A usability test session facilitated by a researcher who introduces tasks, observes reactions in real time, and probes for deeper reasoning. The moderator's presence enables follow-up questions that reveal the why behind user behaviour, though it requires scheduling and can subtly influence participant responses.
Common contexts
- Testing a complex new workflow where the team needs to understand not just what fails but why users interpret it incorrectly
- Running sessions with screen reader users where an unmoderated setup cannot accommodate assistive technology variability
- Conducting moderated sessions with enterprise users to probe workflow context that automated tools cannot capture
Use when
Choose moderated testing when you're in early-stage discovery or testing genuinely novel interactions — the ability to ask follow-up questions is worth the scheduling cost when you need to understand reasoning, not just behavior.
Avoid when
Don't run moderated testing when you need volume quickly to validate a straightforward usability hypothesis — unmoderated testing produces the same behavioral data faster and without the moderator influence risk.
The moderator's greatest skill is knowing when to stay silent — the most valuable insights in a session almost always come from the pause you resist filling.
Real-world examples
- Google's UX research team runs weekly moderated usability sessions in their London and Mountain View labs, with remote observers watching via livestream and submitting questions through a moderator.
- Healthcare.gov conducted intensive moderated testing after its 2013 launch failure; the sessions revealed that insurance-plan comparison was the single most confusing interaction, driving a complete redesign of that flow.
- Apple's Human Interface team runs 'think-aloud pairs' in moderated testing — two participants solving a task together — to surface rationalisation patterns invisible when a single user works silently.