Talk to Customers Without Lying to Yourself: The Mom Test
Your customers will lie to you. Not out of malice, out of kindness. Your job is to ask questions they cannot lie to.
You take a prospect for coffee, describe your idea, and they light up. "That's amazing, I'd definitely use that." You walk out floating. You just learned nothing, and worse, you think you learned something. That is the most dangerous kind of feedback there is.
Rob Fitzpatrick named the cure in The Mom Test. The premise: ask questions so anchored in the person's real life that even your mum, who loves you and wants you happy, could not skew the answer to flatter you. The trick is simple. Stop talking about your idea. Start talking about their life.
Opinions are worthless. Facts about the past are gold.
The three rules
- Talk about their life, not your idea. The moment you pitch, you bias every answer. Keep the idea in your pocket.
- Ask about specifics in the past, not generics about the future. "Would you use a tool that..." invites fantasy. "Walk me through the last time you dealt with this" surfaces reality.
- Talk less, listen more. You are there to learn, not to convince. If you are talking more than a third of the time, you are pitching, not researching.
Good questions versus compliment-bait
The difference between useless and decision-grade interviews is entirely in the questions.
"What's the hardest part about doing this today?" · "Tell me about the last time that happened." · "What did it cost you, in time or money?" · "What have you already tried to fix it?" · "What are you using instead right now?" · "Where does the money for this come from?"
"Would you use this?" · "Do you think this is a good idea?" · "How much would you pay for it?" · "Would you buy a product that did X?" These invite the polite lie. Hypotheticals about the future are worthless because people are genuinely bad at predicting their own behaviour.
Why behaviour beats opinion, with evidence
This is not just folk wisdom. Behavioural research has long shown a gap between what people say they will do and what they actually do, sometimes called the intention-behaviour gap. Stated intentions are weak predictors of future action. Past behaviour, especially behaviour that already cost the person time or money, is a far stronger predictor. So when someone tells you they already cobbled together a spreadsheet, paid for three tools, and assigned an intern to a problem, that is real signal. When they say "sounds cool," that is noise.
If you are talking more than a third of the time, you are pitching, not learning. Ask about their past behaviour, then shut up and let the customer fill the silence.
End with a commitment, not a thank you
A warm chat that ends in "thanks, this was helpful" gave you opinions. A real conversation ends in advancement: an introduction to the budget holder, a follow-up meeting, a request to be a pilot customer, or money. If someone genuinely has the problem, they will give you something that costs them something. If they will not, the enthusiasm was theatre. This connects straight back to Day 2: the truest customer research is a sale.
The takeaway
- People lie to be nice. Ask about their past behaviour so they cannot.
- Talk about their life, never your idea. Specifics over hypotheticals. Listen more than you speak.
- Money and time already spent on the problem is the strongest signal you can collect.
- End every conversation with a real commitment, not a compliment.
Frequently asked questions
What is the Mom Test?
A set of rules from Rob Fitzpatrick's book for customer conversations that produce honest answers. Ask questions so grounded in the person's real past that even your mother could not lie to flatter you. Talk about their life and problems, never your idea.
Why are most customer interviews useless?
Founders pitch the idea and ask hypotheticals like "would you use this," which invite polite, encouraging lies. People do not want to hurt your feelings and are poor at predicting their own behaviour, so you get false validation.
What questions should I ask?
Past and specific: What's the hardest part today? Tell me about the last time it happened. What did it cost you? What have you tried? What are you using now? These reveal real behaviour and spending.
How many interviews do I need?
Usually ten to twenty focused conversations within one segment before clear patterns emerge. Stop when you reliably hear the same problems and language. If every story differs, your segment may be too broad.
"You were the customer" is a cause of death.
Building for yourself instead of the market gets its own autopsy in Kill My Startup. Avoid the trap.
Buy on Amazon →Sources
- Fitzpatrick, R. The Mom Test (2013).
- Sheeran, P. "Intention-Behavior Relations," research on the intention-behaviour gap.
- Blank, S. customer discovery, The Four Steps to the Epiphany (2005).