What is Tree testing?
Tree testing is a usability method that evaluates the findability of topics in a website’s information architecture by asking users to locate items in a text-only hierarchy stripped of all visual design.
How it works
In a tree test, participants see only the navigation structure as expandable text labels (no visual design, page content, or search). They are given tasks (‘Find where to apply for a parking permit’) and navigate the tree by clicking category labels. Metrics include success rate (did they find the correct location?), directness (did they go straight there or backtrack?), and time to complete. By stripping away visual design, tree testing isolates whether the IA structure itself works. If users cannot find items in the tree, no amount of visual design will fix the navigation. Tree testing is conducted using tools like Optimal Workshop’s Treejack.
Applied example
A municipal government tree tests its website navigation. The task ‘Find how to report a pothole’ has only 30% success: users look under ‘Transportation’ but the item is under ‘Public Works.’ This IA problem is invisible in the full website because most users fall back on search. Restructuring the tree and adding cross-links increases findability to 75%.
Why it matters
Tree testing isolates the structural foundation of navigation effectiveness, identifying whether the information architecture itself works before visual design and content confound the picture.
