Lesson 3 of 6

Talk to Customers Without Being Lied To

Asking would you buy this gets polite lies. The Mom Test gets the truth by asking about the past.

8 min read

People will lie to you about your idea, and they will do it to be kind. Ask a friend, a family member, or a polite stranger whether your idea is good and they will say yes, because the truth feels rude and they have no stake in your success. The Mom Test, from Rob Fitzpatrick, is a simple rule that gets around this: ask questions even your mom could not lie to you about, by talking about their life instead of your idea.

Stop pitching, start interviewing

The instinct is to describe your idea and watch their face light up. This is the worst thing you can do, because the moment you pitch, the conversation becomes about your feelings, and the polite human across from you will protect them. They will say I would totally use that, which is a compliment, not a commitment, and you will walk away falsely encouraged.

Flip it. Do not mention your idea at all for as long as you can. Ask them to walk you through the last time they hit the problem you care about. What were they doing, what did they try, what did it cost them, what happened next. You are a detective collecting facts about their past behavior, not a salesperson collecting reactions to your future product.

Ask about the past, not the future

Anything someone says about what they would do, or would pay, or would love, is a guess about a hypothetical, and people are terrible at predicting their own behavior. Anything they tell you about what they already did is a fact. So aim every question at the past. Not would you use a tool that does this, but what did you do the last time this happened. Not would you pay for this, but what do you currently spend to deal with this.

The best signal is evidence they have already tried to solve it: a spreadsheet they built, a freelancer they hired, three apps they downloaded and abandoned. Someone who has spent time or money on a problem has shown you it is real and that they will reach for a fix. Someone who has done nothing about it, no matter how much they nod, is telling you it does not hurt enough.

Good: talk me through the last time this happened.
Good: what do you use today, and what is annoying about it?
Good: what does this problem cost you, in time or money?
Bad: would you use an app that solved this? (a hypothetical they will answer kindly)

Dig for the commitment, not the compliment

Compliments feel great and mean nothing. That is so cool, I love it, you should build it, are all fluff, and when you hear them you should get suspicious, not happy. Real interest shows up as commitment: they ask when they can have it, they introduce you to a colleague with the same problem, they offer to pay, they give you their email and actually open the follow-up. Anything that costs them something is signal. Anything free is noise.

If the conversation stays warm and complimentary but produces no commitment, that is your answer, and it is a useful one. Push gently for a next step. Can I show you what I am building next week. If they will not give you twenty minutes or an email, the enthusiasm was politeness. Better to learn that now than after you build.

Worked example

The same question, asked two ways

Imagine you want to build a meal-planning app for busy parents. The pitch version: I am making an app that plans your week of meals and builds your shopping list, would you use it. The parent, being kind, says oh that sounds great, yes. You leave thrilled and learn nothing, because there was no honest way for them to say no.

The Mom Test version: walk me through how dinner happened this week. When did you decide what to cook. What went wrong. Now you hear the truth: honestly we ordered takeout three times because I forgot to shop. That is a real, expensive, repeated problem, told as a fact about last week. You did not even have to mention your idea to find out it is worth building.

Learn by doing

Paste these into ChatGPT or Claude and run them against your own idea. The model will answer happily. Olune goes further and checks the answer against real Reddit threads, competitor maps, and keyword volume.

Tip: grab the downloadable playbook to run every play with fill-in worksheets →

Prompt 1 · Turn your idea into Mom Test interview questions.

My idea is [describe it]. The problem I think it solves is [problem], for [who]. Write me a 20-minute customer interview script that follows The Mom Test: every question is about their past behavior and current workarounds, none of them pitch my idea or ask about hypotheticals. Include a few follow-up probes for when they mention a workaround.

What a good output looks like

You get an opener that does not mention your product, questions like when did you last run into this and what did you do, what do you use now and what made you pick it, and probes like how much time did that take or what did you try before that. A script you could run on a call tomorrow.

Prompt 2 · Pressure-test what you heard in an interview.

Here is what a potential customer told me in an interview: [paste your notes]. Act as a skeptical coach. Separate the facts about their past behavior from the compliments and hypotheticals. Tell me what is real signal, what is politeness, and what specific commitment I should have asked for but did not.

What a good output looks like

It flags I would definitely use that as a compliment to discount, highlights I built a spreadsheet for this as strong signal, and points out that you never asked for the next step. Then it suggests the ask: a follow-up call, an intro, or a pre-order.

Key terms in this lesson

Takeaways

People lie to be kind. Never ask if your idea is good. Ask about their life.
Questions about the future get guesses. Questions about the past get facts. Always aim at the past.
Evidence they already tried to solve it, a spreadsheet, a freelancer, three abandoned apps, is the strongest signal there is.
Compliments are noise. Commitment, their time, an intro, or money, is signal.

Now run your own idea through it.

You have the method. Olune does the legwork: an honest build-or-kill verdict on live Reddit signals, competitor maps, and keyword volume, in about eight minutes. Free to start.

Test your idea →

Common questions

How many customer interviews do I need?

Fewer than you think to spot patterns, more than one to trust them. By interview five or so in a tight audience you will start hearing the same problems in the same words. Ten to fifteen good conversations usually tell you whether a problem is real and sharp. A consistent audience matters more than raw count.

Where do I find people to interview if I have no audience yet?

Go where they already gather. The subreddits, Slack and Discord groups, and forums for your audience are full of people with the problem. Be a human, reference a specific thing they posted, and ask for fifteen minutes to learn, not to sell. The Mom Test question generator can also turn your idea into a ready interview script.

Should I record the interviews?

With permission, yes, because you will miss the exact words otherwise, and the exact words are the gold. If recording makes it stiff, take rough notes and write up the facts right after while they are fresh. Capture what they did, not your interpretation of what it means.