"Would you use this?" is close to the least useful question you can ask a potential customer. People are polite, they want to be encouraging, and hypothetical enthusiasm costs them nothing. The goal of a good customer interview isn't a nice reaction — it's evidence of past behavior, which is far harder to fake and far more predictive.
Ask about the last time, not the general pattern
"Tell me about the last time you tried to solve this problem" beats "how do you usually handle this?" every time. General patterns get answered with idealized self-descriptions. Specific recent memories get answered with what actually happened, including the messy, unglamorous workaround that reveals the real problem.
Ask what they tried before, and why it didn't stick
If someone has already tried three other tools or workarounds and abandoned all of them, that's a goldmine — you now know exactly what "not good enough" looks like from lived experience, instead of guessing at a feature list.
Watch for the moment they get specific and slow down
Vague, fast, enthusiastic answers are usually politeness. When someone suddenly slows down, gets specific, and starts describing an actual frustrating moment in detail — a Tuesday afternoon, a specific spreadsheet, a specific colleague — that's when you've hit something real. Lean in and ask follow-ups there, not everywhere.
Never pitch during a discovery interview
The moment you describe your solution, the interview is over as a source of truth — the person is now reacting to your idea, not describing their reality, and their politeness kicks back in immediately. Save the pitch for after you've learned everything you can.