A New Livestream Retail Analytics Framework to Assess the Sales Impact of Emotional Displays

Abstract

At the intersection of technology and marketing, this study develops a framework to unobtrusively detect salespeople’s faces and simultaneously extract six emotions: happiness, sadness, surprise, anger, fear, and disgust. The authors analyze 99,451 sales pitches on a livestream retailing platform and match them with actual sales transactions. Results reveal that each emotional display, including happiness, uniformly exhibits a negative U-shaped effect on sales over time. The maximum sales resistance appears in the middle rather than at the beginning or end of sales pitches. Taken together, the results show that in one-to-many screen-mediated communications, salespeople should sell with a straight face. In addition, the authors derive closed-form formulae for the optimal allocation of the presence of a face and emotional displays over the presentation span. In contrast to the U-shaped effects, the optimal face presence wanes at the start, gradually builds to a crescendo, and eventually ebbs. Finally, the study shows how to objectively rank salespeople and circumvent biases in performance appraisals, thereby making novel contributions to people analytics. This research integrates new types of data and methods, key theoretical insights, and important managerial implications to inform the expanding opportunity that livestream e-commerce presents to marketers to create, communicate, deliver, and capture value.