{"site":{"name":"Koji","description":"AI-native customer research platform that helps teams conduct, analyze, and synthesize customer interviews at scale.","url":"https://www.koji.so","contentTypes":["blog","documentation"],"lastUpdated":"2026-06-10T00:52:28.611Z"},"content":[{"type":"documentation","id":"94746df6-9e92-428a-867c-ad3bad7dbfba","slug":"mom-test-methodology","title":"The Mom Test: How to Talk to Customers Without Being Misled","url":"https://www.koji.so/docs/mom-test-methodology","summary":"This guide covers Rob Fitzpatrick's Mom Test methodology for conducting customer conversations that surface genuine insights. It explains the three core rules, the distinction between compliments and commitments, five common anti-patterns, and how to apply these principles to any research context.","content":"# The Mom Test: How to Talk to Customers Without Being Misled\n\nImagine you're building a new app. You describe your idea to your mom and ask, \"Do you think this is a good idea?\" She says, \"Oh, that sounds wonderful, honey!\" You walk away feeling validated.\n\nBut your mom wasn't being honest — she was being kind. And this happens in every customer conversation where you ask the wrong questions. That's the core insight of Rob Fitzpatrick's *The Mom Test*: it's your job as the interviewer to ask questions so good that even the people who love you most can't lie to you.\n\n## The Three Rules of The Mom Test\n\nFitzpatrick distills his methodology into three deceptively simple rules:\n\n### Rule 1: Talk About Their Life, Not Your Idea\n\nThe moment you start describing your product idea, you've contaminated the conversation. People will respond to your pitch, not to their own reality. They'll be polite, encouraging, and almost entirely unhelpful.\n\n**Instead of:** \"We're building a tool that helps teams track their decisions. Would you find that useful?\"\n\n**Ask:** \"When a decision gets made on your team, how do people who weren't in the room find out? Walk me through a recent example.\"\n\nThe first question invites a polite yes. The second reveals the actual problem (or lack thereof) in their real life.\n\n### Rule 2: Ask About Specifics in the Past, Not Generics or the Future\n\nPeople are terrible at predicting their future behavior. Research from Morwitz, Steckel, and Gupta published in the Journal of Marketing Research found that stated purchase intentions are poor predictors of actual behavior, with accuracy varying from 20% to 50% depending on the product category.\n\nBut people are reasonably good at recounting things that actually happened. Past behavior is the best predictor of future behavior.\n\n**Instead of:** \"How often would you use something like this?\"\n\n**Ask:** \"When was the last time this problem came up? What did you do about it?\"\n\n**Instead of:** \"Would you pay $20/month for this?\"\n\n**Ask:** \"Have you looked for solutions to this before? What did you try? Did you pay for anything?\"\n\nIf they haven't bothered to search for a solution, that tells you the problem isn't painful enough. If they've tried and failed with three competitors, that tells you the market is real but solutions are inadequate. Both are more valuable than a hypothetical \"yes, I'd pay for that.\"\n\n### Rule 3: Talk Less, Listen More\n\nIf you're talking more than 20% of the time, you're doing it wrong. Every minute you spend talking is a minute you're not learning.\n\nThis is harder than it sounds. When someone describes a problem you're solving, the temptation to jump in and explain your solution is almost irresistible. Resist it. Let them keep talking. The gold is in the details they share when they think you're just listening.\n\n## Signals That Matter\n\nFitzpatrick makes a crucial distinction between **compliments** and **commitments**. Compliments are worthless; commitments are gold.\n\n| Signal | Type | Value |\n|---|---|---|\n| \"That's a great idea!\" | Compliment | Worthless — social politeness |\n| \"Let me know when it launches.\" | Weak signal | Costs them nothing to say |\n| \"Can I be part of your beta test?\" | Commitment (time) | Moderate — they're offering real time |\n| \"Could you send me more info? I'd like to show my boss.\" | Commitment (reputation) | Strong — they're staking their credibility |\n| \"We'd pay for that. Can we do a pilot?\" | Commitment (money) | Very strong — money is the ultimate validation |\n\nDon't count compliments. Count commitments. The most encouraging words in customer research aren't \"That sounds amazing\" — they're \"When can I start using it?\"\n\n## Anti-Patterns to Avoid\n\nThe Mom Test is as much about what not to do as what to do. Here are the most common mistakes:\n\n### 1. Pitching Instead of Learning\n\n**The trap:** You describe your idea and then ask \"What do you think?\"\n\n**Why it fails:** You've anchored the conversation around your solution. Everything they say is now a reaction to your pitch, not a reflection of their reality. According to a study in Organizational Behavior and Human Decision Processes, anchoring effects are so powerful that even experienced professionals are significantly influenced by initial reference points.\n\n**The fix:** Don't mention your product until the very end of the conversation, if at all. The best interviews feel like you're learning about someone's life, not selling something.\n\n### 2. Fishing for Compliments\n\n**The trap:** \"Doesn't that process sound really annoying? Don't you wish someone would fix that?\"\n\n**Why it fails:** You're telegraphing the answer you want. Of course they'll agree — it costs them nothing and makes you happy.\n\n**The fix:** Ask neutral questions and let the participant's energy guide you. If they describe a process matter-of-factly without frustration, it might not actually be a problem worth solving.\n\n### 3. Accepting Fuzzy Data\n\n**The trap:** \"Yeah, that happens to us sometimes.\" You write it down as validation.\n\n**Why it fails:** \"Sometimes\" could mean once a year or once a day. Without specifics, you have no idea.\n\n**The fix:** Always drill into specifics. \"When was the last time that happened? Walk me through it.\" \"How often does this come up? Give me a rough number.\"\n\n### 4. Talking to the Wrong People\n\n**The trap:** You interview people who are easy to access rather than people who have the problem.\n\n**Why it fails:** Someone without the problem will give you theoretical answers. Someone living with the problem will give you visceral, detailed ones.\n\n**The fix:** Screen ruthlessly. Before the interview, verify that the participant has recent, relevant experience with the topic. See [finding research participants](/docs/finding-research-participants) for guidance.\n\n### 5. Not Seeking Commitments\n\n**The trap:** You have great conversations but never ask for a next step.\n\n**Why it fails:** You accumulate encouraging quotes but no evidence of real demand.\n\n**The fix:** End conversations by asking for a commitment appropriate to the stage: a referral to a colleague, a follow-up meeting, agreement to test a prototype, or a letter of intent.\n\n## Applying Mom Test Principles With Koji\n\nThe Mom Test principles apply to every interview you conduct, whether in person, over video, or through an asynchronous research platform. When running interviews through Koji, the same rules hold: your interview guide should focus on past behavior and real situations, not hypothetical scenarios or reactions to your idea.\n\nKoji's AI analysis can help you identify when participants are giving you compliments versus commitments — flagging generic positive statements versus specific evidence of real demand. This makes it easier to separate signal from noise across multiple interviews.\n\n## Mom Test Quick Reference\n\n| Do This | Not This |\n|---|---|\n| Ask about their life and behavior | Pitch your idea and ask for opinions |\n| Ground questions in specific past events | Ask about hypothetical future behavior |\n| Listen more than you talk (80/20 rule) | Dominate the conversation with your vision |\n| Dig into specifics when answers are vague | Accept \"sometimes\" or \"usually\" at face value |\n| Count commitments, not compliments | Feel validated by \"That sounds great!\" |\n| Seek advancement (next steps, intros, money) | End with \"Thanks, that was really helpful!\" |\n\n## Further Reading\n\n- [User Interview Guide](/docs/user-interview-guide) — the full interview methodology\n- [Writing Interview Questions](/docs/writing-interview-questions) — tactical question design that complements Mom Test thinking\n- [Choosing a Methodology](/docs/choosing-a-methodology) — when to use Mom Test versus other interview approaches\n\n## Further reading on the blog\n\n- [The Mom Test for Customer Interviews: How to Ask Questions That Get Real Answers (2026)](/blog/mom-test-customer-interviews-2026) — The Mom Test is the founder's guide to customer interviews that actually work. This complete guide covers Rob Fitzpatrick's three rules, rea\n- [Best User Research Tools for Startups in 2026: The Complete Buyer's Guide](/blog/best-user-research-tools-for-startups-2026) — 43% of startups fail because of poor product-market fit. The fix isn't more guesswork — it's better customer research. Here are the best use\n- [Concept Testing: The Complete Guide for Product Teams (2026)](/blog/concept-testing-guide-2026) — Concept testing validates whether your idea is worth building before you build it. This guide covers methods, question templates, analysis a\n\n<!-- further-reading:blog -->\n","category":"Research Methods","lastModified":"2026-05-13T00:25:38.788654+00:00","metaTitle":"Mom Test Methodology — Koji Docs","metaDescription":"Learn Rob Fitzpatrick's Mom Test rules for customer conversations that produce reliable data instead of polite encouragement.","keywords":["Mom Test","Rob Fitzpatrick","customer discovery","customer interviews","product validation","startup research","user research"],"aiSummary":"This guide covers Rob Fitzpatrick's Mom Test methodology for conducting customer conversations that surface genuine insights. It explains the three core rules, the distinction between compliments and commitments, five common anti-patterns, and how to apply these principles to any research context.","aiPrerequisites":["user-interview-guide"],"aiLearningOutcomes":["Apply the three core rules of The Mom Test to customer conversations","Distinguish between compliments and commitments in user feedback","Avoid the five most common anti-patterns in customer discovery","Ask questions grounded in past behavior rather than hypothetical futures","Evaluate whether a conversation produced reliable data"],"aiDifficulty":"beginner","aiEstimatedTime":"8 min read"}],"pagination":{"total":1,"returned":1,"offset":0}}