The first 3 interviews are free · Start free or see the prices

Customer Discovery Interview: Discussion Guide and Questions

Use this guide before you write a line of code. It finds how people deal with a problem today and what it costs them. It also checks if they would pay to fix it.

Length
30 min
Questions
15
People
12 to 20
Sections
5
Run this study

What this guide finds out

  • How people do the task today, step by step, in their own words.
  • Which part of the task causes the most pain, and how often it happens.
  • Which workarounds people use, and what these cost in time or money.
  • Where the problem ranks against the other problems people have.
  • Who would decide to pay for a fix, and from which budget.

When to use it

  • You have an idea for a product but no proof that the problem exists.
  • You want to choose between two or three possible target markets.
  • Your team argues about the problem, and nobody has talked to customers in the last month.

Who to talk to

People who do the task that your product idea is about, at least once a month. Talk to 12 to 20 people. Each interview takes about 30 minutes.

The discussion guide

Each question has a probe note. The note tells the moderator, a person or an AI, what to do with the answer.

01

Warm-up and context

3 min

Learn the role of the person and where the task fits in their work.

  1. Tell me about your role and what a normal week looks like for you.

    ProbeIf the answer is general, ask what they did on Monday of this week.

  2. Which part of your work takes more time than you think it should?

    ProbeRecord the answer. Do not lead the person toward the problem of the study.

02

The last time it happened

10 min

Get a full story of one real occurrence of the problem.

  1. Tell me about the last time you had to do this task. Start from the beginning.

    ProbeAsk "What happened next?" until the story ends. If they describe what they usually do, ask about the last specific time.

  2. Which tools, files or people were part of it?

    ProbeAsk for the names of the tools. Ask what each person did.

  3. Which part of that was the hardest?

    ProbeAsk why it was hard. Then ask what happened as a result of the problem.

  4. When did it happen before that time, and how often does it come up?

    ProbeAsk for a count over the last month. Do not accept "often" without a date or an example.

03

Workarounds and cost

8 min

Learn what people do about the problem now and what it costs them.

  1. What do you do today to get around that problem?

    ProbeAsk them to describe the workaround step by step. A spreadsheet, a script or a manual check is a strong signal.

  2. How much time or money does the current way cost you?

    ProbeAsk for an estimate in their own units, such as hours per week. Do not suggest a number.

  3. Have you tried to fix this problem before? What happened?

    ProbeIf they tried a product, ask its name and why they stopped. If they never tried, ask why not.

  4. Who else in your team or company has this problem?

    ProbeAsk for the roles, not the names. Ask if those people feel it more or less than they do.

04

Priority and budget

6 min

Learn if the problem is important enough to pay for.

  1. Where does this problem rank among the other problems you have at work right now?

    ProbeAsk which problems rank higher and why. A problem that is not in the top three seldom gets a budget.

  2. If this problem were gone tomorrow, what would change for you?

    ProbeListen for a concrete result, such as time saved or fewer errors. If the answer is vague, ask for one example.

  3. If you decided to pay for a fix, who would approve it, and from which budget?

    ProbeAsk if they have bought a tool for a similar problem. Ask how long that took.

05

Close

3 min

Find gaps in the guide and more people to talk to.

  1. What should I have asked you about this topic that I did not ask?

    ProbeIf they give a new topic, ask one follow-up about it.

  2. Which other type of person should we talk to about this problem?

    ProbeAsk for a role or a type of company, not a name or contact details. Thank them and end the interview.

Example finding

ExampleA spreadsheet workaround

Operations managers rebuild the weekly report by hand in a spreadsheet. The problem costs them time each week.

“Every Friday I export three files and paste them into one sheet. It takes me most of the afternoon.”
In a User Evaluation study report, each claim links to the clip where a participant says it.

Mistakes to avoid

  • Do not describe your product idea. When people hear the idea, they start to give polite feedback instead of facts.
  • Do not ask "Would you use a tool that…?" People say yes to future questions and then do not buy.
  • Do not count compliments as proof. Count past actions: money spent, workarounds built, tools tried.
  • Do not stop after five interviews that agree with you. Talk to people who do not have the problem too.

Analyze the answers

Analysis templateOpportunity Solution Tree Template: From Interviews to ExperimentsUse this template to connect a product outcome to what you heard in interviews. Each node type has a rule for what belongs in it, so opportunities do not turn into solutions in disguise.

Questions

A customer discovery interview is a conversation with a possible customer before you build a product. It finds out if a problem is real, how often it happens and what people do about it now.

Start with 12 to 20 people in one market. If you test more than one market, plan about 8 people for each one. Stop when new interviews do not show new problems.

No. Keep the idea for a later concept test. In discovery, the goal is to learn the problem. An idea in the conversation makes people focus on your solution.

Yes. The questions ask about past events, which an AI moderator can follow with probe notes. Because it asks the same questions each time, it is easy to compare different markets.

Let the AI moderator ask these questions

The AI moderator interviews screened people by voice or text. Each claim in the study report links to its clip. The first 3 interviews are free.

Run this study