Analyzing Viewers’ Responses to Exposés of Corporate Social Irresponsibility on Social Media: An Emotion Analysis

Abstract

Listen

Translate article

This study examines consumers’ emotional responses to YouTube videos that exposed corporate social irresponsibility incidents. A total of 67,117 comments were collected and screened to identify those that mentioned the brand name. The analysis was divided into two distinct phases. In phase one, BART, a deep learning algorithm, was used to extract emotions expressed through comments in response to exposure to CSI incidents. In the second phase, this study employed negative binomial regression and analyzed a conceptual model to examine the relationship between emotion and supportive behavior by fellow consumers on YouTube. Results of phase one suggested that consumer responses consist of negative and positive emotions toward the brand, such as joy, trust, surprise, and anticipation. The phase 2 findings indicate that anger, anticipation, trust, and sadness garner more support from other consumers, while surprise, disgust, and joy are negatively associated with supportive behavior. This study extends the developing body of literature on consumers’ emotional responses to corporate social irresponsibility by taking a novel context. Moreover, it also showed the importance of different discrete emotional responses in increasing the supportive behavior of fellow consumers. Practically, this study provides directions to promotion managers for managing social media crises due to CSI.