Moderated Usability Testing

Facilitated session where a researcher observes users attempting tasks on a prototype or product, documenting barriers, errors, and qualitative feedback in real time.

Process

Key Fields

Question it answers	Can users complete critical tasks? Where do they get stuck? What mental models drive their behavior?
Participants & timing	5-8 participants · 50-60 min per session · 3-4 weeks including prep and analysis
AI compatibility	AI transcribes think-aloud protocol and segments video into key moments; synthesis of user reasoning requires human judgment.
Output	Session videos, task completion metrics, critical errors documentation, usability issues prioritized by severity, recommendations

Use when

Testing detailed interactions, workflows, and decision points
Qualitative feedback on user reasoning is needed
Design is complex enough to require facilitated explanation
Time and budget allow for 5-8 sessions plus analysis

Skip when

Large-scale metric data is needed (use unmoderated testing)
Design is still at the rough wireframe stage
Budget does not allow for facilitation and analysis

Common Mistakes

Facilitator talks too much

Explaining the prototype or asking leading questions prevents observing real behavior. Stay silent until the user asks for help or becomes completely stuck.

Too many tasks

Five or six tasks feels quick but creates 25-40 data points to analyze. Do 2-3 core tasks deeply and observe all behaviors, not just task success.

Testing with non-target users

Friends, colleagues, or "anyone available" will not interact with the product like target users. Recruit the actual target segment; pay for recruiting if necessary.