Events

Prof. Yifan Yu
Assistant Professor
Department of Information, Risk, and Operations Management
McCombs School of Business, The University of Texas at Austin

On online shopping websites, outstream video ads target search queries and autoplay on shoppers’ screens. We investigate the effectiveness of outstream video ads by conducting a large-scale, query-level study on experimental click-stream data using video analytics, machine learning, and econometric analysis. We find that video ads attract consumer attention and are more effective when products are less differentiated from each other in a market. In addition, various thumbnail features are significantly linked to attention. Contingent upon attention, the effectiveness of outstream video ads is heterogeneous across various visual features of the video content. Successful video ads leverage visual complexity in thumbnails to attract attention but leverage visually simple and conceptually meaningful content features to facilitate product evaluation and offer quality signals. Two laboratory studies further strengthen our conclusions. Our work contributes to the literature on advertisement in E-commerce and visual analytics. It provides important implications for practitioners to understand consumer decisions and advertising effectiveness, create effective outstream video ads, and improve video ad recommendation systems.