Article
Customer Interview Questions to Ask Before You Build
Updated · Team Fonda
Short answer
The best customer interview questions ask about the past, not the future: "Tell me about the last time this happened", "What did you do about it?" and "What did that cost you?" Use five to seven of them per call, add follow-ups such as "why?" and "what happened next?", and keep your own idea out of the conversation.
Key takeaways
- Ask about specific past events. Opinions about future products are unreliable.
- Group your questions by goal: context, the last time, workarounds, cost and next steps.
- Follow-up questions do more work than the main questions.
- Skip any question that mentions your idea, asks for a compliment or asks for a price guess.
- Pick five to seven questions per call. A short list lets the person talk.
What makes a good interview question
A good interview question gets a story, not a verdict. "Do you like the idea?" gets a polite yes. "Walk me through the last time you had to plan a team trip" gets facts you can use: who was involved, which tools they opened, where it went wrong and what they did next.
Three rules cover most cases. First, ask about the past, because people remember what they did far better than they predict what they will do. Second, ask about them, not about your product. Third, keep questions open, starting with what, how, when and who instead of do, would or is.
This article is a question bank. If you still need to find people and plan the calls, start with how to do customer interviews. If the terms are new, the glossary entries for customer discovery and customer interview explain the basics.
Questions to ask in the first five minutes
Open with context. You want to know who this person is and how the problem sits inside their day, without naming the problem yet.
- What does a normal week look like in your role?
- How long have you been doing this?
- Which parts of your work take the most time?
- Which parts do you wish were easier?
- What tools or habits do you rely on most?
Do not rush this part. People relax when they talk about their own work, and the answers tell you whether this person belongs to the group you meant to study. If they do not, thank them, finish politely and note that your target group may need a sharper definition.
Questions about the last time the problem happened
This is the core of the call. Pick one recent event and go through it in order, like a replay.
- When was the last time you ran into this? What happened?
- What were you trying to get done at that moment?
- What did you do first? And then?
- Where did you get stuck or lose time?
- Who else was involved, and what did they do?
- How did it end? Were you happy with the result?
If the person cannot remember a recent time, that is data. A problem that has not come up in months is usually not urgent enough to build a business on. Move on to the questions about how often it happens, and note the answer.
Questions about workarounds and alternatives
A real problem leaves marks. People build spreadsheets, hire help, buy tools or put up with a clumsy process. These questions find the marks.
- What have you already tried to fix this?
- Why did that not work, or why did you stop?
- Are you using anything for it today? How did you pick it?
- Have you ever paid for something to solve this? What was it?
- If you could not solve it at all, what would happen?
Listen for emotion and effort. Someone who built a three-tab spreadsheet to track the problem cares more than someone who shrugs. The competitors you hear about here also feed your competitive analysis.
Questions about cost and priority
You want to know whether the problem is big enough to act on. Ask for facts about past cost, not a guess about future willingness to pay.
Scroll sideways to see the whole table.
| Goal | Question to ask | What a strong answer sounds like |
|---|---|---|
| Time cost | How many hours did this take you last month? | A specific number and a story behind it. |
| Money cost | What did this cost you the last time, in money or lost work? | A concrete loss, not "not sure". |
| Frequency | How often does this come up? | Weekly or monthly, with a recent date. |
| Priority | Where does fixing this rank among your other problems? | Top three, with a reason. |
| Existing budget | Do you already have money set aside for this? | A named tool, line item or person. |
These are example questions, not data about any group of customers.
Avoid "how much would you pay?" at this stage. Answers are guesses. Past spending is a much better signal, and you can test price later with a real offer. The free startup cost calculator can help you work out what you need to charge once you know the problem is real.
Follow-up questions that dig deeper
The main questions open the door. Follow-ups get you inside. Keep a short set ready and use them whenever an answer feels thin.
- Why? (Ask it twice if needed.)
- Can you give me an example?
- What happened next?
- How did that make you feel, or what did it cost?
- You said it was frustrating. What was the worst part?
- Is that typical, or was that time unusual?
Silence is also a follow-up. After someone answers, wait three or four seconds before you speak. Many people fill the gap with the detail you were hoping for.
Closing questions that give you your next interview
End on questions that help you keep learning, not on a pitch.
- Is there anything I should have asked that I did not?
- Who else deals with this and would be willing to talk?
- Can I come back to you if I have a follow-up question?
- Would you want to hear what I learn from these conversations?
The last question matters. A person who asks to see the results or to try something is giving you a stronger signal than someone who says "sounds great".
Questions to skip
Some questions feel natural but produce bad data. Replace them before you go into a call.
Scroll sideways to see the whole table.
| Skip this | Why | Ask this instead |
|---|---|---|
| Would you buy this? | Invites a polite yes. | What did you do the last time this came up? |
| Don't you think it is hard to do X? | Leads the witness. | How do you do X today? |
| Would a feature like Y be useful? | People say yes to any feature. | What is the hardest part of the process? |
| Do you think people need this? | Asks for a guess about others. | Who around you deals with it? |
| What should I build? | Customers know problems, not solutions. | What would a good outcome look like for you? |
Customers are experts in their own problems. You are the one who designs the solution, so do not ask them to do your job.
A worked example
Say you want to build a tool for people who organize recurring community events. This is an illustration, not a real case. You write down six questions: what a normal month of organizing looks like, the last time something went wrong, what they did about it, what tools they use, what it costs in hours and who else I should talk to.
In the third call, the organizer describes the last event. The venue canceled late, and she spent a weekend phoning attendees one by one. You ask "what happened next?" and "what did you try afterward?" She says she now keeps a spreadsheet of phone numbers and a backup venue list. That is a workaround, and it points to a specific problem: reaching attendees fast when plans change.
You notice that this was not the problem you came in with. You had assumed that finding venues was the pain. Say that by the eighth call, five of ten people mention late changes and none mentions venue search. You now have a better problem to test, and a short list of people who might try a first version.
How to use the answers
After each call, write the answers to your key questions in one row of a simple table. Compare rows after every five calls.
- Did this person have the problem recently?
- Did they spend money or real time on a workaround?
- Did they describe it in words similar to other people?
- Did they offer an introduction or ask for updates?
If most rows say yes, test demand next with a small offer, such as a simple page and a clear request, as described in how to validate a startup idea. If the rows are mixed, change the group or the problem and run another small round. Fonda, an AI co-founder for people starting a business, puts Interviews inside its Validate phase, before the Verdict, and your co-founder Marty remembers what you learned. It is free to start, and details on paid use are on the pricing page.
Frequently asked questions
How many questions should I ask in one interview?
Five to seven main questions is plenty for a 20 to 30 minute call. Leave room for follow-ups, because the best material usually comes from the second or third question you ask about a single story.
Should I send the questions in advance?
Usually not. Tell the person the topic and how long it will take, but keep the questions to yourself. Prepared answers sound polished and hide what actually happened.
What if the person only gives short answers?
Ask for a specific example, then use follow-ups such as "what happened next?" and "how did you handle that?". Short answers often mean the question was too general.
Can I use the same questions for B2B and B2C ideas?
The structure is the same: context, the last time, workarounds, cost and next steps. In B2B, add questions about who decides, who pays and what approval is needed, because the user and the buyer may be different people.
When should I show my idea?
Wait until the end, if at all. Treat any reaction as one data point. The conversation is meant to teach you about their problem, not to test your pitch.
Keep reading
- How to do customer interviews
- What is customer discovery?
- What is a customer interview?
- How to validate a startup idea
- Roast my idea
- Startup cost calculator
- Fonda vs IdeaProof (2026): Which AI Idea Validator Can You Actually Trust?
- Fonda vs Claude: Can You Build a Startup with a General AI? (2026)
- How to Do a Competitor Analysis for a New Business
- How to Find a Business Idea That Fits You
- How to Find Your Target Market