42 Interviews, Zero Customers
Every single interview went great. Then she opened the payment page and heard crickets. Tell her what her interviews actually measured.
You're advising Sana, founder of Nestled, an app that helps new parents batch-plan and share childcare logistics with grandparents and babysitters. Before writing a line of code, she did what every startup guide told her to: talked to customers. She ran 42 interviews over two months with parents from her prenatal class group chat, her sister's mom-friend network, and a parenting Slack she's active in.
"I did it right," she tells you. "I didn't just ask my mom. I asked real parents, and 39 out of 42 said they'd definitely use something like this, several said they'd pay, one dad said it was 'exactly what our family needs.' That's overwhelming validation. So I built it."
She launched two weeks ago: a free tier plus a $9/month plan. She's showing you the numbers, confused.
| Metric | Value |
|---|---|
| Discovery interviews conducted | 42 |
| Interviewees who said "I'd definitely use this" | 39 / 42 (93%) |
| Interviewees who said "I'd pay for this" | 14 / 42 (33%) |
| Interviewees asked to pre-order or commit a card | 0 / 42 |
| Interviewees who were personal contacts or friends-of-friends | 40 / 42 |
| Signups since launch (2 weeks) | 61 |
| Signups who upgraded to paid | 1 |
| Signups from the original 42 interviewees | 9 (all on free tier) |
| Typical interview question used | "Would you use an app like this?" / "Would you pay $9/month for it?" |
Looking back at her notes, almost every interview followed the same shape: she'd describe the idea, the parent would say something like "oh that's such a good idea, we struggle with this so much," and Sana would write down 'validated pain point' and move to the next person. Nobody was asked what they currently use, how much time or money the current mess actually costs them, or to do anything besides answer a hypothetical.
Data snapshot
Sana believes her 42 interviews were strong validation and is confused why the launch numbers don't match. Explain what actually went wrong with how she ran those interviews — be specific about the mistakes, not just 'talk to more people.' Then tell her what she should have asked instead, and give her a concrete plan for the next 20 conversations that would produce real evidence instead of compliments.
Rubric
- Names the specific interview failures. Identifies at least two concrete Mom Test violations: pitching the idea and asking for opinions/hypotheticals ('would you use/pay') instead of digging into current behavior, and never asking for any real commitment (pre-order, deposit, card) to separate talk from action.
- Diagnoses the sampling bias. Flags that 40 of 42 interviewees were personal network (friends/friends-of-friends/a community she's active in), who are primed to be polite and compliment her rather than give honest signal — distinct from the question-wording problem.
- Explains why compliments misled her. Explains that 'I'd definitely use this' and 'I'd pay for this' are unfalsifiable future promises/compliments, not data, and connects this directly to the outcome: 14 said they'd pay, 0 of those 14 actually paid when given the real chance.
- Gives corrected interview technique. Provides concrete better questions/behavior for the next 20 conversations — asking about current specific past behavior ('walk me through the last time this was a problem,' what they use today, what it costs them) and deflecting compliments instead of writing them down as validation.
- Requires a real commitment test. Proposes replacing hypothetical questions with an actual ask for commitment (pre-order, deposit, waitlist with a real cost) as the next validation step, ideally with strangers/target segment rather than warm network.