AI Second that Emotion - Using Natural Language Processing to Study the Impact of Non-Stereotyped Video Advertising on Consumers’ Emotions & Online Consumer Engagement

Abstract

This paper aims to provide a deeper understanding of the emotional and online engagement behavioral responses to non-stereotyped gender role depictions in video advertisements. The consumer response to two video ads that portray non-stereotyped gender roles by the well-known brands Gillette and Always was analyzed. A total of 44 382 YouTube comments were classified and analyzed using transformer models for text-based emotion detection. Additionally, a topic analysis and a sentiment analysis were conducted. These analyses aim to classify and examine the underlying topics in the comment section, providing a deeper understanding of these topics’ part in eliciting emotions. The findings indicate that the underlying topics that contribute to online consumer engagement, in the two specific cases, consist of (1) marketing-related discussions, (2) stereotypes in general and gender role stereotypes, (3) product-related discussions, and (4) discussions on feminism. Furthermore, the findings suggest that the online consumer engagement stemming from the non-stereotyped ads (i.e., where target behavior was incongruent with predominant stereotyped gender roles) was mainly negative. This is further supported by the additional findings indicating that non-stereotyped ads lead to intense discussions in the comments section, where people disagree. Previous attempts to analyze online consumer engagement have been limited to a fraction of available consumer-generated comments. Hence, this study fills an academic gap by identifying the emotions that bring about online consumer engagement toward non-stereotyped gender role depictions in video advertisements while also considering the entirety of the available data. Moreover, the findings ultimately question the academic commendation of the role of non-stereotyped gender role depictions in advertisements, as it indicates that actively challenging established gender stereotypes can incite negative emotional - and online consumer engagement behavioral responses.