card-sort-open

An open card sort reveals user mental models by having participants group 40–80 unlabeled items into categories they define themselves, rather than fitting them into predefined groups. This technique informs site map structure during new product launches or IA redesigns.

Use an open card sort when launching a new product or redesigning information architecture without predefined categories. It reveals user mental models by letting participants define their own groupings, whereas closed sorts validate existing category structures.

An effective open card sort requires 40–80 content items and sessions with 15–20 participants. This scale ensures enough data variation to accurately generate similarity matrices and dendrograms for identifying common mental model groupings.

A similarity matrix and dendrogram analyze open card sort results by calculating how often participants group the same items together. This visual clustering identifies common groupings and informs the final site map structure and information architecture documents.

Open card sorting limitations include the requirement for a large participant pool of 15–20 users and 40–80 items. It is not suitable for validating existing categories, as the method relies entirely on users generating their own groupings from scratch.