Participants group labeled physical or digital cards into clusters that make logical sense to them. The resulting mental model patterns guide intuitive website navigation and menu design.
Think of it as unpacking groceries with your family. Everyone naturally groups items into the pantry, the fridge and the spice rack based on where they expect to find food.
Types
Card sorting studies are conducted in three structured formats depending on research goals.
- Open card sorting allows participants to group cards into custom clusters and invent their own category labels freely.
- Closed card sorting provides predefined category labels and asks participants to sort cards into the established buckets.
- Hybrid card sorting combines both by providing predefined categories while letting participants add new buckets if needed.
When to use it
Deploy card sorting when structuring navigation hierarchies, taxonomy tags or content hubs.
- Designing website navigation identifies intuitive menu categories and labels that reflect user expectations rather than internal org charts.
- Consolidating messy intranets groups hundreds of disparate policy documents into scannable departments and topics.
- E-commerce category restructuring organizes product catalogs so shoppers find items in expected departmental aisles.
- Validating taxonomy terminology uncovers whether users understand category terminology before committing code.
Common mistakes
Researchers often compromise card sorting validity through poor card preparation and wording.
- Using category names on the cards, which gives away the intended category and prevents authentic mental grouping.
- Including too many cards, exhausting participants with sixty items and degrading sort quality toward the end.
- Relying solely on open card sorts, forgetting to follow up with closed sorts or tree testing to validate the structure.
Frequently asked questions
Key takeaways
- Card sorting reveals how target users naturally group concepts and label categories.
- Open card sorts generate exploratory ideas; closed card sorts validate existing taxonomic structures.
- Keep card decks between thirty and fifty items to prevent cognitive fatigue and rushed answers.
- Always validate card sorting outcomes with tree testing to verify real-world findability.