The Mom Test: From Reading It at SFO to Using It in Every Customer Interview
Published:
Some books leave behind a few ideas. The Mom Test gave me a method I continue to use: do not ask whether someone likes your idea. Ask what they have already done, how often the problem occurs, what it has cost them, and whether they will commit to a next step.
Author: Koutian Wu; GitHub: ktwu01
I have preserved the complete condensed slide deck on this site:
- Open the English Mom Test slide deck
- Open the Chinese slide deck
- View the original page on Tacite AI
- Visit Rob Fitzpatrick’s official book site
The slides summarize the book; they do not replace it. The official site provides authorized links for the paperback, Kindle, audiobook, and PDF editions.
Why I started reading it
This was not a random item from a reading list.
First, Kehan strongly recommended The Mom Test in an entrepreneurship class. Then my cofounder ycao strongly recommended it again. Two independent recommendations pointed to the same book. That suggested customer interviewing was not a vague instruction to “talk to users,” but a skill that founders must train deliberately.
In May 2026, I read a condensed version at San Francisco International Airport (SFO). An airport is not an ideal environment for deep reading, but the shortened version exposed the central failure clearly: many “customer interviews” are founders presenting their solution and collecting politeness, encouragement, and speculation.
After reading it, I put the material into NotebookLM and asked it to quiz me. Reading alone makes recognition feel like understanding. Answering questions forced me to distinguish praise from evidence, interest from commitment, and future speculation from past behavior. After that exercise, the book became an interview checklist rather than something I had merely finished.
I have continued using it since.
How it changed my questions
The difference fits in one sentence: do not ask people to evaluate your idea; investigate the reality in which they already live.
| Question likely to produce a false signal | A Mom Test-style alternative |
|---|---|
| What do you think of this product? | When did you last encounter this problem? |
| Would you use this feature? | How do you handle it now? Walk me through the process. |
| Would you pay $20 per month? | What do you currently spend—in money or time—to solve it? |
| Is this problem important to you? | What happened the last time it caused trouble? |
| Do you like our solution? | What will you commit next: time, data, an introduction, a trial, or payment? |
The first column asks people to predict the future, protect my feelings, or respond to a pitch. The second requests events, processes, costs, and actions. The first can produce sentences that excite a team. The second can change a product decision.
Four rules I kept
1. Ask about the past instead of requesting a prediction
“I would use it” and “I might buy it” cost nothing. Behavior that has already occurred is harder to invent. When a problem matters, people often use a spreadsheet, manual workflow, contractor, competing tool, or an awkward workaround to address it.
2. Talk less about the solution
Once I explain the product, the conversation turns from investigation into selling. The other person begins evaluating my presentation and idea instead of revealing their workflow. The interview is not for helping them understand me. It is for helping me understand them.
3. Push toward a specific event
Words such as “usually,” “everyone,” and “often” convert evidence into generalization. Better follow-ups ask: When did this last happen? Who was involved? What did you do first? Where did the process stop? What did the failure cost?
4. Use commitment to separate demand from politeness
A compliment is not a commitment. A stronger signal carries a cost: scheduling another meeting, introducing a decision-maker, sharing data, testing an unfinished product, signing a letter of intent, or paying. The more specific the commitment, the more information it contains.
Why I am preserving the HTML here
I do not want this learning history to collapse into “I read The Mom Test.” Keeping the slide deck available gives me a quick review before interviews and gives collaborators a shared reference: are we collecting facts, or are we collecting reassurance?
It also preserves the path by which the method entered my work. Kehan recommended it in an entrepreneurship class; ycao recommended it again; I read the condensed version at SFO in May 2026, used NotebookLM questions to test my understanding, and carried the method into later interviews.
A book enters practice not when I finish its last page, but when I am about to ask, “What do you think?” and replace it with:
Tell me exactly what happened the last time you dealt with this problem.