This article examines usability evaluation methodologies, then presents a non-conscious behavioral indicator based on user eye movements and pupil dilation. The authors test how gender and online buying history affect the behavioral index's usability scores. This study uses three Iranian online food retailers. Thirty participants were asked to add things from predetermined grocery stores to virtual shopping carts before the experiment took them to the other two websites in a random order to collect eye movement data. Each group's presentation order was randomized. The number of fixations, number of saccades, total duration of fixations, scan-path length, pupil size, and task time were inversely linked with self-report usability measures. This research evaluates groups with different levels of online shopping expertise and gender based on experienced usability. Differences between groups suggest that user demographics affect usability.