Designers employ visual hierarchy to predict what users will notice within a display. Faraday’s visual hierarchy theory often ranks interface elements based on motion, size, images, color, text style, and position. Unfortunately, these heuristic-based approaches for determining where users will look first are ineffective. Instead, computational models can be used to improve the predictions of initial fixations. Computational models provide a means to capture those unconscious properties that are not available through introspection. The model’s effectiveness is increased by considering users’ previous experience. By quantifying visual salience and conventional patterns, designers gain better insight into user actions. This talk explores some initial discoveries that help designers understand unconscious processes’ unique influence on search efficiency across various interfaces.