Affinity Mapping
Grouping raw research observations until themes emerge from the data rather than from the team's assumptions.
Affinity mapping is the step between having research and having findings. Every observation goes on its own note — one observation, quoted where possible, not a summary — and then notes get grouped by what they have in common. Groups are named only after they exist. Naming first is the failure mode: the team writes up the categories it expected and sorts evidence into them, which produces a tidy artefact confirming the starting assumption.
The discipline is in the notes. “Users want better search” is a conclusion and contaminates the grouping. “Searched for ‘invioce’, got nothing, gave up and opened the spreadsheet” is an observation, and three of those next to each other become a finding nobody had written down in advance. Keep a reference from each note back to the session it came from, so a finding can be traced when someone challenges it — and they will.
In practice
A team analysing 14 interviews expected themes around pricing. Mapped from raw quotes, the largest cluster was about not knowing whether a colleague had already actioned a request — mentioned by 11 of 14 participants, and absent from the discussion guide entirely. It became the next quarter's roadmap item.
Where teams get it wrong
- Writing conclusions on the notes instead of observations.
- Naming groups before grouping, so the map confirms what the team already believed.
- Losing the link back to the source session, which makes findings unchallengeable and therefore unbelievable.
- Mapping alone. The value is partly in the team arguing about where a note goes.
- Stopping at the map. A wall of clusters is not a finding until somebody writes the sentence.
( Learn more )
Related terms
All terms- User ResearchThe systematic study of users' goals, behaviour and context — through interviews, observation and analysis — to inform what gets designed and built.
- User InterviewA structured conversation with someone who uses or might use a product, aimed at understanding their situation rather than collecting opinions.
- Research RepositoryA searchable store of research findings and their evidence, so what a team already learned can be found again.
- Design ThinkingA problem-solving approach that starts from user needs and moves through framing, ideation, prototyping and testing before committing to a solution.