Consumer values and word-of-mouth: an empirical study of the second-hand fashion retail market

The fashion industry has been in the spotlight for years as one of the sectors that generates the greatest amount of negative environmental impacts: it is responsible for 10% of carbon emissions, consumes large amounts of water resources and uses huge quantities of toxic chemicals (Earth, 2022). This, coupled with the effect that rapid changes in trends have had on reducing the life cycle of garments (Zhang et al., 2025), as well as the exponential growth of production and sales in the last decade (Earth, 2022), has led the industry to rethink business models, reorienting them towards more circular models committed to the principles of sustainability. The transition to circular models has great potential to take hold, driven by growing consumer support (Ortega et al., 2025). A clear example of this support is the rise of consumer-to-consumer (C2C) platforms such as Vinted and Wallapop, which have normalised the mass resale of clothing. Likewise, leading retail brands have integrated their own circularity programmes: Patagonia, through its Worn Wear platform, encourages repurchasing and repairing; Levi’s has implemented its second-hand programme for the resale of used jeans; and Zara has launched pre-owned, a service that facilitates the repair, donation and sale of its own garments among customers. This shift in purchasing behaviour has allowed the second-hand fashion market to grow at a faster rate than other retail sectors. According to the ThredUp report (2025), the global second-hand fashion market is expected to reach £367 billion by 2029. The report further indicates that, as consumers increasingly prioritise second-hand goods, the retail sector is adopting new and effective resale channels, thereby accelerating growth in the resale market.

However, for this to become a reality in the near future, it is essential that consumers support this type of market through their purchasing decisions. To evaluate this, research studies consumer behaviour in word-of-mouth (WOM), considered one of the key factors in the success of products (Taheri et al., 2021). In turn, as antecedents of consumer behaviour, the value system of consumers is considered. Taking as a reference, the model proposed by Sheth et al. (1991), composed of functional, conditional, emotional, epistemic and social value, is extended with the evaluation of environmental value and economic value. The justification for evaluating the value system lies in the fact that, in the second-hand fashion retail market, it is essential to explain the behavioural traits underlying consumers’ purchasing decisions (Vidal-Ayuso et al., 2023). More specifically, Chakraborty and Sadachar (2023) argue that it is precisely personal values and the identity between the brand and the consumer that drive the growth of this business model. These values are particularly relevant because they are linked to consumers’’ lifestyles and remain stable over time (Atherton et al., 2021). This research seeks to go a step further. To date, no research has explicitly focused on jointly measuring the effect of these values on consumers’ behavioural responses, especially WOM, despite recent calls for such analysis (Wako et al., 2025). However, the study of WOM in second-hand fashion purchases is relevant because it measures the success of products (Taheri et al., 2021) as well as could contribute to the proliferation of more responsible consumption behaviours (Hong et al., 2020). Finally, it is important to distinguish this study from the deep literature on second-hand luxury retail. While luxury resale often focuses on status and brand exclusivity (Zhang and Zhao, 2019), this research focuses on the general second-hand fashion retail market. In this context, economic and environmental values play a more critical role, and WOM depends on a broader evaluation of perceived quality and value-for-money, beyond the intrinsic prestige of a luxury label (Wako et al., 2025).

Consumer value theory was proposed by Sheth et al. (1991) and has been widely used to predict consumer behaviours and attitudes (Sweeney and Soutar, 2001). The theory is based on the fact that the consumption experience depends on multiple evaluations, which has been widely supported by the academic community, as each value makes a different contribution depending on the context (Sheth et al., 1991), providing greater predictive power versus the unidimensional value perspective (Gonçalves et al., 2016). The theory is composed of five types of value: functional, conditional, emotional, epistemic and social, although it has often been extended depending on the research context. For instance, in the consumption of luxury goods, social and emotional values typically exert the greatest predictive power due to the search for status and self-expression (Zhang and Zhao, 2019). In contrast, in the context of essential services or commodities, functional value (related to reliability and price) tends to be the primary driver of choice (Zeithaml, 1988). Within the specific field of sustainable or second-hand products, while functional and economic values remain relevant, epistemic value (the desire for novelty or knowledge) and environmental value (altruistic concern) emerge as critical determinants that distinguish this behaviour from conventional retail (Wako et al., 2025). In our case, we additionally evaluated the economic and environmental value of the consumer, being two fundamental variables to explain the purchase behaviour of sustainable products.

Word-of-mouth (WOM) can be considered as “verbal, informal communication occurring in person, by telephone, email, mailing list, or any other communication method” (Goyette et al., 2010, p. 9). Thus, WOM captures an experience created by consumers and subsequently shared with others (Engel et al., 1969). In the specific context of the second-hand fashion retail market, WOM is not merely a consequence of satisfaction but a strategic driver for market growth (Taheri et al., 2021). Unlike conventional retail, the second-hand fashion sector often faces consumer barriers such as perceived risk regarding garment hygiene, quality or authenticity. WOM acts as a critical mechanism to mitigate these risks; when a consumer recommends a second-hand product, they are providing a “social seal of approval” that reduces the uncertainty of potential new buyers (Silva et al., 2021). Furthermore, as second-hand fashion retail relies heavily on the constant influx of both buyers and sellers to maintain a circular flow (Hong et al., 2020), WOM facilitates the recruitment of new participants at a lower cost than traditional advertising, directly contributing to the sector’s scalability and long-term viability. Consequently, this research seeks to go a step further by explaining the intention to generate WOM based on the consumer value theory of Sheth et al. (1991).

Perceived quality is defined as the degree to which a product or service meets consumer expectations (Zeithaml, 1988). Furthermore, perceived quality has been considered a feeling towards the product or service (Agarwal and Rao, 1996), describing a strong relationship between the product or service and the consumer in the context of sustainable products (Papista and Krystallis, 2013). In the specific context of second-hand fashion, this perception becomes more complex. Because garments have been previously worn, consumers evaluate whether these attributes – particularly design and finish – are still preserved. This evaluation acts as a signal of value that helps reduce uncertainty. When consumers perceive quality to be high, the functional and economic value of the purchase is confirmed, which in turn strengthens the intention to recommend the product (Cavazos-Arroyo and Zapata-Sánchez, 2025). Therefore, perceived quality functions not only as a technical attribute but also as a psychological mediator that validates the consumer’s value system before they risk making a public judgement through word of mouth (Silva et al., 2021).

Epistemic value is understood as the degree to which products generate curiosity or novelty to consumers (Sheth et al., 1991). In the second-hand retail market, garments are very limited, providing exclusivity to their consumers (Machado et al., 2019). In this way, the concern for sustainability, coupled with curiosity, that sustainable products generate for less individualistic consumers, will not only lead them to improve their consumption intentions, but also their recommendation to others (Moon et al., 2021). To evaluate this effect on the second-hand fashion retail market, we propose the following research hypothesis:

As for conditional value, it is understood as the circumstances or conditions that determine decision making, so that the consumer will perceive conditional value if the product meets the necessary conditions (Sheth et al., 1991). Lin and Huang (2012) argue that consumers concerned about sustainability tend to shape their purchasing decisions towards more responsible products. Consumers of green products are often important prospectors, especially in the spread of special discount campaigns and seasonal promotions (Aravindan et al., 2023). To evaluate this effect on the second-hand fashion retail market, we propose the following research hypothesis:

Sheth et al. (1991) defines emotional value as the positive feelings aroused in the consumer by the product or service, being one of the most important predictors of purchase behaviour for sustainable products (Kim et al., 2021). So, Liu et al. (2021) explained the positive effect of emotional value in the sustainable food market based on consumers’ motivation to compensate for their irresponsible actions. To evaluate this effect on the second-hand fashion retail market, we propose the following research hypothesis:

Functional value refers to the overall performance of the product in relation to its price, quality and general level of functionality (Sheth et al., 1991), being an important predictor of sustainable consumer purchasing behaviour (Koller et al., 2011). For Aravindan et al. (2023), in the case of sustainable products, the effect of functional value is positive in WOM when consumers make a positive quality-price evaluation. To evaluate this effect on the second-hand fashion retail market, we propose the following research hypothesis:

Social value measures consumers’ perception of the social acceptance or image it causes in society or in specific social groups (Sheth et al., 1991), playing a very relevant role in WOM (Park et al., 2021). In the case of the second-hand retail market, consumers who purchase these types of products tend to associate it with a greater commitment to environmental issues, so they self-perceive the projection of a more positive social image, thus showing better purchasing attitudes (Kim et al., 2021). Kim et al. (2023) found that consumers who share sustainability concerns are willing to recommend products that contribute to reducing environmental impacts, promoting social inclusion and achieving economic justice. In order to evaluate this effect on the second-hand fashion retail market, we propose the following research hypothesis:

For Zeithaml (1988), economic value represents consumers’ evaluation of the relationship between price and quality of products. In the case of second-hand fashion, academic literature has not yet taken a definitive position on the effect of economic value on purchase intentions. In the context of second-hand retail market, Tangri and Yu (2023) found that consumers are particularly sensitive to the price of second-hand fashion. Yadav and Sijoria (2024) found a positive relationship between economic value, second-hand fashion purchasing behaviour and, indirectly, WOM. Therefore, they relied on classical sequential behaviour theory. To expand upon this theory and study the direct effect on WOM, we propose the following research hypothesis:

Finally, environmental value is understood as the commitment that consumers acquire about sustainability and environmental issues (Biswas and Roy, 2015), being one of the most relevant factors in the purchase of sustainable products. In the fashion industry, consumers’ environmental attitudes have been found to result in a rejection of fast fashion products (Joung, 2014). In this context, Nguyen and Nguyen (2025) argue that consumers concerned about environmental issues will be much more likely to have conversations and recommend sustainable products and practices. To evaluate this effect on the second-hand fashion retail market, we propose the following research hypothesis:

Several studies have analysed the impact of consumer values on the perceived quality of sustainable products (see, e.g., Prados-Peña et al., 2023; Zhang and Zhao, 2019). Additionally, perceived quality is considered one of the most relevant factors when recommending second-hand fashion (Li et al., 2024). For example, Cavazos-Arroyo and Zapata-Sánchez (2025) found that perceived consumer value has a direct effect on consumer trust and WOM, while Li et al. (2024) documented the influence of perceived quality on WOM in the context of sustainable products. These findings support the decision to conceptualise perceived quality as an explanatory mechanism that mediates the relationship between consumer values and WOM.

However, the mediating effect of perceived quality is not examined across all proposed consumer values in this study. Instead, the analysis focuses exclusively on functional, economic value, social value and environmental value. This delimitation is justified by the fact that these four values incorporate an external validation or performance-based dimension. In the case of utilitarian values (functional and economic) and environmental value, perceived quality serves as a prerequisite that confirms the materialisation of the expected benefit (Tymoshchuk et al., 2024). Similarly, social value requires that product quality be evident for individuals to feel confident in recommending the product to their social circle, thereby mitigating the image risk associated with the consumption of second-hand products (Silva et al., 2021). Accordingly, prior research suggests that perceived quality functions as an evaluative cognitive process linking these four values to WOM (Li et al., 2024; Tymoshchuk et al., 2024). Therefore, to test these relationships, we propose the following research hypotheses:

A personal survey was used to obtain the data, which was elaborated through Google Forms and distributed through Prolific ^[1] during the months of March and April 2025. The technique for data collection was convenience sampling. In this research, convenience sampling is appropriate because the population under study has very specific characteristics (Dörnyei, 2007). Three criteria were applied: (1) that the first language of the respondents was Spanish, since the survey was distributed in Spanish, although nationality was left open to any country; and (2) that there was parity between men and women, to avoid possible gender bias. To achieve gender parity, the platform was provided with the necessary requirements. Additionally, the questionnaire included a gender field. As responses were received, the gender profile of the remaining respondents was adjusted. This process was repeated until gender parity was achieved. And (3), that the respondents have purchased second-hand fashion at some occasion, since to value WOM it is necessary to have used the product.

A total of 382 valid surveys were obtained. To evaluate the adequacy of the number of observations, G*Power (Green, 1991) was used, in which the following parameters were introduced: (1) number of exogenous variables, which in our case are 7, given that perceived quality is incorporated as a mediating variable; (2) the effect size, which by default establishes a medium effect (0.15); and (3) the probability of error, which was selected to be 5% (power = 95%). For the parameters introduced, G*Power yielded that the minimum number of observations in the sample should be 153, a value that is amply exceeded by the sample analysed (n = 382 observations). Table 1 shows the sociodemographic characteristics.

The profile of the people surveyed corresponds to a female (50.26%), with university studies (52.36%), between 25 and 40 years old (65.71%), who buys second-hand fashion occasionally (45.29%), in physical stores (81.68%).

For the construction of the measurement scale, three items from Lai et al. (2010) were used to measure perceived quality, and four items from Park and Choo (2015), to evaluate WOM were used. Meanwhile, the following variables were used to measure the values of second-hand fashion consumers: three items from Sweeney and Soutar (2001) for functional value, four items from Lin and Huang (2012) for conditional value, five items from Sweeney and Soutar (2001) for emotional value, four items from Dholakia (2001) for epistemic value, six items from Sweeney and Soutar (2001) for social value, four items from Sweeney and Soutar (2001) for economic value and four items from Prados-Peña et al. (2023) for environmental value. All constructs were considered reflective or mode A, in the same way that these were formulated from the studies from which they were adapted. All items were evaluated on a Likert scale from 1 (strongly disagree) to 5 (strongly agree).

Only two adaptations were made to the original items and variables: (1) the research context: second-hand fashion and (2) translation into Spanish. For the translation of the items, the procedure proposed by Brislin (1970) was applied. Four native Spanish professors specializing in economics and business translated all the items into Spanish. Additionally, a pretest (n = 25) was conducted in which respondents were asked to answer the questions and indicate any errors or anomalies. No errors were detected in the pretest; therefore, the measurement scale is available in Table 2.

The applied research methodology consists of structural equation modelling (SEM), which is valuable for analysing the established theoretical relationships and testing the defined research hypotheses. More specifically, we used the SEM approach based on variance through partial least squares (Hair et al., 2021). The PLS-SEM method is a causal-predictive approach whose fundamental strength is prediction in model estimation, being very relevant for evaluating complex models (Hair et al., 2021). PLS-SEM over CB-SEM was selected for three reasons (Hair et al., 2019): (1) it is particularly suitable for the analysis of smaller sample sizes; (2) it is appropriate for research focused on theory development, rather than for studies aimed towards theory confirmation; and (3) its ability to handle non-normal data distributions and to produce robust results further supports its appropriateness. Accordingly, the methodology was applied at two levels (Gefen et al., 2000): (1) the outer (measurement) model to evaluate the relationships between latent variables and their items; and (2) the inner (structural) model to study the relationship between the different constructs.

To assess potential bias, Harman’s single-factor test was conducted, whereby no single factor should account for more than 50% of the total variance (Fuller et al., 2016). In our sample, the first component captured 36.429% of the variance, thus demonstrating that there is no bias.

Four criteria were applied to evaluate convergent validity: (1) the standardized loadings of each item and their significance, which must be statistically significant (p < 0.01) and their values greater than 0.7 (Chin, 2009); (2) Cronbach’s alpha, which must yield values greater than 0.7 (Martínez, 2014); (3) internal consistency through Rho A and Rho C, which must be greater than 0.5 (Fornell and Larcker, 1981) and (4) average variance extracted (AVE), which should yield values greater than 0.5 (Fornell and Larcker, 1981). Table 2 shows the results obtained.

Regarding discriminant validity, the two usual criteria were applied: (1) the Fornell and Larcker (1981) criterion, which states that the shared variance between constructs should not exceed their AVE, (2) the heterotrait-monotrait test, considered the best to evaluate discriminant validity (Henseler et al., 2016), whose values, according to Gold et al. (2001), should be less than 0.9. Table 3 show the results obtained.

Several aspects are analysed in the evaluation of the structural model. The first is the goodness of fit of the model. Williams et al. (2009) state that SRMR should be < 0.1, so the model fit is really good (SRMR = 0.063).

The second analysis refers to hypothesis testing, for which the bootstrap method of 10,000 subsamples was applied (Streukens and Leroi-Werelds, 2016), obtaining the path coefficients of each of the established relationships. Table 4 shows the hypothesis testing analysis for both direct effects (first part) and mediating effects (second part).

The consumer values evaluated present different effects on WOM. EPV showed a direct and significant effect on WOM (β=0.140; p value=0.000) and also EMV (β=0.306; p value=0.000). Therefore, H1 and H3 are accepted. CV shows a direct and significant effect on WOM (β=0.251; p value=0.000). Therefore, H2 is accepted. FV and ENV, however, show positive effects on WOM, although these effects are not significant. For this part, SV and ECV show negative and not significant effects. Therefore, H4, H5, H6 and H7 are rejected, and it is necessary to evaluate the mediating effect of perceived quality between these values and WOM.

The effect between QUA and the other variables in the model is positive and significative, so we analysed QUA’s mediating role (Table 4). QUA fully mediates the relationship between FV and WOM (β _{FV→ QUA → WOM} = 0.141; p < 0.001), SV and WOM (β _{SV → QUA → WOM} = 0.042; p value = 0.001) and ECV and WOM (β _{ECV→ QUA → WOM} = 0.037; p value = 0.007). However, QUA does not mediate the relationship between ENV and WOM. Consequently, H8, H9 and H10 are accepted, while H11 is rejected.

Our findings show that the consumer values evaluated present different effects on WOM. Compensation for irresponsible actions (Liu et al., 2021) and concern for one’s own sustainability (Moon et al., 2021) could be the key reasons to explain why second-hand fashion consumers’ emotional and epistemic value drive them to recommend this type of garment. For its part, the positive and significant effect of the conditional value on WOM is in line with the findings of Aravindan et al. (2023) and could show that consumers tend to recommend value for money of garments and become prospectors of product discount campaigns. Regarding the mediating effect of QUA, these findings cannot be directly compared with previous academic literature. Given that this specific mediation structure has not been examined before. However, it has already been demonstrated that there is a positive effect between consumer values, perceived quality and WOM (see, e.g., Cavazos-Arroyo and Zapata-Sánchez, 2025; Li et al., 2024), so this mediating effect of perceived quality could be intuited. Additionally, Sheth et al. (1991) considered that functional value referred to the overall performance of the product. Therefore, it is consistent to expect that the consumer will recommend it if they perceive its quality as acceptable. Similarly, it is reasonable to think that a positive perception of the quality of second-hand fashion would encourage consumers to recommend it within social groups where it enjoys good social acceptance. Likewise, it is coherent that a perceived quality, combined with an accurate assessment of the value for money of second-hand fashion, would lead consumers to recommend its purchase. Furthermore, this mediating role suggests that high perceived quality serves as a “social safeguard” for the consumer. In the second-hand fashion context, recommending an item involves a significant risk of social stigma if the product is perceived as old, worn or in poor condition. Therefore, quality acts as a critical prerequisite that mitigates potential social embarrassment, allowing consumers to share their functional, social or economic gains without risking their personal reputation or social standing.

Finally, environmental value has no effect on WOM. However, although the second-hand fashion market is widely promoted as a component of the circular economy and environmental sustainability, empirical evidence suggests a divergence between consumers’ environmental awareness and their actual purchasing decisions. Previous research has shown that economic, social and hedonic motivations often outweigh environmental concerns in the adoption of second-hand fashion (Hur, 2020), and that even consumers with high sustainability awareness do not consistently exhibit more sustainable behaviours, indicating the presence of attitude-behaviour gaps and potential moral licencing or rebound effects (Mizrachi and Sharon, 2025).

This study contributes to the theoretical framework of sustainable consumption by examining how consumer value systems drive WOM intention in the second-hand fashion market retail. In contrast to prior research that has primarily focused on purchase intention, the present analysis advances understanding of the mechanisms that support the growth and scalability of the sector.

First, the study has proposed an extension of the consumer value theory (Sheth et al., 1991) by including economic and environmental value, thus proposing an update more aligned with the most recent social, economic and environmental reality. The research highlights the reliability and validity of consumer value theory for assessing WOM, even though it has traditionally been used to explain purchase intention. The study strengthens our understanding of how consumers’ value systems trigger the intention to generate WOM, rather than purchasing behaviour itself. This distinction is fundamental, as WOM is not only a result of satisfaction but a process driven by a complex configuration of values, which acts as a strategic driver for the scalability of the sector by facilitating the acquisition of new customers through social validation and the reduction of perceived risk.

Secondly, the findings reveal the critical, albeit differentiated, role of perceived quality. Quality is shown to act as a necessary bridging mechanism for values such as functional, economic and social to be transformed into recommendations. However, path coefficients suggest that, although perceived quality is relevant (especially for functional value), there are other factors not explained in this model that could be influencing the decision to recommend, which opens up a priority line for future research.

A particularly revealing finding is the absence of a significant relationship between environmental value and WOM intention. Although the literature often links second-hand fashion with environmental awareness, our results suggest that, for the sample analysed, sustainability is not a primary argument that consumers use to recommend these shops to others. This implies that the growth of the sector may depend more on pragmatic (quality and savings) or social arguments than on mere environmental awareness. This “mismatch” between ecological values and communicative behaviour (WOM) represents a relevant theoretical contribution, suggesting that sustainability in second-hand fashion is perceived as a personal or private benefit, but not necessarily as a driver of social influence.

The results of this study provide strategic guidelines for second-hand fashion resale platforms and retailers seeking to scale their business models through WOM.

First, managers of second-hand fashion companies are advised to prioritise the enhancement of emotional, conditional, and epistemic values. Given that emotional value (30%), conditional value (25%) and epistemic value (14%) emerge as the main direct predictors of WOM, firms should design consumer experiences that go beyond the mere purchase–sale transaction and actively reinforce these dimensions. To enhance emotional value, brands may employ storytelling strategies that highlight the uniqueness of garments, framing the purchase as a “treasure hunt” that generates excitement and enjoyment – emotions that consumers are inclined to share. To activate a conditional value, companies should create and communicate specific contexts of use, such as marketing campaigns oriented towards events (e.g. weddings, themed parties or festivals), in which second-hand fashion is positioned as an appropriate and logical solution to a situational need. Regarding epistemic value, actions aimed at stimulating consumer curiosity and interest may be particularly effective.

Second, the findings indicate that perceived quality strengthens utilitarian benefits, namely functional and economic value. Accordingly, retailers are encouraged to implement rigorous cleaning, inspection and restoration processes. The managerial implication is straightforward: if functional benefits such as durability and good condition are not clearly ensured, consumers are unlikely to risk their social reputation by recommending the retailer.

Third, a particularly relevant implication concerns the role of social value, which activates WOM only when perceived quality is high. This result suggests the need to design and implement strategies aimed at mitigating social stigma through product excellence. For consumers to derive social value from sharing their purchase with their social circle, garments must appear “like new” or exhibit superior design quality, closely resembling the experience of buying new clothing. Eliminating visible traces of prior use is therefore essential to reducing potential social rejection and fostering positive WOM.

Finally, to operationalize these strategies, retailers should foster community-based initiatives. The creation of customer clubs and the appointment of brand ambassadors can serve as powerful tools to amplify the social and epistemic values identified in this study. For instance, exchange events and social dynamics on digital platforms allow ambassadors to showcase the high quality of curated items, providing the necessary “social proof” to mitigate stigma. These community actions transform the individual act of buying second-hand fashion into a shared experience, reinforcing the trust required for consumers to engage in positive WOM.

This study presents several limitations that should be acknowledged in order to lay the groundwork for future research. The first limitation concerns the fact that the recommendations analysed in this study were limited to word-of-mouth (WOM) communication. It is possible that such recommendations are less commonly used today, particularly among younger consumers. This opens several avenues for future research. First, the mediation results for perceived quality suggest that additional factors not captured by the present model may be influencing recommendation behaviour, thereby opening a priority avenue for future research. Second, future studies should examine electronic word-of-mouth (eWOM) to compare findings with traditional WOM. Third, they should investigate the role of social media in shaping consumers’ purchasing decisions. Fourth, given the growing influence of social media in today’s consumption patterns, future research should explore the impact of influencers on the marketing of second-hand fashion.