Path replay
Replay makes participant paths easy to read at a glance, so teams can see where participants converge, scatter or backtrack without decoding an aggregate chart.
A tree testing tool for...
-
Validating a new structure
Test labels and grouping while they are still easy to change, before design and engineering commit to them.
-
Building new products or sections
Check whether a user seeing it for the first time can find each new area.
-
Diagnosing navigation problems
When analytics reveal drop-off, use tree testing to audit navigation and identify weak labels or structural problems.
Tree testing in action
Participants navigate a text-only version of your structure.
-
1 · The task
“You need to change the delivery address for an order.”
- 2 · They navigate the structure
-
3 · Your results
- First click
- Orders
- Path
- Orders → Change delivery details
- Time
- 8 seconds
- Outcome
- Direct
See where navigation breaks
- Direct
- Correct destination, no backtracking.
- Indirect
- Correct destination, reached with a detour.
- Elsewhere
- An unexpected destination was chosen.
- Success rate
- How often participants reached an accepted destination.
- Directness
- How often they got there without detours or backtracking.
First clicks
Where participants first expected the answer to be, before exploring anywhere else.
Final destinations
Where participants decided the content belonged, including the incorrect destinations they settled on.
Navigation paths
The route each participant took, including backtracking and repeated wrong turns.
Task outcomes
See direct, indirect and elsewhere outcomes at a glance, then open any segment to inspect the participant paths.
Time per task
How long each navigation choice took.
Participant-level results
Every participant's attempts across the whole study.
Only in UXbeam
UXbeam goes beyond task success: catch biased tasks, replay the paths participants took, and get suggestions for improving the next version.
01
Only in UXbeam
Study-quality checks, built in
-
Start with a strong structure
A stakeholder-drawn structure may reflect assumptions participants don’t share.
If revisions stall, bring in a card-sort sitemap built from participant groupings.
Explore card sorting →Why this matters
It is easy to get stuck with a structure that stays passable through sensible tweaks and miss the better one you never tested. When revisions stop improving findability, bring in a card-sort sitemap and test an alternative structure based on participant groupings.
-
Limit task bias
A task that repeats a tree label turns findability into word matching.
UXbeam flags revealing wording as you write and suggests alternatives.
Writing task instructions →Why this matters
Leading tasks are a major source of bias. They can turn a findability test into a reading task, where participants can match words instead of navigating by meaning. To avoid artificially increasing success rates, UXbeam can flag the task and overlapping terms and suggest rewordings.
-
Know which results to trust
Leading wording can make a task look easier than the navigation really is.
UXbeam checks completed results too and shows the overlapping terms.
Task-quality checks →Why this matters
Catching leading wording before launch is not foolproof. Some slips through, and a task can score direct for the wrong reason. UXbeam can run the check again on completed results and flag any task that shares wording with a top-level label.
02
Go beyond standard metrics
Traffic view
Make the routes participants actually used stand out. Branch thickness shows traffic. Low-traffic branches can be hidden to clear the noise.
Full-scenario completion
Every task can look acceptable while few participants finish the whole journey. UXbeam counts who does, and names the task where the rest diverge.
03
Card sort, tree test, optimize navigation
- Card sort Turn participant groupings into an editable sitemap.
- Sitemap Send sitemaps straight to tree testing, or start from outline, CSV or TXT.
- Tree test Set up in one screen, with task rewording and pretest review built in.
- Diagnose Diagnose with advanced metrics, path visualizations and AI-assisted analysis.
- Iterate Revise labels or structure, then test if pathfinding improved in one click.
UXbeam connects card sorting, editable sitemaps and iterative tree testing in one workflow. Start with participant groupings, test the resulting navigation, diagnose where it breaks, then revise to improve findability.
The complete tree testing tool, free
- Unlimited studies
- Unlimited tasks
- Unlimited responses
- Unlimited participants
- Full analytics included
- Advanced exports (CSV, SVG)
- Card sorting ↔ tree testing
- Version comparison
Complete analysis and handoff
Export your studies in formats you can reuse and share.
-
Full participant evidence
Every response stays inspectable next to the aggregate, with the complete analysis.
-
Participant CSV
Every participant-by-task attempt, ready for Excel or statistics tools.
-
Tree & task exports
The complete tree and task set as plain text, for tickets or version history.
-
Sitemap presentation cards
SVG menu cards for stakeholder review; any path view copies as an image.
Frequently asked questions
What does a tree test tell you that product analytics cannot?
Product analytics can show where users drop off or struggle. A tree test isolates the navigation so you can see whether labels or structure are causing the problem, including participants’ first choices, paths, backtracking and final destinations.
Tree testing or card sorting: which should you run?
Use card sorting when you need to learn how participants group the content. Use tree testing when you have a candidate structure and need to know whether users can find what they need. Many teams run a card sort first, build a candidate sitemap from it, and tree test that.
How does UXbeam compare with other free tree testing tools?
Free tree testing tools often limit response counts or access to results. On UXbeam there is no cap on studies or on participant responses, every collected response is viewable, the complete analysis is included, and you can export participant-level results at no cost.
How do I import a navigation tree?
Paste an indented outline, upload a CSV or TXT file, or import a candidate structure from a UXbeam card-sorting study.
What counts as success in a tree test?
UXbeam separates five outcomes rather than scoring every arrival the same way: Direct, the correct destination with no backtracking; Indirect, the correct destination reached after a detour; Elsewhere, an unexpected destination; Bounced, where the labels did not communicate where to go; and Gave up, where the task was abandoned. A high total that is mostly indirect is a different finding from one that is mostly direct.
How does UXbeam catch leading task wording?
A task that repeats a label from your tree turns findability into word matching. UXbeam checks your task wording against your own labels while you write and suggests less revealing alternatives. The same check runs again on completed results, marking any task whose wording overlaps the destination participants chose.
Can a task have more than one correct destination?
Yes. Mark several destinations as acceptable when a task has more than one valid answer, and designate a parent category when you want to learn which child item participants associate with the task rather than forcing one final destination.
Can you compare two versions of a structure?
Yes. Create a revision from the same tree and tasks instead of rebuilding the study. Each version keeps its own results, and UXbeam marks the versions where more tasks were found directly, so you can see what a change actually moved.
Do participants need an account?
No. Participants open the study link in their browser and complete the tree test without installing anything or creating a UXbeam account.
Can participants take a tree test on a phone?
Yes. The participant experience works in mobile and desktop browsers.
What can you export?
Participant-by-task results as CSV, the complete tree and task set as plain text, and proposed sitemap screens as SVG menu cards. Every individual participant attempt stays inspectable in the results alongside the aggregate.
