Analysing user interviews: from notes to insight
To analyse user interviews, organise notes and recordings, code observations with consistent tags, group them through affinity mapping, identify recurring themes and patterns, write clear insight statements supported by evidence, prioritise them by impact and frequency, and share findings with recommendations. Involving several team members reduces bias and builds shared understanding.
Why analysis matters
After a round of user interviews, teams often have hours of recordings, pages of notes and dozens of memorable quotes. Without careful analysis, it is easy to remember only the most striking moments, focus on findings that confirm existing beliefs or produce a long list of observations that nobody can act on. Structured analysis turns raw data into clear insights that guide design and product decisions. Our companion post on planning and running user interviews covers the earlier stages.
Organise your data
- Gather materials: Recordings, transcripts, notes and summaries in one place.
- Write quick summaries: Key points from each interview, ideally the same day.
- Note participant details: Role, experience, segment and context.
- Remove personal data: Anonymise where possible to protect privacy.
- Set up your workspace: A whiteboard, spreadsheet or research tool for analysis.
Code observations
Coding means tagging pieces of data, such as quotes or observed behaviours, with short labels that describe what they are about. Codes might relate to goals, pain points, workarounds, emotions, journey stages or specific features. Start with a few expected codes based on your research questions, then add new ones as themes emerge. Keep code names consistent so patterns are easy to spot across interviews.
| Observation | Possible codes |
|---|---|
| "I export everything to Excel every Monday." | Workaround, reporting, weekly task |
| Participant hesitated at the pricing page | Pricing confusion, hesitation |
| "I never know if my payment went through." | Uncertainty, payment, feedback |
| Called support twice to change an order | Support contact, order changes |
Affinity mapping
Affinity mapping is a common technique for grouping observations into themes. Write each observation on a separate sticky note, physically or digitally, then group related notes together. Name each group with a short description of the theme. Doing this as a team, with designers, product managers and engineers, brings different perspectives and builds shared understanding of users.
- One observation per note
- Group notes by similarity, not by interview
- Name groups with clear, descriptive labels
- Look for groups within groups
- Discuss surprising or conflicting notes
Identify themes and patterns
Themes are recurring patterns across several participants: shared goals, common frustrations, repeated workarounds or similar emotional reactions. Note how many participants mentioned each theme and whether it differs between user groups. Pay attention to contradictions between what people say and what they do, which often reveal deeper needs.
Comparing user groups
When interviews include different types of users, such as new and experienced customers or Arabic and English speakers, compare themes across groups. A frustration that affects only new users suggests onboarding problems, while one shared by everyone may point to a core product issue. Tagging each observation with the participant’s group makes these comparisons easy and helps teams decide which audiences to prioritise.
Write insight statements
An insight goes beyond describing what happened; it explains why it matters. A useful format is: observation, plus reason, plus implication. For example: "Small business owners export invoice data to spreadsheets every week because the reporting view does not show overdue payments, which means they spend hours on manual work and miss late payers." Support each insight with quotes and evidence from several participants.
- Observation: What people did or said.
- Reason: Why it happens.
- Implication: Why it matters for users and the business.
- Evidence: Quotes and how many participants mentioned it.
Prioritise insights
Not every insight is equally important. Rank them by how many users are affected, how severely, and how closely they relate to business goals. A simple impact and frequency grid helps teams agree what to address first. Link the top insights to "how might we" questions or design opportunities so they lead naturally into ideation.
Share findings
Present the most important insights clearly, with short video clips and quotes that bring them to life. Tailor the format to the audience: a one-page summary for leaders, more detail for designers and product managers. Store findings in a searchable repository so they can be reused. Our guides to sharing research insights and atomic research repositories offer practical approaches.
Common analysis mistakes
- Relying on memory instead of notes and recordings
- Focusing only on the most dramatic quotes
- Treating one participant’s opinion as a pattern
- Writing insights without evidence
- Producing findings with no recommendations or owners
Avoiding these mistakes keeps findings credible and makes it far more likely that the team will act on them.
Using AI tools in analysis
AI tools can speed up transcription, suggest codes and draft summaries of themes. They are helpful for handling large volumes of interviews, but they can miss nuance or invent quotes. Always verify AI-generated themes and quotes against the original recordings, and keep researchers responsible for interpretation. Our post on AI tools in user research covers safe practices.
Frequently asked questions
How long does interview analysis take?
Often as long as the interviews themselves, or longer, depending on the number of sessions and depth required.
What is affinity mapping?
A technique for grouping individual observations into themes by clustering related notes together.
How many participants must mention something for it to count?
There is no fixed rule. Look for patterns across several participants, and note severity even when a problem affects few people.
Should analysis be done alone or as a team?
As a team where possible. Different perspectives reduce bias and build shared understanding.
What tools help with interview analysis?
Digital whiteboards for affinity mapping, spreadsheets or research repositories for coding, and AI-assisted transcription, always checked by researchers.
How do we turn insights into design work?
Rewrite top insights as how-might-we questions, prioritise them and link each to an owner and next step.