The web advertising on electronic commerce web attracts customer’s attention as a new communication channel, and web advertising could provide relevant advertising of web page content. But web advertising is randomly given information content, therefore, providing a personalized advertising is the best way. In this study, we developed an adaptive system to predict a user’s brand preference according his/her eye fixations on the editorial unit of webpages and then provided personalized banners to the user in a real-time manner. In order to examine the effectiveness of the system, an eye-tracking lab experiment was conducted. The results show that our system has a better prediction capability than by chance.