The Mom Test, Rewritten: 20 Validation Questions That Don't Lie

Customer interviews are the most reliable validation tool founders have — and the most reliably misused. Rob Fitzpatrick's The Mom Test named the core failure in 2013: people lie to be nice, especially to founders they like. The fix isn't asking better-intentioned questions. It's asking questions about the past, where politeness has no leverage. Below are 20 questions rebuilt around that rule, grouped by what they're designed to detect, followed by the 5 questions that will quietly wreck your data no matter how many interviews you run.

Why do customer interviews give founders false positives?

Because almost everything founders ask is about the future. "Would you use this?" "Do you think this is a good idea?" "Would you pay $20 a month for it?" Every one of those questions hands the interviewee a polite fiction to perform — and they will perform it. CB Insights' analysis of startup post-mortems found that 43% of failures come down to poor product–market fit — the most-cited root cause — which means many dead startups had founders who believed customer conversations proved demand. The interviews weren't evidence. They were courtesy.

The Mom Test's rule is one sentence long: ask about past behavior, never future intentions. Everything below follows from it.

It matters more now, not less. With Claude, Cursor and no-code tools, a solo founder can ship an MVP in a weekend, so it's tempting to skip the conversations and let the launch be the test. And the politest interviewee of all is the AI you're building with — ask it whether your idea is good and it will usually find a way to say yes. Twenty past-behavior questions to real strangers are still the cheapest evidence you can get.

Problem questions: is the pain real? (1–6)

Use these to open interviews. Your only job is to get a concrete story, not a verdict.

1. "Tell me about the last time you ran into this problem." — The highest-signal opener in validation. A real problem produces a story with specifics: when, where, what it cost.
2. "What did you do about it?" — The most important follow-up. If the answer is "nothing," the problem may be real but not painful enough to pay to solve.
3. "How often does this happen?" — Frequency determines business model. Weekly pains support subscriptions; yearly pains don't.
4. "What have you already tried to solve it?" — Past solutions attempted = past money spent. Both are proof of demand.
5. "How much did that solution cost you — in money, time, or hassle?" — This prices the pain in their own words before you name a price.
6. "What was the worst instance of this you can remember?" — Pushes past the generic case to the emotional peak, where urgency actually lives.

Digging questions: is the story true? (7–13)

Politeness survives into storytelling. These questions test whether the story holds weight.

7. "Walk me through exactly what you did step by step." — Vague stories collapse under step-by-step requests. Real memories have sequence.
8. "Who else was involved, and what did they do?" — Reveals whether the pain is solo or organizational — which changes who your buyer is.
9. "What happened right before that?" — Finds the trigger. Triggers are where marketing messages and purchase moments come from.
10. "Why did you choose that workaround instead of something else?" — Surfaces the competitive landscape you actually face, including "do nothing."
11. "What would happen if you just ignored it?" — The flip side of urgency. "Nothing" is a critical data point.
12. "Can you show me?" — Screenshares, photo rolls, and inboxes don't lie. Artifacts beat adjectives.
13. "Have you ever complained about this to anyone? What did you say?" — People repeat their own words about real pains verbatim. Note the exact phrasing — it's your future landing-page copy.

Commitment questions: will they act? (14–17)

Interest is free; commitment isn't. These convert a positive interview into real evidence.

14. "If this existed today, what would you do next?" — Vague enthusiasm produces vague answers. Listen for a specific next action with a time attached.
15. "Can I come back and show you a version of this next week?" — The classic Mom Test ask. A calendar yes is worth ten verbal yeses.
16. "Who should I talk to next about this?" — People volunteer their network's time only for things they genuinely care about.
17. "Would you be willing to pay a deposit / join the waitlist / do a paid pilot?" — Money, even $10, is the only commitment that never lies. Frame it as optional — a forced yes is worthless.

Numbers questions: is there a business? (18–20)

18. "What do you currently pay for anything adjacent to this?" — Anchors your pricing to an existing budget line instead of inventing one.
19. "How much time does this take you per week?" — Time is money; their answer lets you calculate the value of your solution in their terms.
20. "How many other people do you know who have this problem?" — A rough top-down check on market size, from inside their world.

The 5 questions that always lie

Cut these from your interview script. Every one manufactures false positives.

• ❌ "Would you buy this?" — Future intention, zero cost to say yes.
• ❌ "Do you think this is a good idea?" — You're asking for reassurance, and you'll receive it.
• ❌ "How much would you pay for this?" — Hypothetical pricing is fiction. Get a real pre-order instead.
• ❌ "Would you use an app that...?" — "An app" is free to imagine and free to abandon.
• ❌ "Can I email you when it's ready?" — Email addresses cost nothing. Deposits cost something.

The approved questions earlier in this post are the direct replacements:

Notice the pattern: every banned question is about the future, the abstract, or the free. Every approved question is about the past, the specific, or the costly.

Score your idea in 3 minutesOur free AI validator interviews you, researches real competitors, and scores founder fit, Why Now, tarpit risk, and more.

Scoring what you hear

After 10–15 interviews, patterns beat anecdotes. If fewer than 3 of 10 interviewees told a vivid, unprompted story with real cost attached, the problem isn't painful enough — no matter how warmly they smiled. Our 12-point startup idea validation checklist turns interview evidence into a go/no-go score, and the full guide to validating a startup idea covers what to do when interviews pass but you're still not sure.

Where to find people worth asking

The best interview subjects are strangers with the problem, not friends with encouragement. If you're short on candidates, the AI business ideas, startup ideas for beginners, and startup ideas with no money hubs each list ten curated ideas with the communities where their users cluster — those communities are also where you recruit interviewees.

And if you'd rather skip the cold-start problem entirely, Pikari delivers fresh, personalized startup ideas to your feed every day — problem, audience, and validation angle included — so your interview calendar fills with candidates instead of brainstorming sessions. Try it without signing up. Going deeper on one idea is what Pikari Premium is for: personalization that matches ideas to your actual skills and constraints.