Feasibility of Using Electrooculography-Based Eye-trackers for Neuromarketing Applications

Abstract

Eye-tracking has been increasingly utilized in neuromarketing applications to quantitatively investigate consumers’ subliminal attention and preferences. Eye-tracking technology in neuromarketing has been developed mainly using camera-based methods; however, camera-based eye-trackers are generally expensive for some customers, hindering widespread use of eye-tracking technology. Electrooculography (EOG)-based eye-tracking could be a promising alternative because it can be implemented at a relatively lower cost even less than 100 USD; however, EOG-based eye-tracking has not yet been employed in neuromarketing applications. In this study, we developed a method to estimate the absolute coordinates of eye gaze from EOG and investigated its reliability over time. The EOG signals were acquired in three separate experimental sessions: calibration, instructed eye-tracking test, and free eye-tracking test (FET) sessions. A linear regression model was constructed using the calibration data to estimate the absolute coordinates on the screen. The reconstructed eye-gaze results were visualized in the form of a heat map to be applied to practical neuromarketing scenarios. The results of the instructed eye-tracker test session revealed that the reliability of the eye-tracking results gradually decreased over time but did not drop drastically for up to 8 s. Moreover, the results of the neuromarketing test session analyzed with 6-s-EOG-data revealed clear gender differences among the heat maps of six advertisement pictures, coinciding with previous camera-based study results. Our results suggest that the proposed EOG-based eye-tracker has great potential to be utilized in practical neuromarketing applications.