The Mom Test: customer interviews that can't lie to you
Fitzpatrick's three rules, the compliment-deflection moves, and why commitment — not enthusiasm — is the only interview signal that predicts a sale.
The theory in one paragraph
Everyone tells founders to talk to customers; almost nobody warns them that customer conversations default to lies — polite, well-meant, fatal lies. Fitzpatrick's frame: your mom would praise any idea you pitch because she loves you, and strangers do the same because it's easier than arguing. The Mom Test makes lying impossible by removing the thing people lie about. Three rules: talk about their life instead of your idea, ask about specifics in the past instead of opinions about the future, and talk less while listening more. Done right, the interviewee never learns what you're building — and everything they say is data about the problem, which is the only thing interviews can validate anyway.
How it works
The mechanics — as Rob Fitzpatrick defined them, not the folklore version.
Rule one and two: their life, in the past tense
'Would you use an app that…' invites fiction; 'walk me through the last time you dealt with…' retrieves facts. Past behavior has details — what they tried, what it cost, who else was involved — and details are where the real problem shows its economics. Futures and hypotheticals are free to inflate; the past already happened at a specific price.
Deflect the three false positives
Compliments ('sounds great!'), fluff ('I would totally use that'), and ideas ('you should add…') all feel like validation and predict nothing. The moves: ignore compliments, anchor fluff to specifics ('you say you'd use it — when did you last hit this problem?'), and dig under feature requests for the motivation ('what would that let you do?'). The generic 'do you have this problem' yes is the most dangerous sentence in customer development.
Advance or it didn't happen
A good meeting ends with the other person giving up something that costs them: time (a scheduled follow-up with their team), reputation (an intro to their boss or a peer), or money (a pre-order, a paid pilot, a letter of intent). Fitzpatrick's cold read: 'they keep saying nice things but never commit' doesn't mean ask better — it usually means they're not a customer, and politeness is how they're telling you.
If they know your idea, the data is already contaminated
The book's deepest move is structural, not conversational: the moment you pitch, every subsequent answer is about you — your feelings, their politeness, the social script of encouragement. Founders resist this because interviews double as ego fuel; the Mom Test asks you to walk out of a great conversation with zero praise collected and three specific stories about the problem. In MENA especially, where hospitality culture makes direct criticism rare, engineering the conversation so honesty requires no rudeness isn't a nicety — it's the whole method.
The person behind it
Rob Fitzpatrick
Founder three times over · YC alum turned author-teacher
Fitzpatrick burned years and investor money on products that interview subjects had enthusiastically endorsed — then watched none of them buy. The Mom Test is the self-published post-mortem that became required reading in accelerators worldwide: less a research methodology than a field manual for not getting lied to, written by someone who collected every lie personally.
Lineage — The tactical layer under Customer Development's interview stages →
How to apply it this week
Each step maps to a field in the Assumptions Tracker tool — finishing the read means finishing the work.
Write your three riskiest problem assumptions
Not product assumptions — problem assumptions: who has it, how often, what it costs them, what they do about it today. Each interview exists to attack these, and an interview without a target assumption is coffee.
Assumptions Tracker · assumption listDraft questions that pass the test
For each assumption, questions about past specifics: 'when did this last happen', 'what did you try', 'what did that cost', 'who else was involved'. Strike any question containing your idea, the word 'would', or a price you haven't charged.
Assumptions Tracker · validation methodInterview without pitching
Introduce the topic, not the product. Then the hard discipline: shut up. Every time you explain, you teach them the answers you want; every silence you survive, they fill with the truth.
Log facts and commitments, discard praise
After each conversation, record only: specific past behaviors, numbers, names, and what they gave up (time, intro, money). A journal entry that reads 'loved it!' records nothing; 'currently pays cousin 500 EGP/month to do this in Excel' is a business model clue.
Assumptions Tracker · evidence logPush for advancement, read refusals as data
End every good conversation with a concrete ask that costs them something. Track your advancement rate across interviews — it's the closest thing pre-product founders have to a conversion metric, and a falling one is an early pivot signal.
Feeds your Readiness Score · Foundation
Build it, don't just read it
The steps above are the Assumptions Tracker tool's structure. Open it and work through them with your own startup — your readiness score starts building from the first field.
Free account · no card required
See it in the wild
Teardowns from our benchmarks library where this framework is doing real work.
Benchmark teardown
Airbnb
Founders who lived with their users — literally — and learned from photographed apartments what surveys would never have said
Read the teardown
Benchmark teardown
Vezeeta
Healthcare booking validated on an opaque market's real behavior: patients already paying costs nobody would admit to in the abstract
Read the teardown
Cautionary tale
Swvl
Enthusiasm at subsidized prices is the ultimate false positive — riders' compliments never tested willingness to pay the real fare
Read the teardown
Frequently asked questions
What is the Mom Test in customer development?
It's Rob Fitzpatrick's method for customer interviews that produce truth instead of politeness: talk about the customer's life instead of your idea, ask about specific past events instead of hypothetical futures, and listen more than you talk. It's named for the test case — questions so grounded in facts that even your mom, who wants to encourage you, couldn't answer with a comforting lie.
What are the three rules of the Mom Test?
One: talk about their life, not your idea — the interviewee shouldn't need to know what you're building. Two: ask about specifics in the past, not opinions or generics about the future — 'when did this last happen' beats 'would you use this'. Three: talk less and listen more — every explanation you give teaches them what you want to hear.
What are bad customer interview questions?
Anything hypothetical ('would you use…', 'how much would you pay…'), anything generic ('do you have problems with…'), and anything that pitches before it asks. These invite the three false positives — compliments, fluffy affirmations, and drive-by feature ideas — all of which feel like validation and predict nothing about purchase behavior.
How do you know if a customer interview went well?
Not by enthusiasm — by facts and commitment. A good interview yields specific stories about past behavior (what they tried, what it cost) and ends with the person giving up something real: scheduled time, an introduction, a pre-order or paid pilot. Fitzpatrick's rule of thumb: if they're saying nice things but won't commit anything, the meeting failed regardless of how it felt.
How many customer interviews should you do before building?
Fitzpatrick's guidance is to keep going until you stop hearing new things — commonly a few dozen conversations within one specific segment, not a token five. The count matters less than the design: three interviews that surface past behavior and end in commitments teach more than thirty pitch-and-praise sessions.
Related frameworks
Customer Development
Steve Blank · 2005
Jobs to be Done
Clayton Christensen · 2003
The Lean Startup
Eric Ries · 2011
Sources
Independent educational summary written by StartupKit from public sources. The Mom Test is the work of Rob Fitzpatrick; this page is not affiliated with or endorsed by the author.