What Is Affinity Diagramming?
Affinity diagramming is a collaborative analysis method in which a research team sorts individual data points, observations, quotes, ideas, or findings, into groups based on their natural relationships and themes. Each piece of data is written on a separate card or sticky note, and the team physically or digitally clusters related items together, then labels each cluster. The method was developed by Japanese anthropologist Jiro Kawakita in the 1960s (and is sometimes called the KJ method) as a way to make sense of large amounts of unstructured qualitative data. It's now a standard practice in UX research, design thinking, market research synthesis, and product development, used whenever a team needs to transform raw qualitative input into organized, actionable themes.
Why Affinity Diagramming Matters in Research
Qualitative research generates enormous volumes of unstructured data, interview transcripts, open-ended survey responses, observation notes, usability findings. Without a systematic way to organize this material, teams default to cherry-picking quotes that confirm existing assumptions. Affinity diagramming provides a bottom-up structure: themes emerge from the data rather than being imposed on it, and the collaborative process builds shared understanding across team members who may have observed different parts of the study.
How Affinity Diagramming Works
The process is deliberately simple, but its effectiveness depends on doing each step carefully.
Data Preparation
Start by breaking down your qualitative data into individual, atomic observations. Each sticky note or card should contain one finding, one quote, or one observation, not a paragraph. A good affinity note is specific enough to be meaningful on its own. "User couldn't find the search bar" is better than "Navigation was confusing." Depending on the study size, you might work with 50 to 300+ individual data points.
Silent Sorting
Team members read through the notes and begin grouping related items together without discussion. This silent phase is important, it prevents one person's interpretation from anchoring the group's thinking. Everyone moves notes simultaneously, clustering items that feel related. If someone disagrees with a placement, they move the note; consensus emerges through action rather than debate.
Group Discussion and Labeling
Once the initial clusters stabilize, the team discusses each group, refines boundaries, splits oversized clusters, and merges groups that overlap. Each cluster gets a descriptive label that captures the theme, not just a category word, but a statement that conveys the insight. "Users expect search to work like Google" is more useful than "Search."
Hierarchy Building
Large affinity diagrams often have two or three levels. Individual notes group into sub-themes, sub-themes group into major themes, and major themes may group into overarching categories. This hierarchy provides both detail (the individual observations) and structure (the big-picture patterns) in a single view.
Documentation
The final diagram is documented, photographed, digitized, or captured in a collaboration tool, along with the labels and the team's interpretation of each theme. This becomes the foundation for insight reports, design recommendations, or strategy decisions.
When to Use Affinity Diagramming
- After qualitative research rounds (interviews, usability tests, ethnography) to synthesize findings into themes before reporting
- During workshops and brainstorming sessions to organize ideas generated by cross-functional teams
- Combining findings from multiple studies or data sources into a unified framework
- Prioritizing product or service improvements by clustering user pain points and identifying which themes have the most evidence behind them
- Onboarding team members to a research area, the collaborative sorting process builds shared understanding faster than reading a report
Common Mistakes to Avoid
- Writing notes that are too vague or too compound: each note should contain exactly one atomic observation; "The signup flow is confusing and the pricing page lacks detail" is two notes, not one
- Letting one team member dominate the sorting: the method's strength comes from multiple perspectives; enforce the silent sorting phase and ensure everyone participates in labeling
- Stopping at categories instead of insights: labels like "Onboarding" or "Pricing" don't tell you anything; push for labels that express what you learned, like "Users abandon onboarding when asked for payment information before seeing value"
How Quali-Fi Supports Affinity Diagramming
Quali-Fi's AI-powered thematic analysis can serve as a starting point for affinity diagramming by automatically clustering open-ended survey responses and qualitative data into preliminary theme groups. Research teams can export these AI-generated clusters alongside the raw verbatims, then refine and reorganize them collaboratively, bringing human judgment to the synthesis while letting the platform handle the initial sorting of high-volume data.
Frequently Asked Questions
How is affinity diagramming different from thematic analysis?
Thematic analysis is a formal analytical methodology with defined coding procedures, typically applied by individual researchers to transcripts. Affinity diagramming is a collaborative workshop technique focused on physical or visual grouping of data points by a team. They share the goal of finding patterns in qualitative data, but affinity diagramming emphasizes team participation and visual organization, while thematic analysis emphasizes systematic coding rigor.
How many people should participate in an affinity diagram session?
Three to eight people is the productive range. Fewer than three limits the diversity of perspectives; more than eight makes the physical space crowded and the discussion unwieldy. Include team members who participated in data collection (they bring contextual knowledge) alongside stakeholders who didn't (they bring fresh eyes).
Can affinity diagramming be done remotely?
Yes. Digital whiteboard tools like Miro, FigJam, or MURAL replicate the sticky-note experience online. Remote affinity sessions work best with clear facilitation, assign a facilitator to manage the process, use timers for the silent sorting phase, and use breakout groups if the team is large. The physical energy of in-person sessions is harder to replicate, but the outputs are comparable.
Related Topics
- Card Sorting
- Concept Mapping
- Thematic Analysis
- Qualitative Data
- Participatory Research
- Research Methodology
Drowning in qualitative data? See how Quali-Fi's AI-powered analysis gives you a head start on thematic synthesis.