The Mom Test: Rob Fitzpatrick's rules for customer conversations, wit...

The Mom Test is a short book by Rob Fitzpatrick about how to talk to potential customers without being misled by their politeness. Its subtitle describes it as a guide to learning whether your business is a good idea "when everyone is lying to you" (The Mom Test). The core advice fits in two sentences. Ask people about their own lives and what they have actually done, not about your idea or what they might do someday. Then judge their interest by what they commit to, not by what they say.

A review by the venture firm Unusual Ventures notes the book was published in 2013 and revised in 2014 (Unusual Ventures). The author's site says it takes about two hours to read. This guide summarizes its approach in our own words and applies it to a hypothetical example. Read the book itself for the full argument and its example conversations.

Where the name comes from

If you ask your mother whether your business idea is good, she will probably tell you what you want to hear. Fitzpatrick takes the point further: the question is badly framed, and everyone you ask it will give you some degree of misleading feedback (Unusual Ventures). People are kind, they don't want to discourage you, and predicting your own future behavior is hard.

The name suggests the standard to aim for: questions so grounded in the other person's own experience that even your mother couldn't give you a misleading answer.

The three rules, in plain terms

The book reduces good questions to three rules. Paraphrased:

Keep the conversation on their life rather than your idea. Once you describe your product, the conversation tends to become about being nice to you.
Ask about specific things that already happened. "When did this last happen?" is more reliable than "Do you usually…?" or "Would you…?"
Let them do most of the talking. Every minute you spend explaining is a minute they spend reacting to you instead of describing their world.

A worked example

This example is hypothetical. A founder is considering software that fills last-minute cancellations at dental practices by texting patients on a waitlist. She has three calls booked with office managers at independent practices.

Her first draft of questions sounded reasonable, but each one invited a guess or a compliment. Here is how the three rules change them:

First draft	Rewritten
Would you use an app that fills cancelled appointments automatically?	Tell me about the last time a patient cancelled on the same day. What happened next?
How much would you pay for something like this?	What does an empty chair cost the practice? Have you tried anything to fill them, and what did that cost?
Do cancellations bother you?	How many same-day cancellations did you have last week? How do you know?
Don't you think patients would like getting a text?	When an opening comes up today, how do you contact patients? Who does it?

The rewritten questions can produce bad news, which is what makes them useful. If an office manager says the last same-day cancellation was filled within ten minutes from a paper list at the front desk, the founder has learned something important before writing any code. If she says the practice loses two or three appointments a week and nobody has time to call around, the next questions follow naturally: who notices, what they have tried and who would decide to pay for a fix.

Customer interview questions you can use on your next call has more questions built the same way.

Three kinds of answers to treat with suspicion

The book's second chapter covers what the author's teaching notes call the three major causes of bad and biased data (The Mom Test: teacher resources). In our words, with what to do about each:

Compliments. "That sounds great" tells you the person is polite. Thank them and steer back to facts: "How do you handle that today?"
Generalities and predictions. Answers built on "usually," "always," "never" or "would" describe a pattern they believe in or a guess about the future. Unusual's summary of the book's advice is to anchor these back to specifics from the past and to what the person is already doing.
Ideas, requests and strong opinions. Treat a feature request as a clue to an underlying problem. Ask what the feature would let them do and why that matters.

How to tell whether a conversation went well

A pleasant conversation and a useful one are different things. According to Unusual's summary, the book treats a meeting as a success when it ends with a commitment or a clear next step, where commitment means the other person gives up something they value: their time for a follow-up with a set purpose, their reputation by introducing you to a colleague or decision maker, or their money (Unusual Ventures). The more they give up, the more seriously you can take what they said.

In the dental example, strong outcomes after a first call might be:

The office manager agrees to share a count of last month's same-day cancellations, with no patient details.
She introduces the founder to the practice owner, who would approve any purchase.
The owner agrees to pay for a one-month trial.

By that standard, "let me know when it's ready" ends the call politely without moving anything forward. Market validation shows how to rank these signals before you build.

Prepare, keep it casual and write things down

Three habits from the book, as Unusual summarizes them:

Decide your three most important questions before each conversation. At least one should be a question whose answer could sink the idea. For the founder, that might be "Do practices already fill most same-day cancellations without help?"
Keep it informal. Learning about someone's problems often works better as a quick, casual chat than as a formal meeting. If it feels like they are doing you a favor by talking to you, the setup is probably too formal.
Take notes and review them with your team promptly. Record exact words, strong emotions and specific names and companies, then update your three big questions based on what you heard.

How to take customer interview notes covers a format that keeps what people said separate from what you think it means.

What the book doesn't try to do

The Mom Test is about conversations. It will help you find out whether a problem exists and how people deal with it now. It is not built to tell you how common the problem is across a market or whether people can use what you build. Those questions need other methods, such as a count from a larger sample or a usability test with a working prototype.

Two other cautions. The rules govern what you ask, and whom you ask matters as much: careful conversations with people who don't have the problem will still mislead you, so choose interviewees with care. And keeping your idea out of early conversations does not mean never discussing it. Once you understand the problem, a solution interview tests your approach, and asking for a commitment tells you whether people want it.

The book is a required text in some university courses, including London Business School's MBA, according to the author's teaching page. His personal site says he is working on a second edition (robfitz.com).

Your next step

Write down the three questions whose answers matter most to your idea, including one that could prove you wrong. Rewrite every question on your list that mentions your product or starts with "would you," so that it asks about a recent, specific event instead. Then decide what commitment you will ask for at the end of each conversation.

If you can't find the people to have these conversations with, Instant Expert can find people who match a description you write, such as "office managers at independent dental practices." You review who it finds, it sends your invitations, and you pay only for calls that get booked. The operations professionals in dental services directory page is one place to start. If you pay, pay for their time rather than a favorable answer: paying for participation without paying for positive feedback explains how.