Cluster titles must say what is inside, not stamp a category

What it is

A cluster titled Trust, Pain points, or Communication could sit on almost any study and still not say what happened inside it. An affinity title should be a descriptive affinity label: reading it should let someone foresee the events or quotes underneath—for example, “ID photo rejected as blurry on the third try.” A stamp recovers an abstract category. A description recovers the content of this material.

Why it happens

Abstract words are already circulating bins. Naming a cluster Trust reopens a chute that will accept any hesitation; the title is no longer constrained by the notes and cannot be falsified by a note that does not belong. A descriptive name keeps the constraint on the surface: if the cluster contains no third-try photo, the name must change. Naming also decides what later readers see—decision-makers often read titles only. Empty stamps let downstream treat the diagram as a mood taxonomy rather than an index of specific failures.

Studying it

Give final titles to people who did not cluster, ask them to list expected contents from the name alone, and score hits against the actual notes plus how vacuous the name is. Compare category words (usability, efficiency, emotion) with event phrases on whether later citations of the cluster still return to source quotes. Track how often a title is rewritten emptier in reports and roadmaps. Outcomes: name-to-content prediction accuracy, within-cluster heterogeneity, and semantic drift as titles travel.

Where it stops holding

Higher-order synthesis names may be shorter, but leaf clusters still need descriptive titles underneath; otherwise the synthesis is a pyramid with no contents. Slide character limits allow compression if the wall or appendix keeps a testable description. Category words written by participants in a card sort are their language evidence and should not be “professionalized” by researchers; that is a different kind of material, not a failed affinity title. In multilingual teams, translation often washes events into category words; check back against the original quotes.

Applying it

• Draft titles from verbs and objects that already appear in the cluster; ban opening with Experience, Trust, or Process.
• Test: hide the notes, leave the title, and ask someone who did not place that cluster to guess the contents. If they cannot guess, or if anything could fit, rename.
• If a title could be copied onto another project unchanged, treat it as a stamp and rewrite.
• In external readouts keep one descriptive title plus one representative quote; do not ship the category word alone.

Related

• Same group: Q4.01.1 Cluster from the notes; do not pour them into a ready taxonomy · Q4.01.2 The clustering work is the analysis
• Adjacent: Q4.02 Coding and thematic analysis · Q4.17 Communicating findings and acting on them
• Search terms: descriptive affinity labels · cluster naming · category dumping