Skip to content
Design Academy My progress

Template

Usability findings report

Report issues with evidence, severity and recommendations people can act on.

Usability findings report

Product / version tested: Dates: Participants: (number and who; no personal details) Method:

Summary

Three to five sentences: what we tested, the most important findings and the top recommendations.

Task results

TaskCompletedCompleted with helpNot completedNotes
1
2

Findings

Rate severity: Critical (prevents task completion), Serious (major delay or frustration), Minor (small friction), Suggestion.

Finding 1: (short, specific title)

  • Severity:
  • Seen in: x of y participants
  • What happened:
  • Evidence: (quote or observation, with participant number)
  • Why it matters:
  • Recommendation:

Finding 2:

  • Severity:
  • Seen in:
  • What happened:
  • Evidence:
  • Why it matters:
  • Recommendation:

What worked well

Keep these during changes. -

Recommended next steps

RecommendationFinding(s)PriorityOwner

Limitations

Small sample, prototype fidelity, recruitment bias, anything that limits how far the findings generalise.

View the Markdown source
# Usability findings report

**Product / version tested:**
**Dates:**
**Participants:** (number and who; no personal details)
**Method:**

## Summary
Three to five sentences: what we tested, the most important findings and the top recommendations.

## Task results
| Task | Completed | Completed with help | Not completed | Notes |
|---|---|---|---|---|
| 1 | | | | |
| 2 | | | | |

## Findings
Rate severity: **Critical** (prevents task completion), **Serious** (major delay or frustration), **Minor** (small friction), **Suggestion**.

### Finding 1: (short, specific title)
- **Severity:**
- **Seen in:** x of y participants
- **What happened:**
- **Evidence:** (quote or observation, with participant number)
- **Why it matters:**
- **Recommendation:**

### Finding 2:
- **Severity:**
- **Seen in:**
- **What happened:**
- **Evidence:**
- **Why it matters:**
- **Recommendation:**

## What worked well
Keep these during changes.
-

## Recommended next steps
| Recommendation | Finding(s) | Priority | Owner |
|---|---|---|---|
| | | | |

## Limitations
Small sample, prototype fidelity, recruitment bias, anything that limits how far the findings generalise.

All templates