Consumers’ attitudes toward virtual try-on technology: an extended TAM model Open Access

The diffusion of digital technologies has reshaped consumers’ behaviour and retail strategies (Yang and Hu, 2024). In 2024, 72% of Europeans aged 16–74 made online purchases, consolidating trends accelerated by the COVID-19 pandemic (Centre for Market Insights, 2024, p. 7). Consistently, e-commerce has expanded rapidly, driven by managerial innovations and advanced technologies that enhance consumers’ engagement through personalization and responsiveness (Quinones et al., 2023). As a result, retailers have increasingly adopted multisensory technologies, particularly augmented reality (AR; Cao, 2021), to simulate sensory cues online (Kim and Forsythe, 2007) and narrow the sensory gap between online and in-store shopping (Bonfanti et al., 2023).

Among these, virtual try-on (VTO) has gained popularity in online apparel, enabling customers to simulate product fit and appearance (Batool and Mou, 2023; Xue et al., 2020), thereby facilitating product evaluation, increasing consumer engagement (Romano et al., 2022) and boosting adoption (Sekri et al., 2024). Early studies framed VTO through utilitarian lenses, focusing on both affective and behavioural outcomes (Kowalczuk et al., 2021; Chidambaram et al., 2024), in line with the classical technology acceptance model (TAM; Davis, 1989) - valued for its parsimony in explaining users’ behaviour across diverse end-user technologies (2D, 3D and VTO applications), populations and usage conditions (Kim and Forsythe, 2007; Merle et al., 2012; Burgess et al., 2023).

These studies emphasized VTO’s ability to reduce perceived product risk (Kumar and Srivastava, 2022), with a strong reliance on structural equation model (SEM)-based approaches. More recent research has begun to explore VTO hedonic and experiential value (Sengupta and Cao, 2022), with a growing focus on emotional engagement (Hasan et al., 2021; Holdack et al., 2022). However, most contributions continue to adopt a utilitarian or affective lens, often overlooking the interplay between cognitive, emotional and value-driven dimensions in shaping consumer attitudes toward VTO (Mollel and Chen, 2025). For instance, while constructs such as perceived enjoyment (PE) (Yoon and Oh, 2022; Nguyen et al., 2023; Taufique et al., 2024) have gained increasing traction, they are rarely examined as mediators linking functional perceptions to attitudes (Zhang et al., 2019; Hasan et al., 2021). Similarly, although personal innovativeness (Bhatt, 2022; Gupta and Mukherjee, 2025) and sustainability-related perceptions (Butt et al., 2023; Trivedi, 2023) – such as perceived environmental benefits (PEBs) – have been acknowledged as relevant antecedents of consumer behaviour (e.g. Sekri et al., 2024), they are often studied in isolation, rather than within a unified explanatory model. In that regard, the few studies that have attempted to integrate multiple perspectives within a single framework have resulted in partial overviews (Noh et al., 2016; Chauhan et al., 2021), falling short of capturing the complexity of emerging technologies like VTO. These technologies require the convergence of diverse drivers – utilitarian, hedonic, individual and sustainability-related factors – which remain largely unaddressed within the existing literature. To fill these gaps, this study proposes an extended TAM framework that integrates functional, utilitarian, individual and sustainable drivers of VTO acceptance. Specifically, this study draws upon constructs from the value-based adoption model (VAM) and the innovation diffusion theory (IDT) to integrate two external variables – personal innovativeness and PEBs – into TAM. Additionally, we examine the mediating role of PE in the relationship between utilitarian drivers and consumer attitudes. To our knowledge, this study is among the first to provide a multidimensional extension of TAM tailored to such immersive technologies, addressing the call for more comprehensive models that capture the evolving dynamics of digital technology adoption. Accordingly, this study addresses the following research questions:

To empirically test the proposed framework, our analysis relies on a sample of 228 Italian consumers, using an SEM approach with ordered Probit estimation, which handles ordinal data and latent constructs while capturing heterogeneity in consumer perceptions and responses to VTO technologies, thus addressing identification and convergence issues.

This study advances the literature by integrating utilitarian, hedonic, individual and sustainability-related dimensions of VTO adoption through an extended TAM-based framework. Practically, it helps retailers understand how immersive technologies are evaluated and adopted in online shopping environments and better strategize marketing activities.

VTO is a transformative tool in digital retail industries like fashion, beauty and eyewear, where product appearance and fit are central to the purchasing process (Kumar and Srivastava, 2022; Lee et al., 2022). By allowing users to visualize and customize products before purchases, VTO improves perceived product accuracy and decision-making efficiency (Christiawan et al., 2024). From a utilitarian perspective, VTO enhances consumers’ perceptions of usefulness and ease of use, especially when embedded in mobile-friendly platforms, strengthening perceived product quality and decision comfort (Ivanov et al., 2023), as well as trust, satisfaction and repurchase intentions (Sekri et al., 2024). Beyond functionality, VTO is recognized for its hedonic value, shaping both cognitive and affective consumer responses (Kim and Forsythe, 2008b; Zhang et al., 2019). Complementing this, recent studies on presence-related constructs (e.g. telepresence, self and social presence) link VTO to stronger emotional engagement (Lee et al., 2022), perceived authenticity and psychological ownership (Javornik et al., 2021; Nayak et al., 2022). Notably, these affective responses depend not only on technological aspects (Liu, 2024) but also on users’ satisfaction with their virtual body representation, body esteem and appearance-related anxiety, which mediate VTO effectiveness (Merle et al., 2012; Ivanov et al., 2023). At the same time, personal traits like innovativeness influence how consumers engage with VTO (Kim and Forsythe, 2008a; Plotkina and Saurel, 2019), with innovative consumers being more inclined to perceive it as useful and enjoyable (Lyu et al., 2023). However, since the literature remains fragmented and lacks a unified TAM (see Appendix A: Table A1 for details), the following sub-sections critically review how functional, personal and value-based drivers have been addressed in the context of VTO and propose a model that extends TAM to incorporate these perspectives.

Perceived ease of use (PEOU) and perceived usefulness (PU) are long-standing predictors in technology acceptance (Davis, 1989; Davis et al., 1989; Kumar, 2022) and remain widely used also in digital contexts to explain user evaluations and preferences. In VTO contexts, studies such as Chidambaram et al. (2024) and Kumar and Srivastava (2022) show that when AR systems are designed as intuitive and responsive, they are more likely to favour positive attitudes and intentions (Lee et al., 2022; Mollel and Chen, 2025). These effects are particularly strong in immersive environments where real-time product interaction and fit simulation enhance the perception of informational accuracy and streamline decision-making (Merle et al., 2012; Kowalczuk et al., 2021; Ivanov et al., 2023). Additionally, Sengupta and Cao (2022) found that highly useable VTO systems maximize consumers’ relative advantage by reducing uncertainty during decision processes. This is especially true as repeated use often leads users to discover additional functionalities, further reinforcing PU and deepening engagement with the technology (Silva et al., 2023). Similarly, Oyman et al. (2022) find that PEOU enhances users’ sense of control and trust, reinforcing technology acceptance.

That is, PEOU and PU are not just antecedents of attitude but also interacting and enabling mechanisms that influence consumer value-based responses. Accordingly, the following hypotheses are presented:

In the context of hedonic and immersive technologies like VTO, PE has become a critical driver of user engagement. Grounded in the VAM, PE reflects the intrinsic pleasure and emotional satisfaction embedded in the technology’s perceived value (Yoon and Oh, 2022; Sekri et al., 2024), beyond its functional utility. In VTO environments, where users interact with avatars, simulated garments or real-time filters, PE complements utilitarian predictors (Hasan et al., 2021; Kowalczuk et al., 2021) and is deeply linked to entertaining and experiential value, interactivity and personalization (Voicu et al., 2023; Xue et al., 2024). Additionally, recent studies demonstrate that PE is not merely an outcome of playfulness but also stems from user cognitive evaluations: technologies perceived as easy to use and functionally helpful are also more likely to be perceived as enjoyable (Taufique et al., 2024). For instance, De Canio et al. (2021a) demonstrate that gamification elements and emotional involvement can enhance both enjoyment and purchase intention. Similarly, Krasonikolakis et al. (2021) show that 3D online stores perceived as informative and easier to navigate than traditional websites generate stronger affective responses.

While early works often focused on either utilitarian predictors or affective responses, emerging studies have explored the influence of PE as a mediator in AR wearables and mobile learning platform adoption (Holdack et al., 2022; Huang and Liu, 2024), but this role is rarely tested in apparel-based VTO. Notably, the evidence is mixed: while Hasan et al. (2021) confirm that PE mediates both PEOU and PU in shaping consumers’ shopping intentions, Sekri et al. (2024) report no significant link between PU and PE, suggesting the need to explore technology-specific dynamics and further justifying the role of PE as a mediator of attitudinal responses, rather than a direct predictor of perceived value. Furthermore, despite the growing evidence on how enjoyment enhances technology’s perceived value (Kim and Forsythe, 2008a, b; Chidambaram et al., 2024), few studies have explored the channels through which PEOU and PU translate into hedonic engagement and how the latter, in turn, influences attitudes. Therefore, the following hypotheses are proposed:

PEBs have emerged as a meaningful extension to traditional TAM, as consumers become more environmentally conscious in their online shopping behaviour (Chauhan et al., 2021; De Canio et al., 2021b). In the context of VTO, such benefits represent a non-functional value component rooted in the belief that digital try-on technologies reduce product returns, overconsumption and waste (Trivedi, 2023). Joerss et al. (2021) show that such immersive technologies increase consumer awareness of sustainable options and promote eco-conscious decisions. Similarly, Lavoye et al. (2025) show that when VTO systems provide realistic virtual outputs, users perceive a “sustainable fit”, enhancing symbolic and functional value linked to eco-conscious behaviours. While sustainability is often cited as a macro-level advantage in digital technology adoption, few studies examine it as a direct antecedent of the TAM construct. However, early evidence suggests that environmentally motivated consumers are more likely to perceive such tools as beneficial and practical, particularly when they visibly provide and communicate their sustainable value, ultimately boosting PEOU and usefulness (Steg et al., 2014; Park and Ha, 2012). Notably, Chauhan et al. (2021) found that PEOU, PU and personal innovativeness significantly influence green purchase intentions, indirectly suggesting a pathway for extending TAM with value-based constructs that have been shown to be more effective when mediated by cognitive-affective appraisals of technology. This is even more true considering that environmentally motivated consumers may feel discouraged if they seem too complex or cognitively demanding (Fujii, 2006). In such cases, the sustainable potential of the technology remains latent or unexpressed. Addressing this, PEBs should not be treated as independent variables but rather as inputs that influence technology adoption through their impact on PEOU and usefulness, ultimately aligning with calls for a broader value-driven model of technology adoption (Joerss et al., 2021; Papagiannidis and Marikyan, 2022). Accordingly, the following hypotheses have been formulated:

Among individual traits, personal innovativeness (PersInn) has gained relevance in the extended TAM model, especially in technology-intensive retail environments (Bhatt, 2022; Sekri et al., 2024).

Grounded in Rogers’ IDT, it reflects an individual openness to experimentation, showing higher tolerance for uncertainty and less anxiety about unfamiliar systems (Rogers, 2003; Mollel and Chen, 2025). In online apparel, such consumers are more likely to adopt new technologies earlier (Rosen, 2005; Kim and Forsythe, 2008a), perceiving them as intuitive (Jeong and Choi, 2022), useful (Qasem, 2021) and more emotionally engaging (Castillo and Bigne, 2021). Recent studies found that both personal and technological innovativeness influence attitudes toward new technologies and PEOU and usefulness, as innovative consumers are better positioned to recognize new technologies’ functional values (Almaiah et al., 2022; Lyu et al., 2023).

Koivisto et al. (2016) and Noh et al. (2016) further suggest that innovative users are more likely to interpret technology complexity as a challenge rather than a barrier, reinforcing adaptability and enthusiasm. Acknowledging that personal innovativeness can act as an amplifier of both instrumental and experiential aspects of immersive technologies, we argue that personal innovativeness is not merely a control variable but a key antecedent of usability and functionality in immersive retail (Chauhan et al., 2021). Accordingly, the following hypotheses are proposed:

Attitude towards VTO technology (ATT) plays a central role in determining whether individuals will integrate such technologies into their shopping routines (Davis et al., 1989). Prior research confirms that attitudes mediate the influence of core cognitive and affective perceptions – such as usefulness and enjoyment – on behavioural intentions in only shopping (Kowalczuk et al., 2021; Nguyen et al., 2023), also in the context of impulsive buying behaviour (Kumar and Srivastava, 2022). However, most of these studies rely on unidimensional models centred on either cognitive or hedonic factors. Conversely, the recent study of Mollel and Chen (2025) has begun to model attitudes as shaped by functional (e.g. PEOU, PU) and personal (e.g. body esteem, fit confidence) dimensions within VTO environments in fashion-related contexts. Building on this evolving view, this study theorizes attitudes as the central evaluative outcome of multiple influences that drive digital try-on adoption. Accordingly, the following hypothesis is proposed:

The comprehensive conceptual model, along with the research hypotheses, is presented in Figure 1. See Appendix B (Table B1) for key constructs, their definitions, theoretical foundations and expected relationships with TAM dimensions.

To explore consumers’ behaviour towards VTO technology, this study employs a Structural Equation Model with an Ordered Probit approach. Analyses were conducted in RStudio with the lavaan package (Rosseel, 2012), following standard SEM steps. Preliminary screening confirmed data suitability, and Hoelter’s critical N indicated an adequate sample size (N = 228, p < 0.05).

Data were collected through an online-based survey, administered via Qualtrics platform to ensure methodological validity, as VTO is typically used in digital shopping contexts (Kim and Forsythe, 2008a). These platforms improved data quality through built-in features like randomization, device optimization and missing response management, ultimately minimizing response bias and enhancing reliability (Braun et al., 2021).

Participation was voluntary, and the survey was distributed via social media (WhatsApp, Instagram and Facebook) and email. All respondents were informed about the research’s purpose and provided informed consent before proceeding. Convenience and snowballing sampling were employed (Goodman, 1961; Golzar et al., 2022), as they are well-suited to exploratory research on emerging digital technologies (Etikan et al., 2016).

Participants were first introduced to a simulated online shopping interface illustrating VTO functionality to standardize respondents’ frame of reference. Once contextualized, participants completed the questionnaire, which included items adapted from previously validated scales. Specifically, PU (4 items), PEOU (3 items), PE (4 items), attitudes (ATT, 4 items) and personal innovativeness (PersInn, 3 items) were adapted from Kim and Forsythe (2008b), PEBs (3 items) from Mugge et al. (2017) and behavioural intention to buy (BI, 3 items) from Dodds et al. (1991). All items used a seven-point Likert scale, from “1 = strongly disagree” to “7 = strongly agree” (see survey items in Table 1).

The sample included 228 participants (56.14% female). The mean age was 38.04 years (SD = 14.81), with a strong representation of younger cohorts: Gen X (39.04%), Gen Z (31.14%), Millennials (20.61%) and Baby Boomers (9.21%). Regarding participants’ educational levels, 53.07% had a university degree, 33.33% had post-graduate education, 13.16% had secondary education and one participant completed only primary school.

Data adequacy was assessed to ensure validity and reliability, using descriptive statistics. All mean scores exceeded the midpoint of 3, ranging from 3.69 to 5.58; standard deviation ranged between 1.53 and 1.80; and skewness and kurtosis indices meet Kline’s (2023) acceptable thresholds not exceeding |3| and |10|, confirming data normality and suitability for structural modelling.

The measurement model was evaluated for reliability and construct validity (convergent and discriminant) through exploratory factor analysis (EFA) and confirmatory factor analysis (CFA). Reliability was assessed through Cronbach’s alpha and composite reliability (CR) indices. As shown in Table 1 all alpha values exceed 0.70, ranging from 0.869 (ATT) to 0.948 (PE), and CR values are also greater than 0.70, confirming internal consistency. Convergent validity, assessed via average variance extracted (AVE), shows values above the 0.50 cut-off, ranging from 0.706 (BI) to 0.876 (PEBs), indicating that the constructs explain a significant portion of the variance in their items. According to Ping’s (2004) criteria, the square root of AVE should be greater than its corresponding correlation coefficient – a condition met in this study – confirming that all constructs are empirically distinct and the model has satisfactory discriminant validity.

Model validity was assessed using SEM, following Hair et al. (2010)’s recommendations to evaluate the model through multiple goodness of fit indices: $χ2$⁠, the ratio between $χ2$ and the degree of freedom $(χ2/df)$⁠, Tucker–Lewis index (TLI), comparative fit index (CFI), root mean square error of approximation (RMSEA) and the standardized root mean residual (SRMR). The results indicate a strong model fit. The chi-square value (χ² = 59.835, p > 0.05) was non-significant, suggesting no substantial difference between the observed and model-implied covariance metrics. The χ²/df ratio (1.19) was well below the conventional threshold of 3 (Kline, 2005), suggesting a good balance between model complexity and parsimony. TLI (0.982) and CFI (0.985) exceeded the 0.90 threshold (McDonald and Ho, 2002; Sahoo, 2019). RMSEA was 0.072 (<0.08) and SRMR was 0.041 (<0.05), suggesting minimal residual differences between the observed and predicted correlations (McDonald and Ho, 2002).

Figure 2 and Table 2 (panel A) show the hypothesis testing results and path coefficients. Ten out of the eleven hypotheses were supported, indicating strong empirical support for the extended TAM framework integrating PE, PersInn and PEBs. Among TAM-related relationships, H2 (PU → ATT) and H3 (PEOU → ATT) were significant and positive (β = 0.332, p > 0.001; β = 0.108, p < 0.01), confirming the role of utilitarian perceptions in shaping attitudes toward VTO. However, H1 (PEOU → PU) was not supported, suggesting that consumers do not necessarily PU as a direct consequence of ease of use. PE showed a central role in VTO acceptance: H4 (PEOU → PE) and H5 (PU → PE) were both significant (β = 0.367 and β = 0.598, both p < 0.01), confirming that hedonic perceptions stem from usability and functional value. The mediating effect of PE was assessed following Zhao et al. (2010): as shown in Table 2 (panel B), all indirect effects were significant, supporting H6. As both direct and total effects were also significant and positive, PE exerts complementary mediation in the PEOU–PE–ATT and PU–PE–ATT relationships.

The exogenous constructs also show strong predictive power. PEBs significantly influence both PEOU and PU (H7: β = 0.379, p < 0.01 and H8: β = 0.479, p < 0.01). Similarly, PersInn strongly predicted PEOU (H9: β = 0.440, p < 0.01) and PU (H10: β = 0.304, p < 0.01), confirming the relevance of innovation-based early adoption traits. Finally, as expected, attitude significantly influenced behavioural intention to buy (H11: β = 0.914, p < 0.01), confirming affective evaluation as a key predictor of VTO adoption.

This study extends previous findings on the adoption of VTO technologies by integrating individual, hedonic and value-based drivers into a unique framework (Perannagari and Chakrabarti, 2020; Jayaswal and Parida, 2023). While previous research has established the relevance of PEOU and PU (H2 and H3) in shaping consumer attitudes (Davis, 1989; Kim and Forsythe, 2008a), our results nuance this relationship. Specifically, the non-significant effect of PEOU on PU (H1) diverges from standard TAM assumptions and prior evidence (e.g. Sengupta and Cao, 2022), suggesting that in hedonic contexts like VTO, experiential value may outweigh usability in influencing usefulness perception. This indicates a shift in how consumers evaluate perceived technological value, also in light of the significant mediation role of PE (H4-H6) on attitude (Hasan et al., 2021; Huang and Liu, 2024). Unlike previous studies that consider enjoyment as a standalone predictor or moderator (Zhang et al., 2019; Taufique et al., 2024), our findings highlight its mediating role, consistent with Hasan et al. (2021) and Holdack et al. (2022), thus positioning enjoyment as a link between utilitarian and affective evaluations, expanding TAM in hedonic contexts.

Our study also reinforces the importance of personal innovativeness in shaping both PEOU and PU (H9, H10). Even if consistent with previous findings (Kim and Forsythe, 2008a; Almaiah et al., 2022; Lyu et al., 2023), our model offers a deeper integration by treating innovativeness as a primary antecedent rather than a moderating or contextual factor. Moreover, by showing that PEBs significantly predict both PU and PEOU (H7, H8), our study adds a value-based perspective to TAM, contributing to a better understanding of how consumers evaluate VTO utility and its alignment with ethical and environmental goals (Chauhan et al., 2021; De Canio et al., 2021b).

Finally, the strong effect of attitude on behavioural intention (H11) reinforces the idea that consumers’ overall evaluation of the VTO experience is a reliable predictor of technology adoption (Kowalczuk et al., 2021; Mollel and Chen, 2025). These results underscore the importance of a multidimensional, integrated TAM framework addressing functional and value-based dimensions of acceptance in retail (Bigne and Bigne, 2021).

Despite several studies exploring consumers’ behaviour toward VTO, this study offers a more refined theory-driven extension of the TAM framework by integrating hedonic, personal and sustainability-related factors. Specifically, the contribution is threefold.

First, this study confirms and expands the mediating role of PE between PEOU, PU and attitude. Unlike earlier works considering enjoyment as a direct or parallel predictor (Zhang et al., 2019; Chidambaram et al., 2024), our findings show that enjoyment acts as a complementary mediator, boosting the impact of utilitarian predictors on attitudes. This refines and expands the classical TAM by incorporating hedonic motivations as a functional mechanism that amplifies the effects of core TAM variables (Hasan et al., 2021; Holdack et al., 2022).

Second, the inclusion of personal innovativeness as an antecedent of both PEOU and PU advances TAM by embedding individual-level traits into technology adoption, directly responding to calls for more heterogeneous frameworks in digital contexts (Kim and Forsythe, 2008a; Lyu et al., 2023). Grounded in IDT, our findings show that early adopters perceive emerging technologies like VTO as more useful and intuitive.

Third, by introducing PEBs as a driver of functional evaluations, this study positions TAM within a value-based acceptance perspective (Chauhan et al., 2021; De Canio et al., 2021b), beyond purely utilitarian predictors. Taken together, these contributions offer a multidimensional extension of TAM that better captures the complexity of consumer behaviour in immersive retail environments.

This study offers relevant insights for managers, developers and marketers aiming to improve consumer engagement with VTO in online settings.

First, the significance of PEOU and PU reinforces the need for intuitive and functionally robust VTO systems. Developers should prioritize user-centric design, integrating responsive interfaces, real-time fit simulation and product personalization to enhance perceived value, especially in visually driven sectors like fashion, cosmetics and luxury.

Second, the mediating role of PE reveals that functionality alone does not guarantee consumer acceptance, suggesting developers and UX designers prioritize engaging and emotional experiences through visual realism, inclusivity and interactivity.

Third, our findings suggest that innovative users may respond better to campaigns emphasizing novelty and experimentation, while less innovative users may require educational content, reassuring messages and guided experiences that reduce cognitive load and anxiety. Retailers can leverage this insight to personalize VTO marketing strategies based on consumers’ digital readiness and openness to innovation.

Fourth, the role of PEBs in shaping functional perceptions of VTO indicates an opportunity for firms to reinforce their ethical brand identity: by framing VTO as a sustainable solution, retailers can appeal to the growing segment of environmentally conscious consumers. Finally, managers should avoid treating VTO as a standalone tool, but they should integrate it within a broader omnichannel ecosystem – linked to AR mirrors, loyalty initiatives and post–purchase services – to create hybrid experiences for digitally native consumers. Additionally, inclusive VTO design that accommodates multiple body types, gender identities and visual preferences can enhance accessibility and emotional comfort, aligning technological innovation with social values around diversity, ethics and sustainability.

While this study provides valuable insights into factors affecting consumer attitudes and behaviour toward VTO, some limitations should be acknowledged. First, the sample of 228 Italian respondents, even though appropriate, limits generalizability across cultural or geographical contexts. Since hedonic and sustainability-driven behaviours may vary internationally, cross-cultural validations are needed for the model’s robustness.

Second, since the cross-sectional design does not capture temporal dynamics of VTO adoption, longitudinal research should assess how attitudes and intentions evolve over time, also exploring post-adoption outcomes, including user satisfaction, loyalty and continued use.

Third, although the model integrates personal innovativeness, PEBs and PE, other constructs – like trust, privacy concerns, social influence or technology anxiety – may further enhance its exploratory power. Future research might explore moderating effects of age, digital literacy or prior AR experience to better capture heterogeneity in VTO evaluations.

Finally, since this study focuses specifically on VTO, to develop a more unified theory of immersive technology adoption, future studies should examine its applicability across other augmented reality and virtual reality retail applications.

The supplementary material for this article can be found online.