How Do Virtual Influencers Affect Consumer Brand Evangelism in the Metaverse? The Effects of Virtual Influencers’ Marketing Efforts, Perceived Coolness, and Anthropomorphism

Abstract

Modern commercial organizations have started to embrace the metaverse platform as a new channel for marketing their products and services. As prominent brand representatives, virtual influencers are responsible for strengthening consumer–brand connections in the metaverse. However, the ways in which virtual influencers contribute to consumers’ brand fidelity and evangelistic behaviors remain unknown. To address this gap, this work explores the essential factors that impact consumers’ intention to trust virtual influencers in the metaverse, purchase the products that they promote, and engage in brand evangelistic behaviors. Specifically, a new framework is developed that integrates metaverse virtual influencers’ marketing efforts, perceived coolness, anthropomorphism, brand evangelism, and brand fidelity into a comprehensive conceptual research model. Survey results obtained from 713 respondents in the US demonstrate that marketing efforts, perceived coolness, and anthropomorphism positively influence brand evangelism. The findings also indicate that consumers’ brand fidelity encourages brand evangelism and mediates the relationship between virtual influencers’ features, marketing efforts, and consumers’ brand evangelistic behavior. This study’s significance lies in its focus on the evolving marketing dynamics in the metaverse. It considers how virtual influencers’ features and marketing efforts impact consumers’ brand-related attitudes and behaviors in the metaverse, offering valuable insights for the advancement of metaverse marketing research and practice.

1. Introduction

The metaverse is an emerging platform that enables users to engage with one another and participate in diverse activities in 3D virtual reality environments [1]. It is revolutionizing digital commerce [2] and has the potential to create a new global economy, being projected to reach USD 5 trillion in value by 2030 [3]. Predictions indicate that “approximately one-quarter of the population will dedicate daily at least one hour to engaging with the Metaverse by 2026” [4] (p. 1); users view it as a parallel virtual world where they can live a “digital life” through avatars [5]. Currently, most business concepts related to the metaverse refer to the commercial and entertainment domains, yet the possibilities are enormous [6]. Gaming and virtual worlds, such as Roblox, Fortnite, and Sandbox (along with older platforms, e.g., Second Life), are integral to the metaverse, in which multiple consumers immerse themselves in a shared virtual space to engage in virtual commerce [7,8]. Major corporations, such as Nike and McDonald’s, view the metaverse as a space for shopping, dining, working, socializing, gaming, cultural events, and more [9]; it represents a new marketing approach through which they can maintain their relevance among existing customers and attract new ones [10]. Traditional marketing methods in digital environments, such as the use of human influencers and banner ads, have relied on real-life personas to connect with audiences, but the metaverse enables the creation of entirely virtual marketing agents with distinct advantages. The metaverse has led to a new type of virtual influencer with the ability to increase brand promotion in the community: a computer-generated digital persona with a human-like appearance and personality [11]. These personas influence their audiences in unique and innovative ways and serve as brand ambassadors [12,13]. One of the most recent examples can be found in Maybelline New York’s campaign, which uses a virtual influencer called May, alongside the supermodel Gigi Hadid, to promote a new product; May will play a significant role in Maybelline’s future metaverse campaigns [14]. Unlike human influencers, virtual influencers are fully customizable, can operate 24/7, and provide brands with complete control over their messages and interactions, positioning them as a distinct and potentially more effective alternative tool for targeted marketing [15,16]. Furthermore, their ability to be seamlessly integrated into metaverse environments allows them to engage audiences in ways that traditional digital marketing methods cannot achieve, such as offering immersive, interactive experiences [17]. Considering the expansive market prospects of the metaverse, bringing a new era of marketing innovation and engagement [18], it is crucial to closely examine the features of virtual influencers that impact consumers’ purchasing choices within these specific virtual worlds.

Virtual influencers have recently attracted significant scholarly attention. Studies have investigated the differences between virtual and real influencers [14,16] and examined how human-like virtual influencers impact consumer satisfaction and purchase behavior on traditional social media platforms [15,19]. In particular, several scholars have studied virtual influencers in the metaverse. For instance, Koles et al. [20] interviewed consumers and industry experts to determine virtual influencers’ authenticity. Aw and Agnihotri [21] reported on the development of the virtual influencer industry and proposed several enlightening research directions focused on virtual influencers’ marketing efforts. However, no research has yet focused on the impact of virtual influencers’ marketing efforts on consumers in the metaverse.

This research gap becomes particularly noticeable when companies seek to develop brand evangelism among metaverse consumers using virtual influencers. According to Cestare [22] (p. 8), brand evangelism entails a more engaged and dedicated approach to “propagating positive opinions and convincing others to interact with the specific brand”. Recent marketing studies have indicated that brand evangelism, as a highly impactful construct, contributes significantly to sales expansion, procurement, and customer loyalty [23]. As prominent brand representatives, virtual influencers have the opportunity and responsibility to establish strong connections between consumers and brands, further fostering consumers’ behavior regarding the co-creation of brand value. However, research on the ways in which virtual influencers’ features and marketing efforts affect consumers’ brand evangelism is limited. This raises a pivotal question: How do virtual influencers’ features and marketing efforts impact consumers’ brand evangelism in the metaverse? At present, the marketing potential of virtual influencers in the metaverse remains unexplored.

This research also explores the underlying mechanism through which virtual influencers affect brand evangelism among consumers in the metaverse. Grounded in the stimulus–organism–response (SOR) framework [24], this study proposes brand fidelity as an important mediator in the impacts of virtual influencers’ features and marketing efforts on consumers’ brand evangelism. The SOR framework has been extensively used in marketing studies to analyze customer behavior [2,25]. It provides a structured approach to understanding how marketing stimuli generate emotional and cognitive psychological organisms and motivate their behaviors in response. Brand fidelity is an emotional and cognitive state that mediates the relationship between stimuli (i.e., virtual influencers’ features and marketing efforts) and responses (consumers’ brand evangelistic behaviors). Therefore, utilizing the SOR framework, we propose the mediating role of brand fidelity. Although brand fidelity is a relatively new concept in the study of brand–consumer relationships, it aids in elucidating consumers’ intentions to remain faithful to a brand, their willingness to invite others to interact with it, and their engagement in behaviors that maintain this relationship [26]. Our research model is supported by analyzing the data collected from a survey of 713 respondents in the US.

This study offers three significant contributions to the literature on branding, metaverse marketing, and virtual influencers. Firstly, previous research has mainly focused on the efficacy of virtual influencers in the domain of conventional social media marketing [15], but this study focuses on virtual influencers’ impact on consumers’ brand-related behaviors in the metaverse. Secondly, it identifies brand evangelism as one of the most impactful marketing variables and offers a novel perspective on how virtual influencers can impact consumers’ brand evangelism while creating value for brands in the metaverse. Thirdly, we examine how the mediating role of brand fidelity between virtual influencers’ features, their marketing efforts, and consumer brand evangelism. These findings offer valuable insights for future research on metaverse marketing.

2. Theoretical Background

2.1. Brand Evangelism in the Metaverse

Distinguishing them from brand influencers, Becerra and Badrinarayanan [27] define brand evangelists as those who admire a brand and frequently engage with it, primarily driven by their enthusiasm for it. As Nkoulou Mvondo et al. [23] (p. 3) state, brand influencers might not personally endorse or have an affinity for the brand; “their endorsement is often driven by financial incentives, leveraging their authority or fame to persuade others to buy or try the brand”. By contrast, brand evangelism suggests “a strong connection between the brand and consumers, demonstrated by spreading positive word-of-mouth and convincing others to connect with the specific brand” [28] (p. 1). In this manner, these consumers become messengers for others, as in traditional media.

Scholars generally accept Becerra and Badrinarayanan’s [27] (p. 1) statement that “brand evangelism consists of three dimensions: brand purchase intention (BPI), positive brand referrals (PBR), and oppositional brand referrals (OBR)”. BPI quantifies the degree of consumers’ intention to choose a brand for future purchases. Brand evangelists promote brands by consistently purchasing products [29]. PBR represent brand evangelists’ active support through positive opinions, product recommendations, and their encouragement of brand engagement. OBR reflect a negative attitude towards rival brands, driven by attachment and loyalty to the leading brand [23]. Evangelism marketing aims to inspire strong brand belief among buyers and compels them to spread the word. As Panda et al. [30] state, an engaging brand experience motivates customers to share their enthusiasm. Kang et al. [31] highlight that researchers are becoming increasingly interested in how people connect with brands and how this connection affects the value of these brands. In today’s digital world, this connection is essential because it allows consumers to share information quickly and broadly, and this can influence how other consumers behave.

Numerous studies have explored the factors that lead to consumer brand evangelism in offline and online marketing environments. Early research focused on individual personality traits and examined how they affected consumers’ brand evangelism. In their empirical study, Matzler [32] found that extraversion can directly and indirectly enhance brand evangelism through brand passion. Similarly, Doss and Carstens [33] demonstrated that there were significant correlations among individuals’ extraversion, neuroticism, openness, and brand evangelism. Subsequently, researchers began exploring the antecedents of brand-related psychological factors and brand features. For example, Harrigan et al. [34] discovered that active customer engagement with a brand and collaborative value creation led to brand-related evangelical behaviors. However, research on the antecedents of brand evangelism remains limited, and the present study is the first to explore these in the context of metaverse marketing.

Research on the metaverse can be categorized into three primary themes: metaverse adoption, applications, and characteristics. In terms of metaverse adoption, several studies have examined the drivers and barriers by employing various behavioral theories, including the unified theory of acceptance and use of technology, the theory of planned behavior, and social cognitive theories [35]. Several studies have explored its use in business [2,36], the workplace [37,38], and education [39]. Studies have also focused on defining the metaverse itself or its components [15,40]. These works provide different definitions of the metaverse; thus, there is no consensus regarding its nature [7]. Some scholars have also conducted literature reviews on metaverse virtual influencers [21]. However, researchers have still not explored the impact of such influencers’ features and marketing efforts on the occurrence of brand evangelism in consumer behavior.

2.2. Characteristics of Metaverse Virtual Influencers

Existing studies on human influencers (HIs) suggest that they significantly enhance brand engagement in online settings by leveraging their extensive networks of connections with other social media users, as well as through their perceived expertise via the content that they create [41]. With the rapid development of media and technologies, virtual influencers (VIs) have emerged as a novel type of social media influencer that is computer-generated using advanced software and lacks a physical human presence. They mimic human influencers by sharing content, seeking to appear authentic and credible, and ultimately, promoting brands [20,42]. Virtual influencers (VIs) and human influencers (HIs) play distinct roles in influencer marketing, each with unique strengths and challenges. HIs are perceived as more authentic and trustworthy due to their real-life experiences and their emotional connections with their followers, which are difficult to replicate artificially [14,15]. Conversely, VIs offer brands complete control over their content creation and messaging, eliminating the possibility of misbehavior [43]. Nevertheless, the perceived genuineness of HIs is often low [20]. VIs also present a cost advantage, as they do not incur expenses related to travel and accommodation, and they can operate 24/7 across multiple campaigns [21]. However, while HIs are valued for their perceived expertise in their specific domains, VIs may struggle to establish similar levels of credibility due to their lack of real-life experience [16]. Nonetheless, VIs are novel and innovative; they can attract attention and cause brands to appear forward-thinking and innovative [44]. Ethical and legal concerns further differentiate these two types of influencers; HIs are subject to established regulations and ethical standards, while VIs raise new questions about transparency and the potential to mislead consumers [45]. These distinctions are crucial for marketers to consider when choosing between VIs and HIs, particularly in the evolving context of digital platforms, including the metaverse. By understanding the strengths and limitations of each type, brands can align their influencing strategies with their marketing objectives and address audience preferences. These distinctions are summarized in

Table 1. Distinctions between human influencers (HIs) and virtual influencers (VIs) in influencer marketing.

Aspect	Human Influencers (HIs)	Virtual Influencers (VIs)
Authenticity and Trust	Perceived as more authentic due to real-life experiences.	May be seen as less genuine, affecting consumer trust.
	Independent content creation, leading to potential unpredictability.	Brands have complete control over content and behavior.
Content Control
Engagement and Novelty	Engage personally with audiences; preferred by consumers.	Their novelty can attract attention; may enable brands to appear innovative.
Cost and Efficiency	Incur expenses such as travel and accommodation costs; limited by personal schedules.	Cost-effective over time; available 24/7 without logistical constraints.
Perceived Expertise	Seen as experts through personal experience.	Lack real-life experiences; may struggle to establish expertise.
Ethical and Legal Concerns	Subject to existing regulations and ethical standards.	Raise unique ethical and legal questions; potential to mislead consumers.

.

Virtual influencers’ marketing efforts contribute to the creation of immersive brand experiences while influencing consumers’ purchase intentions. Previous studies have identified five essential dimensions of social media marketing that can increase the number of positive consumer responses: interaction, entertainment, trendiness, customization, and communication quality [46]. However, it is unknown whether such categorization can be applied to explain the marketing efforts of virtual influencers in metaverse marketing. Interaction involves virtual influencers’ direct engagement with consumers through personalized conversations and experiences. Phua et al. [47] stated that such interactions create individualized attention and a sense of intimacy, enhancing consumers’ emotional connections with the influencer and, in turn, with the endorsed brand. Advancements in technology have enabled brands to leverage social media influencers for informal interactions aimed at cultivating and reinforcing consumer relationships while providing information [48]. Successful brands understand the importance of integrating entertainment components into their products, services, and communities. Virtual influencers create engaging narratives and experiences, elicit emotional responses from consumers, and foster memorable associations with brands. Consequently, enjoyment, fun, satisfaction, and relaxation are pivotal in determining how customers respond to virtual influencers [49]. Virtual influencers enhance brand perceptions by acting as trendsetters, associating brands with trendiness, novelty, and a forward-thinking approach. Many consumers seek brands and products that align with their modern lifestyles. The degree of customization reflects how well a service is tailored to individuals’ preferences [50]. Therefore, customized products satisfy individual consumers’ preferences and help to build stronger brand fidelity and affinity [46]. Currently, virtual influencers offer tailored recommendations and content that enable consumers to feel valued and understood, creating a sense of inclusion in certain virtual groups. According to Phua et al. [47], communication quality pertains to the clarity, authenticity, and effectiveness of the messages conveyed by virtual influencers. Effective communication enhances consumers’ understanding of a brand’s attributes, fosters trust in influencers’ endorsements and increases satisfaction [51]. When information is accurate, customers trust in its reliability and comprehensiveness. Furthermore, according to Yuan et al. [52], customers view information as convincing and credible if it is obtained through positive relationships with communicators who actively listen to them, identify their concerns, and provide the required information.

According to Sundar et al. [53], perceived coolness is a multidimensional phenomenon in technological product judgment, where a product is perceived as attractive while simultaneously being capable of creating an original subculture. A product is perceived as attractive due to its aesthetically pleasing features and functionalities. Originality refers to a product’s design and function being unique and innovative [54]. Subcultural appeal refers to a product’s ability to encourage the formation of a subculture through its usage. Given that products with unique functions and appearances are not standard in mainstream society, they are likely to be more noticeable [55]. Perceived coolness significantly influences a product’s success in the marketplace, shaping how consumers perceive and utilize it [56]. However, the perception of coolness varies among consumers. Several researchers contend that “when consumers perceive a product or service as cool, they are inclined to show positive behavioral intentions” [54] (p. 3), including usage intention [56], loyalty [57], adoption [58], and purchase intention [59]. To the best of our knowledge, scholars have not yet explored the perceived coolness of metaverse virtual influencers; therefore, by doing so, this study fills a notable gap in the existing literature.

As noted by Epley et al. [60], anthropomorphism is a psychological concept whereby individuals attribute human-like characteristics and attributes to non-human entities, including robots or animals [61]. Previous research has demonstrated that it significantly affects consumer behavior [62]. Consumers are more emotionally engaged and attached to non-human creatures when they believe that they are human, influencing their decision-making behavior [63]. They perceive products as more trustworthy if they show human-like characteristics (e.g., arm gestures, smile mimicry), making them more endearing and likable, thus positively impacting their willingness to purchase them [64,65]. In the metaverse environment, individuals tend to place greater trust in the recommendations of virtual personas with human-like qualities [66], influencing their brand perceptions and fidelity [63].

2.3. Stimulus–Organism–Response Theory

The stimulus (S)–organism (O)–response (R) theory framework is the theoretical foundation of our research [24]. This theory is ideal for two reasons. Firstly, it has been widely applied in earlier studies across various domains, particularly in marketing and consumer behavior studies, to understand how environmental factors influence individuals’ internal states and subsequent actions [2,64,67]. Its application has proven effective in the analysis of the psychological mechanisms underlying consumer engagement, decision-making, and loyalty. For example, Kim et al. [64] used the SOR framework to analyze consumer behavior in virtual reality tourism, examining how environmental stimuli such as immersive content impacted emotional engagement and satisfaction. Similarly, Jafar et al. [2] employed the SOR framework to investigate how various aspects of the metaverse, such as telepresence and product knowledge, influence purchase intentions. Moreover, Armawan [67] analyzed studies that applied the SOR framework and investigated whether social media marketing affected brands and purchase intentions. This framework’s broad applicability in these contexts demonstrates its effectiveness in capturing the nuanced relationships between stimuli, internal cognitive or emotional processes, and consumer responses. Secondly, this theory offers a useful approach to exploring the effects of virtual metaverse influencers’ features and marketing efforts as external stimuli, as these are factors that affect users’ emotional and cognitive states in relation to brand fidelity. These internal experiences lead to responses to stimulus variables [68], such as the development of brand-evangelistic consumer behavior. By using this established theoretical model [69,70,71], we adopt a structured and validated approach to exploring how virtual influencers influence consumer behavior in the metaverse.

This study focuses on three stimuli: the perceived coolness of influencers, their human-like characteristics (anthropomorphism), and their marketing efforts. These factors impact the development of users’ brand fidelity and brand evangelism towards a particular brand. In the context of the metaverse, influencers’ marketing efforts play a crucial role as stimuli. Interactions with influencers, entertaining content, customization options, and high-quality communication with influencers can significantly shape consumers’ brand perceptions. According to the theory of planned behavior, individuals’ behavioral intentions are determined by their attitudes and norms, as well as behavioral control [72,73]. In influencer marketing strategies, the marketing efforts of virtual influencers shape consumers’ attitudes towards the endorsed brand and build a trusting relationship with them [17]. On the other hand, the perceived coolness and anthropomorphism of virtual influencers affect subjective norms, as individuals may feel more inclined to align with brands that are promoted by influencers who are perceived as “cool” or relatable.

Within the SOR framework, “stimuli influence on behavior is mediated through organismic experiences” [2] (p. 1) or the development of users’ brand fidelity. Therefore, our research model incorporates brand fidelity as a mediator of the characteristics of virtual metaverse influencers in developing brand evangelistic behavior. As noted by Grace et al. [74], consumers actively engage with brands, and these brands have become a means of self-expression. Therefore, the role of influencers as brand managers obliges them to communicate effectively and position brands in ways that foster strong connections with consumers. Brand fidelity is a novel construct in the marketing literature and deserves greater attention among scholars. Garg and Joshi [75] define it as a consumer’s faithfulness to a brand; it leads to behaviors that maintain a stable and lasting relationship between consumers and brands. These behaviors include the “derogation of alternatives, positive illusions, and cognitive interdependence, consisting of accommodation, forgiveness, and willingness to sacrifice, making brand fidelity multi-dimensional” [75] (p. 5). Grace et al. [26] further reduce these dimensions to the following: accommodation/forgiveness based on price and performance, cognitive interdependence, and the derogation of alternatives.

Metaverse influencers offer personalized digital interactions, leading to higher levels of consumer loyalty and emotional connection. They play a crucial role in developing brand fidelity, acting as aspirational and relatable personas, and aligning with consumers’ desired identities.

3. Research Model and Hypotheses

This study combines the concept of perceived coolness with dual process theory, as introduced by Sundar et al. [53]. Dual process theory posits that the decision-making process of the brain involves two reaction modes. In the first mode, the brain analytically processes information to form judgments [76]. When consumers perceive metaverse influencers as “cool”, this perception becomes part of a conscious brand analysis. In the second mode, the brain intuitively selects relevant content from its memories and previous experiences [76].

In the metaverse, where uniqueness and subcultural appeal are crucial [55], perceived coolness heightens the perceived significance of a brand promoted by influencers. This intuitive–heuristic process might cause consumers to develop a stronger sense of brand fidelity towards products or services promoted by influencers who are perceived as “cool” [77]. Based on this, we propose the following hypothesis.

H1. 

There is a positive correlation between the perceived coolness of metaverse influencers and consumers’ brand fidelity.

According to Epley et al. [60], anthropomorphism is a psychological concept in which individuals assign human traits or behavior to non-human animals, nature, or inanimate objects. It affects consumer behavior and can lead to higher degrees of emotional attachment [78]. According to social identity theory, individuals aim to increase their self-concept and self-esteem by identifying with specific social groups [79]. Followers are more likely to be influenced by virtual influencers who exhibit anthropomorphic features due to their relatability and perceived similarity [80]. Thus, consumers are more inclined to connect with and support the brands that these influencers represent, mirroring their digital aspirations and aligning with their digital persona’s identity. This suggests that the more human-like and relatable metaverse influencers appear, the stronger the fidelity that consumers will exhibit towards the brands that these influencers endorse [15]. Based on the above, we propose the following hypothesis.

H2. 

There is a positive relationship between metaverse influencers’ anthropomorphism and consumers’ brand fidelity.

Expanding on previous work [46], this study also proposes that metaverse influencers’ marketing efforts, characterized by multidimensional variables, significantly shape consumers’ perceptions and brand fidelity. The continuous interaction between consumers and virtual influencers in the metaverse fosters a sense of reciprocity, encouraging consumers to engage in brand fidelity behaviors. This hypothesis is grounded in the elaboration likelihood model, a dual process model that is widely used to understand audience evaluations of persuasive communication [81]. The ELM proposes that there are central and peripheral routes in information processing, with the persuasiveness of a message depending on the argument’s strength and credibility [82]. In the central route, a persuasive message requires the following dimensions: interaction, entertainment, customization, and communication quality. These demand cognitive engagement. By contrast, the peripheral route involves the evaluation of messages using environmental cues. Trendiness is a peripheral cue that facilitates quick judgment without extensive cognitive processing [83].

These dimensions collectively influence consumers’ perceptions and brand fidelity, reflecting the impact of influencer marketing on brand fidelity in the metaverse. This notion aligns with previous ELM research in the context of advertising and online reviews [84], extending its application to understanding the implications of metaverse influencer marketing on purchase intentions [85]. Thus, in this study, we aim to elucidate how these dimensions shape consumers’ behavioral intentions in the metaverse, emphasizing brand fidelity as a metric associated with influencer marketing’s effectiveness. Based on the above, we hypothesize the following.

H3. 

There is a positive relationship between the following dimensions of metaverse influencers’ marketing efforts: (a) interaction, (b) entertainment, (c) trendiness, (d) customization, (e) communication quality, and consumer brand fidelity.

Brand evangelism represents a strong emotional connection between brands and consumers, including active brand support and spreading positive word of mouth. According to Becerra and Badrinarayanan [27], brand evangelism enables consumers to extend their digital influence in the metaverse beyond their immediate social circles. As Shoaib et al. [86] note, an inspiring brand experience motivates customers to spread their enthusiasm to others, thus amplifying the brand’s reach.

The SOR framework is instrumental in understanding this dynamic, where brand fidelity functions as a response shaped by external stimuli. Stimuli, such as the perceived coolness of virtual influencers, their anthropomorphic qualities, and their marketing efforts, evoke emotional and cognitive reactions in consumers, leading to heightened brand fidelity. The emotional connection fostered by brand fidelity is a powerful motivator in encouraging consumers to actively advocate for the brand, as consumers with strong brand fidelity are more likely to engage in brand evangelism behaviors [87]. Therefore, brand fidelity serves as a mediator because it represents the emotional and cognitive commitment that connects external stimuli with consumer responses, such as brand evangelism, ensuring that these stimuli translate into sustained fidelity. The mediation of brand fidelity in these interactions reflects this theory’s applicability in understanding consumer behavior in the metaverse. Building on this analysis, we hypothesize the following.

H4. 

There is a positive correlation between brand fidelity and consumers’ brand purchase intention in the metaverse.

H5. 

There is a positive correlation between brand fidelity and consumers’ positive brand referrals in the metaverse.

H6. 

There is a positive correlation between brand fidelity and consumers’ oppositional brand referrals in the metaverse.

H7. 

Brand fidelity mediates the relationship between perceived coolness and brand evangelism.

H8. 

Brand fidelity mediates the relationship between anthropomorphism and brand evangelism.

H9. 

Brand fidelity mediates the relationships between the following dimensions of metaverse influencers’ marketing efforts: (a) interaction, (b) entertainment, (c) trendiness, (d) customization, (e) communication quality, and brand evangelism.

Figure 1 illustrates the research model applied in this study. This model visually represents the key constructs and hypotheses of the study, providing a clear framework for the exploration of the examined relationships.

4. Methodology

4.1. Measures

The questionnaire adopted in this study employed items and instruments that were adapted from previous research to guarantee both its validity and reliability. Six items from Sundar et al. [53] were adapted to measure PC. AN was measured using three items taken from Nowak and Rauh [88] and Powers and Kiesler [89]. MIME was measured using twenty-one items: four for IN, three for EN, three for TR, and three for CU. These were derived from the following works: Kim and Ko [48], Lee and Choi [90], and Coelho and Henseler [91]. For CQ, eight items were adapted from O’Reilly [92] and Stohl and Redding [93]. To measure BF, we used three items derived from Grace et al. [26]; for BE, we used ten items: four for BPI, adapted from Becerra and Korgaonkar [94], and three items each for PBR and OBR, derived from Power et al. [95]. Minor adjustments were made to all items to ensure that they fit the metaverse context. For instance, the original item “Having this product will make me look cool” was modified to “Interacting with metaverse influencers makes me look cool”. Before the questionnaire was distributed, it was evaluated by two marketing professors to ensure its content and face validity. After incorporating these professors’ feedback, a 5-point Likert scale was used for the items’ measurement.

4.2. Data Collection

This research focuses on how virtual influencers impact the development of brand evangelism (BE) and brand fidelity (BF) behaviors among metaverse consumers. To conduct this study, we created a structured three-section questionnaire, which is presented in

Table A1. Questionnaire.

Construct	Measurement Item on 5-Point Likert Scale “1—Strongly Disagree to 5—Strongly Agree”
Brand purchase intention (BPI)	BPI1: In the near future, I would probably buy from this brand.
	BPI2: In the near future, I intend to buy a specific product from this brand.
	BPI3: In the near future, I would likely buy from this brand.
	BPI4: In the near future, I would possibly buy from this brand.
Positive brand referrals (PBR)	PBR1: I will spread positive word of mouth about buying products from this brand.
	PBR2: I will recommend buying products from this brand to my friends.
	PBR3: If my friends were looking for specific brand, I would tell them to buy products from this brand.
Opposite brand referrals (OBR)	OBR1: When my friends are looking for specific product, I would tell them not to buy from any of the other brands.
	OBR2: I would likely spread negative word of mouth about products of other brands.
	OBR3: If someone tries to decry this brand, I will unquestionably tell them off.
Perceived coolness (PC)	PC1: MI is cool.
	PC2: When I first heard of MI, I remember thinking that it would be cool to interact with them.
	PC3: Interacting with MI makes me look cool.
	PC4: When I interact with MI, my response often is something like “That’s cool!”.
	PC5: MI has some cool features.
	PC6: MI is cooler than human influencers.
Interaction (IN)	IN1: MI have the knowledge to answer customers’ questions.
	IN2: MI are never too busy to answer customers’ requests.
	IN3: MI give customers individual attention.
	IN4: MI are consistently courteous with customers.
Entertainment (EN)	EN1: It is fun and enjoyable to share a conversation with MI.
	EN2: I was absorbed in the conversation with MI.
	EN3: I enjoy choosing products more if they are recommended by MI than if I choose them myself.
Trendiness (TR)	TR1: MI give the newest information.
	TR2: MI provides up-to-date products.
	TR3: It is fashionable to interact with MI.
Customization (CU)	CU1: If I changed MI to interact with human influencers, the products and services would not be as customized as those I have now.
	CU2: I feel that interacting with MI meets my personal needs.
	CU3: MI provide information about products according to my preferences.
Communication quality (CQ)	CQ1: The communication with MI is timely.
	CQ2: The communication with MI is accurate.
	CQ3: The communication with MI is credible.
	CQ4: MI is honest.
	CQ5: MI is trustworthy.
	CQ6: My interactions with MI are more productive than interactions with human influencers.
	CQ7: Interacting with MI is more efficient than other forms of communication.
	CQ8: MI saves a tremendous amount of time.
Anthropomorphism (AN)	AN1: MI behaves like humans.
	AN2: MI behaves lifelike.
	AN3: MI behaves naturally.
Brand fidelity (BF)	BF1: When a brand which is promoted by MI increases its prices, this is well justified.
	BF2: I would feel offended if someone said something bad about this brand.
	BF3: This brand is one-of-a-kind and, in my opinion, there is no competition.

in the Appendix A. The first section outlined the study’s aim, requiring the respondents to reflect on their recent interactions with metaverse virtual influencers and rate their overall experiences and satisfaction on a given scale by answering questions. They were assured that the provided information would remain confidential. The next section gathered demographic data (age, sex, profession, education, income, and daily hours spent online and in the metaverse). The third section included items that referred to the variables of interest. The cross-sectional survey was carried out using a non-probabilistic convenience structured sample and distributed through the Amazon Mechanical Turk (MTurk) online platform, which offers rapid access to many metaverse users. This platform was chosen due to its efficiency and cost-effectiveness, allowing us to reach a large number of respondents within a short timeframe. To ensure the responses’ relevance to the research objectives, the questionnaire link was accompanied by a message specifying that only individuals who had used a metaverse platform at least once in the last three months were eligible to participate. While no monetary compensation was offered to the respondents, their participation was entirely voluntary, which may have encouraged genuine responses. However, the use of MTurk introduced potential biases, as the participants may not have fully represented the general population in terms of demographics and technological literacy. To address this, attention checks were included to ensure the data’s quality; future research could adopt probabilistic sampling or longitudinal designs to capture broader and temporal trends. Despite these limitations, the sampling approach aligned with the exploratory nature of this study and contributed to providing valuable initial insights into metaverse users’ behavior.

The sample included 713 participants from the US, who were recruited within two months. It comprised 431 men (60.4%) and 282 women (39.6%). Most respondents (51.1%) were between 18 and 30 years old. Additionally, 80.7% held a bachelor’s degree, 29.9% were employed in the public or private sector, and 23.7% were self-employed (businesspeople). Regarding the number of hours spent on the Internet, 52.6% of the respondents spent at least six hours online, and 48.7% spent at least six hours on metaverse platforms.

4.3. Data Analysis

The analysis of the demographic data and the identification of common method bias were performed with IBM SPSS 23.0. Subsequently, SmartPLS 4.0.9.9 was used to assess the measurement and structural model. The initial phase of the data analysis focused on the model fit, data reliability, and validity. The subsequent stage focused on the hypothesized relationships.

5. Results and Discussion

5.1. Common Method Bias

In behavioral research, common method bias is considered a significant concern. It occurs if the responses do not reflect the respondents’ real attitudes but are affected by the survey instrument. The questionnaire first outlined the strict study anonymity protocol. The respondents were advised to respond neutrally and truthfully, understanding that there were no correct or incorrect answers. They were reminded to maintain neutrality and honesty throughout the survey completion process. Secondly, after applying Harman’s single-factor approach, it was found that the variance extracted using a single factor was 16.113%. This value, being below the recommended 50% threshold, indicated that this research had no common method bias. A full collinearity test was also conducted [96], which revealed that all variance-inflated factors were below 3.3. Therefore, common method bias was not a concern in this study.

5.2. Assessment of Measurement Model

In this study, in line with Ringle et al. [97], we first evaluated the reliability and convergent validity using standard criteria: factor loadings (>0.5) to evaluate the reliability of the measures, Cronbach’s alpha, the composite reliability (CR) of the constructs (>0.7), and the average variance extracted (AVE) (>0.5). The latent variables ranged from 0.671 to 0.812, indicating a statistically significant loading. Additionally, Cronbach’s alpha varied from 0.776 to 0.898, and the CR varied from 0.776 to 0.897, indicating the reliability of the measurement instruments. Finally, the AVE was higher than 0.5 and ranged from 0.510 to 0.610. The details are presented in

Table 2. Reliability and validity.

Construct	Item	Loading	Alpha >0.7	CR >0.7	AVE >0.5
Anthropomorphism (AN)	AN_1	0.774 ***	0.803	0.803	0.576
	AN_2	0.770 ***
	AN_3	0.733 ***
Brand fidelity (BF)	BF_1	0.790 ***	0.824	0.824	0.610
	BF_2	0.778 ***
	BF_3	0.774 ***
Brand purchase intention (BPI)	BPI_1	0.782 ***	0.832	0.832	0.554
	BPI_2	0.721 ***
	BPI_3	0.698 ***
	BPI_4	0.773 ***
Communication quality (CQ)	CQ_1	0.812 ***	0.898	0.897	0.522
	CQ_2	0.699 ***
	CQ_3	0.710 ***
	CQ_4	0.724 ***
	CQ_5	0.751 ***
	CQ_6	0.697 ***
	CQ_7	0.705 ***
	CQ_8	0.676 ***
Customization (CU)	CU_1	0.727 ***	0.797	0.797	0.567
	CU_2	0.774 ***
	CU_3	0.757 ***
Entertainment (EN)	EN_1	0.713 ***	0.776	0.777	0.537
	EN_2	0.786 **
	EN_3	0.698 ***
Interaction (IN)	IN_1	0.779 ***	0.815	0.815	0.524
	IN_2	0.734 ***
	IN_3	0.708 ***
	IN_4	0.671 ***
Oppositional brand referrals (OBR)	OBR_1	0.768 ***	0.776	0.776	0.536
	OBR_2	0.722 ***
	OBR_3	0.707 ***
Positive brand referrals (PBR)	PBR_1	0.772 ***	0.777	0.777	0.537
	PBR_2	0.742 ***
	PBR_3	0.682 ***
Perceived coolness (PC)	PC_1	0.695 ***	0.862	0.862	0.510
	PC_2	0.672 ***
	PC_3	0.698 ***
	PC_4	0.746 ***
	PC_5	0.755 ***
	PC_6	0.715 ***
Trendiness (TR)	TR_1	0.722 ***	0.798	0.798	0.568
	TR_2	0.749 ***
	TR_3	0.788 ***

Note: * indicates significant paths: * p < 0.05, ** p < 0.01, *** p < 0.001; NS = not significant.

.

Next, the discriminant validity was evaluated using the Fornell–Larcker method and the heterotrait/monotrait (HTMT) ratio. In line with Fornell and Larcker [98], each latent variable’s AVE square root exceeded the coefficients of its correlations with the other variables, confirming the discriminant validity of this study. According to Henseler et al. [99], the HTMT values should not exceed 0.85; thus, in this study, the discriminant validity was confirmed (see

Table 3. Discriminant validity analysis.

Construct	AN	BF	BPI	CQ	CU	EN	IN	OBR	PBR	PC	TR
AN	0.792	0.692	0.544	0.667	0.640	0.630	0.679	0.498	0.583	0.662	0.645
BF	0.600	0.789	0.595	0.687	0.688	0.681	0.710	0.550	0.570	0.697	0.690
BPI	−0.371	−0.356	0.752	0.604	0.569	0.551	0.607	0.487	0.518	0.572	0.571
CQ	0.586	0.591	−0.351	0.788	0.664	0.624	0.675	0.546	0.563	0.683	0.646
CU	0.606	0.584	−0.382	0.626	0.799	0.629	0.682	0.507	0.581	0.658	0.664
EN	0.473	0.466	−0.258	0.487	0.483	0.841	0.631	0.550	0.536	0.645	0.645
IN	0.595	0.578	−0.326	0.596	0.611	0.473	0.795	0.541	0.608	0.703	0.682
OBR	0.584	0.583	−0.382	0.617	0.582	0.467	0.590	0.797	0.494	0.531	0.540
PBR	−0.668	−0.659	0.428	−0.660	−0.666	−0.534	−0.644	−0.632	0.836	0.583	0.547
PC	−0.182	−0.180	0.346	−0.197	−0.180	−0.320	−0.214	−0.204	0.241	0.899	0.640
TR	0.613	0.613	−0.359	0.629	0.620	0.478	0.606	0.631	−0.648	−0.184	0.796

Note: bold values = square roots of the average variance extracted; off-diagonal values = correlations.

, which provides a detailed summary of the discriminant validity analysis). Moreover, the variance inflation factors (VIFs) in our study were below five, indicating acceptable levels of multicollinearity.

5.3. Assessment of Structural Model

In this study, we assessed the structural model and tested the hypotheses through the use of model fit indices, the R-squared (R²) measure, the Q-squared (Q²) measure, the SRMR value, and the path coefficient significance. According to Niu and Mvondo [54], the R² is a statistical indicator used in in regression analysis that shows the proportion of variance in a dependent variable that is explained by the independent variables. Based on the rules provided by Hair et al. [100], R² values of 0.25, 0.50, and 0.75 reflect small, medium, and large effect sizes, respectively. Our analysis revealed that the endogenous constructs BF (0.680), BPI (0.355), OBR (0.303), and PBR (0.325) demonstrated medium effect sizes, implying the acceptability of our regression model. The predictive relevance (Q²) was also assessed to gauge the model’s ability to predict data points not included in the sample. Also known as the Stone–Geisser indicator or redundancy of cross-validity, a Q² value greater than 0 for a specific endogenous construct is indicative of its predictive relevance [101]. We calculated the Q² values of the endogenous constructs BF (0.390), BPI (0.160), OBR (0.132), and PBR (0.141), which indicated that the model possessed sufficient predictive capabilities, confirming its robustness in forecasting the relationships between the latent variables. The R² and Q² values are listed in

Table 4. R-squared (R²) and Q-squared (Q²) values.

Construct	R2	Q2
BF	0.680	0.390
BPI	0.355	0.160
OBR	0.303	0.132
PBR	0.325	0.141

.

In addition, the model fit was assessed using several critical indices, including the standardized root mean square residual (SRMR). The SRMR value for the structural model was found to be 0.055, which fell between the thresholds for a moderate and high fit. Generally, an SRMR value below 0.08 is considered to indicate a good fit [100], although values up to 0.10 may still be acceptable for exploratory models [102]. Given the above, the model demonstrated a good fit.

Following Ringle et al. [97], we applied a bootstrapping technique with a recommended sample size of 5000 in order to test our hypotheses. Testing of the first hypothesis indicated that PC (β = 0.135) positively impacts BF. This finding is reasonable given that the perceived coolness of virtual influencers increases their impact on how the brand that they endorse is perceived [55]. It emphasizes that consumers show increased loyalty to brands promoted by influencers who are perceived as “cool” within the metaverse. This highlights the influential role of perceived coolness in terms of consumer behavior and brand preferences in the metaverse.

The testing of H2 revealed that AN (β = 0.146) directly enhances BF, aligning with Li et al. [80], who found that the anthropomorphic features of virtual influencers and followers’ affection for these influencers are intrinsically related. These findings confirm that consumers are more likely to show greater fidelity to brands endorsed by metaverse influencers who possess more relatable, human-like qualities. Additionally, consumers align their digital aspirations and identities within the metaverse as a result of the consumer–virtual influencer interaction.

The results indicate that the MIME dimensions of IN (β = 0.153), EN (β = 0.159), TR (β = 0.140), CU (β = 0.129), and CQ (β = 0.120) increase consumers’ BF (H3), suggesting that there is a positive relationship between metaverse influencers’ marketing efforts and brand fidelity. When virtual influencers interact more with consumers, provide entertaining content, remain trendy, offer customization options, and maintain high-quality communication, consumers tend to feel more connected to the brand, thus enhancing BF. The more influencers engage with consumers and deliver appealing content, the stronger the relationship between the brand and its consumers becomes, which results in increased brand fidelity [46].

The testing of H4, H5, and H6 showed that BF has a direct impact on BE behaviors (BPI: β = 0.596, PBR: β = 0.570, and OBR: β = 0.550), aligning with Afridawi and Rasool [87], who stated that the strong emotional bond created by brand fidelity motivates consumers to actively advocate for the brand. BF plays a crucial role in captivating consumers and transforming them from passive purchasers to proactive brand evangelists who willingly spread positive brand referrals while expressing negative attitudes towards competing brands. The direct effects are presented in

Table 5. Direct effects.

Hypothesis	Direct Relationships	β	Std. Error
H1	PC ➔ BF	0.135 *	0.064
H2	AN ➔ BF	0.146 *	0.072
H3a	IN ➔ BF	0.153 *	0.068
H3b	EN ➔ BF	0.159 **	0.055
H3c	TR ➔ BF	0.140 *	0.068
H3d	CU ➔ BF	0.129 *	0.061
H3e	CQ ➔ BF	0.120 **	0.045
H4	BF ➔ BPI	0.596 ***	0.048
H5	BF ➔ PBR	0.570 ***	0.052
H6	BF ➔ OBR	0.550 ***	0.051

Note: * indicates significant paths: * p < 0.05, ** p < 0.01, *** p < 0.001; NS = not significant.

, which includes the hypothesized relationships, corresponding beta (β) coefficients, and standard errors.

Table 6. Indirect effects.

Hypothesis	Indirect Relationships	β	Std. Error
H7a1	PC ➔ BF ➔ BPI	0.080 *	0.038
H7a2	PC ➔ BF ➔ OBR	0.074 *	0.035
H7a3	PC ➔ BF ➔ PBR	0.077 *	0.037
H8a1	AN ➔ BF ➔ BPI	0.087 *	0.044
H8a2	AN ➔ BF ➔ OBR	0.080 *	0.041
H8a3	AN ➔ BF ➔ PBR	0.083 *	0.042
H9a1	IN ➔ BF ➔ BPI	0.091 *	0.042
H9a2	IN ➔ BF ➔ OBR	0.084 *	0.039
H9a3	IN ➔ BF ➔ PBR	0.087 *	0.041
H9b1	EN ➔ BF ➔ BPI	0.095 **	0.034
H9b2	EN ➔ BF ➔ OBR	0.087 **	0.032
H9b3	EN ➔ BF ➔ PBR	0.091 **	0.033
H9c1	TR ➔ BF ➔ BPI	0.083 *	0.040
H9c2	TR ➔ BF ➔ OBR	0.077 *	0.037
H9c3	TR ➔ BF ➔ PBR	0.080 *	0.039
H9d1	CU ➔ BF ➔ BPI	0.077 *	0.037
H9d2	CU ➔ BF ➔ OBR	0.071 *	0.034
H9d3	CU ➔ BF ➔ PBR	0.074 *	0.036
H9e1	CQ ➔ BF ➔ BPI	0.071 *	0.028
H9e2	CQ ➔ BF ➔ OBR	0.066 *	0.026
H9e3	CQ ➔ BF ➔ PBR	0.068 *	0.027
H7a1	PC ➔ BF ➔ BPI	0.080 *	0.038

Note: * indicates significant paths: * p < 0.05, ** p < 0.01, *** p < 0.001; NS = not significant.

presents the results regarding the mediation effects; it also includes the hypothesized relationships, corresponding beta (β) coefficients, and standard errors. Our findings indicate that BF mediates the relationship between PC (β = 0.080, 0.074, and 0.077, i.e., BPI, OBR, and PBR) and the BPI, OBR, and PBR dimensions of BE (H7). These results are reasonable because virtual influencers who are perceived as “cool” strengthen customers’ fidelity to the brands that they promote [77] and cause them to engage in evangelistic behaviors [87]. This implies that brands should strategically design their virtual influencers to possess features that resonate with their target audience’s perceptions of coolness, as these features can significantly enhance brand fidelity and inspire greater consumer advocacy.

Additionally, our findings showed that BF mediates the relationship between AN (β = 0.087, 0.080, and 0.083, i.e., BPI, OBR, and PBR) and the BPI, OBR, and PBR dimensions of BE (H8). This highlights the importance of designing virtual influencers with relatable, human-like features so as to effectively build strong emotional connections with consumers, contributing to the development of brand fidelity and encouraging them to engage in brand evangelism behaviors. Finally, our testing revealed that BF also serves as a mediator between the MIME dimensions of IN (β = 0.091, 0.084, and 0.087, i.e., BPI, OBR, and PBR), EN (β = 0.095, 0.087, and 0.091), TR (β = 0.083, 0.077, and 0.080), CU (β = 0.077, 0.071, and 0.074), and CQ (β = 0.071, 0.066, and 0.068) and the dimensions of BE (BPI, OBR, and PBR) (H9). This consistent mediating effect across multiple variables indicates the robustness of brand fidelity as a mediator in influencing consumers’ brand evangelistic behavior in the metaverse. Marketers can leverage this by focusing on long-term engagement strategies that deepen their emotional bonds with consumers, ensuring that virtual influencers are not merely promotional tools but integral figures in the brand’s narrative within the metaverse.

In this study, importance–performance mapping analysis (IPMA) was also used to extend the PLS-SEM model by considering both the importance and performance of each construct [103]. IPMA provides insights into the variables that most influence the dependent variables by evaluating the importance and performance of each construct in the research model [104].

The IPMA results, presented in

Table 7. Importance–performance mapping analysis.

Construct Relationships	Importance	Performance
BF ➔ BPI	0.494	66.060
BF ➔ OBR	0.441	66.060
BF ➔ PBR	0.457	66.060
PC ➔ BF	0.151	67.468
AN ➔ BF	0.140	66.216
IN ➔ BF	0.119	67.298
EN ➔ BF	0.088	67.451
TR ➔ BF	0.095	66.318
CU ➔ BF	0.003	66.995
CQ ➔ BF	0.297	67.677

, revealed that BF is of primary importance in establishing BE, with largely equivalent importance for BPI, OBR, and PBR (0.494, 0.441, 0.457), all showing values of 66.06%. This suggests that improving this factor will significantly impact the development of brand evangelism behaviors. In contrast, PC, AN, and IN in BF had lower importance scores (0.151, 0.140, and 0.119), although their performance was slightly better (67.47%, 66.22%, and 67.298%). These constructs are less influential but still contribute to the model. Lastly, CQ in BF stands out, with moderate importance (0.297) and performance (67.677%), emphasizing the importance of the communication quality between virtual influencers and metaverse users. On the other hand, EN, TR, and CU in BF show lower importance and a minimal influence on BF performance, suggesting that these factors play a minor role in brand evangelism behaviors. Therefore, the IPMA results emphasize the need to prioritize BF and CQ while refining less influential constructs for stronger model performance.

6. Implications and Future Research

6.1. Theoretical Implications

The findings of this study have various academic implications in the branding domain. Firstly, this work contributes to the literature on the development of brand evangelism by exploring how metaverse virtual influencers’ characteristics and marketing efforts can affect consumers’ brand evangelism. Previous research investigating this phenomenon has primarily focused on virtual shopping and branding in the metaverse. For instance, Rane et al. [105] researched strategies adopted in the metaverse to strengthen customer loyalty, whereas Wongkitrungrueng and Suprawan [106] explored how metaverse experiential value affects consumer brand behavior and brand perception in both the virtual and real worlds. Scholars have also examined virtual influencers’ influences on consumer behavior, as well as the potential challenges and opportunities [20]. The present research is innovative in its identification of the factors (perceived coolness, anthropomorphism, and metaverse influencers’ marketing efforts) that influence consumers’ purchase intentions and brand-related behaviors. It enhances the theoretical knowledge by providing a novel perspective on consumer–brand interactions in the metaverse, explaining how virtual metaverse influencers can create value for brands and demonstrating the impact of these factors on brand fidelity and evangelism.

Secondly, this study contributes to the literature on virtual metaverse influencers by showing that influencers’ marketing efforts directly increase brand fidelity. The link between the marketing efforts of influencers, agents, or e-services and consumer brand behavior has been explored in the literature. However, scholars have focused on customers’ satisfaction with brands or service benefits. Chung et al. [107] explain how chatbot services create engaging and interactive encounters between brands and customers. Morra et al. [108] show how social media marketing can increase customers’ responses and enhance brand equity. Our research is unique in its exploration of the impacts of five dimensions of metaverse influencers’ marketing efforts (interaction, entertainment, trendiness, customization, and communication quality) on brand fidelity. In this study, we incorporated the SOR theory to explore how virtual influencers in the metaverse affect brand evangelism, and we identified brand fidelity as an essential link between marketing efforts and brand evangelism. A theoretical model of influencer factors has been constructed in which metaverse influencers’ marketing efforts are critical in achieving consumers’ brand fidelity, which is the primary factor that fosters brand evangelistic behavior.

Thirdly, the results expand social identity theory, as they demonstrate that virtual influencers’ perceived anthropomorphism positively influences brand fidelity. These findings align with previous studies [15,78,109] that have reported that virtual influencers with human-like qualities positively influence consumer satisfaction, particularly through anthropomorphism. Furthermore, this research highlights the positive relationship between brand evangelism behaviors and anthropomorphism in the metaverse.

Fourthly, our research addresses several academic calls to (1) evaluate novel variables that have the most significant impacts on customer purchase intention [110] and (2) expand the literature on the development of brand evangelism as consumer behavior in the metaverse, as marketing researchers demonstrate increasing interest in brand evangelism. Our research demonstrates that brand fidelity is a crucial brand evangelism predictor that “contributes to higher sales and market share” [111] (p. 4). This result indicates that brand fidelity is a significant construct because customers who exhibit fidelity to a particular brand are more inclined to become brand evangelists. They tend to become committed, active customers who recommend the brand to other people while disparaging the competition.

6.2. Managerial Implications

This study has various implications for brand management. Firstly, our findings emphasize the importance of cultivating brand fidelity through strategic partnerships. They highlight the mediating role of brand fidelity, through which virtual influencers’ features and efforts influence consumers’ brand evangelistic behaviors (BPI, OBR, and PBR) in the metaverse. By creating strong emotional bonds with consumers through consistent brand experiences and effective communication, brands can encourage customers to advocate for them and extend their reach through positive word-of-mouth referrals.

Secondly, this study highlights the importance of applying virtual influencers’ marketing strategies in the metaverse. Companies in the luxury fashion sector, such as Louis Vuitton and Chanel, could explore the use of virtual influencers to create exclusive metaverse experiences and achieve brand activation. By focusing on the five marketing effort dimensions described above, virtual influencers can be used to create engaging brand experiences for consumers, strengthening the consumer–brand bond. Implementing marketing strategies that optimize virtual influencers’ interactions with consumers, deliver captivating content, and maintain high-quality communication channels can heighten brand fidelity and increase brand evangelism. For example, virtual influencers could be used to host virtual fashion shows and exclusive metaverse events or even create personalized shopping experiences. Brands such as Louis Vuitton could use avatars to showcase their new collections in virtual worlds such as Decentraland or The Sandbox, engaging users in ways that physical stores cannot.

Thirdly, a crucial insight derived from our study is that brand managers should prioritize collaborations with virtual influencers who exhibit anthropomorphic qualities, which are perceived as “cool” in the metaverse. Brands in the entertainment and gaming industries, such as Epic Games (creators of Fortnite) or Sony PlayStation, could greatly benefit from creating metaverse virtual influencers that connect with gamers’ aspirations in the digital world. These virtual influencers could be used to promote new games and in-game products or even build fan communities to engage with the brand across various virtual platforms. A “cool” virtual influencer could interact with gamers, provide personalized content, and host live events in virtual spaces, encouraging players to form deeper connections with the brand. Therefore, virtual influencers could resonate with consumers’ digital aspirations to encourage brand evangelistic behavior and stimulate positive purchasing experiences. Customers who form strong emotional connections with a brand become brand evangelists, spreading brand information to other people through conventional media operations. Mvondo et al. [111] (p. 23) highlight that “brand evangelism contributes to brand success; it cannot be bought but must be carefully cultivated”. In this context, consumers will engage in various supportive behaviors, such as repeat purchases, valuable feedback, and referrals, to attract more metaverse customers.

Our research underscores the transformative potential of virtual influencer marketing in the metaverse, offering useful insights for brand managers seeking to capitalize on this emerging trend. Future applications of this strategy could be envisaged in industries such as automotive (e.g., Tesla), sports (e.g., Adidas), and even travel (e.g., Airbnb), where metaverse virtual influencers could be utilized to build interactive, engaging campaigns that highlight sustainability initiatives, product innovations, or new experiences. By strategically partnering with virtual influencers, enhancing brand fidelity through interactive marketing strategies in the metaverse, and effectively navigating challenges, brands can cultivate enduring relationships with consumers and drive brand evangelism. As brands embrace the metaverse as a new marketing frontier, adapting to consumers’ preferences and behaviors is crucial for success in this rapidly evolving digital environment.

6.3. Limitations and Future Research

Our study has a few limitations. Firstly, we utilized a structured questionnaire to examine metaverse users through a non-probabilistic sample on Amazon Mechanical Turk, which was chosen for its efficiency and its access to a broad participant pool. To ensure the data’s quality, we implemented multiple attention checks and pre-screening filters. However, we acknowledge the potential demographic and cultural biases associated with MTurk. Future research could address these limitations through the use of alternative sampling methods or by cross-validating results with other data sources to gain more diverse and nuanced insights. Secondly, it is essential to note that our data collection was geographically restricted to participants from the US, which may limit the generalizability of the findings. Consequently, this research offers a limited perspective on the metaverse, lacking international and cross-cultural insights, which could enrich the literature and provide a more comprehensive understanding of this field. Future studies should consider broadening the scope of the research by conducting cross-country analysis to validate the proposed model across different cultural contexts. Thirdly, as this study employed a cross-sectional design, we were unable to capture temporal changes or evolving patterns in consumer behavior. Future research could adopt a longitudinal approach to track trends and examine causal relationships over time, offering deeper insights into the dynamics of consumer–brand interactions. Fourthly, control variables were not considered in this study; therefore, future studies could incorporate metaverse users’ characteristics, such as their age, gender, prior experience with the metaverse, or affinity for digital technologies, which may influence their perceptions and behaviors. Finally, this study investigated brand fidelity (BF) and brand evangelism (BE) as the outcomes of perceived coolness (PC), anthropomorphism (AN), and metaverse influencers’ marketing efforts (MIMEs). Further research could explore other variables, such as impulse buying, brand intimacy, and brand romance.