Tree testing evaluates how easy a website is to navigate. Participants are asked to navigate to a specific page on a website using only page names in a menu.
A good navigation flow leads most participants directly to the right page to complete a task. A high success rate means the information architecture is intuitive, which is essential for a good user experience.
Tree testing is quick to set up because it requires no visual design elements. For this reason, this research method is ideal for testing many versions of a navigation flow with a large number of participants.
Some tips to conduct tree-testing effectively are:
Provide less than 10 tasks.
Write tasks that test a specific part of a website that needs improvement.
Write tasks as realistic scenarios that make sense for typical users.
Don't use exact phrases from page names. Using the exact phrases will lead the participants to the correct answers too easily.
The output of tree testing will typically be percentages of success or failure. The researcher will tally the performance of each participant into four categories:
Direct success is when a participant navigates directly to the correct page with no issues.
Indirect success is when a participant navigates to the correct page after backtracking or taking detours.
Indirect failure is when a participant navigates to the wrong page after backtracking or taking detours.
Direct failure is when a participant navigates directly to the wrong page.
It's important to note that in the real world, what might be an indirect success could result in a user giving up. A large percentage of indirect successes should still be cause for concern.
Example of a tree-testing “pietree” from Optimal