Affinity Mapping

Purpose

Organise a large number of items (ideas, observations, data points) into meaningful groups based on natural relationships, revealing the underlying structure of a data set. In A2.1, affinity mapping transforms 50–100+ raw ideas into 15–25 thematic clusters that expose the solution landscape and enable concept synthesis.

When to Use

Use affinity mapping when:

• ≥30 items need organising (below 30, manual sorting is faster)
• Categories are not known in advance (bottom-up structure discovery)
• Multiple perspectives exist and consensus on grouping is needed

Sample Size and Duration

Items: 30–150 (optimal: 50–100)

Participants: 3–8

Duration: 45–90 minutes (scales with item count)

Expected output: 15–25 named clusters

Prerequisites

• Raw ideas from A2.1 Step~2 (one idea per card/note)
• Large wall, table, or digital board
• 3–8 participants (all who generated ideas, ideally)
• 45–90 minutes (scales with item count)

Complete Procedure

Step 1: Prepare (5 minutes)

1. Post all idea cards on wall / board (random layout)
2. Remind participants: sort by functional similarity (how the idea solves the problem), not surface similarity (what it looks like)

Step 2: Silent Sorting (15–25 minutes)

1. All participants sort simultaneously, silently
2. Move cards into groups; anyone may move any card
3. If two participants disagree on placement, duplicate the card rather than debating
4. Continue until movement slows naturally

Step 3: Name Clusters (10–15 minutes)

1. Review each group as a team
2. Agree on a descriptive label capturing the shared solution mechanism (e.g. “predictive analytics approaches,” “peer support models”)
3. Split groups that contain two distinct mechanisms; merge groups that differ only in wording

Step 4: Identify Outliers (5–10 minutes)

1. Review unclustered cards
2. For each: reclassify into an existing group, create a new singleton group, or flag as a standalone concept candidate
3. Do not discard—outliers often represent the most novel ideas

Step 5: Document (10 minutes)

1. Photograph the physical board or export the digital board
2. Create a cluster list: cluster name, member idea IDs, brief description of shared mechanism
3. Note cluster sizes (large clusters may need splitting in synthesis; tiny clusters may signal under-explored areas)

Quality Criteria

1. 15–25 clusters from 50–100+ raw ideas (fewer than 10 suggests over-aggregation; more than 30 suggests under-aggregation)
2. Cluster labels describe solution mechanisms, not channels or technologies
3. All ideas assigned (no orphans left unaddressed)
4. Outliers explicitly reviewed and documented
5. Cluster map photographed / exported and linked to idea IDs

Theoretical Foundation

Seminal references:

• : Developed the KJ method (affinity diagramming) as a systematic approach to organising qualitative data bottom-up, allowing patterns to emerge from the data rather than being imposed by pre-existing categories.

Contemporary references:

• : Integrated affinity mapping as a core synthesis tool in design thinking, demonstrating its value for making sense of ethnographic data, brainstorming outputs, and user research findings.

Challenges and Solutions

Challenge 1: Sorting by Surface Similarity

• Symptoms: Groups labelled “app ideas,” “email ideas,” “AI ideas” (channel-based, not mechanism-based)
• Solution: Facilitator reframes: “Group by what problem-solving approach the idea uses, not what technology it mentions.” Reshuffle if necessary after naming reveals surface grouping.

Challenge 2: One Giant Cluster

• Symptoms: A single group contains 30+ ideas because participants sorted too broadly
• Solution: Split the mega-cluster: “Within this group, are there sub-approaches that differ meaningfully?” Typical mega-clusters split into 3–5 sub-groups on closer inspection.

Challenge 3: Territorial Sorting

• Symptoms: Participants repeatedly move the same card back and forth between groups
• Solution: Duplicate the card and place it in both groups. This often signals that the idea genuinely bridges two approaches—a valuable insight for concept synthesis.

Relationship to Other Methods

Affinity Mapping receives input from:

• Structured Brainstorming (the referenced method)
• Brainwriting 6-3-5 (the referenced method)
• SCAMPER (the referenced method)
• Analogical Ideation (the referenced method)
• Crazy 8s (the referenced method)

Affinity Mapping provides output to:

• Concept Synthesis (the referenced method): clusters become the basis for synthesised concepts in A2.1 Step~4

Tools and Templates

• Physical: Large wall or table, sticky notes, marker pens for cluster labels
• Digital: Miro, Mural, FigJam (clustering features built in)
• Cluster documentation template (cluster name, member IDs, mechanism description, size)

• J. Kawakita (1991). The Original KJ Method. Kawakita Research Institute.
• T. Brown (2009). Change by Design: How Design Thinking Transforms Organizations and Inspires Innovation. HarperBusiness.