What is tree testing in UX research?
Tree testing is a research method used to evaluate the findability of content within a proposed information architecture. Participants are shown a text-only representation of the navigation hierarchy, a tree, without any visual design, and are asked to find specific items by clicking through the categories. The text-only format isolates the structural and labeling quality of the IA from the visual design, revealing whether users can find items based on the labels and organization alone. Tree testing measures success rate, the percentage of users who find the correct item, and directness, whether users found it without backtracking.
When should you use tree testing?
Tree testing is most valuable after card sorting has established how users naturally group and label content, and before building out the full visual navigation. It validates the proposed structure with real users before design investment is made. Tree testing is particularly useful for identifying specific labels that are confusing, revealing categories that users cannot distinguish from each other, and finding items that users consistently look for in the wrong place. It provides quantitative data on findability that card sorting, being more generative, does not produce.
How does tree testing differ from card sorting?
Card sorting is a generative method: participants create their own categories and labels for content, revealing their mental model of how information should be organized. Tree testing is an evaluative method: participants navigate a proposed structure to find specific items, revealing whether the structure works. Card sorting happens before the structure is defined. Tree testing happens after a structure has been proposed, to validate it. Both methods contribute to information architecture design, with card sorting informing the creation of the structure and tree testing validating it before implementation.