The Mom Test: a founder's practical guide
Rob Fitzpatrick's 130-page book in a 10-minute summary — the 3 rules, what good questions look like, the compliments to ignore, and how to practice them.
This is a fan-made primer. The book itself is the real source — buy it from momtestbook.com or wherever you buy books. We have no affiliation with Rob Fitzpatrick.
What is the Mom Test?
The Mom Test is a method for talking to potential customers in a way that produces useful answers — even when the person you're talking to wants to be polite.
The name comes from a thought experiment: if you described your idea to your mom, she'd tell you it sounds wonderful. Not because she means it, but because she loves you and doesn't want to hurt your feelings. Most prospects do the same — they want to be encouraging, so they answer hypothetical questions hypothetically and overstate their interest.
The Mom Test is a way of asking questions so that even your mom couldn't lie to you. You don't ask "would you use this?" — you ask about specific things they've actually done. Past behavior doesn't lie. Hypotheticals do.
The book was written by Rob Fitzpatrick in 2013. It's 130 pages, dense with examples, and one of the few startup books worth reading twice.
The 3 rules
The whole framework reduces to three principles. Once you internalize them, the specific question wording matters less.
Rule 1. Talk about their life, not your idea
Don't pitch. Don't describe what you're building. Don't ask if they'd use it. Instead, ask about their actual life — what they did last week, what frustrated them, what they paid money to fix. The moment you describe your idea, the conversation becomes about your idea instead of their reality.
Rule 2. Ask about specifics in the past, not generics in the future
"Would you use a tool that helps with X?" gets a polite lie. "Tell me about the last time you tried to solve X — what did you do, what worked, what was frustrating?" gets the truth. Past behavior is data. Future intentions are fiction.
Rule 3. Talk less and listen more
Founders love their ideas. They'll fill silence by explaining. Don't. Ask the question, then shut up. The most useful answers come 10 seconds after the awkward pause when most people would have moved on.
Bad questions vs good questions
The fastest way to internalize the Mom Test is to see good and bad versions side by side. The pattern is always the same: bad questions reach for the future, the hypothetical, or the polite. Good questions reach for the past, the specific, and the awkward.
❌ "Do you think this is a good idea?"
Compliment-bait. The polite answer is yes. You learn nothing.
✓ "What's the hardest thing you dealt with this month in [domain]?"
Forces them to recall a real, specific event.
❌ "Would you pay $20 a month for this?"
Fictional commitment. People say yes to hypothetical money all day.
✓ "How much do you currently spend solving this — in tools, contractors, or your own time?"
Reveals what they actually allocate budget to.
❌ "How often do you face this problem?"
Frequency claims are exaggerated when the problem is on someone's mind. They'll say "all the time" out of politeness.
✓ "When did this last happen? Walk me through what you did."
Replaces opinion with a story you can verify.
❌ "Is this useful to you?"
Yes/no questions invite politeness. You learn whether they're polite, not whether the product is useful.
✓ "What were you doing right before this would have helped you? And what did you do instead?"
Maps the workflow around the problem and shows whether your product even fits in.
❌ "Would you buy this if it existed?"
Verbal commitment is the cheapest currency. Everyone is generous with it.
✓ "Can I follow up with you when there's a beta? Want to put down a $50 deposit to be first in?"
Asking for skin in the game separates curiosity from buyers.
7 "compliments" you should ignore
These phrases sound like validation. They aren't. They're politeness, deflection, or curiosity — and treating them as buying signals is how founders end up building products nobody wants.
- "That's a great idea!" — they're being polite, not analytical
- "I'd definitely use that." — verbal interest, not commitment
- "This is so needed in our industry." — opinion about the industry, not their wallet
- "You should talk to [person who isn't actually a buyer]." — deflection
- "Let me know when it's ready." — politest way to end the conversation
- "My company would love this." — they don't control the budget
- "How much will it cost?" — curiosity, not buying signal
The book's rule: a compliment is feedback you can't act on. If you can't turn it into a TODO or a verifiable fact, ignore it.
How to practice without dialing strangers
Reading the book once doesn't make you good at the Mom Test. The skill is real-time — recognizing when a conversation has drifted into hypotheticals and steering it back to specific past behavior. That muscle takes practice, and most founders skip the practice because cold-calling 20 strangers is hard.
Two ways to practice without phoning strangers, in order of priority:
1. Practice with a friend who'll roleplay
Find another founder. Pitch them as if they're your target customer. Have them deliberately give you compliments and hypotheticals. Your job: catch yourself before chasing the lie. Switch roles. Do this 5 times before your first real interview.
2. Use a voice AI as a synthetic interviewer
This is what we built GoNoGo for. You describe your idea naturally, an AI strategist asks Mom Test–style questions, and you get a transcript showing where you fell into hypotheticals. It's 30 minutes, free, and gives you a feel for the question rhythm before you do it with real prospects. Not a replacement for real conversations — a way to warm up.
30 min · No credit card · Then go talk to humans
When you're ready for real interviews
Synthetic practice and AI sessions are warm-ups. The Mom Test is fundamentally a tool for human conversations — and at some point you have to actually call strangers. Here's how to make that less painful:
- Start with people one degree removed. Friends-of-friends in your ICP, not strangers from cold lists. Easier to book, still useful.
- Don't schedule "feedback calls." That sets the expectation that they'll comment on your idea. Schedule "learning calls" — you're learning about their world.
- Aim for 20–30 conversations. Patterns emerge around the 15th. Stopping at 5 gives you anecdotes, not signal.
- Take notes by hand or recording. Don't rely on memory. Specifics decay within hours.
- End every interview with the commitment ask. "Can I follow up when I have something to show?" Their willingness to commit is the truest signal you'll get.
Frequently asked questions
Should I read the actual book?+
Why is it called the Mom Test?+
How many interviews do I need to do?+
Can AI replace Mom Test interviews?+
What's the biggest mistake people make with the Mom Test?+
The real source
Buy The Mom Test by Rob Fitzpatrick
This page is a primer. The book is the real thing — 130 pages of dense, specific examples that make the principles stick. Worth every minute.
Get the book at momtestbook.comMom Test deep dives
Related guides