“When was the last time this problem actually cost you something, and what did you do about it?”
From Rob Fitzpatrick · Author of The Mom Test
Why it works
The Mom Test's core rule is that you cannot trust opinions about the future — only facts about the past. 'Would you pay for this?' invites flattery; 'what did you do the last time it cost you something?' forces a concrete account that either contains evidence of pain (time spent, money paid, a hacky workaround built) or reveals there was none. Asking what it *cost* is the lever: if the person can't name a real consequence, the problem isn't worth solving yet — and you've learned that without building anything.
When to ask
Any time you need to know whether a problem is real and urgent — validating a new idea, qualifying a segment, or pressure-testing a feature request. Reach for it the instant a conversation turns to 'I would definitely use that.'
Good follow-ups
- How much time or money did that end up costing you?
- Have you looked for a way to fix it? What did you try?
- What made you stop looking, or settle for the workaround?
Watch out for
Accepting a hypothetical dressed up as a fact ('I'd probably just…'). If there's no concrete last time, that itself is the finding — don't rescue the idea by supplying the answer for them. Equally, don't pitch mid-question; the moment they sense what you want to hear, the data is contaminated.
Where to ask
- User interviewgreat
The Mom Test in its native habitat — live, you can press 'how much did it cost?' until a real consequence surfaces or it's clear there isn't one.
- Sales callworkable
A strong discovery question on a sales or win-loss call; mind the power dynamic — buyers under a pitch over-state pain.
- In-product surveypoor
Asynchronous, the cost half goes unanswered; narrow it to one factual past event so a single line can still carry evidence.
Reworded for this context: “Last time this problem cost you time or money, what did you actually do about it?”
- Long-form surveypoor
No way to reject a hypothetical dressed up as a fact; you can't tell a real last time from a polished 'I'd probably…'.
Pairs well with
Stage: Consider · A question popularized by Rob Fitzpatrick