Designers
ToolsDocsProAll the Possibles

All 141 tools / Research and usability testing

Score tree test results per task with success, directness, first clicks and confidence intervals

Tree test scorerFreeFor researchers and designers

20 min

saved on a typical run. Our estimate, against tallying success, directness and first clicks per task in a spreadsheet.

Before: what you give it

The example input, as typed in. You paste text, CSS, markup or notes.

results:
  task,participant,success,direct,first_click,time
  Find the refund policy,p1,yes,yes,Support,12
  Find the refund policy,p2,yes,no,Products,31
  Find the refund policy,p3,no,no,Products,44
  Find the refund policy,p4,yes,yes,Support,15
  Compare laptop prices,p1,no,no,About,40
  Compare laptop prices,p2,yes,yes,Products,10
  Compare laptop prices,p3,yes,yes,Products,9
  Compare laptop prices,p4,yes,no,Support,28
expectedFirstClicks:
  Find the refund policy = Support
  Compare laptop prices = Products

After: what comes back

The first lines of a real run on that input. Nothing here is edited by hand.

8 attempts across 2 tasks. Overall success 75%, directness 50%.
Bands (convention): 80% and up strong, 60 to 79 mixed, under 60 weak. Ranges are 95% Wilson intervals.

Find the refund policy: 75% success (30 to 95), mixed, n=4, direct 50%, median 31s.
  First clicks: Support 50%, Products 50%; expected first click 50%.
  - only 4 participants: treat the rate as a rough signal (the 95% range is wide)
  - many successes needed backtracking; the right branch is findable but not obvious
Compare laptop prices: 75% success (30 to 95), mixed, n=4, direct 50%, median 28s.
  First clicks: Products 50%, About 25%, Support 25%; expected first click 50%.
  - only 4 participants: treat the rate as a rough signal (the 95% range is wide)
  - many successes needed backtracking; the right branch is findable but not obvious

Limits: needs one row per participant per task with columns task, success, and optionally direct, first_click, time (and expectedFirstClicks for first-click scoring). Small samples give wide ranges and the bands are rules of thumb, not benchmarks. It cannot judge whether the task wording was fair.
... full output on the tool page

What you get

Per task: success rate with 95% Wilson interval and band, directness, first-click spread, and median time

What it cannot do

Needs CSV with task and success columns; the strong, mixed and weak bands are a rule of thumb, and small samples give wide intervals.

Who it is for

  • Researchers
  • Designers

Free and Pro

Free25 runs a day per IP address, up to 200,000 characters a run.
ProNo daily limit, 300 requests a minute, up to 200,000 characters a run.

What each plan includes

Use it from your assistant

The tool is called tree_test_scorer. Connect the server once, then ask for it by name or just describe the job. Install steps for each app.

claude mcp add --transport http designers https://designers.allthepossibles.com/mcp
curl -X POST https://designers.allthepossibles.com/api/tree-test-scorer -H 'content-type: application/json' -d '{"results":"task,participant,success,direct,first_click,time\nFind the refund policy,p1,yes,yes,Support,12\nFind the refund policy,p2,yes,no,Products,31\nFind the refund policy,p3,no,no,Products,44\nFind the refund policy,p4,yes,yes,Support,15\nCompare laptop prices,p1,no,no,About,40\nCompare laptop prices,p2,yes,yes,Products,10\nCompare laptop prices,p3,yes,yes,Products,9\nCompare laptop prices,p4,yes,no,Support,28","expectedFirstClicks":"Find the refund policy = Support\nCompare laptop prices = Products"}'

Try it here

The form starts with the same example input, so you can run it first and then swap in your own.

More in research and usability testing

All research and usability testing tools