Five-second rule
also: five-second test, 5 second test
The five-second rule in UX holds that visitors form their impression of a page within the first few seconds of viewing. The five-second test operationalizes it: show a design for five seconds, hide it, then ask what people remember — what the page offers, and what they were supposed to do.
The five-second test was popularized in usability practice (notably by UIE researchers in the mid-2000s) as a cheap way to test first impressions: if a viewer cannot say what the page sells and where to act after five seconds of exposure, the layout has failed at its first job, whatever its aesthetics.
Those five seconds are governed by attention, not reading. A viewer manages only a handful of fixations in that window, and they go to whatever the visual hierarchy makes dominant. If the hierarchy points at a stock photo, the stock photo is what gets remembered — the value proposition never entered the field of sharp vision.
Classically you run it with a panel: recruit viewers, flash the design, collect recall answers. Saliency prediction attacks the same window from the other side — models like UNISAL are trained on the first seconds of free viewing, so a predicted heatmap shows which elements will occupy those first fixations before you recruit anyone. Heatpoints scores exactly that first-seconds window on any URL or mockup.
Use both directions of the test: prediction to fix the obvious misfires (headline cold, CTA invisible) in minutes, and a human five-second test when you need to verify comprehension — that what people recall is what you meant.
RELATED TERMS