🎯 Goal: Learn to interview customers WITHOUT leading them, so you get truth (past behaviour) instead of polite praise.
The book The Mom Test (Rob Fitzpatrick) teaches one strange but powerful rule: ask in a way that even your own mom — the person who loves you most — couldn’t lie to make you feel good. The trick is to never talk about your idea; ask about their PAST BEHAVIOUR: when they last hit the problem, what they did, how much money and time it cost, whether they’ve ever paid for a solution. The question “do you like this idea?” almost always yields useless answers, because people are polite and praise you to be kind. You’ll leave with a question set that pulls the truth out of a 15-minute chat.
Lesson content
🎯
Praise kills startups. When you show off your idea, the listener’s brain switches to “polite mode”: they nod, they compliment, they say “sounds great, let me know when it launches”. You go home thinking you validated demand, but all you got was good manners. The Mom Test flips it: you don’t ask for an opinion on your solution, you investigate their life like a journalist. Truth lives in what they DID, not in what they SAY they’ll do.
🧭
3 rules + turning bad questions into good ones: • Talk about their life, not your idea. Bad: “Is my spending-tracker app good?” → Good: “How did you track your spending last month?” • Ask about what happened, not opinions or predictions. Bad: “Would you use it?” → Good: “When did you last need this?” • Talk less, listen more. After each answer, keep digging with “and then what?”, “why?”, “walk me through that time” instead of rushing to close.
🧰
TOOL — The 8 Mom Test questions (mine behaviour, not opinions):
Question
What it uncovers
When did you last hit this?
Anchors on past fact, not a guess
What did you do about it then?
Real behaviour reveals seriousness
How much money / time did it cost?
Measures the pain in numbers
What else have you tried?
Reveals rivals and current spend
What’s most annoying about how you do it now?
Finds the core pain to target
When did you last pay for this?
Having paid = strongest proof of demand
Who decides on the spending?
Identifies the real buyer
Who else struggles with this?
Opens the path to the next customers
Tip: every time a customer offers an opinion (“I’d probably use it”), bounce it back with a behaviour question (“when did you last use something similar?”).
🌏
Case study — Coolmate (step by step): 1) Instead of asking friends “is an online menswear idea good?”, the Coolmate team asked customers about real buying behaviour: where they buy, what they hate most, how often they repurchase. 2) They heard a recurring pain: dreading the store, hard to pick sizes, wasting time re-buying basics. 3) Because they listened to behaviour instead of praise, they fixed the real pain (fast delivery, easy returns, reliable sizing) and reached a very high repeat rate (illustratively over 50%). Contrast — the classic failed startup: asks 20 friends “is the idea good?”, all 20 nod, builds for 6 months… nobody buys. The only difference: one asked about behaviour, the other fished for praise.
⚠️
5 interview traps: • Pitching first, then asking “what do you think” → you only get praise. • Trusting future promises (“I’ll use it”, “I’d pay”) as evidence — those are just hypotheticals. • Talking too much, convincing the customer instead of letting them tell their story. • Asking vague questions (“do you often run into this?”) instead of specific ones (“when was the last time?”). • Ignoring money signals: forgetting to ask how much they’ve spent / would spend, so you never learn if the pain is big enough.
✅
Checklist for a solid interview: ① Did I avoid pitching my idea in the first 10 minutes? ② Did every question anchor on a specific PAST event? ③ Did I ask about the MONEY / TIME they already spent? ④ Did I listen more than I talked? ⑤ Afterwards, can I underline at least 3 pieces of real BEHAVIOUR (not praise)?
Practice exercise
🔬 APPLIED EXERCISE (1 week): (1) Pick 3 real people from your target group. (2) Interview each using the 8 Mom Test questions — the question “do you like my idea?” is strictly banned. (3) After each session, split your notes into two columns: real BEHAVIOUR (did it / paid for it) and Opinions / praise. Only the left column is evidence.
Worked template: FILLED EXAMPLE: Instead of “is my vocab app good?”, ask “when did you last try to learn vocabulary, what did you use, did you quit?”. Person A says: “Last month I downloaded 2 apps, deleted them after 3 days, then went back to writing words in a notebook.” → That’s real BEHAVIOUR: the need is real, the old fix wasn’t good enough — worth more than any praise.
Quick quiz
1. What is the core principle of The Mom Test (Rob Fitzpatrick)?
→ Ask cleverly so even someone who loves you can’t lie to please you
The Mom Test asks so that even someone who loves you can’t lie to please you — getting truth, not praise.
2. Which question fits the Mom Test spirit?
→ “When did you last hit this problem and what did you do?”
Good questions ask about specific past behaviour (“when you last… what did you do”), not opinions or future guesses.
3. Why does “do you like my idea?” usually give useless answers?
→ Because people are polite and tend to praise, and praise isn’t proof of demand
People are polite and tend to praise; praise isn’t proof of demand and can trick you into thinking you validated.
4. What is the STRONGEST evidence that a pain is real?
→ They once spent money or time to solve it themselves
The strongest evidence is that they once spent money/time self-solving — real behaviour, not a promise.
5. A customer says “if that app existed I’d probably pay 100k/month”. What kind of information is this?
→ A future promise (a hypothetical), to be validated by real behaviour like a deposit
A “I’d pay…” statement is a future promise (hypothetical); turn it into behaviour like a deposit before trusting it.
6. Your ideal “talk : listen” ratio in an interview should be?
→ Listen more than you talk, constantly asking “and then what?”, “why?”
A good interview means listening more than talking, digging with “and then what?”, “why?” instead of rushing to close.
Advanced
A deeper framework
At an advanced level, sort everything a customer says into three evidence tiers: opinions (“sounds cool”) — nearly worthless; future promises (“I’ll use it / I’d pay”) — hypotheticals that must be turned into behaviour to be validated; and past behaviour with a cost attached (“last month I paid 300k for something similar”) — the strongest evidence.
The advanced move beyond the Mom Test is the commitment ladder: after the interview, invite the customer to take one real step — give an email, spend time on a trial, place a deposit, refer someone. Each rung costs them money, time or reputation; the more expensive the commitment, the more real the demand signal.
Tiering the answers of 10 interviewees
8/10 praise “great idea”
Opinion — doesn’t count as evidence
3/10 say “I’d probably pay”
Future promise — needs validation
1/10 once paid 300k on a workaround
Behaviour + cost — strongest evidence
8 compliments don’t equal 1 person who actually spent money. Count behaviour, not praise.
Common trap: The “compliment trap”: collecting every “sounds cool”, “good idea” as confirming data. Praise is social noise, not a demand signal — always convert each statement into past behaviour with money/time attached before believing it.
Advanced questions
1. You interview 10 people: 8 praise it (“great, I’d definitely use it”), but none has ever spent money or time solving the problem. What should you conclude?
→ Demand is NOT proven: praise isn’t evidence, and nobody ever paying/self-solving signals the pain may not be big enough
True Mom Test spirit: praise comes easily out of politeness. Nobody spending money/time to self-solve means real behaviour hasn’t confirmed demand — seek behavioural evidence (paid, built a workaround) before building.
2. A customer says “if such an app existed I’d probably pay 100k/month”. The best way to validate it is?
→ Turn the promise into behaviour now: ask for their email, a prepaid signup, or a small deposit
The Mom Test warns that future statements are hypotheticals. Validate by climbing the commitment ladder — turning the promise into costly behaviour (deposit, prepayment) — to see if it’s real.
3. How does the commitment ladder measure demand?
→ By how much money, time or reputation the customer will spend at each real commitment rung
A real commitment always costs the customer something (money/time/reputation). The more expensive the commitment, the more trustworthy the demand signal — that’s how you get past empty praise.
🎯 Real-life mission
REAL-LIFE MISSION: Interview 5 people in the spirit of the Mom Test — NEVER ask “do you like my idea?”. Ask about the LAST TIME they hit the problem, what they DID, how much money/time it cost, and whether they’ve ever paid for a solution. With at least 1 person, try the commitment ladder: ask for an email or a small deposit. Write up: 5 pieces of real BEHAVIOUR (not praise) + who actually committed.