Investigating Consumer Responses to Brand Activism and Artificial Intelligence-driven Tools

Abstract

Current political and technological trends have enormous impact on marketing strategy and consumer behavior. To guide marketing managers and advance theoretical and practical knowledge in these turbulent times, this dissertation focuses on two highly relevant and cutting edge issues, by investigating consumer responses to brand activism (paper 1 and 2) and artificial-intelligence driven tools (papers 3-5).Many societies around the globe experience an increasing political polarization and many consumers expect brands to engage in brand activism, i.e., publicly taking a stand on divisive socio-political issues (Mukherjee & Althuizen, 2020). As a response, brands increasingly advocate for controversial issues such as Black Lives Matters, gun laws or LGBTQIA-rights. Although it seems tempting for brands to actively influence public opinion and to differentiate themselves from competitors at the same time, managers need to be aware about possible backlashes from those consumers who disagree with their stance. Extremely negative consumer reactions might follow, such as people burning their Nike shoes as response to the brands engagement for Black Lives Matter. Empirical evidence has already established that brand activism is a risky strategy (Bhagwat et al., 2020) and scholars call for research to help managers understand the psychological mechanisms influencing the effects of brand activism on consumer responses. Consequently, paper 1 of this dissertation investigates the role of moral emotions (such as anger or gratitude) as mediating factor shaping their reactions when they (dis-)agree with the company’s stance. Thereby, we focus on both company- and issue related responses. Moreover, paper 2 examines the role of consumer-brand identification (CBI) and political ideology related to consumers’ responsiveness to brand activism. In addition, we assess perceived marginalization as further relevant mediating factor, which explains consumer reactions in case of their disagreement with the brand’s position. As further mega-trend in marketing, the proliferation of Artificial intelligence (AI) driven tools is strongly transforming marketing activities and customer experiences (Longoni & Cian, 2022). Both scholars and practitioners acknowledge the immense and often even disruptive potential of AI-infused applications such as self-driving cars, precise customer-screening and demand-forecasting tools, or service robots (Davenport et al., 2020; Osburg et al., 2022). As one of the most pervasive and prevalent examples, the release of the AI-driven content-generation tool ChatGPT has triggered a real hype. In just five days, it has attracted more than 1 million users, making it the fastest diffusion of a new technology ever recorded (Anderson & Rainie, 2023). Several research studies (including our study in paper 3 which relies on the predecessor model of ChatGPT) revealed that AI-generated content is hardly discernible from human-authored content. Given this high performance and expected efficiency gains for marketing automation, managers are increasingly tempted to use AI as an autonomous content creator. However, the understanding of consumer expectations and responses to AI as a content author remain limited in the marketing domain. Therefore, using the example of a highly relevant marketing text related to talent attraction, paper 3 of this dissertation compares a human-authored text with an AI-generated text. Our research 3 investigates potentials of AI-authored texts for branding activities and explores readers’ reactions to AI disclosure. Moreover, the impact of matched or violated expectations on the company’s image as an employer and the role of feelings of betrayal as a mediating variable are examined. Studies in various contexts and the results of paper 3 revealed that people tend to have an algorithm aversion, leading to negative effects when AI is disclosed. However, as transparency is going to be a mandatory legislative requirement (e.g., as regulated in the European AI Act (European Parliament, 2023)), managers are increasingly faced with the question how to use AI without risking negative consumer reactions. As a possible solution, paper 4 investigates whether human-AI collaboration could serve as an escape from consumers’ algorithm aversion. Furthermore, this research examines the effectiveness of two distinct collaboration forms (i.e., “AI supporting a human author” vs. “AI author controlled by a human”), and the moderating impact of people’s (general) morality perceptions of a company’s AI use. ChatGPT and similar tools could also be integrated as digital conversational agents to fully automate various consumer-firm interactions and service processes. Despite an increasing prevalence and high potential for efficiency gains, these chatbots still often fail and recovery strategies are urgently needed. Consequently, paper 5 evaluates the effectiveness of two prominent failure recovery messages to maintain consumer satisfaction and loyalty. In addition, effects of situational factors such as different failure attributions or a double failure are taken into account.