Emotion-Induced Engagement in Internet Video Ads

This study shows how advertisers can use emotion and attention patterns to engage consumers in watching internet video ads. In an experiment, joy and surprise are assessed through automated facial expression detection for a sample of ads. Concentration of attention is assessed through eye-tracking, and retention of viewers through zapping. This allows tests of unexplored predictions about the interplay of these emotions and attention at each point in time during exposure. The authors find that surprise and joy are effective in concentrating attention and retaining viewers. But importantly, the level rather than the velocity (i.e. change in level) of surprise impacts attention concentration most, whereas the velocity rather than the level of joy impacts viewer retention most. The latter effect is asymmetric: higher gains for increase than loses for decrease. Based on these findings, we develop a measurement tool to compare ads on the basis of their predicted ability to evoke surprise and joy effectively to concentrate attention and retain viewership. Benchmark emotional profiles are developed to show which emotion to evoke when and how intensely, to maximize viewer engagement.