Build a FigJam Affinity Map from Research
Put one sourced observation on each sticky, group related observations and name themes only after reviewing the evidence. FigJam provides the workspace; the affinity method below is a manual research practice, not an automatic validation of findings.
Before you start: Have research notes you are permitted to share and a short participant/context key. Remove identifying details that the workshop does not need before adding material to a collaborative board.
Worked example: Five interviews contain complaints about teammate invitations. Separate difficulty locating the action, uncertainty about permissions and delayed email delivery rather than collapsing everything into onboarding is confusing.
Step-by-step guide
Create observation cards
Write one observation per sticky in plain language. Include a source identifier and enough context to locate the original note without exposing unnecessary personal information.
Separate evidence from interpretation
Keep direct observations distinct from proposed causes. A participant asking for help is evidence; assuming they lack product knowledge is an interpretation that needs checking.
Group for similarity
Move related observations together without forcing every card into a cluster. Keep outliers visible because an unusual case may reveal an important boundary.
Name specific themes
Describe the shared problem rather than the preferred feature. Use permission uncertainty instead of build a better dialog so the theme does not prejudge the solution.
Challenge the clusters
Read the original context for representative and contradictory notes. Count distinct participants separately from repeated comments by the same person.
Write the synthesis
For each useful theme, state the observation, affected context and uncertainty. Assign a next question or test before carrying the theme into product prioritization.
Common mistakes
Quote count equals user count
One participant can contribute many notes. Track distinct sources so repetition does not masquerade as prevalence.
Solution labels too early
A cluster called new dashboard hides the underlying need. Name the problem before discussing a feature.
Discarding outliers
Contradictory evidence helps define boundaries. Keep it visible and explain whether it changes the theme.
Tips
Keep source identifiers readable at normal board zoom.
Use an uncertainty area for notes that do not fit.
Ask a second researcher to challenge a theme before sharing it.
Sources and verification
Based on product documentation checked on 2026-10-24. This is a documentation-based guide, not a hands-on benchmark. Features and availability can change.