Tree testing
We run tree testing: we give people a task 'where would you look for X?' and see whether they find what they need in your section structure — on a bare menu tree, without design and images. So you check whether the navigation and section logic are clear on their own. Honestly upfront: this is a narrow method about findability in the structure, not about design or conversion; it gives data and hypotheses, not proof and not growth.
Tree testing — overview
Tree testing is a project: we take your site's section structure (the navigation tree in text form), frame tasks to find typical items, give them to relevant participants and measure findability — the share of successful paths, where people turn the wrong way, at which level they get lost. The method deliberately removes design and visual cues to isolate the LOGIC of the structure. Honestly about the scope, this is key: a tree test checks ONLY findability in the hierarchy (whether section names and logic are clear), NOT visual design, page texts or conversion; it is a narrow but precise slice of information architecture. Honestly about the nature: it is data from a sample of participants (findability in %), useful and fairly concrete, but still on a limited sample — hypotheses on problem sections, not a final verdict; participant relevance matters. Honestly about the essence: this is structure diagnostics and hypotheses, NOT redoing navigation and NOT growth — changing the structure and validating it is separate; the test itself does not improve findability. Honestly about pairing: a tree test complements card sorting well (first learn users' mental model, then check the structure) — but they are separate methods. The tree-test platform/recruiting (if paid) is separate. An important boundary: this is tree testing (findability in the structure), not usability testing of the finished interface (with design — separate), not card sorting (forming the structure — separate) and not redesign. If the structure is small and simple, the method is overkill. Picture this: instead of 'it is logical to us that it is in this section' you see that 70% of people look for a product not where it is. The base price starts from 12,000 ₽ per test; it depends on the tree size and the number of tasks.
Problems we solve
- You are not sure people find what they need in your navigation.
- Section names and logic seem obvious only to you.
- You do not know at which menu level users get lost.
- You designed the site structure without testing it on people.
What's included in the Tree testing service
- Preparing the navigation tree in test form
- Framing tasks to find typical items
- Recruiting relevant participants
- Measuring findability (success, wrong turns, levels)
- Identifying problem sections and names
- Hypotheses for the structure (change — separately)
- A report with findability metrics
- Reviewing results with you
What you get
- You know whether people find what they need in the structure
- Problem sections and unclear names are visible
- Hypotheses for navigation (rework/growth — separately)
- Data for information-architecture decisions
How the work goes: steps
- We clarify the structure, search tasks, audience; prepare the tree
- We run the test, measure findability and errors
- We compile a report with problems and hypotheses, review with you
Why PDV Expert
- Fixed price and timeline — no surprises on the invoice.
- Report and recommendations in plain language — clear without a technical background.
- In touch at every step and answering questions about the result.
FAQ
Will a tree test raise conversion?
Not directly. It checks findability in the structure and gives hypotheses; changing navigation and validating the effect is separate. The test itself does not improve findability but helps you not lose people on an unclear structure.
Why without design?
Deliberately: by removing design and images we isolate the logic of the structure and names — we check exactly that, not visual cues. So a tree test is narrow: it is about findability in the hierarchy, not the whole interface.
How is it different from card sorting?
Card sorting helps BUILD a structure from users' mental model; a tree test CHECKS whether they find what they need in an already-defined structure. They are often used as a pair, but they are different methods (card sorting — separate).
About the provider
The «Tree testing» service is provided by PDV Expert — a team specialising in «Conversion & analytics». We work under contract and deliver a written report with recommendations.