This study examines how key experiential elements, telepresence, social interaction, immersion, realism and augmented reality elicit the “wow” effect (a strong emotional response) and, in turn, shape willingness to purchase luxury goods.
A questionnaire survey generated 409 valid responses, which were analyzed using partial least squares structural equation modeling (PLS-SEM) and necessary condition analysis (NCA).
Grounded in appraisal theory and hedonic consumption perspectives, the results show that all five experiential stimuli (telepresence, immersion, social interaction, perceived augmentation and perceived realism) significantly enhance the “wow” effect, which subsequently increases willingness to purchase luxury goods. The effect is mediated by adventurous, rather than merely satisfying, shopping motives, underscoring the primacy of exploration. Furthermore, goal orientation exerts a negative moderating effect on this relationship, with the positive influence of wow effect on luxury purchase intentions being more pronounced among consumers with lower goal orientation.
This research extends affective consumption theory to virtual luxury contexts and offers practical guidance for designing immersive, socially interactive virtual environments that heighten consumer engagement.
Background
The rapid emergence of the metaverse has fundamentally transformed how consumers experience, evaluate and engage with brands. No longer limited to passive digital interfaces, contemporary virtual environments enable real-time avatar interaction, immersive exploration and symbolic self-expression through persistent and socially embedded digital worlds (Hadi et al., 2024; Hennig-Thurau et al., 2023). For luxury brands in particular, the metaverse represents a strategic frontier where experiential intensity, emotional resonance and symbolic value increasingly substitute for physical materiality (Bakeshloo et al., 2025; Gupta et al., 2023). As luxury consumption has long been grounded in emotion, imagination and identity construction, understanding how consumers feel within immersive environments is central to explaining digital luxury adoption and engagement (Belk, 2013, 2018; Rosendo-Rios and Shukla, 2023).
Despite growing academic interest in metaverse marketing, existing research has predominantly focused on technological affordances such as telepresence, immersion, realism and interactivity, as well as platform acceptance and behavioral intention outcomes (Pantano and Viassone, 2015; Zhang et al., 2025). Recent work further suggests that metaverse brand experiences are increasingly designed as holistic, immersive narratives that integrate social interaction, identity expression and experiential engagement (Yao and Ren, 2025). Emotional responses are often treated as generic mediators or secondary mechanisms, typically operationalized as enjoyment, pleasure or overall positive affect (Hollebeek et al., 2022; Lacher and Mizerski, 1994). While this stream of research has advanced understanding of whether immersive technologies influence consumer responses, it offers limited insight into how intense experiential emotions are generated and why certain digital encounters lead to extraordinary engagement while others do not. This limitation is particularly salient in luxury contexts, where consumption decisions are disproportionately shaped by emotional arousal, symbolic meaning and extraordinary experiences rather than functional utility alone (Gupta et al., 2023; Kauppinen-Räisänen et al., 2020).
One experiential emotion that has gained increasing attention in practitioner discourse, yet remains under theorized in academic research, is the so called wow effect. The wow effect is conceptualized in this study as a high-arousal experiential state arising from consumers' cognitive appraisals of immersive stimuli. Rather than being a direct emotional reaction to environmental features, the wow effect reflects a configuration of appraisals characterized by perceived novelty, heightened self-relevance and the need for cognitive accommodation. This distinguishes the wow effect from general positive affect, positioning it as an awe-like emotional response that involves both affective intensity and cognitive restructuring.
This gap is theoretically consequential. Psychological research indicates that high intensity emotions such as awe differ qualitatively from everyday positive emotions like pleasure or satisfaction, as they involve sudden cognitive accommodation, heightened arousal and motivational reorientation toward exploration and meaning making (Keltner and Haidt, 2003; Yaden et al., 2017). Awe related experiences are associated with representational renewal and schema disruption, which can fundamentally alter perception, attention and behavioral orientation (Hinsch et al., 2020). In immersive virtual contexts, where users interact through avatars, experiment with digital identities and navigate socially rich environments, such high intensity emotional responses may be especially salient (Ki et al., 2025; Shang et al., 2012). Yet current metaverse and digital luxury research have not adequately theorized how specific experiential stimuli escalate into wow experiences, nor how these experiences shape downstream consumption motives in a differentiated manner.
Despite increasing recognition of the wow effect in immersive environments, existing research has largely treated it as a descriptive outcome rather than a theoretically grounded mechanism. In particular, prior studies have not sufficiently explained how specific experiential features are cognitively processed to generate such high-intensity emotional responses. Addressing this limitation requires a shift from structural frameworks toward a micro-level account of emotional generation. Accordingly, the present study adopts an appraisal-based perspective to explicate the cognitive and affective processes through which immersive stimuli are transformed into the wow effect.
To address this gap, the present study develops and tests a theory-driven model explaining the antecedents and consequences of the wow effect in metaverse-based luxury consumption. This study adopts appraisal theory as the central explanatory framework to explain how immersive metaverse environments generate high-arousal emotional responses. Rather than treating emotional reactions as direct outcomes of technological stimuli, appraisal theory suggests that emotions arise from individuals' cognitive evaluations of experiences along dimensions such as novelty, self-relevance and cognitive congruity. Within this perspective, the “wow” effect is conceptualized as an appraisal-driven emotional state that emerges when immersive experiences exceed existing cognitive schemas and require accommodation.
To provide a deeper explanation of this process, the present study adopts an appraisal-based perspective to explicate how consumers cognitively evaluate immersive stimuli and how these evaluations culminate in the emergence of the wow effect. Against this backdrop, this study pursues three interrelated research aims. First, it aims to conceptualize and empirically examine the wow effect as a distinct, high intensity experiential emotion in metaverse-based luxury consumption, rather than subsuming it under general positive affect. Second, it aims to identify how key metaverse-specific experiential stimuli, namely telepresence, social interactivity, immersion, perceived realism and perceived augmentation, jointly give rise to the wow effect. Third, it aims to explain how and when the wow effect translates into luxury purchase intention by distinguishing between adventure-oriented and gratification-oriented shopping motivations and by examining the moderating role of achievement goal orientation. Together, these aims position the wow effect as a central mechanism linking immersive experiential design to motivationally differentiated luxury consumption outcomes in virtual environments.
By framing the wow effect as an appraisal-driven experiential narrative rather than a simple emotional reaction, this study responds to recent calls for stronger storytelling in theoretical development within interactive marketing research (Wang, 2025). Specifically, it offers a coherent explanation of how immersive metaverse environments are cognitively interpreted, emotionally experienced and behaviorally enacted, thereby advancing understanding of consumer experience formation in digital luxury contexts.
Relevant literature and theoretical underpinning
The wow effect in virtual marketing and the metaverse
The wow effect has long attracted scholarly attention across sociology, philosophy and psychology as a distinctive form of intense experiential response (Pathak and Prakash, 2023). Early conceptualizations by Moore (1996) describe the wow effect as comprising three interrelated components: cognitive affective reactions such as surprise and pleasure, heightened physiological arousal triggered by stimulus exposure and downstream behavioral tendencies including curiosity, attention and exploratory engagement. This formulation positioned the wow effect not merely as momentary affect, but as an experiential configuration that links perception, emotion and action.
Subsequent research has extended the concept into digital and immersive environments. In augmented reality and virtual contexts, the wow effect is commonly used to describe users' awe like reactions to unexpected, vivid and technologically enhanced stimuli (Javornik, 2016). Drawing on emotion theory, scholars have associated the wow effect with awe, an emotion that arises from perceived vastness and the need for cognitive accommodation (Keltner and Haidt, 2003). Awe related experiences are known to trigger representational renewal, whereby existing mental schemas are disrupted and reconstructed, allowing individuals to perceive the world from a broader and more imaginative perspective (Hinsch et al., 2020; Yaden et al., 2017). In this sense, the wow effect extends beyond simple pleasure or enjoyment by involving heightened arousal, cognitive disruption and motivational redirection.
Importantly, prior research also suggests that the wow effect is not uniformly positive. While it often signals experiential success in immersive environments, excessive or poorly calibrated wow inducing stimuli can generate perceptual burden, cognitive overload and emotional volatility, ultimately undermining consumer responses (Arghashi, 2022). This dual nature highlights the wow effect as a powerful but complex experiential mechanism whose consequences depend on contextual and motivational factors. Despite its relevance, existing virtual marketing research has largely treated the wow effect descriptively or subsumed it under broad positive affect, leaving its theoretical boundaries and behavioral implications underdeveloped. Table 1 presents an overview of the key literature surrounding this concept.
Literature review for the concept of wow-effect/awe
| Source | Summarize elements | Theory/Procedure | Research theme |
|---|---|---|---|
| Arghashi (2022) | Novelty, Wow-effect, Inspiration, Information overload, Distraction, Hedonic shopping motivation, Purchase intention | Stimulus-organism-response (SOR) framework | The application of augmented reality technology in e-commerce retailing |
| Hinsch et al. (2020) | Psychological inspiration, Perceived hedonic benefits, Perceived augmentation quality, Perceived ease of use, AR expertise, Nostalgia, Wow-effect, Behavioral inspiration | Consumer inspiration | The use of augmented reality in marketing |
| Pathak and Prakash (2023) | Decision Comfort, Environmental Embedding, Immersion, Mental Tangibility, Psychological Ownership, Purchase Intention, Product Likeability, Specificity, Stimulated Physical Control, Wow feeling | Flow theory | The use of augmented reality in shopping scenarios |
| Rodrigues et al. (2022) | Awe, Behavioral intentions, Tourist satisfaction | concept of awe | The utilization of astronomical tourism destinations |
| Kautish et al. (2023) | eWOM, awe experience, purchase intentions, Cognitive MCI, Social MCI, Hedonic MCI, Functional MCI | Awe experience | The application of artificial intelligence voice assistants in shopping scenarios |
| Kim and Kim (2025) | Awe, Volunteer trip intention, Mindfulness, Conservation message, Travel motivation | Voluntourism and Awe | The use in volunteer tourism |
| Source | Summarize elements | Theory/Procedure | Research theme |
|---|---|---|---|
| Novelty, Wow-effect, Inspiration, Information overload, Distraction, Hedonic shopping motivation, Purchase intention | Stimulus-organism-response (SOR) framework | The application of augmented reality technology in e-commerce retailing | |
| Psychological inspiration, Perceived hedonic benefits, Perceived augmentation quality, Perceived ease of use, AR expertise, Nostalgia, Wow-effect, Behavioral inspiration | Consumer inspiration | The use of augmented reality in marketing | |
| Decision Comfort, Environmental Embedding, Immersion, Mental Tangibility, Psychological Ownership, Purchase Intention, Product Likeability, Specificity, Stimulated Physical Control, Wow feeling | Flow theory | The use of augmented reality in shopping scenarios | |
| Awe, Behavioral intentions, Tourist satisfaction | concept of awe | The utilization of astronomical tourism destinations | |
| eWOM, awe experience, purchase intentions, Cognitive MCI, Social MCI, Hedonic MCI, Functional MCI | Awe experience | The application of artificial intelligence voice assistants in shopping scenarios | |
| Awe, Volunteer trip intention, Mindfulness, Conservation message, Travel motivation | Voluntourism and Awe | The use in volunteer tourism |
Digital luxury consumption and the extended self in the metaverse
Within the metaverse, digital fashion, including clothing, accessories and branded items designed exclusively for avatars, functions as a symbolic medium for self-expression, identity construction and social signaling (Crepax and Liu, 2024; Peng et al., 2025). These virtual goods are not merely aesthetic enhancements but carry symbolic meanings that parallel, and in some cases extend beyond, those of physical luxury products.
Extending Belk's theory of the extended self (Belk, 1988, 2013), digital possessions can serve as identity markers through which individuals express values, status and aspirations. Historically, tangible goods such as clothing, automobiles or jewelry have functioned as extensions of the self (Belk, 2018). However, the transition from Web 2.0 to Web 3.0 has enabled the dematerialization of possessions and the re-embodiment of identity through avatars and virtual representations (Belk, 2013; Yang, 2024). In metaverse environments, users actively detach, reconfigure and reintegrate aspects of their identity through digital fashion choices, thereby imbuing virtual goods with personal and symbolic meaning (Ki et al., 2025).
Luxury digital fashion amplifies this process by embedding prestige, exclusivity and uniqueness into virtual goods that are visible within socially interactive environments. Beyond enhancing avatar appearance, luxury digital items facilitate social connectivity, recognition and community engagement, strengthening consumer brand relationships in the metaverse (Huang et al., 2025; Kumari et al., 2024). These characteristics make luxury digital consumption particularly conducive to emotionally intense experiences, positioning the metaverse as a context in which the wow effect is likely to emerge and exert meaningful behavioral influence.
Appraisal-based generation of the wow effect
Understanding how immersive metaverse stimuli translate into a high-intensity experiential state such as the wow effect requires a micro-level theory of emotional generation. To address this, the present study integrates appraisal theory as the core explanatory framework. Appraisal theory posits that emotions do not arise directly from external stimuli, but from individuals' cognitive evaluations of those stimuli along key dimensions such as novelty, self-relevance, goal congruence and coping potential. These appraisals determine both the intensity and the qualitative nature of emotional responses.
Within immersive digital environments, experiential features such as telepresence, immersion, social interactivity, realism and augmentation do not inherently produce emotional arousal. Rather, they shape how consumers interpret and evaluate their experiences. Specifically, telepresence and immersion enhance perceptions of self-relevance by creating a sense of psychological embeddedness within the virtual environment. Social interactivity introduces unpredictability and social significance, heightening evaluative attention. Perceived realism contributes to cognitive plausibility, enabling consumers to reconcile virtual experiences with existing mental schemas. In contrast, augmentation introduces elements of novelty and incongruity by extending beyond the constraints of physical reality.
The wow effect emerges when these appraisal dimensions converge in a manner that simultaneously signals high novelty and high self-relevance, while also exceeding existing cognitive expectations. This configuration triggers a process of cognitive accommodation, whereby individuals must adjust their mental schemas to make sense of the experience. Drawing on research on Awe, such moments of schema disruption are associated with heightened arousal, perceptual expansion and a reorientation toward exploration. In this sense, the wow effect can be understood as an appraisal-driven emotional state characterized by experiential escalation, rather than as a direct reaction to environmental stimuli.
By integrating appraisal theory, the present study moves beyond a purely structural explanation of stimulus–response relationships and provides a mechanistic account of how immersive experiences generate high-intensity emotional responses. This perspective allows us to theorize not only whether metaverse stimuli influence consumers, but precisely how and why they give rise to the wow effect.
To further clarify this mechanism, the generation of the wow effect can be understood as a sequential appraisal process. First, immersive stimuli introduce variations in novelty, complexity and experiential intensity. Consumers initially evaluate these stimuli in terms of whether the experience deviates from expectations and whether it holds personal relevance. When stimuli are perceived as both novel and self-relevant, they trigger heightened attentional engagement. Second, consumers assess whether the experience can be readily understood within existing cognitive schemas. When the experience exceeds existing mental frameworks, a state of cognitive incongruity emerges. Third, this incongruity initiates a process of cognitive accommodation, whereby individuals adjust their mental representations to reconcile the experience. It is this moment of accommodation, characterized by simultaneous arousal and perceptual expansion, that gives rise to the wow effect. In this sense, the wow effect is not a direct reaction to immersive features, but the outcome of an appraisal sequence that culminates in schema disruption and cognitive restructuring.
Hedonic consumption theory and experiential escalation
Hedonic consumption theory offers this complementary perspective by conceptualizing consumption as a multisensory, imaginative and emotionally driven process (Hirschman and Holbrook, 1982). Rather than viewing consumers as primarily rational decision makers, HCT emphasizes experiential engagement, fantasy and emotional arousal as central drivers of consumption value.
In immersive digital environments, hedonic processes are intensified through multisensory stimulation, narrative immersion and identity experimentation. Virtual fashion platforms provide visually rich and spatially immersive experiences that heighten telepresence and perceptual vividness (Hollebeek et al., 2022). Avatars enable fantasy enactment and identity play, allowing consumers to explore idealized or aspirational selves in ways that are difficult or impossible in physical settings (Hirschman, 1983; Wu and Holsapple, 2014). These processes are closely associated with emotional arousal, amazement and engagement, which have been observed in a range of immersive and interactive consumption experiences (Hollebeek et al., 2022; Lacher and Mizerski, 1994).
From an HCT perspective, the wow effect represents an escalation of hedonic engagement rather than a simple affective reaction. It arises when experiential stimuli not only attract attention but also amplify fantasy, symbolic meaning and intrinsic motivation. As such, HCT provides the theoretical grounding needed to explain why certain metaverse stimuli generate awe like, high arousal emotional states that motivate exploration and symbolic consumption.
Appraisal theory as the core mechanism of emotional generation
The present study positions appraisal theory as the central explanatory mechanism underpinning the wow effect. Appraisal theory posits that emotional responses arise from individuals' cognitive evaluations of environmental stimuli rather than from the stimuli themselves. Within immersive metaverse environments, features such as telepresence, immersion, social interactivity, realism and augmentation shape how consumers interpret their experiences along dimensions of novelty, self-relevance and cognitive congruity.
Other perspectives, including hedonic consumption and extended self, are incorporated as complementary lenses that help explain the experiential and symbolic consequences of these appraisal processes. Hedonic consumption highlights how immersive environments amplify experiential engagement, while extended self-perspectives explain how such experiences become integrated into identity expression. However, these perspectives do not independently explain emotional generation; instead, they operate as interpretive extensions of the appraisal process.
Hypotheses development
Consistent with appraisal theory, the influence of metaverse-specific experiential stimuli on the wow effect is understood to operate through consumers' cognitive evaluations of these stimuli. While the appraisal dimensions are not explicitly modeled as separate constructs in the empirical analysis, they provide the underlying theoretical mechanism through which immersive features are expected to generate high-intensity emotional responses.
Metaverse-specific experiential stimuli and the wow effect
Drawing on appraisal theory, the present study conceptualizes the relationship between metaverse-specific experiential stimuli and the wow effect as fundamentally grounded in consumers' cognitive evaluations of their experiences. Appraisal theory posits that emotional responses arise not directly from environmental stimuli, but from individuals' assessments of those stimuli along dimensions such as novelty, self-relevance and cognitive congruity. In immersive digital environments, features such as telepresence, immersion, social interactivity, realism and augmentation shape these appraisal processes by influencing how consumers interpret, engage with and make sense of virtual experiences. The wow effect is therefore theorized to emerge when these stimuli induce appraisal configurations characterized by heightened novelty, strong self-relevance and the need for cognitive accommodation. While these appraisal processes are not explicitly modeled as separate constructs in the empirical analysis, they provide the underlying theoretical mechanism through which metaverse stimuli are expected to generate high-intensity emotional responses. Telepresence refers to the subjective sensation of “being there” in a mediated environment (Steuer, 1992). In metaverse contexts, heightened telepresence allows users to experience virtual spaces as psychologically real, thereby amplifying emotional engagement and susceptibility to awe-like responses (Cummings and Bailenson, 2016; Kim et al., 2023).
Immersion refers to a state of deep mental involvement in which users' attention shifts away from the physical world and toward the virtual environment (Agrawal et al., 2020). High levels of immersion reduce awareness of the physical world and facilitate experiential flow, conditions that are conducive to emotional escalation rather than reflective appraisal (Hoffman and Novak, 2009; Hollebeek et al., 2022). In metaverse environments, immersion is further intensified through persistent virtual worlds and continuous avatar embodiment, making wow experiences more likely to emerge (Dwivedi et al., 2022; Ki et al., 2025).
Social interactivity captures the extent to which users can interact with others in real time within a virtual environment (Yadav and Varadarajan, 2005). Unlike earlier online shopping environments, the metaverse enables synchronous social presence through avatars, allowing consumption experiences to unfold in socially visible and performative ways (Shen et al., 2021; Zhang et al., 2025).
Perceived realism refers to the extent to which virtual environments are experienced as authentic and lifelike (Lee, 2004). Realistic visual, spatial and behavioral cues enhance cognitive accommodation by blurring the boundary between virtual and physical experience (Javornik, 2016; Keltner and Haidt, 2003).
Augmentation reflects the enhancement of reality through virtual overlays, transformations or extensions that go beyond physical constraints (Azuma, 1997). In metaverse environments, augmentation enables experiences that are not merely realistic but extraordinary, amplifying novelty, surprise and perceived vastness (Arghashi, 2022; Javornik, 2016). Such experiential amplification closely aligns with the defining characteristics of the wow effect as an awe-like emotional response triggered by unexpected and impactful stimuli (Hinsch et al., 2020).
Taken together, these metaverse-specific experiential features are expected to influence the wow effect through their impact on consumers' cognitive appraisals. Telepresence and immersion enhance perceptions of self-relevance by embedding the individual within the virtual environment, thereby increasing the personal significance of the experience. Social interactivity introduces unpredictability and social meaning, amplifying evaluative attention and emotional intensity. Perceived realism facilitates cognitive plausibility, enabling individuals to reconcile virtual experiences with existing mental frameworks, while augmentation introduces novelty and schema incongruity by extending beyond physical constraints. When these appraisal dimensions converge, they are more likely to trigger cognitive accommodation, resulting in a high-arousal experiential state characterized as the wow effect.
(a) Telepresence, (b) immersion, (c) social interactivity, (d) augmentation each have a positive effect on the wow effect and (e) perceived realism.
The mediating role of shopping motivation
From an appraisal perspective, the wow effect represents an emotional state arising from cognitive evaluations characterized by novelty and schema disruption. Such appraisal configurations are associated with cognitive accommodation processes that expand attention and promote openness to new experiences. As a result, the wow effect is expected to influence not only emotional intensity but also the direction of consumer motivation. Specifically, appraisal-driven high-arousal states are more likely to activate exploratory and novelty-seeking tendencies rather than comfort-oriented or satiation-driven motivations.
Adventure shopping is characterized by exploration, novelty seeking, curiosity and experiential discovery (Arnold and Reynolds, 2003). Psychological research on awe and surprise suggests that high-intensity emotions promote openness, curiosity and exploratory behavior rather than emotional satiation or comfort seeking (Keltner and Haidt, 2003; Yaden et al., 2017). In immersive metaverse environments, wow experiences may therefore redirect consumers toward adventure-oriented engagement, motivating them to explore digital fashion offerings as part of an ongoing experiential journey rather than as isolated purchase acts (Milanesi et al., 2023; Xi and Hamari, 2021). This aligns with appraisal-based accounts of awe-like emotions, which suggest that novelty-driven cognitive accommodation redirects individuals toward exploration and experiential engagement.
By contrast, gratification shopping emphasizes immediate emotional comfort, indulgence and sensory pleasure (Arnold and Reynolds, 2003). While such motivation is central to many forms of physical luxury consumption, its relevance may be attenuated in digital contexts where tactile and embodied sensory gratification is limited (Kauppinen-Räisänen et al., 2020; Rosendo-Rios and Shukla, 2023). As a result, wow experiences in the metaverse may be less effective in activating gratification-oriented motivation. From an appraisal perspective, gratification-oriented motivation is less likely to be activated, as the wow effect is driven by novelty and cognitive expansion rather than emotional closure or comfort.
Accordingly, we propose that the wow effect translates into luxury purchase intention primarily through adventure shopping rather than gratification shopping. As such, the following are presented:
The wow effect positively influences luxury purchase intention.
Adventure shopping motivation mediates the relationship between the wow effect and luxury purchase intention.
Gratification shopping motivation does not mediate the relationship between the wow effect and luxury purchase intention.
The moderating role of achievement goal orientation
Although wow experiences may activate adventure-oriented motivation, not all consumers respond to emotional intensity in the same way. Achievement goal orientation reflects individual differences in the extent to which consumers prioritize performance, outcomes and instrumental success over experiential enjoyment (Dweck and Leggett, 1988; Elliot and McGregor, 2001). From an appraisal perspective, highly goal-oriented individuals are more likely to engage in controlled, outcome-focused evaluations, thereby attenuating the influence of novelty-driven emotional responses such as the wow effect.
Consumers with strong achievement-oriented goals tend to evaluate consumption opportunities through a utilitarian or outcome-focused lens and may therefore be less responsive to emotionally driven motivations (Katyal et al., 2022; Li and Kang, 2025). In such cases, even when wow experiences stimulate exploration, their translation into purchase intention may be attenuated. Conversely, consumers with lower achievement goal orientation may be more receptive to experiential and exploratory cues and are therefore more likely to translate the wow effect into a stronger intention to purchase luxury goods. As such, the following is presented:
Achievement goal orientation moderates the relationship between wow effect and luxury purchase intention, such that the relationship is weaker when achievement goal orientation is high.
The research model is shown in Figure 1.
A diagram of a research model illustrating the relationships between different factors and luxury purchase intention. The model includes several key components: Telepresence, Sense of immersion, Social interactivity, Perceived augmentation, and Perceived realism, all of which influence the Wow effect. The Wow effect is further connected to Adventure Shopping and Gratification Shopping, which in turn affect Luxury purchase intention. Additionally, Goal orientation directly influences Luxury purchase intention. Arrows indicate the directional relationships between these components, showing how they interact to impact luxury purchase intention.Research model. Source: Figure by authors
A diagram of a research model illustrating the relationships between different factors and luxury purchase intention. The model includes several key components: Telepresence, Sense of immersion, Social interactivity, Perceived augmentation, and Perceived realism, all of which influence the Wow effect. The Wow effect is further connected to Adventure Shopping and Gratification Shopping, which in turn affect Luxury purchase intention. Additionally, Goal orientation directly influences Luxury purchase intention. Arrows indicate the directional relationships between these components, showing how they interact to impact luxury purchase intention.Research model. Source: Figure by authors
Research methodology
Data collection and sample
An online survey was used in this study to look into how consumers behave in the metaverse. This study employed purposive sampling (Ooi et al., 2025), selecting target respondents who met specific criteria during the research process. The questionnaire was created electronically and distributed via Soudiao.com, China's leading online survey platform. By 2026, Sojump had reached a total of 6.2 million registered users (Sojump, 2026). To ensure the reliability and consistency of the research, we included a series of screening questions at the survey's outset to confirm participants met the study's prerequisites (Wu et al., 2025). These prerequisites required participants to satisfy the following conditions: (1) be at least 18 years of age and (2) have used the metaverse within the past six months. Data were collected in April 2025. We distributed 450 survey invitations by using the metaverse player group on Xiaohongshu. A quick explanation of the metaverse and examples of virtual activities were given at the outset to maintain uniformity, and qualified individuals were filtered using screening questions. A total of 409 valid responses were collected. Data cleaning involved checking for straight-lining patterns and inconsistencies; no major quality issues were found. A post hoc power analysis using the inverse square root and gamma-exponential methods (α = 0.05, power = 0.80) confirmed that the sample size exceeded the minimum requirement of 158 respondents (Kock and Hadaya, 2018). Of those surveyed, 49.9% were women and 50.1% were men. Most participants were aged between 26 and 40 years old (57.9%), held a bachelor's degree (37.4%), Regarding income, 30.3% earned between RMB 5,001 and RMB 8,000. Overall, the sample represents young, educated and digitally active consumers, aligning with typical metaverse user profiles (Ki et al., 2025), as shown in Table 2.
Demographic profiles
| Number | Percentage (%) | ||
|---|---|---|---|
| Gender | Female | 204 | 49.9 |
| Male | 205 | 50.1 | |
| Age | 18–25 years | 59 | 14.4 |
| 26–30 years | 124 | 30.3 | |
| 31–40 years | 113 | 27.6 | |
| 41–50 years | 78 | 19.1 | |
| 51–60 years | 26 | 6.4 | |
| 61 years and above | 9 | 2.2 | |
| Education | Bachelor degree/professional qualification | 153 | 37.4 |
| Diploma/advance diploma | 142 | 34.7 | |
| No College Degree | 82 | 20 | |
| Postgraduate (e.g. Master, Doctorate) | 32 | 7.8 | |
| Income level (per month) | RMB2,001 to RMB5,000 | 45 | 11 |
| Less than RMB2,000 | 18 | 4.4 | |
| RMB15,001 to RMB30,000 | 83 | 20.3 | |
| RMB30,001 and above | 47 | 11.5 | |
| RMB5,001 to RMB8,000 | 124 | 30.3 | |
| RMB8,001 to RMB15,000 | 92 | 22.5 | |
| Number of times you have used the metaverse | 3 to 5 times | 127 | 31.1 |
| Less than 3 times | 107 | 26.2 | |
| More than 5 times | 175 | 42.8 | |
| Number of times you have used ZEPETO | 3 to 5 times | 127 | 31.1 |
| Less than 3 times | 107 | 26.2 | |
| More than 5 times | 175 | 42.8 |
| Number | Percentage (%) | ||
|---|---|---|---|
| Gender | Female | 204 | 49.9 |
| Male | 205 | 50.1 | |
| Age | 18–25 years | 59 | 14.4 |
| 26–30 years | 124 | 30.3 | |
| 31–40 years | 113 | 27.6 | |
| 41–50 years | 78 | 19.1 | |
| 51–60 years | 26 | 6.4 | |
| 61 years and above | 9 | 2.2 | |
| Education | Bachelor degree/professional qualification | 153 | 37.4 |
| Diploma/advance diploma | 142 | 34.7 | |
| No College Degree | 82 | 20 | |
| Postgraduate (e.g. Master, Doctorate) | 32 | 7.8 | |
| Income level (per month) | RMB2,001 to RMB5,000 | 45 | 11 |
| Less than RMB2,000 | 18 | 4.4 | |
| RMB15,001 to RMB30,000 | 83 | 20.3 | |
| RMB30,001 and above | 47 | 11.5 | |
| RMB5,001 to RMB8,000 | 124 | 30.3 | |
| RMB8,001 to RMB15,000 | 92 | 22.5 | |
| Number of times you have used the metaverse | 3 to 5 times | 127 | 31.1 |
| Less than 3 times | 107 | 26.2 | |
| More than 5 times | 175 | 42.8 | |
| Number of times you have used ZEPETO | 3 to 5 times | 127 | 31.1 |
| Less than 3 times | 107 | 26.2 | |
| More than 5 times | 175 | 42.8 |
Measures
The survey comprised two sections: demographic information and construct measurements. Items from well-known research that were suitably adapted for the setting of luxury products in the metaverse were used to measure the constructs, as shown in Table 3. The “Wow effect” measurement is based on Arghashi (2022) and highlights high-arousal emotional indicators, such as surprise or excitement as described in the questionnaire. This approach also helps distinguish differences between emotional outcomes and their cognitive antecedents. By using these emotional labels, we can gauge the intensity of users' emotions.
Measurement items for the constructs
| Construct/Source/Measurement itme |
|---|
| Wow-effect (Arghashi, 2022) WE1: This ZEPETO app amazed me WE2: When I used this ZEPETO app, I often thought “wow” WE3: This ZEPETO app has thrilled me from the very beginning Tele-presence (Park et al., 2023) TP1: When I use ZEPETO app, I feel that the character or object is real TP2: When I play ZEPETO app, I feel like I can touch characters and objects TP3: When I play ZEPETO app, I respond with a smile or talk about a character or object I see in ZEPETO app TP4: When I play ZEPETO app, I feel like I am actually in a virtual world TP5: When I play ZEPETO app, I can easily understand the atmosphere around me Social interactivity (Park et al., 2023) SI1: We also exchange opinions on common interests with users through ZEPETO app SI2: I feel close to other users active in ZEPETO app SI3: Users active in ZEPETO app seem to help each other well SI4: ZEPETO app allows users to have effective meetings SI5: In ZEPETO app, users recognize and acknowledge my value Perceived augmentation (Daassi and Debbabi, 2021) PA1: The service can add products to my ZEPETO body virtually PA2: The way the products were virtually placed on my ZEPETO body seemed real PA3: The online products seemed to be part of my ZEPETO body PA4: The level of reality seemed high on the ZEPETO screen PA5: The ZEPETO products seemed to exist in real time Sense of immersion (Daassi and Debbabi, 2021) SOI1: The ZEPETO app created a new environment that suddenly disappeared at the end of the show SOI2: At ZEPETO, I was unaware of my surroundings SOI3: During the ZEPETO virtual experience, my body was in the room, but my mind was in the world created by the show SOI4: The ZEPETO app made me forget the reality of the outside world SOI5: During the ZEPETO virtual experience, I forgot about things that had happened before the show or that would occur after the show SOI6: The ZEPETO virtual experience made me forget my immediate surroundings Perceived realism (Daassi and Debbabi, 2021) PR1: In comparison with the real world, the ZEPETO augmented environment seemed real PR2: My experience in the ZEPETO augmented environment seems consistent with my real-world experience PR3: The things that happen in the ZEPETO augmented environment look like the things that happen in real life PR4: This ZEPETO augmented reality–based experience was similar to in-store shopping experience Luxury purchase intention (Ostovan and Nasr, 2022) LPI1: I am likely to purchase luxury brands in ZEPETO LPI2: I have a strong possibility to purchase luxury brands in ZEPETO LPI3: I Intend to buy luxury brands in ZEPETO again in the future Goal orientation (Li and Kang, 2025) GO1: No matter what I do, I have the highest standards for myself in ZEPETO GO2: I never settle for second best in ZEPETO GO3: Whenever I am faced with a choice, I try to imagine what all the other possibilities are, even ones that are not present at the moment in ZEPETO Adventure shopping (Coelho et al., 2023) AS1: I find shopping stimulating in ZEPETO AS2: To me, shopping is an adventure in ZEPETO AS3: Shopping is a thrill to me in ZEPETO AS4: Shopping makes me feel like I am in my own universe in ZEPETO Gratification shopping (Coelho et al., 2023) GS1: When I'm in a down mood, I go shopping to make me feel better in ZEPETO GS2: To me, shopping is a way to relieve stress in ZEPETO GS3: I go shopping when I want to treat myself to something special in ZEPETO |
| Construct/Source/Measurement itme |
|---|
| Wow-effect ( |
Data analysis
Study 1A
To examine the hypothesized model, SmartPLS 4.0 was employed. This updated version integrates functionalities from the widely used process macro (Hayes, 2015), enabling robust estimation of conditional indirect effects within the PLS-SEM environment. Consistent with the recommended two-stage analytical procedure (Hair et al., 2021), the analysis proceeded by first validating the measurement model, followed by structural model evaluation.
Common method bias
Common method bias (CMB) were assessed to ensure validity. Regarding procedural aspects, we conducted preliminary testing with experts in relevant fields to ensure questions were formulated clearly, unambiguously and directly. We assured respondents that their answers would remain confidential and anonymous. For statistical methodology, we employed Harman's one-factor test. Results indicated that the first factor accounted for 28.238% of total variance, falling short of the recommended 50% threshold). This study employed the method proposed by Liang et al. (2007) to examine potential CMB. The results showed that the ratio of the average substantive variance to the average method variance was 491.95:1. Therefore, it can be concluded that the threat of CMB to the study results is negligible. The heterotrait–monotrait (HTMT) ratio of correlations was used to evaluate discriminant validity (see Table 4) (Henseler et al., 2015).
Correlation matrix table
| WE | TP | SI | PA | SOI | PR | LPI | GO | AS | |
|---|---|---|---|---|---|---|---|---|---|
| TP | 0.404** | ||||||||
| SI | 0.453** | 0.322** | |||||||
| PA | 0.456** | 0.300** | 0.306** | ||||||
| SOI | 0.409** | 0.265** | 0.264** | 0.295** | |||||
| PR | 0.395** | 0.344** | 0.259** | 0.220** | 0.294** | ||||
| LPI | 0.433** | 0.379** | 0.285** | 0.264** | 0.336** | 0.308** | |||
| GO | 0.456** | 0.338** | 0.373** | 0.316** | 0.303** | 0.322** | 0.252** | ||
| AS | 0.430** | 0.366** | 0.293** | 0.294** | 0.292** | 0.401** | 0.338** | 0.338** | |
| GS | 0.478** | 0.266** | 0.337** | 0.272** | 0.252** | 0.285** | 0.313** | 0.261** | 0.315** |
| WE | TP | SI | PA | SOI | PR | LPI | GO | AS | |
|---|---|---|---|---|---|---|---|---|---|
| TP | 0.404** | ||||||||
| SI | 0.453** | 0.322** | |||||||
| PA | 0.456** | 0.300** | 0.306** | ||||||
| SOI | 0.409** | 0.265** | 0.264** | 0.295** | |||||
| PR | 0.395** | 0.344** | 0.259** | 0.220** | 0.294** | ||||
| LPI | 0.433** | 0.379** | 0.285** | 0.264** | 0.336** | 0.308** | |||
| GO | 0.456** | 0.338** | 0.373** | 0.316** | 0.303** | 0.322** | 0.252** | ||
| AS | 0.430** | 0.366** | 0.293** | 0.294** | 0.292** | 0.401** | 0.338** | 0.338** | |
| GS | 0.478** | 0.266** | 0.337** | 0.272** | 0.252** | 0.285** | 0.313** | 0.261** | 0.315** |
Note(s): ** Correlation is significant at the 0.01 level (2-tailed)
Measurement model
As shown in Table 5, composite reliability (CR) scores for every construct were higher than the suggested cutoff of 0.70, ranging from 0.900 to 0.927, and Cronbach's alpha values ranged from 0.853 to 0.899, confirming satisfactory internal consistency (Hair et al., 2021). As presented in Table 5, all item loadings exceeded 0.70, and the average variance extracted (AVE) values for all constructs were above the recommended minimum of 0.50, ranging from 0.600 to 0.810. These results suggest that a substantial proportion of variance in each construct is explained by its respective items (Hair et al., 2021). Table 6 demonstrates that every value, with the highest being 0.544, was below the cautious cutoff of 0.85. This indicates satisfactory discriminant validity among all constructs.
Reliability, validity and descriptive statistics of constructs
| Indicators | Loading | Cronbach's alpha | CR | AVE | Mean | Std. deviation | Skewness | Kurtosis |
|---|---|---|---|---|---|---|---|---|
| AS | 0.855 | 0.902 | 0.697 | 5.045 | 1.249 | −0.997 | 0.535 | |
| AS1 | 0.821 | |||||||
| AS2 | 0.861 | |||||||
| AS3 | 0.832 | |||||||
| AS4 | 0.824 | |||||||
| GO | 0.853 | 0.911 | 0.773 | 5.024 | 1.154 | −0.585 | 0.430 | |
| GO1 | 0.867 | |||||||
| GO2 | 0.891 | |||||||
| GO3 | 0.879 | |||||||
| GS | 0.883 | 0.927 | 0.810 | 5.352 | 1.155 | −0.963 | 1.309 | |
| GS1 | 0.918 | |||||||
| GS2 | 0.885 | |||||||
| GS3 | 0.896 | |||||||
| LPI | 0.856 | 0.912 | 0.776 | 4.934 | 1.388 | −0.982 | 0.388 | |
| LPI1 | 0.881 | |||||||
| LPI2 | 0.872 | |||||||
| LPI3 | 0.890 | |||||||
| PA | 0.879 | 0.912 | 0.674 | 5.222 | 1.054 | −0.611 | 0.345 | |
| PA1 | 0.799 | |||||||
| PA2 | 0.838 | |||||||
| PA3 | 0.834 | |||||||
| PA4 | 0.843 | |||||||
| PA5 | 0.790 | |||||||
| PR | 0.855 | 0.902 | 0.697 | 4.990 | 1.282 | −0.939 | 0.167 | |
| PR1 | 0.849 | |||||||
| PR2 | 0.835 | |||||||
| PR3 | 0.823 | |||||||
| PR4 | 0.833 | |||||||
| SI | 0.890 | 0.919 | 0.694 | 5.406 | 1.067 | −0.847 | 0.795 | |
| SI1 | 0.844 | |||||||
| SI2 | 0.809 | |||||||
| SI3 | 0.828 | |||||||
| SI4 | 0.831 | |||||||
| SI5 | 0.852 | |||||||
| SOI | 0.867 | 0.900 | 0.600 | 5.284 | 1.002 | −0.637 | 0.229 | |
| SOI1 | 0.756 | |||||||
| SOI2 | 0.773 | |||||||
| SOI3 | 0.796 | |||||||
| SOI4 | 0.771 | |||||||
| SOI5 | 0.789 | |||||||
| SOI6 | 0.762 | |||||||
| TP | 0.899 | 0.925 | 0.711 | 4.873 | 1.349 | −0.890 | −0.204 | |
| TP1 | 0.825 | |||||||
| TP2 | 0.838 | |||||||
| TP3 | 0.858 | |||||||
| TP4 | 0.848 | |||||||
| TP5 | 0.847 | |||||||
| WE | 0.876 | 0.923 | 0.801 | 5.474 | 1.112 | −0.982 | 1.716 | |
| WE1 | 0.900 | |||||||
| WE2 | 0.904 | |||||||
| WE3 | 0.880 |
| Indicators | Loading | Cronbach's alpha | CR | AVE | Mean | Std. deviation | Skewness | Kurtosis |
|---|---|---|---|---|---|---|---|---|
| AS | 0.855 | 0.902 | 0.697 | 5.045 | 1.249 | −0.997 | 0.535 | |
| AS1 | 0.821 | |||||||
| AS2 | 0.861 | |||||||
| AS3 | 0.832 | |||||||
| AS4 | 0.824 | |||||||
| GO | 0.853 | 0.911 | 0.773 | 5.024 | 1.154 | −0.585 | 0.430 | |
| GO1 | 0.867 | |||||||
| GO2 | 0.891 | |||||||
| GO3 | 0.879 | |||||||
| GS | 0.883 | 0.927 | 0.810 | 5.352 | 1.155 | −0.963 | 1.309 | |
| GS1 | 0.918 | |||||||
| GS2 | 0.885 | |||||||
| GS3 | 0.896 | |||||||
| LPI | 0.856 | 0.912 | 0.776 | 4.934 | 1.388 | −0.982 | 0.388 | |
| LPI1 | 0.881 | |||||||
| LPI2 | 0.872 | |||||||
| LPI3 | 0.890 | |||||||
| PA | 0.879 | 0.912 | 0.674 | 5.222 | 1.054 | −0.611 | 0.345 | |
| PA1 | 0.799 | |||||||
| PA2 | 0.838 | |||||||
| PA3 | 0.834 | |||||||
| PA4 | 0.843 | |||||||
| PA5 | 0.790 | |||||||
| PR | 0.855 | 0.902 | 0.697 | 4.990 | 1.282 | −0.939 | 0.167 | |
| PR1 | 0.849 | |||||||
| PR2 | 0.835 | |||||||
| PR3 | 0.823 | |||||||
| PR4 | 0.833 | |||||||
| SI | 0.890 | 0.919 | 0.694 | 5.406 | 1.067 | −0.847 | 0.795 | |
| SI1 | 0.844 | |||||||
| SI2 | 0.809 | |||||||
| SI3 | 0.828 | |||||||
| SI4 | 0.831 | |||||||
| SI5 | 0.852 | |||||||
| SOI | 0.867 | 0.900 | 0.600 | 5.284 | 1.002 | −0.637 | 0.229 | |
| SOI1 | 0.756 | |||||||
| SOI2 | 0.773 | |||||||
| SOI3 | 0.796 | |||||||
| SOI4 | 0.771 | |||||||
| SOI5 | 0.789 | |||||||
| SOI6 | 0.762 | |||||||
| TP | 0.899 | 0.925 | 0.711 | 4.873 | 1.349 | −0.890 | −0.204 | |
| TP1 | 0.825 | |||||||
| TP2 | 0.838 | |||||||
| TP3 | 0.858 | |||||||
| TP4 | 0.848 | |||||||
| TP5 | 0.847 | |||||||
| WE | 0.876 | 0.923 | 0.801 | 5.474 | 1.112 | −0.982 | 1.716 | |
| WE1 | 0.900 | |||||||
| WE2 | 0.904 | |||||||
| WE3 | 0.880 |
The HTMT criterion
| AS | GS | LPI | PA | PR | SI | SOI | TP | |
|---|---|---|---|---|---|---|---|---|
| AS | ||||||||
| GS | 0.363 | |||||||
| LPI | 0.396 | 0.360 | ||||||
| PA | 0.339 | 0.309 | 0.304 | |||||
| PR | 0.469 | 0.329 | 0.359 | 0.254 | ||||
| SI | 0.335 | 0.380 | 0.326 | 0.346 | 0.296 | |||
| SOI | 0.339 | 0.288 | 0.390 | 0.338 | 0.342 | 0.299 | ||
| TP | 0.417 | 0.300 | 0.432 | 0.338 | 0.392 | 0.359 | 0.300 | |
| WE | 0.497 | 0.544 | 0.500 | 0.520 | 0.456 | 0.513 | 0.469 | 0.456 |
| AS | GS | LPI | PA | PR | SI | SOI | TP | |
|---|---|---|---|---|---|---|---|---|
| AS | ||||||||
| GS | 0.363 | |||||||
| LPI | 0.396 | 0.360 | ||||||
| PA | 0.339 | 0.309 | 0.304 | |||||
| PR | 0.469 | 0.329 | 0.359 | 0.254 | ||||
| SI | 0.335 | 0.380 | 0.326 | 0.346 | 0.296 | |||
| SOI | 0.339 | 0.288 | 0.390 | 0.338 | 0.342 | 0.299 | ||
| TP | 0.417 | 0.300 | 0.432 | 0.338 | 0.392 | 0.359 | 0.300 | |
| WE | 0.497 | 0.544 | 0.500 | 0.520 | 0.456 | 0.513 | 0.469 | 0.456 |
Structural model
Before assessing the structural relationships, we examined potential multicollinearity issues among the predictor constructs. Collinearity is not an issue because all variance inflation factor values fell significantly below the cautious cutoff of 3.3, ranging from 1.759 to 2.726 (Hair et al., 2019).
A bootstrapping procedure with 5,000 resamples was performed in SmartPLS 4 (Hair and Alamer, 2022) to estimate the path coefficients, t-values and significance levels. The results are presented in Table 7. The analysis confirmed all five hypothesized relationships between metaverse experiential attributes and the wow effect. Specifically, telepresence (β = 0.144), sense of immersion (β = 0.186), social interactivity (β = 0.238), perceived augmentation (β = 0.247) and perceived realism (β = 0.174) each had a significant positive effect on the wow effect, thereby supporting H1a through H1e. The R2 value for the wow effect was 0.426, indicating that 42.6% of its variance can be explained by these five antecedents. In line with H2, the wow effect significantly influenced luxury purchase intention (β = 0.251), confirming its pivotal role as an emotional driver of digital luxury consumption. The R2 value for luxury purchase intention was 0.243, suggesting moderate explanatory power of the model.
Hypotheses testing
| Hypothesis | Relationship | Std. β | t-value | p-values | Decision |
|---|---|---|---|---|---|
| H1a | TP → WE | 0.144** | 3.219 | 0.001 | Supported |
| H1b | SOI → WE | 0.186*** | 4.400 | 0.000 | Supported |
| H1c | SI → WE | 0.238*** | 5.523 | 0.000 | Supported |
| H1d | PA → WE | 0.247*** | 5.675 | 0.000 | Supported |
| H1e | PR → WE | 0.174*** | 3.896 | 0.000 | Supported |
| H2 | WE → LPI | 0.251*** | 4.051 | 0.000 | Supported |
| H3 | WE → AS → LPI | 0.065* | 2.162 | 0.015 | Supported |
| H4 | WE → GS → LPI | 0.043 | 1.372 | 0.085 | Unsupported |
| H5 | GO × WE → LPI | −0.097** | 2.514 | 0.006 | Supported |
| Hypothesis | Relationship | Std. β | t-value | p-values | Decision |
|---|---|---|---|---|---|
| TP → WE | 0.144** | 3.219 | 0.001 | Supported | |
| SOI → WE | 0.186*** | 4.400 | 0.000 | Supported | |
| SI → WE | 0.238*** | 5.523 | 0.000 | Supported | |
| PA → WE | 0.247*** | 5.675 | 0.000 | Supported | |
| PR → WE | 0.174*** | 3.896 | 0.000 | Supported | |
| WE → LPI | 0.251*** | 4.051 | 0.000 | Supported | |
| WE → AS → LPI | 0.065* | 2.162 | 0.015 | Supported | |
| WE → GS → LPI | 0.043 | 1.372 | 0.085 | Unsupported | |
| GO × WE → LPI | −0.097** | 2.514 | 0.006 | Supported |
As for the mediating effects, adventure shopping significantly mediated the relationship between the wow effect and luxury purchase intention (β = 0.065, 95% bias-corrected CI [0.016, 0.115]), supporting H3. In contrast, the indirect effect of the wow effect on purchase intention through gratification shopping was not statistically significant (β = 0.043, 95% bias-corrected CI [−0.006, 0.095]), a finding consistent with H4. Finally, the moderating analysis revealed that goal orientation significantly weakened the positive association between the wow effect and luxury purchase intention (β = −0.097), supporting H5. The interaction plot (see Figure 2) illustrates this moderation effect, indicating that the positive impact of the wow effect on purchase intention was stronger among individuals with lower goal orientation. The results indicate that consumers with lower goal orientation are more sensitive to the emotional stimulation triggered by the wow-effect, while those with higher goal orientation exhibit more restrained responses to such experiential cues.
A line graph titled 'GO x WE' presents the interaction between goal orientation (GO) and the wow effect (WE) on luxury purchase intention (LPI). The x-axis represents the wow effect (WE) ranging from -1.1 to 1.1, while the y-axis represents luxury purchase intention (LPI) ranging from -0.463 to 0.417. Three lines are plotted: one red line representing GO at -1 standard deviation, one blue line representing GO at the mean, and one green line representing GO at +1 standard deviation. The red line shows a steeper positive slope, indicating a stronger positive impact of the wow effect on luxury purchase intention for individuals with lower goal orientation. The blue line, representing the mean goal orientation, shows a moderate positive slope. The green line, representing higher goal orientation, shows the least steep positive slope, indicating a weaker positive impact of the wow effect on luxury purchase intention. All values are approximated.Moderating effect of goal orientation on the relationship between the wow effect and luxury purchase intention. Source: Figure by authors
A line graph titled 'GO x WE' presents the interaction between goal orientation (GO) and the wow effect (WE) on luxury purchase intention (LPI). The x-axis represents the wow effect (WE) ranging from -1.1 to 1.1, while the y-axis represents luxury purchase intention (LPI) ranging from -0.463 to 0.417. Three lines are plotted: one red line representing GO at -1 standard deviation, one blue line representing GO at the mean, and one green line representing GO at +1 standard deviation. The red line shows a steeper positive slope, indicating a stronger positive impact of the wow effect on luxury purchase intention for individuals with lower goal orientation. The blue line, representing the mean goal orientation, shows a moderate positive slope. The green line, representing higher goal orientation, shows the least steep positive slope, indicating a weaker positive impact of the wow effect on luxury purchase intention. All values are approximated.Moderating effect of goal orientation on the relationship between the wow effect and luxury purchase intention. Source: Figure by authors
This study conducted a controlled variable analysis to examine the relationship between ZEPETO usage frequency and metaverse usage duration. The results showed that ZEPETO usage frequency had no significant effect on the wow effect (β = 0.038, p = 0.159) or luxury purchase intention (β = 0.002, p = 0.486). Furthermore, users' time spent in the metaverse did not have a significant effect on the wow effect (β = 0.009, p = 0.406) or luxury purchase intention (β = −0.042, p = 0.186).
Q2 reflects the structural model's ability to explain endogenous variables (Hair et al., 2019). A Q2 value greater than zero indicates the model possesses predictive relevance, while a Q2 value less than zero indicates the model lacks predictive relevance (Hair et al., 2019). In this study, all Q2 values were positive, indicating the model possesses predictive relevance (see Table 8). In this study, the f2 values indicate that the key structural paths exhibit small to moderate effect sizes. The model explains a substantial amount of variance in the “wow” effect (R2 = 0.426) and a moderate amount of variance in luxury purchase intention (R2 = 0.243).
Path coefficients, predictive relevance (Q2), effect size (F2) and explained variance(R2)
| Endogenous variable | Predictor | β | f2 | R2 | Q2_predict |
|---|---|---|---|---|---|
| WE | TP | 0.144 | 0.028 | 0.426 | 0.403 |
| SOI | 0.186 | 0.050 | |||
| SI | 0.238 | 0.081 | |||
| PA | 0.247 | 0.088 | |||
| PR | 0.174 | 0.043 | |||
| LPI | WE | 0.251 | 0.047 | 0.243 | 0.185 |
| Endogenous variable | Predictor | β | f2 | R2 | Q2_predict |
|---|---|---|---|---|---|
| WE | TP | 0.144 | 0.028 | 0.426 | 0.403 |
| SOI | 0.186 | 0.050 | |||
| SI | 0.238 | 0.081 | |||
| PA | 0.247 | 0.088 | |||
| PR | 0.174 | 0.043 | |||
| LPI | WE | 0.251 | 0.047 | 0.243 | 0.185 |
Studies 1B: replicating previous study
In Study 1B, we replicated the previous study with some modifications. To examine the generalizability of the results, we recruited participants from a different country (South Korea) for Study 1B.
We recruited 145 Korean participants (81.38% aged 18–40, 49.1% female). The overall study design was identical to that of Study 1A. We first examined whether multicollinearity existed among the predictor variables; the results showed that, with VIF values ranging from 1.5 to 2.581, there was no multicollinearity.
CR scores for every construct were higher than the suggested cutoff of 0.70, ranging from 0.794 to 0.930, and Cronbach's alpha values ranged from 0.792 to 0.879, confirming satisfactory internal consistency (Hair et al., 2021). As presented in all item loadings exceeded 0.6, and the AVE values for all constructs were above the recommended minimum of 0.50, ranging from 0.546 to 0.800. This indicates satisfactory discriminant validity among all constructs.
The results show that telepresence (β = 0.153, p < 0.05), sense of immersion (β = 0.125, p < 0.05), social interactivity (β = 0.121, p < 0.05), perceived augmentation (β = 0.157, p < 0.05) and perceived realism (β = 0.145, p < 0.05) had a significant positive effect on the wow effect. The wow effect significantly influenced luxury purchase intention (β = 0.347, p < 0.001), confirming its pivotal role as an emotional driver of digital luxury consumption.
Necessary condition analysis (NCA) results
To complement the sufficiency-based findings obtained through PLS-SEM, this study employed necessary condition analysis (NCA) to explore whether any predictors act as non-compensatory constraints (Dul, 2016). The NCA results are summarized in Table 9. Following the recommendations of Dul et al. (2020), we conducted a critical-effect analysis using R software. The scatterplot shown in Figure 3 illustrates a phenomenon termed “necessary but not sufficient conditions.” This phenomenon can be identified from the CE-FDH line and manifests as the empty area in the upper-left corner of the scatterplot. Sun et al. (2025) indicate that the CE-FDH line is appropriate for analyzing survey data using a seven-point Likert scale. Another critical aspect involves the analysis of statistical significance and the required effect size (d).
Results of NCA
| NCA model | CR-FDH | Bottleneck table (percentile-levels) | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| DV | IV | (d) | P | 0% | 10% | 20% | 30% | 40% | 50% | 60% | 70% | 80% | 90% | 100% |
| WE | PA | 0.042 | 0.016 | NN | 4.10 | 4.10 | 4.10 | 4.10 | 4.10 | 4.10 | 4.10 | 4.10 | 4.10 | 9.10 |
| PR | 0.011 | 0.269 | NN | 0.00 | 0.10 | 0.10 | 0.10 | 0.10 | 0.10 | 0.10 | 0.10 | 8.70 | 8.70 | |
| SI | 0.076 | 0.000 | NN | NN | 0.70 | 0.70 | 0.70 | 0.70 | 0.70 | 0.70 | 16.50 | 37.70 | 37.70 | |
| SOI | 0.033 | 0.000 | NN | NN | NN | NN | NN | NN | NN | 2.50 | 8.80 | 8.80 | 22.50 | |
| TP | 0.022 | 0.061 | NN | NN | 0.40 | 0.40 | 0.40 | 0.40 | 3.40 | 3.50 | 3.50 | 4.70 | 10.80 | |
| LPI | WE | 0.244 | 0.000 | NN | NN | NN | 5.70 | 16.20 | 16.20 | 27.60 | 44.30 | 44.30 | 60.90 | 60.90 |
| NCA model | CR-FDH | Bottleneck table (percentile-levels) | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| DV | IV | (d) | P | 0% | 10% | 20% | 30% | 40% | 50% | 60% | 70% | 80% | 90% | 100% |
| WE | PA | 0.042 | 0.016 | NN | 4.10 | 4.10 | 4.10 | 4.10 | 4.10 | 4.10 | 4.10 | 4.10 | 4.10 | 9.10 |
| PR | 0.011 | 0.269 | NN | 0.00 | 0.10 | 0.10 | 0.10 | 0.10 | 0.10 | 0.10 | 0.10 | 8.70 | 8.70 | |
| SI | 0.076 | 0.000 | NN | NN | 0.70 | 0.70 | 0.70 | 0.70 | 0.70 | 0.70 | 16.50 | 37.70 | 37.70 | |
| SOI | 0.033 | 0.000 | NN | NN | NN | NN | NN | NN | NN | 2.50 | 8.80 | 8.80 | 22.50 | |
| TP | 0.022 | 0.061 | NN | NN | 0.40 | 0.40 | 0.40 | 0.40 | 3.40 | 3.50 | 3.50 | 4.70 | 10.80 | |
| LPI | WE | 0.244 | 0.000 | NN | NN | NN | 5.70 | 16.20 | 16.20 | 27.60 | 44.30 | 44.30 | 60.90 | 60.90 |
The image contains six scatter plots, each depicting relationships between different variables. Each plot includes a green trend line and various data points. Panel A: NCA Plot PA - WE. The x-axis is labeled PA, and the y-axis is labeled WE. The plot shows a green trend line with data points scattered around it. Panel B: NCA Plot PR - WE. The x-axis is labeled PR, and the y-axis is labeled WE. The plot shows a green trend line with data points scattered around it. Panel C: NCA Plot SI - WE. The x-axis is labeled SI, and the y-axis is labeled WE. The plot shows a green trend line with data points scattered around it. Panel D: NCA Plot SOI - WE. The x-axis is labeled SOI, and the y-axis is labeled WE. The plot shows a green trend line with data points scattered around it. Panel E: NCA Plot TP - WE. The x-axis is labeled TP, and the y-axis is labeled WE. The plot shows a green trend line with data points scattered around it. Panel F: NCA Plot WE - LPI.The ceiling line rationale. Source: Figure by authors
The image contains six scatter plots, each depicting relationships between different variables. Each plot includes a green trend line and various data points. Panel A: NCA Plot PA - WE. The x-axis is labeled PA, and the y-axis is labeled WE. The plot shows a green trend line with data points scattered around it. Panel B: NCA Plot PR - WE. The x-axis is labeled PR, and the y-axis is labeled WE. The plot shows a green trend line with data points scattered around it. Panel C: NCA Plot SI - WE. The x-axis is labeled SI, and the y-axis is labeled WE. The plot shows a green trend line with data points scattered around it. Panel D: NCA Plot SOI - WE. The x-axis is labeled SOI, and the y-axis is labeled WE. The plot shows a green trend line with data points scattered around it. Panel E: NCA Plot TP - WE. The x-axis is labeled TP, and the y-axis is labeled WE. The plot shows a green trend line with data points scattered around it. Panel F: NCA Plot WE - LPI.The ceiling line rationale. Source: Figure by authors
For the wow effect, none of the five experiential features met the established threshold for necessity (d ≥ 0.1 and p < 0.05). Although social interactivity yielded a statistically significant result (d = 0.076, p < 0.001), its effect size falls below the recommended cutoff, suggesting that SI, while influential, does not qualify as a necessary condition in the strict NCA sense. Similar conclusions apply to perceived augmentation (d = 0.042, p = 0.016) and sense of immersion (d = 0.033, p < 0.001), both of which were statistically significant yet demonstrated small effect sizes. Telepresence (d = 0.022, p = 0.061) and perceived realism (d = 0.011, p = 0.269) did not reach statistical significance and were also far below the necessity threshold. In contrast, the wow effect itself was found to be a strong necessary condition for luxury purchase intention, with a large effect size (d = 0.244, p < 0.001). This finding suggests that without sufficiently strong emotional motivation through the wow effect, consumer willingness to purchase high-end digital fashion products is unlikely to reach high levels, regardless of other factors. Please see Table 10 for the detailed information.
SEM and NCA results
| SEM results | NCA results | Relationships (X → Y) | Interpretations |
|---|---|---|---|
| sig | n.s | PA → WE | On average, an increase in X will increase Y; no minimum level of the construct is needed to ensure that Y will manifest |
| PR → WE | |||
| SI → WE | |||
| SOI → WE | |||
| TP → WE | |||
| sig | sig | WE → LPI | On average, an increase in X will increase Y. However, a certain level of X is necessary for Y to manifest |
| SEM results | NCA results | Relationships (X → Y) | Interpretations |
|---|---|---|---|
| sig | n.s | PA → WE | On average, an increase in X will increase Y; no minimum level of the construct is needed to ensure that Y will manifest |
| PR → WE | |||
| SI → WE | |||
| SOI → WE | |||
| TP → WE | |||
| sig | sig | WE → LPI | On average, an increase in X will increase Y. However, a certain level of X is necessary for Y to manifest |
Furthermore, these findings indicate that regardless of differences in individual characteristics or experiential configurations, a certain level of wow effect is essential to generate high luxury purchase intent, thereby further validating the reliability of the core mechanisms identified in the model.
Discussion
The findings empirically support the proposed appraisal sequence, suggesting that immersive stimuli influence behavior only as far as they trigger cognitive accommodation processes that culminate in the wow effect. This study set out to examine how immersive experiential stimuli in the metaverse give rise to the wow effect and how this high-intensity emotional response translates into luxury digital fashion purchase intention. Interpreting these findings through the lens of appraisal theory provides a deeper understanding of the underlying mechanism. Specifically, the results suggest that the wow effect emerges not simply from exposure to immersive stimuli, but from consumers' cognitive evaluations of these stimuli along dimensions such as novelty, self-relevance and cognitive congruity. When these appraisal dimensions converge, they trigger cognitive accommodation, resulting in a high-arousal, awe-like emotional state that fundamentally reshapes consumer motivation.
A central finding of this study is that adventure shopping, rather than gratification shopping, mediates the relationship between the wow effect and luxury purchase intention. From an appraisal perspective, this pattern can be explained by the specific configuration of cognitive evaluations underlying the wow effect. High levels of perceived novelty and the need for cognitive accommodation are known to expand attentional scope and increase openness to new experiences. As such, the wow effect reflects not merely heightened emotional intensity, but a shift in cognitive orientation toward exploration and discovery.
This interpretation aligns with research on Awe, which suggests that high-arousal emotions triggered by schema disruption promote curiosity, perceptual expansion and exploratory behavior rather than emotional closure. In this context, adventure shopping emerges as a natural downstream consequence of appraisal-driven emotional escalation, as consumers seek to further engage with, understand and extend the novel experience. By contrast, gratification-oriented motivation, which is associated with emotional comfort and satiation, is less compatible with appraisal configurations characterized by novelty and cognitive expansion. As a result, the wow effect does not translate into gratification-driven consumption pathways.
Beyond the immediate context of metaverse-based luxury consumption, these findings point to a broader theoretical implication regarding the role of high-arousal emotional states in consumer behavior. Specifically, the results suggest that appraisal configurations characterized by novelty and cognitive accommodation do not simply intensify affect but fundamentally redirect motivational processes. Based on this interpretation, several generalizable propositions can be derived.
First, high-intensity emotional responses arising from novelty-driven appraisals are more likely to activate exploration-oriented motivations than consumption-oriented or gratification-based motivations. This suggests that not all positive emotions operate in the same manner, and that emotions associated with cognitive expansion may produce qualitatively different behavioral outcomes compared to those associated with emotional satisfaction.
Second, appraisal-driven emotional states such as the wow effect may function as transitional mechanisms that shift consumers from evaluative processing to experiential engagement. In this sense, the role of emotion is not merely to amplify existing motivations, but to reconfigure the direction of consumer behavior.
Third, in immersive and technologically mediated environments, the amplification of novelty and self-relevance may systematically bias consumer responses toward exploration, experimentation and identity construction. This implies that digital consumption contexts may not simply replicate traditional consumption processes, but instead operate through distinct motivational structures driven by appraisal-based emotional dynamics.
From a theoretical standpoint, these findings extend hedonic consumption theory by demonstrating that hedonic motivation in immersive digital environments is not monolithic. Rather than uniformly enhancing gratification-based pleasure, high-intensity emotional experiences such as the wow effect selectively activate exploratory and adventure-oriented motivations. This suggests that hedonic value in the metaverse is structured around experiential escalation and symbolic engagement rather than sensory indulgence.
The wow effect does not merely intensify positive affect but redirects motivation toward exploration and meaning-seeking behaviors. In doing so, it functions as a bridge between immersive stimuli and adventure-oriented consumption outcomes, while leaving gratification-oriented pathways relatively under activated. This interpretation aligns with psychological accounts of awe as an emotion that promotes curiosity, openness and cognitive expansion rather than comfort or satiation.
The absence of a significant mediation effect through gratification shopping also underscores a broader implication for luxury consumption in virtual environments. Digital luxury does not simply replicate the experiential logic of physical luxury consumption. Instead, it operates through a distinct motivational structure in which novelty, exploration and symbolic engagement take precedence over immediate emotional satisfaction. This distinction reinforces the view that metaverse-based luxury consumption represents a qualitatively different form of luxury experience rather than a digital substitute for physical ownership.
Taken together, these findings suggest that the wow effect functions as an appraisal-driven mechanism of motivational redirection rather than a simple intensifier of positive affect. By linking immersive experiential stimuli to cognitive accommodation and exploratory motivation, this study provides a theoretically grounded explanation of how high-intensity emotional states shape consumer behavior in digital luxury contexts. More broadly, the results highlight the importance of distinguishing between different types of emotional responses in consumer research, particularly in environments where novelty, immersion and symbolic engagement play a central role.
Concluding comments
Theoretical implications
This study makes three primary theoretical contributions.
First, it advances consumer research in metaverse contexts by establishing appraisal theory as a central mechanism for explaining emotional arousal. While prior studies have predominantly focused on technological affordances or general affective responses, this research demonstrates that high-intensity emotional states such as the wow effect arise from cognitive evaluations of novelty, self-relevance and schema incongruity. In doing so, the study shifts the focus from stimulus-driven explanations to a process-based understanding of emotional generation.
Second, the study reconceptualizes the wow effect as a distinct appraisal-driven emotional state that redirects consumer motivation rather than merely intensifying positive affect. The findings show that the wow effect activates exploration-oriented motivations, specifically adventure shopping, while not supporting gratification-based pathways. This distinction extends existing understanding of hedonic consumption by demonstrating that not all positive emotions lead to similar behavioral outcomes.
Third, the study contributes to the emerging literature on metaverse consumption by offering a narrative explanation of how immersive experiences translate into meaningful consumer responses. In line with recent calls for storytelling-driven theorization (Wang, 2025), the research develops a coherent account linking immersive stimuli, cognitive appraisal processes, emotional arousal and behavioral outcomes. This narrative perspective highlights the dynamic and interpretive nature of consumer experiences in virtual environments.
Societal implications
Beyond theoretical advancement, these findings carry broader societal implications. As immersive digital environments become increasingly integrated into everyday life, the amplification of high-arousal emotional experiences may influence how individuals allocate attention, construct identity and engage with consumption practices. In particular, environments designed to maximize novelty and emotional intensity may encourage continuous exploration and prolonged engagement, potentially reshaping perceptions of value, satisfaction and well-being. Understanding these dynamics is therefore critical not only for marketing practice but also for assessing the broader impact of immersive technologies on consumer behavior and quality of life.
Practical implications
The findings of this study offer several important implications for luxury brand managers seeking to operate effectively in immersive virtual environments. First, the results underscore the importance of deliberately designing for the wow effect. Luxury brands should prioritize immersive features such as telepresence, realism, social interactivity and perceived augmentation, as these elements are central to generating strong emotional responses in consumers. Strategic partnerships with virtual world platforms can further enhance these effects by improving avatar realism, refining image quality and enabling more expressive and responsive forms of social interaction. Similarly, brands may benefit from creating proprietary luxury spaces or hosting exclusive, highly immersive brand events within virtual worlds. Such initiatives can stimulate intense emotional engagement and, in turn, increase consumers' willingness to purchase digital luxury products.
Second, this research highlights the centrality of exploration as a motivational driver in virtual consumption contexts. The evidence shows that adventure shopping, rather than gratification shopping, plays a more significant mediating role between the wow effect and purchase intentions. For managers, this implies that consumer excitement is derived less from instant emotional satisfaction and more from opportunities to discover, take risks and explore. Accordingly, luxury brands should design discovery-based experiences that actively encourage novelty-seeking behaviors. Examples include gamified scavenger hunts, exclusive mystery drops and narrative-driven fashion experiences that unfold progressively, thereby encouraging sustained exploration and repeated interaction. These approaches capitalize on consumers' desire for stimulation and uncertainty, translating immersive curiosity into luxury purchase intention.
Finally, the moderating effect of goal orientation points to the need for personalized positioning strategies. The results show that consumers differ in their responses to immersive experiences depending on their motivational orientation. Hedonically oriented consumers are more responsive to symbolic narratives, emotionally rich storytelling and experiences that emphasize pleasure, identity play and fantasy immersion. Conversely, consumers with higher goal orientation require rational justifications to complement emotional engagement. For these consumers, luxury brands can emphasize aspects such as asset exclusivity, blockchain-verified authenticity or investment value. By tailoring strategies to the motivational profile of different consumer segments, brands can optimize emotional impact and enhance conversion rates. This reinforces the importance of consumer heterogeneity in digital luxury marketing and suggests that a “one-size-fits-all” approach to virtual engagement is unlikely to be effective.
Limitations and future research
Despite its contributions, this study has several limitations that warrant consideration. First, the sample was drawn from users of a social media–linked metaverse community, which may limit generalizability. While these users represent an important and relevant segment of early metaverse adopters, they may exhibit higher levels of digital literacy, novelty seeking and openness to immersive experiences than the broader population. Replication across different cultural contexts, platforms and user segments would strengthen confidence in the robustness of the findings.
Second, the proposed framework focuses on experiential stimuli and motivational mechanisms but does not account for other potentially relevant boundary conditions, such as brand familiarity, price sensitivity or prior luxury consumption experience. While the present study adopts an appraisal-based perspective to explain how high-intensity emotional responses emerge in immersive environments, future research could extend this framework by incorporating additional theoretical lenses. For instance, identity-based motivation may provide further insight into how consumers align immersive experiences with self-concept, while broader perspectives on digital consumption may help explain variability in emotional responses across contexts. Incorporating such factors would enable a more nuanced understanding of how appraisal-driven emotional escalation unfolds in virtual luxury consumption.

