Your mom will always tell you your business idea is great. Not because it is — but because she loves you and doesn’t want to see you hurt. This is the fundamental insight behind Rob Fitzpatrick’s “The Mom Test,” and it took me an embarrassingly long time to internalize it.
The rule is simple: you should be able to talk to your mom about your business and get useful data. But only if you never mention the business itself. The test isn’t about your mom, of course. It’s about the fact that even the most biased person in the world can give you truthful data if you ask the right questions.
After applying these principles across dozens of validation rounds with startups in the Startup Burgenland accelerator and in my own ventures, I’ve collected a library of real scenarios where founders got it right — and more often, where they got it catastrophically wrong. Let me walk you through both.
Scenario 1: The Meal Planning App
A founder I advised — let’s call her Sarah — wanted to build a meal planning app for busy parents. She’d been a busy parent herself and hated the nightly “what’s for dinner?” panic. Classic scratching-your-own-itch scenario.
How she originally conducted interviews:
Sarah would describe the app — weekly meal plans, automated grocery lists, budget tracking — and ask parents, “Would you find this useful?” Every single person said yes. She had 15 interviews and 15 positive responses. She felt validated.
What went wrong:
She built the app. It was decent. She launched it. Almost nobody signed up, and those who did stopped using it within two weeks. The problem? Parents already had solutions. They used Pinterest boards, saved Instagram recipes, or just… bought the same five meals every week. The “pain” was real but the motivation to switch to an app was almost zero.
What the Mom Test approach would have revealed:
Instead of describing the app, Sarah should have asked: “Walk me through what happened last Tuesday night around dinner time.” Then: “Have you tried anything to make dinner planning easier? What?” Then: “How much would you say you spend per month on food that goes to waste because you didn’t plan?”
These questions would have surfaced that most parents have found workable (if imperfect) solutions, that the emotional pain is high but the actual behavioral drive to change is low, and that food waste — not dinner panic — was the money problem parents would actually pay to solve.
The Mom Test doesn’t tell you the idea is bad. It tells you what the actual problem landscape looks like so you can make informed go/no-go decisions.
Scenario 2: The Freelancer Invoice Tool
A developer named Marcus wanted to build an invoicing tool specifically for freelance developers. His pitch: existing tools like FreshBooks were too complex, and simple options like Wave didn’t handle the specific quirks of developer billing (hourly vs project-based, retainers, milestone payments).
The bad interview:
“I’m building an invoicing tool just for developers. It handles hourly, project, and retainer billing in one place. What do you think?”
Response from every developer: “Oh cool, yeah, invoicing is a pain.”
Marcus took this as validation. It wasn’t. “Invoicing is a pain” is a universal truth. The sky is blue. Water is wet. Invoicing is a pain. This tells you nothing about whether someone would switch tools and pay money for your specific approach.
The Mom Test version:
“How do you currently handle invoicing? Walk me through what you did for your last invoice.”
What Marcus discovered in proper interviews: most freelance developers had already solved invoicing. They had templates. They had systems. Some used Notion, others used plain email with a PDF attached. The pain wasn’t invoicing — it was chasing late payments. That’s a completely different product.
The developers who did express frustration? They were early-career freelancers who hadn’t built systems yet. That’s a different market segment with different willingness to pay. This distinction only emerged when Marcus stopped talking about his solution and started asking about behavior.
He eventually pivoted to a payment reminder tool with automatic follow-ups. It was a smaller idea than the original invoicing tool, but people actually paid for it. Starting with the right problem matters more than starting with a big vision.
Scenario 3: The B2B Compliance Dashboard
This one comes from the accelerator. A team wanted to build a compliance dashboard for Austrian SMEs to track GDPR, tax, and labor law requirements. Noble idea. Real problem. And they validated it wrong in a textbook way.
What they did:
They created a beautiful pitch deck with mockups and presented it to 10 small business owners. Nine out of ten said, “Yes, compliance is a nightmare. I’d pay for something that simplifies it.”
The team celebrated. They started building.
What they should have done:
Ask: “When was the last time you had a compliance issue? What happened?” Then: “How do you currently track your compliance requirements?” Then: “How much did you spend on compliance-related services last year?”
What they would have found: most SME owners handle compliance by ignoring it until their accountant yells at them once a year. They don’t track it. They don’t want to track it. They want someone else to handle it entirely. A dashboard that helps you do something you’re determined to not do is a product without a customer.
The real opportunity was a done-for-you compliance service, not a dashboard. Same problem space, completely different solution — and one that people had already demonstrated willingness to pay for (through accountants and lawyers).
The Five Questions That Always Work
Across all these scenarios, I’ve found five questions that consistently produce useful, honest data regardless of the industry or problem space.
Question 1: “Tell me about the last time you dealt with [problem area].”
This anchors the conversation in reality. If they can’t think of a specific instance, the problem isn’t top-of-mind enough to build for.
Question 2: “What did you do about it?”
This reveals existing solutions and behavioral patterns. If they did nothing, the problem isn’t urgent. If they cobbled together a workaround, there’s opportunity.
Question 3: “What’s the hardest part about how you currently handle this?”
This identifies the specific friction within the broader problem. Your product doesn’t need to solve everything — it needs to solve the hardest part.
Question 4: “How much time or money does this cost you?”
This quantifies the problem. If someone spends 30 minutes a week on something, that’s about 25 hours a year. Is that enough pain to justify switching to a new tool? Usually not. If they spend 5 hours a week, that’s a different story.
Question 5: “If you could fix one thing about this situation, what would it be?”
This reveals priorities. Often, the thing they want to fix isn’t what you’d expect. Your assumptions about what matters most are usually wrong. Their answer to this question is usually right.
Notice what these questions have in common: none of them mention your idea. None of them ask about the future. All of them focus on behavior, not opinions. That’s the Mom Test in five questions.
I keep these taped to my monitor when I’m doing interview rounds. It’s embarrassing how often I still catch myself drifting into pitch mode, even after years of doing this.
When Positive Signals Are Actually Noise
Here’s the part that makes founders uncomfortable: even with good interview technique, you’ll get false positives. Here’s how to spot them.
The Enthusiastic Non-Buyer. This person loves your concept, asks great questions, suggests features, and gets excited during the call. But when you say “Would you be willing to pre-order?” they suddenly need to “check with their partner” or “think about it.” Enthusiasm without commitment is noise.
The Niche Outlier. One person in your interview batch has an extreme version of the problem. They’d pay anything. They need this yesterday. But they’re the only one. Building for one person’s extreme pain doesn’t make a business unless you can find 99 more like them.
The Feature Request Trap. When people start suggesting features, it feels like validation. “Oh, and it should also do X, and Y, and what about Z?” But this is brainstorming, not buying. People will enthusiastically design a product they’ll never purchase. Feature suggestions only matter from people who’ve already committed money.
The “My Friend Would Love This” Deflection. When someone says “I know someone who’d be perfect for this,” they’re politely telling you they’re not the customer. Don’t get excited about secondhand referrals. Get excited about first-person buying behavior.
The counter to all of these is the same: ask for commitment. “Will you pay for this today?” cuts through every form of false positive. It’s uncomfortable to ask. Ask anyway. Being comfortable with discomfort is the price of accurate validation.
Tracking and Interpreting Interview Data
After each interview, I fill out a simple scorecard:
| Signal | Score |
|---|---|
| Described specific, recent instance of problem | +2 |
| Currently spending money on a bad solution | +3 |
| Expressed emotional frustration about the problem | +1 |
| Agreed to a follow-up call or test | +2 |
| Offered to pre-order or pay | +5 |
| Couldn’t recall specific instance | -2 |
| Said “interesting” but no commitment | -1 |
| Changed subject when discussing money | -3 |
After 10-15 interviews, I total the scores. If the average is below +3, the opportunity is weak. Between +3 and +6, it’s worth a small experiment to test further. Above +6, you probably have something worth building.
This is obviously reductive — you can’t turn qualitative research into a single number and pretend it’s science. But the scorecard forces you to evaluate evidence rather than feelings. And feelings are what lead founders astray.
The best founders I’ve worked with treat customer interviews the way scientists treat experiments: they go in with a hypothesis, collect data with as little bias as possible, and update their beliefs based on what they find. The worst founders treat interviews like focus groups for their existing idea — looking for confirmation, ignoring disconfirmation.
Don’t be the second type. The truth might be uncomfortable, but building on a lie is far more painful.
Key Takeaways
- The Mom Test is about question design, not about avoiding your mom. You can get truthful data from anyone if you never mention your idea and only ask about their behavior.
- Five questions work in any context: What happened? What did you do? What’s hardest? What does it cost you? What would you fix?
- False positives are more dangerous than false negatives. Learn to recognize enthusiasm without commitment, niche outliers, feature brainstorming, and referral deflections.
- Use a scorecard to force evidence-based decisions. Feelings aren’t data. Past behavior, money spent, and commitment made — those are data.
- The goal of interviews is to find the truth, not to confirm your idea. If the truth kills your idea, that’s the best possible outcome at this stage.
