Last quarter I watched a Stanford student conduct her first user interview. She had a beautiful script—twelve questions, carefully worded, organized by theme. She was prepared.
But when she did the actual interview, I realized she was completely missing the point of in person research.
Her interviewee mentioned, almost in passing, that she “dreaded opening the app.” My student nodded, checked something off, and moved to the next question on her list.
That moment—dreaded—was the point of the whole interview. That one word contained frustration, maybe anxiety, probably a workaround she’d invented. And my student drove right past it like it was a speed bump instead of a treasure chest.
This is the most common mistake I see in user research. People treat interviews like surveys with eye contact. They have their questions. They ask them. They get answers. Done.
But here’s what a survey can’t do: follow up.
Three little words
When someone tells you something interesting, you don’t need a clever question. You need three words:
Tell me more.
The temptation—especially when you’re nervous—is to prompt with specifics. Your interviewee says they hate their commute, and you jump in: “Is it the traffic? The length? The unpredictability?”
Don’t. You’ve just handed them a multiple choice test. You’ve constrained their answer to your imagination of what the problem might be.
Instead: Tell me more.
Maybe they’ll say it’s the traffic. Or maybe they’ll tell you their commute is the only time they have to themselves all day, and they dread it ending. Maybe the problem isn’t the commute at all—it’s what’s waiting when they arrive.
You cannot anticipate what you don’t know. That’s literally the point of research.
Actually listen
The second half is actually listening. Not listening-while-thinking-about-your-next-question. Not listening-for-the-thing-you-expected-to-hear. Actual listening.
People can tell when you’re really paying attention. When someone feels genuinely heard, they keep talking. They go deeper. They tell you things they didn’t plan to share.
The best interviews I’ve conducted felt like conversations where I barely said anything. A few “hmms.” Some nodding. An occasional “and then what happened?” The interviewee did 90% of the talking, and I learned things I never would have thought to ask about.
Your script is a safety net, not a cage
Your interview script should be a topic map, not a railroad track. When something interesting surfaces, abandon it. Follow the thread. You can always come back to your prepared questions.
The insights that change products—the ones that make everyone in the debrief say “I never thought of that”—they almost never come from prepared questions. They come from the second and third follow-up, from the moment you stopped performing “researcher” and started being genuinely curious.
Next interview, when something catches your attention, resist the urge to narrow the aperture with specific prompts.
Just say: Tell me more.
Then shut up and listen.