Affinity diagramming is a method for sorting a pile of research notes, observations, or ideas into clusters based on natural relationships, so themes rise out of the data instead of being forced onto it. Teams use it to make sense of messy qualitative data after interviews or tests.
Use cases
Skip the synthesis and you'll cherry-pick the quotes that fit your hunch and call it insight.
300 sticky notes, no meaning: After 15 interviews you have a wall of observations and no idea what they say. Clustering them turns the pile into a handful of themes you can act on.
The loudest voice wins: A team argues over "what users want" from memory. Sorting the actual notes together builds shared understanding and shuts down the opinion contest.
Findings stuck in one head: Research lives in one researcher's notebook. A diagram externalizes it so anyone who wasn't in the room can read the result and get it.
How it's used in practice
One observation per note: Keep each note raw and in the user's words, not your interpretation, so the data stays honest.
Cluster in silence first: Let people sort without talking, then discuss. Clusters should form bottom-up, not get assigned to columns you set up in advance.
Name clusters after they form: The label is the insight. Write it as a short sentence ("users abandon checkout when shipping cost appears late"), not a one-word bucket.
Mind the outliers: Look hard at small clusters and stray notes. The odd one out often hides the real opportunity.
🪄
Pro-tip: Don't bring your categories to the wall. The whole point is that themes emerge from the notes, not from buckets you arrived with. Label the columns first and you're just sorting evidence into your existing assumptions. And run it with the team that did the research, not solo, or you lose the shared "aha" that makes people act on it.
Challenges & limitations
Time and space hungry: Big studies produce hundreds of notes, and the sort takes real hours and a large surface, physical or virtual.
Bias in the notes: Who writes the notes and how they're worded quietly shapes the clusters. Garbage notes, garbage themes.
Over-clustering: It's easy to force tidy groups that flatten real nuance, so you end up with neat themes that lose the messy truth.
Commonly Used Tools:
This is a method (rooted in the KJ Method), not a named framework, so the "tools" are the boards teams sort on:
Miro — an online whiteboard with sticky notes and affinity templates, good for remote and large teams.
FigJam — Figma's whiteboard, handy when research lives next to the design work.
Mural — a collaborative whiteboard built for workshops and synthesis sessions.
