The Mom Test for Startup Ideas: How to Get Honest Feedback

Imagine you tell your mom about your startup idea.

She listens. She smiles. She says: "That sounds great, honey. I'd definitely use that."

You leave feeling validated. You have confirmation. Your mom said she'd use it.

Except you haven't learned anything. Because your mom loves you. She wants you to succeed. She has no idea what a good startup idea looks like, and even if she did, she'd still tell you yours was great.

This is the problem Rob Fitzpatrick identified in his book The Mom Test. The feedback most founders collect is useless -- not because the people giving it are lying, but because the questions being asked make honest answers almost impossible.

The Mom Test is a way to fix that. Here's how it works.

The Core Idea

The Mom Test gets its name from a simple principle: ask questions in a way that even your mom can't answer dishonestly.

This sounds counterintuitive. The problem isn't that people want to mislead you. It's that certain types of questions -- the ones founders naturally gravitate towards -- make it nearly impossible for anyone to give you useful information.

When you ask "Would you use this?", the honest answer requires the person to predict their own future behavior, weigh how much they like you, and decide whether being encouraging is kinder than being accurate. Most people, most of the time, will default to encouraging. Especially people who care about you.

The fix is to stop asking people to evaluate your idea. Instead, ask them about their life.

The Three Rules of the Mom Test

Fitzpatrick distills the method into three simple rules.

Rule 1: Talk about their life, not your idea.

Instead of describing your product and asking for reactions, ask about how they currently handle the problem your product would solve. Ask about the last time they encountered it. Ask what they did. Ask what was annoying about it.

This takes the attention off your idea entirely. There's nothing to be polite about. People can just describe their actual experience honestly.

Rule 2: Ask about specifics in the past, not generics about the future.

"Would you ever use a tool that helped you track client pitches?" is a future-hypothetical question. It's useless.

"Walk me through the last time you had to follow up on a pitch you sent a client" is a specific-past question. It's invaluable.

Past behavior is real. Future hypotheticals are fantasy. People are very bad at predicting what they'll do in situations they haven't experienced. But they can describe exactly what they actually did last Tuesday.

Rule 3: Listen, don't talk.

This is the hardest rule. Most founders go into customer conversations with an agenda. They want to tell people about their idea. They want to describe the features. They want to see the person's face light up.

Resist this. Your job in a Mom Test conversation is to listen. Ask questions, then be quiet. Let silence sit. Follow up on the things they say, not the things you want to talk about. Take notes.

Every time you spend 30 seconds describing your idea, you're wasting 30 seconds you could have spent learning something.

What Good Questions Look Like vs. Bad Questions

The difference between a useful customer question and a useless one is almost always the same: useful questions are about the past, bad questions are about the future or the idea.

Here are some examples:

Bad: "Do you think people would pay for an easier way to manage freelance invoices?" Good: "How do you handle invoicing right now?"

---

Bad: "Would this feature be useful to you?" Good: "Have you ever tried to solve this problem before? What happened?"

---

Bad: "How much would you pay for something like this?" Good: "How much does this problem cost you right now -- in time, in tools, in what you've already tried?"

---

Bad: "What do you think of my idea?" Good: "What's the hardest part of [the thing your idea helps with]?"

---

Notice that none of the "good" questions mention your idea at all. You're collecting data about their reality, not their opinion of your concept. Their reality is honest data. Their opinion of your concept is influenced by how much they want to encourage you.

The Signs You're Getting Bad Data

Even when you're not explicitly asking people to evaluate your idea, bad data sneaks in. Here are the signals to watch for.

Compliments: "That's so cool." "I love the concept." "You're so smart for thinking of this." These feel great and mean nothing. When you hear a compliment, say thank you and move on. Don't let it inflate your confidence. Follow up with: "What would make you actually use something like this?"

Hypotheticals: "I would definitely do X if Y." Whenever a sentence contains "would" or "if," treat it with skepticism. It's a prediction, not a fact. Respond by anchoring to reality: "Has there been a time when you actually ran into that situation?"

Vague agreements: "Yeah, that's definitely a problem." Probe deeper. "When did it last happen? What did you do about it?" Vague agreement is cheap. Specific stories are gold.

Enthusiasm without action: Someone who says they love the idea but doesn't sign up for your waitlist, doesn't introduce you to a colleague with the same problem, and doesn't ask when it's launching -- that's a fan, not a customer. Be wary of people who are enthusiastic in conversation and passive everywhere else.

The Three Listening Modes

When talking to potential customers, there are three things worth listening for. Fitzpatrick calls them the three types of data.

Facts: What is actually true about their current situation? What tools are they using? How much time are they spending? What have they already tried? These are your most reliable signals.

Emotions: Not just what they say, but how they say it. Did they sit forward when they described a problem? Did they sound tired or resigned? Authentic frustration -- the kind where they've been dealing with something for years and haven't found a good solution -- is a strong signal. Mild inconvenience is not.

Behaviors: What have they already done about the problem? Have they paid for something? Have they tried to build a workaround? Have they asked around for recommendations? Prior behavior is the closest proxy you have for future action.

A person who has tried three different tools and is still unhappy is a far better prospect than a person who agrees the problem sounds annoying but has never done anything about it.

How to Structure a Mom Test Conversation

You don't need a rigid script. But having a loose structure helps.

Opening (2 minutes): Set context and lower the stakes. "I'm doing some research on [problem area] and I'd love to learn about your experience. There's no pitch -- I just want to understand how you currently handle this."

This signals: I'm not here to sell you something. You can be honest.

The main conversation (15-20 minutes): Ask open questions about their current process. Follow up on anything interesting. Let them talk. Take notes.

A good opener: "Walk me through how you currently handle [the thing your product does]."

Then follow threads. If they say "it's kind of a mess right now," ask: "What makes it messy?" If they say they've tried different tools, ask: "What did you try? What happened?"

The close (5 minutes): Ask two things. First: "Is there anyone else you think I should talk to about this?" A referral is a sign they found the conversation valuable. Second: Show them your landing page or waitlist (if you have one) and observe their reaction. Watch whether they sign up without prompting.

The whole thing is 20-30 minutes. Keep it short. Respect their time.

A Common Mistake: Asking Too Early

Many founders do customer interviews after they've already built something. By then, they're not doing a Mom Test -- they're doing damage control.

The Mom Test works best before you've written a line of code. Before you've built anything. Before you've gotten attached to a solution.

At that stage, you're genuinely curious. You have no product to defend. You can follow the conversation wherever it goes. And what you learn will shape everything about what you build.

Once you've spent three months building something, it's very hard to hear feedback that suggests you built the wrong thing. The emotional investment makes it harder to absorb honest information.

Do the conversations first. Build second.

The Simplest Takeaway

Here's the Mom Test distilled to one sentence: stop asking people what they think of your idea, and start asking them about their life.

Their opinion of your idea is shaped by social dynamics, politeness, and their desire to encourage you. Their life is just their life. It's honest by default.

Ask about the last time they faced the problem. Ask what they did. Ask what didn't work. Ask what they wish existed.

And then listen. Really listen.

That's where the useful signal lives -- not in the reaction to your pitch, but in the unfiltered description of their daily reality.

If you hear the same frustrations from enough strangers, in specific enough terms, with enough emotion behind them -- that's the closest thing to a green light you can get before building. And it costs nothing but time and curiosity to collect.