Co-Creating Values with Customers Through AI-Powered Branded App: The Mediating Role of Perceived Authenticity

The exponential growth of artificial intelligence (AI) has a profound influence on the marketing industry. Artificial intelligence technology becomes an independent touchpoint to “touch” users throughout their customer journey while enhancing brand identity and improving customers’ satisfaction. Despite of the high investment of brands new channels and new technology of AI-powered branded apps, customers quickly abandon the app. Previous research has explored the motives of branded apps’ continuance usage, including functional and emotional factors. Moving beyond utilitarian and hedonic benefits, consumers increasingly require more authentic brand delivery; they may question whether this virtual experience is like what they experience at the store and whether the app’s developer cares for their needs. Previous scholarly research still questions the authenticity of user experience during their interactions and calls for further empirical investigation to systematically explore the role of authenticity in determining preference for, or resistance to, AI-powered branded apps. Drawing from Regulatory Engagement Theory, we conceptualize authenticity as the key construct in customers' value experience process, triggering customers' behavioural intentions. The Parasocial Interaction Theory is adopted to explain the influence of media factors (e.g., media richness) and brand factors (e.g., co-branding fit) on authenticity and the influence of authenticity on behavioral intention such as usage intention and value co-creation intentions. The research employs a post-positivist paradigm with a convergent mixed-method design to allow an in-depth comprehension of AI-powered branded apps and customer perceptions. We conducted topic modeling and content analysis to explore the authenticity dimensions (Study 1). Then, the found dimensions are validated with structural equation modeling (Study 2) and further explored among different types of customers with Q-methodology (Study 3). The topic modeling and content analysis of 11,400 reviews reveal four possible dimensions of authentic AI-powered branded apps, including social presence, credibility, integrity and symbolism. Then, a quantitative study (n = 640) validates the dimensions above with Confirmatory Factor Analysis. The Structural Equation Modeling results reveal that authenticity is an important mediator between the app's characteristics (media richness and co-branding fit) and desirable outcomes (value co-creation and intention to use). Norms and age groups are also found as moderators. The qualitative study employing Q-methodology has further provided three types of customers based on their authenticity perception, including the Ethical Precisionist, the Reflective Care-Seeker, and the Pragmatist. Theoretically, this study extends the effort to measure authenticity from a multidimensional perspective by conceptualising that an AI-powered branded app’s authenticity includes credibility, integrity, symbolism, and social presence. This study provides empirical evidence for the impact of ‘proper means’ on engagement strength and value experience; thus, contributing to the application of regulatory engagement theory in the marketing technology context. The current study broadens the parasocial interaction theory by examining it from the perspective of authenticity and virtual media character (e.g., AI agents in branded apps). From an industry perspective, the study offers practical recommendations for marketing professionals to optimize AI functionalities, thereby enhancing authenticity in customer experiences. Industry projects that aim to leverage digital health marketing to support the marginalized community can also take advantage of these research findings. In terms of social impact, this research supports sustainable economic growth and productive employment through the AI technology market. By offering unique insights into authenticity, it assists marketing professionals in achieving efficient workflows, leading to more effective investments in AI technologies.<p></p>