The Mom Test is a short, practical book by Rob Fitzpatrick on how to talk to customers properly. The title comes from the idea that even your mom would lie to you about your business idea if you ask the wrong questions. The solution is asking questions she can't lie about - ones grounded in her actual life, not your idea.
It's the most useful thing I've read on customer discovery. Short, direct, and practical.
The Mom Test itself
Stop asking people what they think of your idea. That's the trap. Instead, ask about their life, their problems, and their past behaviour. Facts don't lie, opinions do.
Three rules:
- Talk about their life, not your idea
- Ask about specifics in the past, not generics or hypotheticals
- Listen more than you talk
Bad: "Would you pay for an app that does X?" - They'll say yes to make you feel good.
Good: "How do you currently handle X? How much did that cost you last time? How often does it come up?" - This reveals whether the problem is real and painful enough.
The three types of bad data
Most customer conversations produce useless data. Learn to spot it:
- Compliments - "That's a great idea!", "I love it!" - Worthless. Treat all compliments as red flags.
- Fluff - "I would definitely use that", "Everyone I know needs this" - Hypothetical future behaviour is not signal.
- Ideas - Customers will throw feature requests at you. Don't build from these directly. Dig into the problem behind the request instead.
When you hear any of these, redirect: "When was the last time you dealt with this? What did you do?"
Asking about the past, not the future
The most reliable signal comes from what people have already done, not what they say they will do.
- "Have you tried to solve this before?" - if not, the problem might not be painful enough
- "What did you use? How much did you pay?" - reveals real spend and effort
- "Why did you stop using it?" - usually the most revealing question
If someone has never tried to solve the problem even with a hacky workaround, that's a warning sign.
Don't pitch early
Founders pitch when they should be listening. The moment you reveal your solution, people stop telling you the truth and start trying to be helpful or polite.
Hold the pitch as long as possible. Your goal in the first conversation is to learn, not to sell. You'll know enough to pitch later. You won't get that learning back.
Commitment signals over compliments
A conversation that ends with "sounds great, let me know when it launches" is a failure. Real validation is a commitment that costs something:
- Time - they agree to a follow-up, a test, a pilot
- Reputation - they offer to intro you to others
- Money - a pre-order, a deposit, a signed LOI
If someone won't give any of these, their enthusiasm means nothing. Push for advancement, not compliments.
Finding the right people to talk to
Don't interview everyone. Find the people who have the problem acutely right now. Signs you've found them:
- They've already tried to solve it (even badly)
- They can name the last time it happened and what it cost them
- They're visibly frustrated when talking about it
Warm intros are better than cold outreach. Ask at the end of every conversation: "Who else should I be talking to?"
Keeping conversations casual
Formal interviews put people in performance mode. Casual conversations get you the truth.
You don't need to announce "I'm doing customer research." Just talk to people. Bring it up naturally. The best conversations don't feel like interviews to either party.
Keep notes. Review them afterwards with the team. The patterns across conversations - not any single one - are what actually matter.
The point of it all
You're not trying to prove your idea works. You're trying to find out if it does, as cheaply and quickly as possible. That means actively looking for reasons it won't work, not collecting reassurance.
A single "no" backed by specifics is worth more than ten enthusiastic "yes"es with no commitment attached.