Customer Discovery Interviews: Ask Better Questions, Get Better Evidence

What is a customer discovery interview?

A customer discovery interview is a conversation that helps you understand a potential customer's circumstances, behavior, problems, and buying process. It informs a business hypothesis; it is not a product pitch disguised as research.

Ask about a specific recent event and what the person actually did. Hypothetical praise is weaker evidence than a costly workaround or a concrete commitment. Y Combinator's user-interview guidance emphasizes specifics and listening.

Learn about	Evidence to seek
Problem	A recent incident and its consequences
Urgency	Why the person acted, delayed, or accepted it
Alternative	The actual process, tool, or workaround
Purchase	Who owns the budget and approves change
Next step	A specific action the participant agrees to take

Why it matters

Good interviews can expose a weak customer hypothesis before you invest heavily in features or acquisition. They also reveal words buyers use to describe their situation, helping you improve positioning and experiments.

An interview can establish that a problem deserves investigation. It cannot, by itself, prove that your solution works, that the interviewee will pay, or that the market is large.

Worked example: interest versus evidence

This scenario is invented for teaching. You are considering scheduling software for pet groomers. Compare two conversations:

Statement	Interpretation	Follow-up
“An easier calendar sounds nice.”	General interest	Ask how bookings work today
“Two staff members booked the same slot last Friday.”	A specific event	Reconstruct the process and consequence
“I spent an hour rearranging appointments.”	Reported cost	Ask how often it happens and what has been tried
“Our owner approves new subscriptions.”	A buying-role clue	Ask about the actual approval process

Suppose ten qualifying interviews reveal six recent conflicts, three pilot requests, and one paid pilot. Report those separately. Do not call six problem reports six sales or use this small convenience sample to estimate market-wide demand.

Copyable interview script

Use this original script as a starting point. Adapt it to the participant and leave room for unexpected details.

1. Opening: “I am researching how [customer type] handles [job]. I want to understand your current process. Is it okay to take notes?” Agree on recording separately if used.
2. Context: “What is your responsibility for this work, and who else is involved?”
3. Recent event: “Walk me through the most recent time you handled it, from the trigger to the result.”
4. Friction: “Where did the process slow down or go wrong? What happened afterward?”
5. Frequency: “Which other recent occasions were similar? Are there times it works well?”
6. Alternative: “What have you changed or purchased to improve this? What made you choose that approach?”
7. Buying process: “How was the last related purchase approved? Who participated?”
8. Close: “What did I misunderstand? Would you be open to [specific next research step]?”

Spend most of the discussion on their experience. If you show a concept afterward, label that as a separate solution test so enthusiasm about the demo does not rewrite the interview evidence.

Replace leading questions

Leading prompt	More useful prompt
Wouldn't automated reminders save you time?	How do reminders happen today?
Do you hate your current software?	What works well, and where does it fall short?
Would you pay $100 for this?	How do you budget for this job, and what have you bought before?

Test an actual priced offer separately when appropriate. A question about past spending helps you understand context; it does not establish willingness to pay your proposed price.

Copyable evidence worksheet

Copy one record per conversation. Use a participant code when names are unnecessary.

Field	Fill in
Date, participant code, segment	[Enter]
Hypothesis being investigated	[Enter]
Recent event and current process	[Enter]
Reported frequency and cost	[Enter; label estimates]
Observed artifact or behavior	[Enter; distinguish from self-report]
Current alternative and purchase history	[Enter]
User, buyer, approval steps	[Enter]
Evidence against our hypothesis	[Enter]
Agreed next action and date	[Enter]
Interpretation and remaining uncertainty	[Enter]

Separate direct notes from your interpretation. Collect only the detail needed for the research and agree on how any customer materials will be used.

Turn interviews into a decision

Compare patterns within a defined segment. Include people who declined, did not buy, or use an alternative successfully. Record how you recruited participants and the limits of the sample.

Choose one next test: narrow the customer, investigate a repeated problem, test a prototype, or make a concrete paid offer. Set a decision date and write what outcome would change your plan.

What investors are looking for

Explain whom you interviewed, what they did about the problem, what contradicted your assumptions, and what happened next. Show a traceable path from interviews to product and commercial decisions. Counts alone are less useful than clear evidence and learning.

Common mistakes

• Pitching before understanding the current process.
• Asking only friends or enthusiastic supporters.
• Steering participants toward your preferred answer.
• Treating estimates, observed facts, and commitments as equivalent.
• Collecting notes without making a decision.

Founder checklist

Define the segment and hypothesis. Prepare neutral prompts and a notes format. Capture recent behavior, buying roles, contrary evidence, and an agreed next action.

Your next action: run a small interview round, compare evidence, and choose one focused experiment.

Sources and further reading

• Y Combinator: User Interview Guidance
• SparkLaunch: Ideal Customer Profile

References checked November 10, 2026. The script, worksheet, and scenario are original illustrative teaching material.