Usability Testing Scenario Template: Design & Document

Five participants catch approximately 80% of usability issues (Nielsen, 2000). This applies to qualitative usability testing where the goal is finding problems, not statistical significance. If you are testing multiple distinct user segments (e.g., admins vs. members), recruit 3-5 per segment. For quantitative metrics (task success rates with confidence intervals), you need 20+ participants.

Moderated testing is better for complex flows, prototype testing, and when you need to ask follow-up questions. Unmoderated testing is better for simple tasks, large sample sizes, and when you need results faster. Start with moderated testing for early-stage discovery. Switch to unmoderated for validation at scale. The [product discovery handbook](/discovery-guide) covers when to use each method in the research lifecycle.

Usability testing observes real users attempting real tasks. A [UX audit](/templates/ux-audit-template) is an expert review against heuristics. They catch different problems. Audits catch technical violations (contrast, keyboard traps, inconsistency). Usability tests catch mental model mismatches (users expect the feature to work differently than it does). Run both for thorough coverage.

Sometimes one user loves a feature and another hates it. Look for patterns in user segments. A power user and a novice may have opposite reactions to the same interface. Document both perspectives and note the segment. If the finding is split 50/50 with no segment pattern, flag it as "inconclusive" and recommend A/B testing with the [experimentation](/glossary/a-b-testing) approach.

Both. The written report is the system of record. Video clips are the persuasion tool. A 30-second clip of a user struggling to find the "Add Data Source" button is more convincing than a paragraph describing the problem. Include 2-3 clips in your readout presentation. Link to full session recordings for stakeholders who want more context. ---