A customer discovery interview script is a structured set of open-ended questions founders use to test whether a real problem exists, who feels it, and whether they would pay to solve it. This article gives you a complete, ready-to-use script:opening questionsompliments instead of honest signal. Use it before you write a line of code.
Key Takeaways
- Ask about past behavior, never future hypotheticals like "would you use this?"
- Compliments are noise; specific stories, budgets, and workarounds are signal.
- The Mom Test rule: talk about their life, not your idea.
- Ten interviews reveal patterns; three reveal almost nothing.
- Always ask what they have already tried and what it cost them.
- End every interview with an ask: a referral, a pre-order, or a follow-up.
What Are the Goals of a Discovery Interview?
The goal of a customer discovery interview is to validate or invalidate your riskiest assumption about a problem, not to pitch your solution. A good interview answers three questions: is the problem real and frequent, does the person already spend time or money on it, and who else feels it. If you leave an interview feeling encouraged but with no new facts, the interview failed.
Customer discovery is the systematic process of testing whether a problem is real. Unlike surveys or focus groups, discovery interviews are one-on-one conversations that dig into a person's actual past behavior: what theydid last time the problem occurred. The output is not opinions or compliments but evidence:frequency of the problemmeone finally act. Founders typically run 10 to 20 interviews per segment before deciding whether to build, pivot, or stop.
Keep the interview to 30-45 minutes. Anything longer and the conversation drifts into speculation. Anything shorter and you never reach the budget questions, which is where most founders chicken out.
How Do You Open a Customer Discovery Interview?
Open with context-setting questions that make the person comfortable and establish that you are researching a problem, not selling a product. The first five minutes decide whether you get honest answers or polite ones. Never mention your idea until the very end, and only if they ask.
Opening and context-setting questions
- "Before we start, can you walk me through what your role looks like day to day?"
- "How long have you been dealing with [problem area]?"
- "When did this last come up for you?"
- "Who else on your team touches this?"
- "What does a typical week look like when this goes wrong?"
These questions feel small, but they anchor the conversation in reality. "When did this last come up?" is the single most useful opener in any startup customer interview script, because it forces a specific memory instead of a general opinion. If the person cannot recall a recent instance, the problem is probably not urgent enough to build for.
What Problem and Workflow Probes Should You Ask?
Problem deep-dive probes move from "this is annoying" to "here is exactly how it hurts." The goal is a narrated story of the last occurrence, with numbers attached:time lost. This is the core of any founder customer discovery template.
Problem and workflow deep-dive probes
- "Walk me through the last time this happened, step by step."
- "What did you do first? Then what?"
- "How long did that take? How often does it happen?"
- "What did it cost you — time, money, missed opportunities?"
- "What have you tried to fix this? Why did it fail?"
- "What are you using now instead? How much does it cost?"
- "If this never got solved, what happens?"
The phrase "walk me through the last time" is doing heavy lifting. Past behavior is the only reliable predictor of future behavior. Contrast that with "would you use a tool that...?" — a question that reliably produces enthusiastic, meaningless yeses. These are the mom test interview questions in action: Rob Fitzpatrick's The Mom Test argues that people lie to be nice, but they rarely lie about what they actually did last Tuesday.
The Mom Test rules in practice
Fitzpatrick's three rules, widely cited in founder circles, are simple: talk about their life, not your idea; ask about specifics in the past, not generics or the future; and talk less, listen more. If you catch yourself pitching, stop and ask another question. A useful ratio is roughly 80% listening, 20% talking.
How Do You Ask About Budget, Urgency, and Switching Costs?
Budget and urgency questions separate people who complain from people who buy. Ask what they already spend on the problem today —tools— because existing spend is the strongest evidence of willingness to pay. Then ask what would make them switch.
Budget, urgency, and switching-cost questions
- "What are you spending on this today — tools, time, or people?"
- "Have you ever paid for a solution to this? What happened?"
- "If a fix existed tomorrow, what would it need to do to be worth paying for?"
- "Who signs off on purchases like this, and how long does that take?"
- "What would make you switch from what you use now?"
- "What is the cost of doing nothing for another six months?"
- "On a scale of 1 to 10, how urgent is this compared to everything else on your plate?"
Notice that none of these ask "would you pay for my product?" They ask about current spend, decision processes, and urgency relative to competing priorities. A founder who hears "we'd definitely pay for that" but learns the person has never spent a dollar on the problem should treat that as a polite lie, not validation.
| Signal | Weak (polite lie) | Strong (real intent) |
|---|---|---|
| Problem frequency | "Sometimes, I guess" | "Every week, usually on Mondays" |
| Existing spend | "Nothing really" | "We pay $400/month for a workaround" |
| Past attempts | "I looked into it once" | "We tried two tools and a contractor" |
| Urgency | "Someday would be nice" | "We need this fixed before Q3" |
| Commitment | "Send me the link" | "Can I pre-order / pilot this?" |
What Are the Red Flags and Polite-Lie Traps?
The biggest red flag is enthusiasm without specifics. Compliments, feature requests, and "that sounds cool" are all ways of ending the conversation kindly. Your job is to notice them and redirect to facts.
Common traps and how to defuse them
- The compliment: "I love it!" → "Thanks — what would you do with it on Monday morning?"
- The hypothetical: "I'd definitely use that." → "What are you using for this today?"
- The feature wishlist: "You should add X." → "What happened the last time you needed X?"
- The vague pain: "It's frustrating." → "Tell me about the last time it frustrated you."
- The friend discount: Friends and family over-validate. Interview strangers in your target segment first.
One honest caveat: even perfect interviews produce biased samples. People who agree to talk to founders are often more engaged than average users. Treat discovery as directional evidence, not proof, and pair it with behavioral tests like pre-orders or landing-page signups before committing serious capital.
How Do You Synthesize 10 Interviews?
Synthesis turns ten conversations into one decision. After each interview, write down the three most surprising facts within an hour, while memory is fresh.After ten interviews— patterns, not anecdotes.
A step-by-step synthesis framework
- Tag every quote by theme:pain.
- Count how many of the 10 mentioned each theme unprompted.
- Identify the segment with the sharpest pain — usually one subgroup, not everyone.
- Write a one-sentence problem statement using their words, not yours.
- Decide: build, pivot the segment, or stop. Pre-commit to the threshold before you start.
A practical rule of thumb: if fewer than 6 of 10 interviewees describe the same pain with similar urgency and existing spend, you do not yet have a problem worth building for. If 8 or more do, you have a wedge — and you should test willingness to pay immediately with a pre-order or pilot. GoWithFund Startup Studio can help here: its pitch profiles and pre-order tools let you turn validated demand into early commitments and track milestones as you move from interviews to a first cohort.







