What customer-discovery questions actually pass The Mom Test?
Questions about the past and the concrete pass; questions about the future and the hypothetical fail. Ask "tell me about the last time this happened," "what did that cost you," and "what have you already tried." Avoid "would you use," "do you think," and "how much would you pay."
Good questions point at the past
A good discovery question is answerable from memory, not imagination. It asks what already happened, so the person reports a fact rather than predicting their own behaviour. These travel well across almost any market:
- "Talk me through the last time you ran into this. What happened?"
- "What did that cost you — in time, money, or stress?"
- "What have you tried to fix it? Why did that not stick?"
- "Who else deals with this? Who pays when it goes wrong?"
- "How are you handling it today, even if it's a hack?"
Bad questions point at the future
The questions that feel most natural to ask are the ones that contaminate your data. They invite a polite guess about an imagined future, and people answer to end the conversation pleasantly.
- "Would you use a product that did X?" — a hypothetical, not a commitment.
- "Do you think this is a good idea?" — an opinion about you, not their life.
- "How much would you pay for this?" — a number invented on the spot.
Turn answers into commitment
When someone says the problem is real, test it with a small ask of their time, reputation, or money — an intro to a colleague, a follow-up with their real data, a pre-order. Talk is free; commitment is the signal. If the problem doesn't justify a tiny commitment now, it won't justify a purchase later.
Frequently asked questions
Why are hypothetical questions bad in user interviews?
People are poor predictors of their own future behaviour and tend to give the answer that ends the conversation pleasantly. "Would you use this?" almost always returns a yes that means nothing.
What is a good first question in a discovery interview?
"Tell me about the last time you dealt with [the problem]." It anchors the conversation in a specific real event and surfaces facts — what happened, what it cost, what they tried — instead of opinions.
How do I know if a customer problem is real?
Look for evidence of past behaviour: they've already spent time, money, or effort trying to solve it. A real problem has a paper trail of workarounds. A nice-to-have has only enthusiasm.
Last reviewed 2026-06-15.