| Author: | Qu, Ying |
| Title: | AI avatars and virtual influencers : understanding trust dynamics in digital marketing |
| Advisors: | Lo, K. Y. Chris (LMS) |
| Degree: | Ph.D. |
| Year: | 2025 |
| Department: | Department of Logistics and Maritime Studies |
| Pages: | ix, 198 pages : color illustrations |
| Language: | English |
| Abstract: | The rapid advancements in artificial intelligence (AI) have revolutionized the digital marketing landscape, leading to the proliferation of avatars—digital entities with an anthropomorphic appearance, controlled by humans or software, that are capable of interaction. While avatars have long been utilized, their roles have evolved from service agents to multifaceted tools across social media and beyond. For example, virtual influencers—AI-powered, computer-generated characters that operate on social media in lieu of human influencers—have gained significant attention from academia and industry. However, despite their growing popularity, building consumer trust in virtual influencers remains challenging. Virtual influencers can be freely modified and are highly adaptable, capable of appearing in diverse environments (i.e., virtual or real), paired with any companion (i.e., virtual or human), and expressing a broad spectrum of emotions, from positive to negative. Yet, little is known about how to leverage this adaptability to build trust in virtual influencers. This dissertation addresses these gaps through three interconnected chapters, offering a comprehensive examination of avatar marketing with a specific focus on virtual influencers. Chapter One provides a systematic literature review of avatar marketing by synthesizing insights from 203 Web of Science-indexed publications from the past fifteen years. It employs bibliometric network visualization to chart the evolution of avatar research and uses citation network analysis to identify seven distinct research domains, with main path analysis mapping the knowledge structure within each. The chapter concludes by outlining future research directions across these domains and proposing a four-phase, seven-domain avatar marketing framework. Chapter Two investigates the effect of contextual cues on shaping consumer trust and overall attitudes toward virtual influencers. Three experiments reveal that human-like virtual influencers are perceived as less trustworthy than either their anime-like counterparts or human influencers. However, trust in virtual influencers improves when they are presented in a virtual (vs. real) environment or accompanied by a virtual (vs. human) companion, resulting in more positive attitudes. Chapter Three examines the asymmetric effects of emotional expressions on consumer perceptions of virtual influencers. Through an analysis of Instagram field data combined with three controlled experiments, the research demonstrates that negative (vs. positive) emotional expressions elicit stronger feelings of uncanniness and reduce consumer trust, thereby diminishing marketing outcomes (i.e., product attitudes and purchase intentions). This dissertation provides a comprehensive exploration of virtual influencer marketing, integrating insights from consumer psychology, human-avatar interaction, and marketing communication. The findings offer valuable implications for both researchers and practitioners, providing guidelines for effectively incorporating virtual influencers into digital marketing strategies. The dissertation concludes by discussing the theoretical and managerial implications and suggesting directions for future research. |
| Rights: | All rights reserved |
| Access: | open access |
Copyright Undertaking
As a bona fide Library user, I declare that:
- I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
- I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
- I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.
Please use this identifier to cite or link to this item:
https://theses.lib.polyu.edu.hk/handle/200/13933