Mom TestCAUTION

Wondel.ai Agent Skills — Business, Marketing, UX & Coding Frameworks from Bestselling Books. 50 skills + 12 guided journeys for Claude Code, Codex, Cursor & other agentskills.io agents.

The instruction file, verbatim from the audited commit — this is the text the model reads, and the surface the audit's instruction layer examines. Quoted here so you can judge it without cloning anything.

---
name: mom-test
description: 'Talk to customers without leading them using Mom Test rules: discuss their life not your idea, ask about specifics in the past, and talk less. Use when the user mentions "customer interviews", "validate my idea", "users say they want it but dont buy", "leading questions", "The Mom Test", "customer feedback bias", or "interview script". Also trigger when preparing user-research questions, interpreting ambiguous feedback, or designing customer-discovery that avoids false positives. Covers commitment and advancement, avoiding compliments, and extracting signal from noise. For product-market fit, see jobs-to-be-done. For rapid prototype testing, see design-sprint.'
license: MIT
metadata:
  author: wondelai
  version: "1.4.0"
---

# The Mom Test Framework

Framework for customer conversations that won't lead you astray, based on a fundamental truth: everyone is lying to you -- not maliciously, but because you're asking the wrong questions. The Mom Test provides rules for asking questions so good that even your mom can't lie to you.

## Core Principle

**Good customer conversations are about their life, not your idea.** The moment you mention what you're building, people switch from sharing truth to performing politeness. Talk about their problems, their lives, and their existing behavior instead of pitching, and ask about specifics in the past, not hypotheticals about the future. Above all, talk less and listen more.

## Scoring

**Goal: 7/7.** Score a conversation (or interview plan) by the seven-row Quick Diagnostic below: **1 point per row that passes.**
- **6-7** = focused on their life and past behavior, concrete facts captured, a real commitment (time/reputation/money) secured, they talked 80%+, and beliefs got updated.
- **4-5** = some past-behavior facts but leaking into hypotheticals, compliments accepted as signal, or ending without an ask.
- **<=3** = a pitch in disguise: leading questions, opinions and fluff, a polite zombie lead, no learning.

Always state the current score out of 7, name the failing diagnostic rows, and give the specific fix for each.

## Framework Sections

### 1. The Mom Test Rules

**Core concept:** Three rules that make it impossible for even your most supportive loved ones to give you false validation, shifting conversations from opinion-gathering to fact-finding.

**Why it works:** People are unreliable predictors of their own future behavior, so opinions are worthless. Past behavior is the only reliable data and can genuinely inform product decisions.

**Key insights:**
- Rule 1: Talk about their life, not your idea -- never mention your solution until the end, if at all
- Rule 2: Ask about specifics in the past, not generics or hypotheticals about the future
- Rule 3: Talk less, listen more -- aim for them to speak 80% of the time
- A question fails the Mom Test if the answer is always "yes" regardless of whether the business will succeed
- Good questions could potentially destroy your currently imagined business

**Product applications:**

| Context | Application | Example |
|---------|-------------|---------|
| Idea validation | Ask about the problem, never the solution | "Tell me about the last time you tried to [problem area]" not "Would you use an app that does X?" |
| Feature prioritization | Discover what people do vs. what they say | "Walk me through how you handled this last week" |
| Pricing research | Anchor to existing spending behavior | "What are you currently paying to solve this?" not "Would you pay $X?" |

**Copy patterns:**
- "Tell me about the last time you..."
- "What else have you tried?"
- "Why does that bother you?"

See: [references/question-patterns.md](references/question-patterns.md) when drafting an interview script -- a 5-tier question hierarchy, domain-specific question banks (SaaS/consumer/marketplace), and four formulation exercises.

### 2. Good vs Bad Questions

**Core concept:** Most interview questions are broken because they ask people to predict the future, evaluate hypothetical products, or confirm your assumptions. Good questions anchor in observable past behavior and extract concrete facts.

**Why it works:** Asking "would you buy this?" is like asking "will you go to the gym next week?" -- the answer is always yes, the follow-through rarely there. Behavior that already happened can't be rationalized away.

**Key insights:**
- Bad: "Do you think it's a good idea?" -- always gets a yes
- Bad: "Would you buy a product that does X?" / "How much would you pay?" -- hypothetical, anchored to please you
- Good: "How are you dealing with this problem today?" -- reveals actual behavior
- Good: "What have you tried before and why did you stop?" -- reveals past decisions
- Good: "Where does the money come from for solutions like this?" -- reveals real budgets
- The scariest questions -- ones with the power to change what you're building -- produce the most useful data

**Product applications:**

| Context | Application | Example |
|---------|-------------|---------|
| Problem validation | Confirm the problem exists and matters | "When did this last come up? What did you do? What didn't work?" |
| Market sizing | Check if enough people share the problem | "Who else in your industry deals with this? How do they handle it?" |
| Competitive analysis | Find real alternatives already in use | "What tools/processes do you currently use for this?" |

**Copy patterns:**
- "What's the hardest part about [doing this thing]?"
- "How often does this come up?"
- "Walk me through what happened the last time this came up"

### 3. Avoiding Compliments and Opinions

**Core concept:** Three types of bad data feel like progress but actively mislead: compliments ("That's a great idea!"), fluff (hypotheticals, maybes, future promises), and ideas (feature requests disconnected from real problems). Deflecting these and digging for truth is the core skill.

**Why it works:** Compliments are the fool's gold of custome

CAUTIONgrade B · trust 89/100 Install with care. The audit found things worth knowing before you trust its output.

What the source does

Filesystem

none-observed

Network

none-observed

Shell

none-observed

Dependencies

pinned

Secrets in source

none-found

Findings (5)

.agents/skills/37signals-way

.agents/skills/architecture-optimization

.agents/skills/blue-ocean-strategy

.agents/skills/clean-architecture

.agents/skills/clean-code

Gates applied: no_behavioural_pass.