How to Conduct User Interviews for UX Research
Practice3 minInnovialabs editorial team
In short
A good interview is about what people have actually done, not what they think or would do; recruit from your real audience, ask open questions without hints, and synthesize notes into themes backed by evidence.
The core rule
User interviews exist to understand people's real behavior, context and problems. So ask about what someone did in the past, not what they think of your idea. "Yes, I would use that" is worth almost nothing. A story about how they handled the task last week is valuable data.
Step 1. Define the goal
Before recruiting, write down 2-4 research questions: what the team wants to learn. For example: "How do small businesses track orders today, and where do they lose time?". These are not questions for participants but a frame for checking later whether you met the goal.
Step 2. Recruiting
- Screener. A short survey that filters out poor fits. Ask about behavior, "How often do you order delivery?", not "Do you like delivery?".
- Real audience. Not friends or colleagues. If you have several segments, recruit from each.
- Channels. Current customers, your mailing list, communities, social media, dedicated recruiting services.
- Incentive. Offer something for people's time; it improves attendance.
- How many. Start with a small number of interviews per segment and continue until new conversations stop adding anything new.
Step 3. The interview guide
A guide is a plan, not a script. A typical structure:
- Intro. Introduce yourself, explain the purpose, ask for consent to record. Say there are no right answers and you are not testing them.
- Warm-up. Simple questions about role and context.
- Core. Questions about specific recent situations.
- Wrap-up. "What didn't I ask that I should have?" and thanks.
How to ask questions
| Bad | Better | Why |
|---|---|---|
| "Would paying in the app be convenient for you?" | "Tell me how you paid for your last order." | Past behavior instead of hypothetical |
| "Don't you think search is awkward?" | "How did you find the product you needed last time?" | No hint of the answer |
| "Do you usually compare prices?" | "What did you do before your last purchase?" | Open question instead of yes/no |
| "What features do you want?" | "What was the hardest part of this process?" | Problems instead of solutions |
Useful techniques:
- "Tell me about the last time you..." starts a concrete story.
- "Why?" and "What happened next?" go deeper.
- Silence. Do not fill pauses: people often say the most important thing right after one.
- Do not defend the product or explain how a feature "really" works.
Step 4. Notes and recording
- Work in pairs: one person leads, the other takes notes.
- Record quotes and facts separately from your interpretations.
- Right after the interview, spend a few minutes on a short debrief: key points, surprises, questions for the next sessions.
- Keep recordings and transcripts in one place with clear names, respecting participants' consent.
Step 5. Synthesizing into themes
- Break notes into observations, one fact or quote per card.
- Group similar ones (affinity mapping) on a board.
- Name themes as insights: not "Payment" but "Users delay payment because they are unsure about delivery times".
- Attach evidence: quotes and how many participants mentioned it.
- Link to decisions: what the team should test or change.
A good research output is a short document readable in a few minutes, with themes, quotes and recommendations.
Common mistakes
- Pitching the idea instead of listening.
- Asking about the future: "Would you...?".
- Drawing conclusions from one memorable participant.
- Running interviews without synthesis, so the data stays in people's heads.