How to Run User Interviews That Reveal What People Really Need

Most product mistakes begin with an assumption nobody tested. Teams build features because they seem logical, or because a loud customer asked, and then discover that real people do not behave the way the plan expected. User interviews are one of the cheapest and fastest ways to replace assumptions with evidence. All you need is a few conversations and the discipline to listen.

This guide explains how to plan, run and use interviews well, even if you have never done research before.

What interviews are good for (and not)

Interviews are best for understanding motivations, context and problems: why people do something, how they currently cope, what frustrates them and what they care about. They are less reliable for predicting what people will buy, or for measuring how many people share a behaviour.

Use interviews for	Use other methods for
Discovering problems and needs	Measuring how common a behaviour is (surveys, analytics)
Learning how people currently solve a task	Testing whether a design is usable (usability tests)
Understanding decision-making and context	Comparing two designs at scale (A/B tests)
Generating ideas and hypotheses	Forecasting demand (market data, pre-sales)

Good teams combine methods: interviews to find out what to explore, quantitative data to see how big it is.

Step 1: Define what you want to learn

Start with a clear research goal. “Learn about our users” is too vague. Better examples:

• “Understand how freelancers currently track their invoices, and where it goes wrong.”
• “Find out why new users stop using the app in their first week.”
• “Learn how teams decide which tournament platform to use.”

A good goal leads to a short list of learning questions. Everything you ask in the interview should help answer them.

Step 2: Recruit the right participants

The quality of insight depends on who you talk to.

• Define your target profile: role, behaviour, experience level, region. Focus on behaviour (for example, “has organised an online event in the last six months”) rather than demographics alone.
• Include different types: new and long-term users, people who churned, and people who chose a competitor.
• Aim for roughly five to eight participants per segment to start. Patterns often emerge quickly, and you can add more if answers differ widely.
• Find them through your existing users, communities, social channels, or professional networks. A short screening questionnaire helps you confirm they fit.
• Offer a fair thank-you, such as a gift card or extended subscription, and be clear about the time required.
• Avoid relying only on friends and family. They tend to be polite rather than candid.

Always get informed consent. Explain the purpose, how the information will be used and whether you will record. Respect their privacy and store recordings and notes securely.

Step 3: Write a good interview guide

An interview guide is a flexible outline, not a script. Structure it in a few stages:

1. Warm-up (5 minutes): introduce yourself, explain the purpose, get consent, and ask easy background questions.
2. Context (10 minutes): ask about their role, routines and environment.
3. Core topics (20 to 30 minutes): explore the specific behaviour and problems tied to your goals.
4. Wrap-up (5 minutes): ask what you did not ask but should have, and thank them.

Plan for 45 to 60 minutes, and prepare more questions than you will use.

Ask about the past, not the hypothetical

This is the most important rule. People are poor at predicting their own future behaviour, but they can describe what actually happened.

Weak question	Better question
“Would you use an app that tracks invoices?”	“Tell me about the last time you sent an invoice. What happened?”
“Do you think this feature is useful?”	“When did you last need to do that? How did you handle it?”
“How much would you pay for this?”	“What do you currently spend on solving this problem, in money or time?”
“Do you like our onboarding?”	“Walk me through what you did when you first opened the app.”

Rob Fitzpatrick’s book The Mom Test popularised this idea: ask questions that even your own mother could not lie to you about, by focusing on real events and facts rather than opinions about your idea.

Avoid leading and closed questions

• Don’t hint at the answer you want: “Don’t you find it annoying when…?”
• Prefer open questions that start with “how,” “what” or “tell me about”.
• Use follow-ups like “Why was that?”, “What happened next?” and “Can you show me?”

Step 4: Run the conversation well

The interview is about the participant, not your product.

• Build rapport. Be friendly, relaxed and curious.
• Listen more than you talk. Aim for the participant to speak most of the time.
• Embrace silence. If you wait a moment after an answer, people often add the most valuable detail.
• Dig into stories. Ask for specific examples and walk through them step by step.
• Notice emotions and workarounds. Frustration, hacks and spreadsheets reveal unmet needs.
• Stay neutral. Don’t defend your product or react strongly to criticism. Thank people for honest feedback.
• Don’t pitch. If you start selling, you will hear what you want to hear.
• Take notes or record, with permission, so you can stay present. Having a second person to take notes helps.

If you are interviewing current users of a product, avoid making it a support session. Park individual bug reports to follow up on separately.

Step 5: Capture and analyse what you heard

Raw notes are not insight. Within a day or two of each interview, while memory is fresh:

1. Write a short summary of the key stories, quotes and surprises.
2. Tag observations by theme, such as “onboarding confusion” or “price sensitivity”.
3. Look for patterns across participants. Something said by one person is an anecdote. Something repeated by several is a signal.
4. Separate what people did from what they said they wanted. Weigh behaviour more heavily.
5. Create simple artefacts like journey maps, a list of top problems or a short “how might we” list to share.

Affinity mapping, which means grouping sticky notes or digital cards by theme, is a simple and effective way to see patterns.

Step 6: Turn insight into decisions

Research only matters if it changes something.

• Write insight statements: “Freelancers delay invoicing because gathering project details takes longer than the work itself.”
• Turn them into opportunities and prioritise them with a framework such as RICE, as described in our MVP planning guide.
• Share findings in a short, visual format, with real quotes, and invite the team to watch clips or read excerpts so they build empathy first-hand.
• Test solutions with prototypes before building. Our UI/UX design service is often used for exactly this step.
• Follow up with participants where appropriate, so they know their input mattered.

Common mistakes

• Talking to the wrong people, or only to fans of your product
• Asking leading or hypothetical questions
• Pitching instead of listening
• Treating one person’s opinion as a universal truth
• Not recording or documenting, then relying on memory
• Collecting insights and never acting on them
• Skipping consent and privacy care

Interview checklist

• Clear research goal and learning questions
• Target participant profile and screener
• Five to eight participants per segment recruited, with consent
• Interview guide with warm-up, context, core topics and wrap-up
• Questions about past behaviour, open and non-leading
• Quiet, neutral listening; no pitching
• Notes or recordings stored securely
• Summaries and theme tagging within 48 hours
• Insights shared with the team and linked to decisions
• Follow-up planned

Final thoughts

Great interviews feel like curious conversations. Ask about real experiences, listen for the story behind the answer, and look for patterns across people. A handful of well-run interviews can save months of building the wrong thing. If you would like help with product discovery, prototyping or turning insights into a roadmap, talk to our team.