Interviewing Users

Interviewing users is structured conversations with customers or prospective customers to gather qualitative insights about their needs, experiences, and behaviours. Done well, an interview produces evidence about how people behave in practice. Opinions about how they think they would behave are far less reliable.

Goal

To understand how users solve a problem, what they need, and the context they operate in. Interviews provide depth that surveys and analytics cannot, and they generate the language customers use to describe their problems.

Context

People are not reliable narrators of their own behaviour. Asked "do you cook every night?", most people overstate. Asked "would you pay $10 a month for this?", most people say yes. But when asked to take out a credit card, far fewer do. A good generative interview is designed to get around these gaps by anchoring questions in real evidence the person can show you, and by uncovering the emotional drivers behind decisions.

The principle behind the framework below: past behaviour is the best predictor of future behaviour. We are interested in what people have actually done, not what they say they would do.

The Generative Interview Framework

Five steps, run in order. Each step has a specific job, and skipping any one of them leaves the interview at the level of opinion rather than evidence.

1. The Setup

You want people to open up about their underlying feelings, and that requires trust. Spend the first few minutes establishing it.

• Set the scene. Explain that you're looking for stories and the words they use, not right or wrong answers. Tell them you'll mostly listen.
• Get familiar. Ask for the names of key people in their life early on (friends, family, colleagues, teammates). Use those names in follow-up questions later. It signals you're paying attention and makes the conversation feel less like a script.

2. Find the triggers

Things don't happen by chance. Every purchase, every signup, every change in behaviour was caused by something. The interviewer's job is to anchor questions in real moments rather than abstract preferences.

Use evidence-based questions that ask the person to show you something concrete:

• "Can you show me your recent orders from other fashion sites?"
• "Looking at your browser history, show me products you searched for on our site but ended up buying elsewhere."
• "Open your closet and show me the last complete outfit purchase."

Compare those to the opinion versions like "What kind of fashion do you like?" or "Would you buy from us if we did X?" Both produce hypothetical answers people don't act on.

3. Uncover the emotions

Decisions feel logical to the person making them, but most are emotional. The job-to-be-done framework names four forces that drive behaviour change.

A person switches when push and pull together are stronger than habit and anxiety. Each force needs its own questions, and the questions still need to be evidence-based:

• Push: "Open your closet. What items are you ready to replace, and why?"
• Pull: "Show me the last three fashion items you saved on social media. What drew you to each one?"
• Habit: "Walk me through your typical clothes shopping routine. Let's look at your browsing history from last week."
• Anxiety: "Show me something you almost bought recently but didn't. Talk me through the decision."

4. Identify the trade-offs

Every decision involves a trade-off, even when the person making it isn't aware of one. Ask what they gave up, what they considered and rejected, what they would change. Trade-offs reveal what the person actually values, separate from what they say they value.

5. Summarise and provide alternatives

Two related techniques at the end of the interview:

• Summarise. Paraphrase the person's journey back to them and confirm the trade-offs. This clarifies your understanding, gives them a chance to correct misinterpretations, and often surfaces details they didn't think to mention the first time. Try to summarise badly on purpose: restate the trade-off slightly wrong and watch them correct you. Their corrections are usually the most useful part of the conversation.
• Provide alternatives. Suggest different options the person could have chosen and ask why they didn't. Alternatives expose priorities and the situational factors that drove the decision.

Interviewing tips

A few habits that improve the quality of every interview, regardless of the framework above:

• Watch the energy. When someone gets animated, slow down and dig in. That's where the strongest signal is.
• Don't make people feel stupid. If you've asked a question they don't know how to answer, follow up with "yeah, that's a tough one, I'm not even sure how I'd answer it." You want them to stay open, not close up.
• Use analogies. If you don't understand the process they're describing, ask for an analogy.
• Interview together. Different team members pick up on different things. Having more than one person in the interview catches more signal and saves miscommunication later.
• Good cop, bad cop. When the team disagrees on what something means, argue amongst yourselves rather than with the interviewee.

Inputs

Outputs

Anti-patterns

• Opinion questions: "Would you buy this?" / "Do you like X?" People answer hypothetically and unreliably. Switch to evidence-based questions.
• Leading Questions: Asking questions that suggest a particular answer.
• Confirmation Bias: Only acknowledging information that confirms pre-existing beliefs.
• Under-preparing: Entering interviews without a clear plan or objectives.
• Lack of active listening: Failing to pay close attention to user responses and nonverbal cues, missing the moments where energy spikes.
• Poor documentation: Inadequate or inconsistent note-taking that loses the specific language the person used.