mention your product or idea only in the closing question, if at all, so the interview remains focused on current behavior. If you are still refining who counts as a qualified respondent, How to write screener questions that actually screen explains how to route the opening questions. After collecting a batch of interviews, run theme extraction across the transcripts before drawing conclusions from any single conversation.
Frequently asked questions
Who should you talk to for a customer discovery interview?
Talk to people with recent, direct experience of the problem. Do not recruit solely on hypothetical interest in a future product or general opinions about the problem area.
How is a customer discovery interview different from a feedback survey?
Use a survey when you know what to ask and want to measure it at scale. Use a discovery interview to identify what matters through open-ended, AI-probed questions.
How many discovery interviews do you need before patterns show up?
There is no fixed number. Continue interviewing within one well-defined segment until new interviews stop adding workflows or workarounds.
Should the interview mention your product or solution?
Mention your product or solution only at the end, if at all. Naming it early can pull answers toward what participants think you want to hear instead of their current behavior.
Full reference
Templates & the Library
Related templates
