Skip to content

Moderated testing

By Khyati Ojha. Reviewed by Ishan Vyas.

Moderated testing is usability testing led live by a facilitator, who gives a participant tasks, observes and listens as they work, and asks follow-up questions, in person or remotely. The facilitator's live presence is what separates it from unmoderated testing, where software presents the tasks.

How it works

The facilitator works from a task list and a few planned questions but adapts in the moment, asking a participant to say more, returning to a hesitation or skipping a question already answered. The format works in a lab or remotely over a screen share, where facilitator and participant sit in different places and still talk throughout.

A facilitator can notice a pause, test a hunch with the next question and clear up a misunderstanding before it distorts the data. Nielsen Norman Group notes that doing this without influencing the participant is difficult and takes training. Each session also needs a facilitator at a booked time, so moderated studies cost more per participant and rarely reach large samples.

A worked example

Partway through a yogurt brand's moderated test of its online store, 5 of the first 12 participants pause at the same line on a product page, "30 percent less". The facilitator asks what they think it means. Four of the five read it as less yogurt rather than less sugar, and the claim is rewritten before a quantitative test with 1,000 shoppers.

Automate this in Pulse

Clearing up a misunderstanding before it distorts the data is what a moderated tester's follow-up is for, and when a participant's account of a task goes vague, Pulse's moderator asks what they thought that step would do.

AI moderation