Card sorting

Card sorting is a research project: we prepare a set of cards (your site's items/topics), give relevant participants the task to group them and, if needed, name the groups (open, closed or hybrid sorting), analyze which items people combine, what words they use, where they diverge. The result is material for designing information architecture. Honestly about the nature, this is key: card sorting is an EXPLORATORY, qualitative INPUT method: it shows users' mental models and gives a basis for the structure, but it does NOT output a ready 'correct' navigation — people group differently, agreement is partial, and the final structure is designed by us with you based on the data (not 'as voted'). Honestly about sample and interpretation: on a small sample there is variance; the result needs interpretation, not mechanical application. Honestly about pairing: card sorting is logically linked with tree testing — first gather the model (card sorting), then check the resulting structure (tree test); these are separate methods. Honestly about the essence: this is input and hypotheses for structure, NOT building navigation, NOT redesign and NOT growth — designing, implementing and validating is separate; card sorting by itself does not create the structure. The platform/recruiting (if paid) is separate. An important boundary: this is card sorting (gathering the mental model), not a tree test (checking the structure — separate), not a usability test and not development. If the structure is small and obvious, the method is overkill. Picture this: instead of 'we'll arrange sections as seems logical to us' you see how customers themselves group your products — and build the menu for them. The base price starts from 12,000 ₽ per study; it depends on the number of cards and participants.

• You build the section structure from your logic, not customers'.
• You do not know how users mentally group your products/topics.
• You invent section names 'from inside', unclear to customers.
• There is no basis for designing clear navigation.

• Preparing a set of cards (items/topics)
• Choosing the sorting type (open/closed/hybrid)
• Recruiting relevant participants
• Collecting groupings and names, analyzing matches/divergences
• Conclusions on the mental model and structure hypotheses
• Flagging limitations (variance, interpretation)
• A report with material for IA
• Reviewing results with you

• You know how customers group and name items
• Material and hypotheses for a clear structure
• A base for IA design (design/validation — separately)
• Navigation from users, not internal logic

• We clarify items, goals, audience; prepare cards and the sorting type
• We run the sorting, analyze groupings and names
• We frame conclusions and structure hypotheses, review with you

• Will card sorting give a ready site structure?

  No. It is an exploratory input: it shows users' mental models and gives a basis, but it does not output a ready 'correct' navigation — people group differently. The final structure is designed by us with you based on the data (design — separately).

• Will card sorting raise conversion?

  Not directly. It is material for designing the structure; building navigation, implementing and validating (tree test/usability) is separate. The method itself does not create the structure or raise conversion.

• Is it the same as tree testing?

  No. Card sorting helps BUILD a structure from the mental model; a tree test CHECKS findability in an already-defined structure. It is logical to use them as a pair, but they are different methods (tree test — separate).