Customer discovery interview questions for founders

What should you ask in a customer discovery interview?

Ask someone to describe the last time they faced the problem you want to solve. Follow the sequence of what happened, what they tried, and what it cost them. An interview helps you understand a problem and choose a test. A positive conversation does not show that someone will buy your product.

Start with a question you need to answer. For example: "How do independent consultants follow up when a potential client stops replying?" Speak with people who have recently done that work. A friend who likes your idea may have nothing useful to say about this particular problem.

This guide is a suggested interview plan, not a report of customer research. The GOV.UK guide to planning user research also recommends connecting specific research questions to specific user groups.

Explain the conversation before you begin

Tell the participant why you are talking, roughly how long it will take, and how you will use the notes. Ask permission before recording. Give them room to skip a question or end the conversation.

You can say: "I am trying to understand how consultants handle client follow-up. This is a research conversation. I will take notes for my own planning, and you can skip anything you do not want to discuss. Is that okay?"

Keep private details out of shared examples. Ask separately before publishing an identifiable quote. Use participant labels in your working notes rather than copying unnecessary contact information into every experiment record.

Questions about a recent experience

Use these as prompts, rather than a checklist you must finish:

1. When was the last time you dealt with this problem?
2. What were you trying to get done?
3. What happened first? What did you do next?
4. What made that difficult?
5. What did you use to handle it?
6. What time, money, or other work did it take?
7. Who else was involved in choosing or paying for the solution?
8. What happened in the end?
9. What have you tried that did not work?
10. Is there something about this process I have missed?

Follow a concrete detail. If someone says they use a spreadsheet, ask what they record and when they last opened it. If they are comfortable showing an example, let them remove private information first.

Questions that can bias the answer

"Would an automatic reminder save you time?" introduces your proposed solution. Ask "How do you remember to follow up today?" first.

"Would you pay for this?" asks for a prediction with little commitment. Ask what they already pay for, who approves spending, and what happened when they last chose a tool. You can later test a specific paid offer, with clear delivery terms, as a separate experiment.

"Is this frustrating?" supplies the emotion you hope to hear. Ask which part, if any, causes trouble. If the person says the process works well, record that answer too.

Do not turn every interview into a pitch. Once you explain your product, the participant knows the answer you want. If you show a concept, label those notes separately from their account of existing behavior.

Copy this interview note template

• Research question: What uncertainty does this conversation address?
• Participant context: What makes this person relevant to the question?
• Recent situation: When did the problem happen, and what triggered it?
• Current approach: What did they actually do or use?
• Cost or consequence: What time, spending, delay, or missed work did they describe?
• Exact words: A short quote, with permission and enough context.
• Interpretation: What do you think the observation means?
• Contradicting evidence: What does not fit your assumption?
• Next question or action: What will you check next?

Keep the participant's account separate from your interpretation. "Sent a follow-up after finding an old email" is an observation. "Needs automation" is your proposed explanation.

A worked example and the next decision

Here is a hypothetical example. A consultant describes losing track of two proposals during a busy week. They already have reminders, but cannot tell which proposals still need a response. That points to a possible visibility problem rather than a lack of reminder notifications.

A useful next step could be to ask another consultant how they review open proposals, or offer a manual weekly review to someone who described the same difficulty. Write down what action would count as a useful signal before offering it. A compliment about the idea is a different signal from agreeing to try the review.

Several interviews can reveal patterns worth investigating. They do not estimate the size of the market or prove demand. Keep notes about how you recruited people, whose experiences are missing, and which accounts contradict the pattern.

Use the marketing experiment template to turn one uncertainty into an action. If you have a specific offer ready to show, a landing-page experiment can help you observe the next step people take.