The Metric Tree Testing Is Actually Built Around

The correct answer is "First-click success rate — whether the first click was on the correct path."

Tree testing's key metric is first-click success rate: the percentage of participants whose very first click in the stripped-down, text-only tree was on the correct path toward the target item. This is deliberately narrow — it isolates whether the labels and structure work on their own, independent of visual design, and the pattern of wrong first clicks (which wrong branch absorbs them) is what tells a team exactly which label is competing for the same mental territory as the correct one.

The distractors name plausible-sounding but incorrect metrics: total browsing time isn't the metric tree testing centers on — a participant could wander and still eventually succeed, which the first-click metric wouldn't credit as a clean result; participants don't invent categories in a tree test the way they do in an open card sort — the tree is fixed and given to them; and tree testing isn't measuring product-wide session length at all, only navigation of the specific proposed structure being tested.