A Structural Equation Modeling of Loyalty Toward Sustainability Fashion Product Businesses on Social Media Platforms

Abstract

The objectives of this study are to examine the direct relationships among perceived ethics, perceived sustainability, customer trust, customer engagement, and customer loyalty; and to investigate the mediating roles of customer trust and customer engagement in explaining the relationship between ethical and sustainability perceptions and customer loyalty. Using the Stimulus–Organism–Response (SOR) framework and the Theory of Planned Behavior (TPB) as theoretical foundations, this research examines how ethical and sustainability perceptions within social commerce environments influence customers’ psychological states and behavioral responses. A quantitative approach was used, involving data collection from 360 Thai consumers who had previously bought sustainable fashion items through social media. The proposed model was then evaluated using partial least squares structural equation modeling (PLS-SEM). The results suggest that consumers’ evaluations of seller ethics significantly enhance their perceptions of product sustainability, customer trust, and engagement. Furthermore, perceived sustainability of fashion products affects both trust and engagement. Customer trust subsequently promotes both engagement and loyalty; however, customer engagement exhibits the most substantial direct effect on customer loyalty. Mediation analysis confirms the essential functions of trust and engagement in mediating the impacts of ethical and sustainability perceptions on loyalty. These findings highlight the importance of ethical transparency and proactive customer engagement in fostering trust and long-term customer loyalty within social media-based sustainable fashion commerce. Therefore. This study provides both theoretical and practical insights for sustainable fashion enterprises functioning within digital contexts.

1. Introduction

The global fashion industry is facing increasing scrutiny regarding its environmental and social consequences, encompassing significant water usage, greenhouse gas emissions, labor abuses, chemical contamination, and the escalating problem of textile waste [1,2]. Fast fashion has enhanced production and consumption cycles, making clothing cheaper and shorter-lived and thereby amplifying resource use and waste generation [2]. This shift in perspective has spurred increasing advocacy for a transition away from fast fashion, advocating instead for a more sustainable approach predicated on diminished environmental impacts, ethical manufacturing practices, and prolonged product lifespans throughout the value chain [2,3,4]. Sustainable fashion typically encompasses apparel and associated products conceived, manufactured, distributed, and utilized in manners that mitigate adverse environmental consequences while fostering social accountability and transparency [3,4]. These solutions cover the entire consumption process, including decisions made before purchase, how things are used, and what happens after use [4]. Contrasting many other sustainable product categories, sustainable fashion products are strongly associated with personal identity, self-expression, social visibility, and lifestyle signaling. Consumers often evaluate sustainable fashion not only based on environmental concerns, but also according to ethical production practices, transparency, aesthetic value, and social influence communicated through digital platforms [2,3,4]. In addition, the fashion industry has been widely criticized for issues related to fast fashion, labor exploitation, excessive waste generation, and greenwashing practices, making trust and ethical perceptions particularly important determinants of consumer loyalty in sustainable fashion contexts [3,4].

Previous research indicates that sustainable fashion encompasses a diverse array of strategies and business models, such as eco-design, circular business models, and collaborative consumption. The primary goal of these initiatives was to reduce waste and create closed material loops [1,4,5]. Despite increased awareness, a persistent attitude–behavior gap remains evident in the context of sustainable apparel consumption. Although many consumers express concern for environmental and ethical issues, they often place greater importance on price, design, and ease of use, which hinders the broader acceptance of sustainable options in the marketplace [6,7,8,9]. Prior studies have identified several barriers, including higher costs, limited availability, a lack of trustworthy information, fears of greenwashing, and perceived trade-offs in aesthetics or style [7,8,9,10]. These impediments diminish the conversion of favorable dispositions into actual purchases and repeat transactions, thereby curtailing allegiance to sustainable fashion brands [7,8,9].

For specific consumer demographics, sustainability attributes are essential to their self-perception and ethical frameworks, thus cultivating the potential for strong, value-based allegiance to brands that genuinely embody environmental and social responsibility [7,11]. In contrast, other consumers prioritize sustainability only when it seamlessly integrates with design appeal, prevailing fashion, and affordability [3,4,6]. Previous studies have shown that the perceived sustainability of fashion items elevates perceived quality and value by integrating beliefs about eco-friendly materials, ethical production methods, and overall environmental impact. These elements contribute to a greater likelihood of purchase and a higher acceptance of price increases for sustainable apparel [12,13]. Consumers link sustainable clothing with enhanced quality, durability, and comfort; therefore, they demonstrate increased purchase and repurchase intentions, suggesting that sustainability can develop into a key factor in fostering lasting relational outcomes, rather than remaining a niche attribute [12,13,14].

Although prior studies have examined sustainable fashion consumption, most research has primarily focused on purchase intention, environmental attitudes, or general sustainability perceptions, with limited attention to how ethical perceptions of sellers and perceived sustainability jointly influence customer loyalty through relational mechanisms such as trust and customer engagement, particularly in social media environments [12,13]. In addition, previous findings remain inconsistent regarding the processes through which sustainability perceptions are translated into long-term loyalty behaviors [12,13,14]. Furthermore, empirical evidence from developing countries, especially Thailand, remains limited despite the rapid growth of social commerce and increasing consumer awareness of sustainable consumption in the Thai fashion market [15]. Thailand provides an important context for this investigation because social media platforms play a central role in fashion purchasing decisions, while concerns regarding ethical business practices, greenwashing, and sustainability credibility continue to influence consumer trust and engagement [16,17].

From a marketing perspective, the success of sustainable fashion increasingly depends on effective communication, strategic branding, and the development of consumer trust [18]. Studies of sustainable fashion marketing show that how people feel about sustainability has a positive effect on their attitudes toward brands, and that trust is an important part of brand evaluation and electronic word-of-mouth communication, especially on social media [3,14]. Rapid greenwashing is a big threat to consumer trust, which could hurt the fashion industry’s efforts to adopt circular and sustainable practices [8,14,19]. Many researchers stress how important it is to have clear supply chains, reliable certification processes, and standardized sustainability metrics [15]. These things give consumers the power to check claims and make smart choices [14,19,20]. Social media platforms have emerged as a space for disseminating, debating, and amplifying sustainability narratives, ethical assertions, and consumer experiences [21,22]. Social media platforms can potentially increase the likelihood of consumers purchasing sustainable clothing by fostering environmental awareness, leveraging social influence, and providing exposure to sustainability-related messaging [1,3]. Customer engagement, which includes activities such as liking, commenting, sharing, and collaboratively creating sustainable fashion content, functions as a strategic instrument for reinforcing consumer relationships, influencing perceived value, and promoting behavioral loyalty [1,3,4,23].

Furthermore, current research suggests that consumer loyalty to sustainable fashion brands is likely dependent on a combination of demonstrable sustainability accomplishments, ethical conduct, and communication strategies designed to foster trust and engagement [8,19]. The perception of sustainability and ethical practices can augment perceived value and resonate with consumers’ moral and identity-related motivations; simultaneously, trust and engagement—especially within social media platforms—are anticipated to serve as critical relational mechanisms, converting these perceptions into customer loyalty [1,3,8,11,12]. Nevertheless, empirical investigations that consolidate these constructs within a unified structural model, particularly within the context of social media-driven sustainable fashion enterprises, are scarce, thereby highlighting a distinct requirement for additional theory-driven, quantitative research employing structural equation modeling [1,3,4]. This research addresses the existing gap by creating and testing a combined structural model for social media-based sustainable fashion businesses. The objectives of this study are as follows: (1) to examine the direct relationships among perceived ethics, perceived sustainability, customer trust, customer engagement, and customer loyalty; and (2) to investigate the mediating roles of customer trust and customer engagement in explaining the relationship between ethical and sustainability perceptions and customer loyalty. The findings are expected to add to the current research. The study also has practical implications for businesses. Specifically, it will provide actionable insights for sustainable fashion brands on how to build trust, encourage engagement, and develop long-term customer loyalty through ethical practices and effective sustainability communication on social media.

2. Literature Review

2.1. Underpinning Theories: Stimulus–Organism–Response (SOR) and the Theory of Planned Behavior (TPB)

This investigation’s conclusions are principally derived from two theoretical frameworks: the Stimulus–Organism–Response (SOR) model and the Theory of Planned Behavior (TPB). The SOR model suggests that external stimuli, encompassing perceived ethics of sellers and perceived sustainable fashion products, act as triggers that influence internal psychological processes (the organism), such as customer trust and engagement, which subsequently lead to behavioral outcomes like customer loyalty [24]. The results of this research align with this model, indicating that ethical perceptions function as stimuli that reinforce sustainability beliefs, thus fostering trust and engagement, which in turn enhance loyalty. The serial mediation through customer trust and customer engagement supports the organism’s role as a psychological mechanism that converts ethical cues into loyal behaviors, thereby validating prior research that emphasizes engagement as both a mediator and moderator within sustainable consumption contexts [24,25].

The Theory of Planned Behavior (TPB) offers a more comprehensive perspective by elucidating how attitudes, subjective norms, and perceived behavioral control influence consumers’ intentions and actions regarding sustainable fashion [26,27]. Trust, a critical attitudinal component, is influenced by ethical knowledge and social norms, thus bolstering purchase intentions and cultivating loyalty. This study extends the TPB framework by integrating trust and engagement as key mediators that link ethical perceptions with loyalty outcomes, illustrating that positive attitudes alone are insufficient without relational constructs that promote active consumer involvement [26]. Furthermore, the significant impact of engagement on loyalty supports the Theory of Planned Behavior’s assertion that perceived behavioral control is achieved through active consumer participation in sustainability efforts [27].

Although the SOR framework provides a useful explanation of how external stimuli influence internal psychological states and behavioral outcomes [28]. It primarily emphasizes the stimulus–response mechanism and provides a limited explanation regarding the cognitive and normative decision-making processes underlying sustainable consumption behaviors [24,28]. Therefore, this study integrates TPB as a complementary theoretical perspective to explain how ethical and sustainability perceptions shape consumers’ attitudinal evaluations and behavioral involvement in sustainable fashion contexts [24,29].

Within the proposed framework, seller perceived ethics and perceived sustainable fashion products represent external stimuli under the SOR model. While customer trust and customer engagement represent organism-related psychological states that influence customer loyalty as the behavioral response [28]. From the TPB perspective, customer trust reflects consumers’ positive attitudinal evaluations of sustainable fashion brands, based on ethical credibility and sustainability perceptions [28,30]. Customer engagement reflects consumers’ active behavioral involvement and participatory commitment, which is conceptually associated with perceived behavioral control and consumers’ willingness to participate in sustainability-related brand activities [31,32]. Subjective norms are indirectly reflected through social media interactions and sustainability-related social influence embedded within the digital fashion environment [33].

Rather than replacing the established TPB constructs, trust and engagement are conceptualized as relational extensions of TPB that better capture the interactive and socially embedded nature of sustainable fashion consumption in social media contexts [25,34,35]. The integration of SOR and TPB, therefore, advances the literature by combining external stimulus mechanisms with cognitive and relational behavioral processes, providing a more comprehensive explanation of how ethical and sustainability perceptions translate into long-term customer loyalty.

2.2. Perceived Ethics of Sellers (PES)

The ethical conduct of sales personnel significantly shapes customer trust, satisfaction, and loyalty, thereby establishing a crucial basis for enduring customer relationships. Research within the sportswear sector indicates that ethical sales practices positively affect customer trust and satisfaction, which subsequently fosters loyalty; however, the direct influence of ethics on loyalty may be diminished or negligible in the absence of these mediating factors [29,36]. Additionally, ethical marketing strategies contribute to enhanced perceptions of product sustainability and reinforce the sustainability of customer-brand relationships, ultimately resulting in heightened brand loyalty [37]. In addition, ethical behavior cultivates perceptions of corporate social responsibility (CSR) and green marketing, which mediate the relationship between a firm’s ethicality and brand loyalty, thus emphasizing the wider societal impact of ethical practices on consumer commitment [16]. These observations highlight that ethical seller behavior is not merely a moral obligation but also a strategic instrument for cultivating trust and enduring loyalty by aligning with consumers’ values and expectations regarding responsible business practices [16,29,36,37].

2.3. Perceived Sustainable Fashion (PSF)

Perceived sustainable fashion significantly influences customer relationships, both in terms of transactions and beyond, by shaping attitudes, trust, and engagement levels. Studies show that consumers’ views on sustainability impact their emotional and rational involvement with brands in online shopping, with satisfaction and commitment acting as mediators [31]. Within the fast fashion sector, perceptions of sustainability strengthen brand trust and affective brand commitment, which are essential for sustaining enduring consumer relationships [38]. Furthermore, the implementation of sustainable collections bolsters corporate legitimacy via altruistic attributions, thereby improving perceptions of corporate social responsibility, brand trust, and purchase intentions [39]. Consequently, these studies collectively demonstrate that the perception of sustainable fashion acts as a catalyst, promoting positive consumer attitudes and increasing engagement, which in turn strengthens brand connections grounded in environmental and social responsibility [31,38,39].

2.4. Customer Trust (CT)

Customer trust functions as a crucial intermediary, connecting perceived ethics and sustainability with loyalty outcomes. Sellers’ ethical conduct and their perceived commitment to social responsibility significantly bolster trust, which in turn acts as a primary precursor to customer satisfaction and loyalty across diverse sectors, such as sportswear and fast fashion [29,35,36]. Furthermore, trust mediates the impact of perceived firm ethicality on brand loyalty, particularly through corporate social responsibility (CSR) and green marketing endeavors [16]. Trust is a significant factor in whether consumers build lasting relationships with brands, as it reduces the uncertainty often associated with sustainable products [38]. This highlights trust’s crucial role as a cognitive foundation for emotional engagement, thus promoting lasting consumer–brand relationships in ethically conscious markets [16,29,35,36,38].

2.5. Customer Engagement (CE) and Customer Loyalty (CL)

Customer engagement functions as both a consequence of trust and a catalyst that enhances its impact on customer loyalty. The perception of sustainability affects both the emotional and rational facets of engagement, which, in turn, mediate the relationship between sustainability perceptions and trust; however, the mediation effects on loyalty may differ [31,40]. Engagement signifies active consumer involvement in the co-creation of value that aligns with sustainable practices, thereby proving crucial for maintaining enduring brand commitment that extends beyond mere cognitive assessments [12,31,40]. Customer engagement in this study is conceptualized as a multidimensional psychological state reflecting customers’ cognitive, emotional, and behavioral connection with sellers and their purchasing activities. Although some indicators may appear related to satisfaction or loyalty, the selected items were intended to capture customers’ ongoing interaction, emotional attachment, comfort, and active participation toward the seller within the social commerce environment. Prior studies have suggested that customer engagement may include relational and experiential dimensions that extend beyond transactional behavior and reflect deeper customer involvement with the seller and platform [29,32].

Loyalty is strengthened not only through direct effects of ethics and sustainability but also via these psychological processes involving trust and engagement. Empirical studies indicate that customer loyalty is significantly influenced by satisfaction, trust, and engagement, which together cultivate affective commitment. This commitment is grounded in shared values concerning ethics and sustainability [29,31,36,37,40]. Consequently, this holistic viewpoint underscores the necessity of cultivating engaged communities to secure a lasting competitive edge within sustainable fashion markets.

2.6. Hypotheses Development

Based on existing research on ethical behavior, views on sustainability, trust, engagement, and loyalty, the following hypotheses are proposed in Figure 1. Consumer perceptions of a seller’s ethical behavior often influence how sustainable consumers evaluate the products they consider [29]. When sellers are viewed as honest, socially responsible, and environmentally conscious, customers are more likely to see fashion products as genuinely sustainable, rather than as examples of greenwashing [19,41]. Research indicates that consumer perceptions of a brand’s environmental friendliness and responsibility are positively influenced by ethical conduct and transparent sustainability initiatives [40,42,43]. Therefore,

H1. 

Perceived ethics of sellers has a positive influence on perceived sustainable fashion products.

Ethical behavior is a crucial factor in building customer trust (CT). Research across sectors, including the sportswear industry, indicates that vendors’ ethical behavior significantly enhances customer trust, which in turn strongly influences customer loyalty [22,29]. Similarly, the ethical standards a brand is perceived to have in e-commerce are closely related to trust in that brand [43]. Therefore, we propose that

H2. 

Perceived ethics of sellers has a positive effect on customer trust.

Furthermore, ethical and responsible practices can foster deeper customer engagement (CE) by communicating shared values and a commitment to long-term objectives. Studies combining brand ethics with engagement models show that ethical perceptions improve engagement by enhancing customer experience and building trust [43]. Furthermore, sustainability-related cues and ethical practices increase both cognitive and emotional involvement with brands and platforms [31,40,42]. Therefore, the following hypothesis is proposed:

H3. 

Perceived ethics of sellers has a positive effect on customer engagement.

Furthermore, perceiving a fashion product as sustainable (PSF) can, in turn, build trust. Positive views of product sustainability and companies’ environmental responsibility increase brand or retailer trust and perceived credibility in fashion and retail [22,40,42]. Consequently, the following hypothesis is proposed:

H4. 

Perceived sustainability fashion product has a positive influence on customer trust.

Furthermore, perceived sustainability cultivates heightened customer engagement. In both online and physical retail environments, the perceived sustainability of products and services correlates with increased satisfaction and commitment, which, in turn, fosters both rational and emotional engagement [31,40,42]. In accordance with these studies, the hypothesis is proposed.

H5. 

Perceived sustainable of fashion products has a positive influence on customer engagement.

A substantial body of relationship marketing research underscores that trust serves as a fundamental driver of engagement behaviors. While some research on digital platforms indicates that the direct effect of trust on engagement might be less significant than that of experience, trust still fosters deeper, more lasting brand interactions [40,43]. Therefore, the hypothesis is demonstrated as follows;

H6. 

Customer trust has a positive effect on customer engagement.

Furthermore, trust serves as a potent direct predictor of customer loyalty (CL). Research on sportswear, sustainable brands, and secondhand clothing consistently shows that higher levels of trust are associated with greater customer loyalty and positive word of mouth [29,37,40,44]. Based on previous studies, the hypothesis is presented as follows:

H7. 

Customer trust has a positive influence customer loyalty.

Finally, customer engagement—cognitive, emotional, and behavioral involvement with a brand—strongly predicts loyalty outcomes. Research on sustainability and e-commerce indicates that heightened engagement substantially enhances both customer loyalty and the likelihood of repeat purchases [31,40,43,45]. Therefore, it is hypothesized that

H8. 

Customer engagement has a positive effect on customer loyalty.

3. Methods

3.1. Research Context

This study used a quantitative, cross-sectional survey to examine the connections between how consumers view the ethics of sellers, how they perceive the sustainability of fashion products, and how these factors relate to customer trust, engagement, and loyalty towards sustainable fashion brands on social media in Thailand. Data were collected using a self-administered questionnaire, distributed online through popular social media platforms (like Facebook, Instagram, TikTok, and Line Official), where sustainable fashion brands actively interact with their customers. To test the proposed model and its hypotheses, structural equation modeling (SEM). Partial least squares structural equation modeling (PLS-SEM) was performed using SmartPLS version 4 (SmartPLS GmbH, Oststeinbek, Germany).

3.2. Sampling and Data Collection

The participants in this study were Thai consumers who had previously purchased sustainable fashion items through social media. To ensure participants had relevant experience, we required that they had bought sustainability fashion products on social media at least three times in the year before the study. We used non-probability purposive sampling because the research focused on a specific group of consumers with relevant purchasing habits. The data was collected by identifying popular sustainability fashion brands and sellers on social media in Thailand from August to October 2025. The questionnaire was shared through a survey link in online communities and groups that focus on sustainable consumption, eco-fashion, and ethical lifestyles. The questionnaire began with screening questions to confirm that participants (1) lived in Thailand, and (2) had purchased sustainability fashion products on social media at least three times. For example, sustainable fashion products were identified based on commonly recognized characteristics discussed in the sustainability and ethical consumption literature, including environmentally friendly materials, recycled or organic fabrics, ethically produced apparel, and products marketed by brands with sustainability commitments. In total, 450 questionnaires were distributed, and we received 420 responses. After data cleaning and removal of incomplete or ineligible responses, 360 valid questionnaires remain for further analysis (effective response rate of approximately 85.71%). In accordance with [46], the minimum sample size requisite for PLS-SEM was ascertained through the application of the 10-times rule and a power analysis, predicated on the maximum quantity of structural paths directed toward a latent construct. The maximum number of structural paths (arrows) pointing at a single endogenous construct in the research model is eight. Based on [46] the 10-times rule, the minimum required sample size was therefore 10 × 8 = 80 cases. In addition, a statistical power analysis was conducted using G*Power version 3 (Heinrich Heine University Düsseldorf, Düsseldorf, Germany). The analysis employed the following parameters: significance level (α) = 0.05, statistical power = 0.80, and medium effect size (f² = 0.15), with eight predictors for the most complex endogenous construct. The results indicated that the minimum recommended sample size was eighty. Accordingly, the actual sample size used in this study (n = 360) substantially exceeds both the minimum requirement suggested by the 10-times rule and the power analysis threshold, confirming that the sample size is adequate for PLS-SEM analysis. Regarding the 360 completed questionnaires, 60% were completed by females and 40% by males. Most of the respondents were aged in the range of 20–30 years old (34.6% and 41–50 years old = 33.7%). 87.1% were undergraduates, while only 8.2% had a graduate-level education. The largest group of respondents spent money on sustainability-focused fashion products on social media 3 to 5 times per month (74.8%) and 6 to 8 times per month (21%). Of the 360 questionnaires, 31.5% of respondents reported an average monthly income of USD 800–1200, while 23.6% reported earning less than USD 1200–1600.

3.3. Instrument Development

Each construct within the model was operationalized as a multi-item latent variable, with each variable being assessed via reflective indicators. Measurement items were derived from established scales within the existing literature concerning business ethics, sustainable fashion, trust, engagement, and loyalty, and subsequently adapted to align with the specific context of sustainability-focused fashion products on social media platforms within Thailand.

All items were assessed on a five-point Likert scale, with endpoints 1 (strongly disagree) and 5 (strongly agree). The questionnaire’s design incorporated three principal sections: preliminary screening inquiries regarding purchasing habits linked to sustainability-oriented fashion products on social media, specifically focusing on frequency and platform utilization. Construct measures encompassed the perceived ethics of sellers, which evaluated consumers’ perceptions of the honesty, fairness, transparency, and responsibility demonstrated by sustainability fashion sellers on social media. These items were derived from established scales concerning perceived seller ethics and ethical climate within online retailing [47,48] and were tailored to the sustainability fashion domain. Furthermore, perceived sustainability of fashion products assessed consumers’ evaluations of the environmental and social sustainability attributes of the products they had acquired (e.g., eco-friendly materials, fair labor practices, durability, and diminished environmental impact). These items were adapted from prior measures of perceived product sustainability and sustainable fashion consumption [49,50]. Customer trust was measured by the extent to which consumers trusted sustainability fashion sellers on social media, focusing on reliability, integrity, and benevolence [51,52]. Customer engagement reflected consumers’ cognitive, emotional, and behavioral involvement with sustainability fashion brands and sellers on social media, encompassing interaction, participation [17,53]. Customer loyalty items were generated from previous studies [18,24,54] comprising four items. Furthermore, customer loyalty was treated as a distinct construct representing customers’ long-term commitment and repurchase intention, whereas customer engagement reflects the psychological and behavioral involvement occurring during customer–seller interactions. The measurement items for customer engagement were adapted from previously validated scales in related consumer engagement and social commerce studies, thereby supporting the construct validity of the operationalization [17,53].

The initial measurement items were derived from English-language sources, and a systematic translation process was implemented to ensure both linguistic and conceptual equivalence within the Thai context. A standard forward–backward translation procedure was utilized. The questionnaire was first translated from English into Thai by a bilingual academic with expertise in marketing and consumer behavior. A second, independent bilingual expert, who had not seen the original instrument, back-translated the Thai version into English. The two English versions were then compared to find any differences. These differences were discussed with the translators and the research team to resolve any meaning issues. Minor changes were made to improve clarity, cultural relevance, and readability. Content validity was evaluated by a panel of 4–6 experts in marketing, sustainability, and social media commerce. They assessed each item for its relevance, clarity, and how well it represented the underlying concept. After the expert review, some item wording was improved, and a few redundant elements were either removed or combined. A pilot test was then given to about 30 consumers, all of whom met the target population criteria. Participants were asked to complete the questionnaire and provide feedback on item clarity, length, and overall understanding. A preliminary reliability analysis using Cronbach’s alpha was conducted to identify problematic items. Items with low correlations or that reduced the scale’s internal consistency were either revised or removed before the main data collection began.

3.4. Data Analysis

Data screening and preliminary analyses were conducted employing IBM SPSS Statistics version 29 (IBM Corp., Armonk, NY, USA) or an equivalent statistical software application. This procedure encompassed the evaluation of missing data, the identification of outliers, the assessment of normality, and the examination of common method bias, including the application of Harman’s single-factor test. Descriptive statistics were computed to offer a summary of the participants’ demographic characteristics and their purchasing behaviors concerning sustainability-oriented fashion products on social media platforms. To assess the proposed research model and its corresponding hypotheses, PLS-SEM was utilized, specifically employing SmartPLS (version 4). The choice of PLS-SEM was determined by several considerations, as detailed by [46,55]. This research, which is predictive in nature, seeks to elucidate the factors contributing to variations in customer loyalty within the context of sustainable fashion product enterprises. The proposed model includes several hidden factors and mediating variables, specifically customer trust and customer engagement, which makes it suitable for Partial Least Squares Structural Equation Modeling (PLS-SEM). PLS-SEM is particularly useful for analyzing data that is not normally distributed and complex models, especially when working with samples of a moderate size.

The assessment involved an evaluation of reliability (Cronbach’s alpha, CR), convergent validity (factor loadings, AVE), and discriminant validity (Fornell–Larcker criterion) [56]. Items exhibiting factor loadings between 0.40 and 0.70 may be retained when AVE and composite reliability remain within acceptable thresholds, consistent with established PLS-SEM guidelines. Since no items exhibited critically low loadings requiring deletion, no pre- and post-elimination comparisons were applicable in this study. Furthermore, convergent validity and discriminant validity were established using the retained indicators, as reflected in the reported AVE, composite reliability, Fornell–Larcker criterion, and HTMT results in the revised manuscript. Following the establishment of an acceptable measurement model, the structural relationships among the constructs were subsequently examined. Path coefficients, t-values, and p-values were calculated using a bootstrapping method, incorporating a minimum of 5000 resamples to evaluate the significance of the proposed pathways. The mediating roles of customer trust and customer engagement in the relationship between perceived ethics/perceived sustainability and customer loyalty were evaluated through the analysis of indirect effects and their associated confidence intervals, which were determined via bootstrapping. This methodological approach enabled a comprehensive evaluation of the constructs’ measurement properties and a detailed investigation of the proposed structural relationships within the context of customer loyalty toward sustainability fashion product businesses on social media platforms in Thailand.

4. Results

4.1. Measurement Model

Confirmatory factor analysis confirmed that all model constructs showed good reliability and convergent validity, which is consistent with current best practices for evaluating scales [46,57,58]. All standardized factor loadings were above 0.70 (ranging from about 0.708 to 0.883), indicating that each item was a strong indicator of its underlying construct and met the recommended thresholds for indicator reliability [58,59] as shown in

Table 1. Results of confirmatory factor analysis for measurement model.

Constructs	Items	Factor Loading	t Value	Cronbach’s Alpha (α)	Dijkstra-Henseler’s Rho (ρA)	Jöreskog’s Rho (ρc)	AVE	VIF
Perceived ethics of sellers (PES)	The seller provides accurate and sufficient product information.	0.846	52.712	0.901	0.901	0.927	0.717	2.402
	The seller communicates honestly without exaggeration.	0.883	65.866					2.965
	The product quality matches the description/advertisement.	0.846	46.474					2.499
	The seller treats all customers equally.	0.883	62.953					2.877
	The seller protects customers’ personal data.	0.770	28.595					1.722
Perceived sustainability fashion product (PSF)	This fashion product is environmentally friendly.	0.827	44.511	0.816	0.817	0.879	0.644	2.356
	This product is made using sustainable materials or processes.	0.810	41.092					2.293
	This product is produced in a socially responsible way.	0.798	31.859					1.708
	This product supports long-term environmental sustainability.	0.776	29.775					1.645
Customer trust (CT)	I trust that the product will be of good quality.	0.714	17.261	0.733	0.742	0.832	0.554	1.557
	I trust the seller provides good warranty and after-sales service.	0.708	18.010					1.543
	I trust the seller is careful and responsible in selling.	0.803	41.512					1.569
	I trust that transactions are secure and my data is protected.	0.748	27.097					1.462
Customer engagement (CE)	I am satisfied with my purchase experience.	0.784	32.395	0.837	0.841	0.885	0.606	1.835
	I have a long-term relationship with the seller.	0.804	35.748					1.896
	I appreciate the seller’s ability to provide clear information.	0.824	45.239					2.048
	I tend to buy when the seller offers new products.	0.723	22.049					1.680
	I feel comfortable every time I purchase.	0.751	27.177					1.766
Customer loyalty (CL)	I intend to purchase sustainability fashion products via social media platforms.	0.773	36.217	0.778	0.793	0.856	0.598	1.580
	I share positive purchase experiences with others.	0.844	49.551					1.870
	I encourage my family to buy sustainability fashion products via social media.	0.757	23.342					1.843
	I recommend others (e.g., friends) to buy sustainability fashion products via social media.	0.713	20.932					1.725

Abbreviations: PES = Perceived ethics of sellers; PSF = Perceived sustainability fashion product; CT = Customer trust; CE = Customer engagement; CL = Customer loyalty.

. This suggests that the items related to perceived seller ethics, perceived sustainability of fashion products, customer trust, customer engagement, and customer loyalty all significantly contributed to their respective factors. Internal consistency was found to be satisfactory across all assessed constructs. Cronbach’s alpha coefficients, ranging from approximately 0.73 to 0.90, alongside Dijkstra–Henseler’s rho (ρA ≈ 0.74–0.90) and composite reliability (Jöreskog’s rho, ρc ≈ 0.83–0.93), all exceeded the 0.70 benchmark. This particular threshold is frequently utilized to indicate acceptable or good reliability within the framework of structural equation modeling [46,57,60].

These results imply that the multi-item scales produce stable and consistent scores and are suitable for use in hypothesis testing. Convergent validity was further substantiated, as evidenced by average variance extracted (AVE) values exceeding 0.50 across all constructs (approximately 0.55–0.72). This implies that each latent variable explains more than half of the variance in its related indicators [57,61,62]. The variance inflation factor (VIF) values were all less than 3, demonstrating that multicollinearity did not appear to be an issue. The measurement model results show that the scales measuring ethics, sustainability perceptions, trust, engagement, and loyalty have strong psychometric properties, comparable to established instruments used in recent structural equation modeling (SEM) studies [57,58,59].

4.2. Discriminant Validity (Fornell–Larcker Criterion)

Table 2. Discriminant validity using Heterotrait-monotrait ratio (HTMT) and Fornell-Lacker criterion.

Heterotrait-Monotrait Ratio (HTMT)
Construct	CE	CL	CT	PES	PSF
CE
CL	0.888
CT	0.848	0.854
PES	0.841	0.710	0.792
PSF	0.864	0.751	0.873	0.889
Fornell-Lacker Criterion
Construct	CE	CL	CT	PES	PSF
CE	0.778
CL	0.741	0.773
CT	0.755	0.659	0.744
PES	0.735	0.618	0.659	0.847
PSF	0.719	0.614	0.708	0.780	0.803

The bold and italicized numbers are the square root of the AVE of the construct (√AVE).

shows the evaluation of discriminant validity, using the Heterotrait–Monotrait ratio (HTMT) and Fornell–Larcker criterion [56]. The diagonal elements, highlighted in bold, represent the square roots of the average variance extracted (AVE) for each construct. The off-diagonal elements show the correlations between the different constructs. First, discriminant validity was further assessed using the Heterotrait–Monotrait ratio (HTMT), as recommended in recent PLS-SEM literature. The HTMT values ranged from 0.710 to 0.889, which were below the recommended threshold of 0.90, indicating acceptable discriminant validity among the constructs [63]. Although several construct pairs, such as customer engagement–customer loyalty (0.888) and perceived ethics of sellers—perceived sustainability fashion product (0.889), showed relatively high HTMT values, they remained within the acceptable limit. Furthermore, the HTMT value between customer engagement and customer trust was 0.848, confirming that the two constructs are empirically distinct. Therefore, the results demonstrate adequate discriminant validity for the measurement model.

Discriminant validity is confirmed when a construct’s square root of AVE surpasses its correlations with all other constructs, thereby signifying that the construct accounts for a greater proportion of variance with its own indicators than with other latent variables [62,64,65,66]. Specifically, the square root of AVE for customer engagement (CE = 0.778), customer loyalty (CL = 0.773), customer trust (CT = 0.744), perceived ethics of sellers (PES = 0.847), and perceived sustainable fashion product (PSF = 0.803) all surpass the corresponding correlations within their respective rows and columns.

For example, perceived ethics of sellers has √AVE = 0.847, which is higher than its correlations with customer engagement (0.735), customer loyalty (0.618), customer trust (0.659), and perceived sustainable fashion product (0.780); similarly, perceived sustainable fashion product has √AVE = 0.803, which is greater than its correlations with customer engagement (0.719), customer loyalty (0.614), customer trust (0.708), and perceived ethics of sellers (0.780). The observed pattern satisfies the Fornell–Larcker criterion, thereby suggesting that each construct is empirically separate from the others [62,64]. These results, in conjunction with satisfactory AVE and reliability metrics, corroborate the discriminant validity of the five latent constructs within the measurement model. Consequently, these findings align with the established methods for validating Structural Equation Modeling (SEM) scales used in management and marketing research [57,62,64,65,66].

4.3. Hypotheses Testing

The structural model’s findings indicate that all proposed relationships are statistically significant, as detailed in

Table 3. The results of hypotheses testing.

Hypothesis	Path	Beta (β)	t-Statistic	f-Square	Results
H1	PES → PSF	0.780	32.610 **	1.550	Supported
H2	PES → CT	0.274	3.880 **	0.063	Supported
H3	PES → CE	0.329	5.799 **	0.124	Supported
H4	PSF → CT	0.495	7.156 **	0.205	Supported
H5	PSF → CE	0.163	2.878 *	0.027	Supported
H6	CT → CE	0.422	8.634 **	0.259	Supported
H7	CT → CL	0.232	4.170 **	0.054	Supported
H8	CE → CL	0.565	10.391 **	0.321	Supported
R2 for customer trust = 0.531; R2 for customer engagement = 0.678;
R2 for customer loyalty = 0.572;
Q2 for customer trust = 0.428; Q2 for customer engagement = 0.536;
Q2 for customer loyalty = 0.375.

* Significant at p < 0.05, ** Significant at p < 0.001.

and Figure 2. The coefficient of determination (R²) was assessed to evaluate the explanatory power of the structural model. The results showed that customer trust achieved an R² value of 0.531, indicating that 53.1% of the variance in customer trust was explained by the predictor constructs. Customer engagement reported an R² value of 0.678, demonstrating substantial explanatory power, while customer loyalty achieved an R² value of 0.572, indicating moderate explanatory power. According to the recommended thresholds for PLS-SEM, these findings suggest that the proposed model has satisfactory predictive capability and explanatory relevance. The f-square (f²) results indicate the effect size of each predictor construct on the endogenous variables in the structural model. f² values of 0.02, 0.15, and 0.35 represent small, medium, and large effects, respectively. The results show that PES has a large effect on PSF (f² = 1.550), while CE has a large effect on CL (f² = 0.321), whereas the remaining relationships demonstrate small to medium effect sizes, indicating varying levels of practical significance among the proposed relationships. The Q² values for customer trust (0.428), customer engagement (0.536), and customer loyalty (0.375) are all greater than zero, indicating that the structural model has satisfactory predictive relevance for the endogenous constructs. Perceived ethics of sellers exhibits a robust positive impact on the perceived sustainability of fashion products (H1: β = 0.780, t = 32.610, p < 0.001), thereby suggesting that ethical seller conduct significantly bolsters customers’ perceptions of product sustainability. Additionally, Perceived ethics of sellers affect customer trust (H2: β = 0.274, t = 3.880, p < 0.001) and customer engagement (H3: β = 0.329, t = 5.799, p < 0.001). In addition, perceived sustainability of fashion products demonstrates significant positive effects on customer trust (H4: β = 0.495, t = 7.156, p < 0.001) and customer engagement (H5: β = 0.163, t = 2.878, p < 0.05), implying that when customers perceive fashion products as sustainable, they are more inclined to trust the seller and engage with the brand. customer trust, in turn, positively influences customer engagement (H6: β = 0.422, t = 8.634, p < 0.001) and customer loyalty (H7: β = 0.232, t = 4.170, p < 0.001). Ultimately, customer engagement exhibits the most pronounced direct effect on customer loyalty (H8: β = 0.565, t = 10.391, p < 0.001), suggesting that customers who are actively engaged are more likely to demonstrate loyalty towards sustainable fashion enterprises on social media platforms.

4.4. The Results of Mediating Effects

Table 4. Mediating effects.

Path	Indirect Effect	SE	t Value	p Value	LLCI	ULCI
PES → CT → CE	0.116	0.031 ***	3.747	0.000 ***	0.058	0.178
PES → PSF → CT	0.386	0.057 ***	6.800	0.000 ***	0.280	0.497
PES → CT → CL	0.064	0.022 *	2.943	0.003 *	0.026	0.111
PES → CE → CL	0.186	0.037 ***	5.039	0.000 ***	0.118	0.263
PES → PSF → CE	0.127	0.044 *	2.863	0.004 *	0.043	0.217
PSF → CT → CE	0.209	0.040 ***	5.199	0.000 ***	0.134	0.291
PSF → CT → CL	0.115	0.034 ***	3.414	0.001 ***	0.054	0.188
PSF → CE → CL	0.092	0.033 *	2.777	0.006 *	0.030	0.161
CT → CE → CL	0.239	0.037 ***	6.497	0.000 ***	0.170	0.314
SERIAL MEDIATION
PES → PSF → CE → CL	0.072	0.026 *	2.762	0.006 *	0.024	0.126
PES → CT → CE → CL	0.065	0.019 ***	3.369	0.001 ***	0.031	0.106
PES → PSF → CT → CL	0.090	0.027 ***	3.331	0.001 ***	0.042	0.148
PES → PSF → CT → CE	0.163	0.032 ***	5.081	0.000 ***	0.104	0.230
PSF → CT → CE → CL	0.118	0.025 ***	4.659	0.000 ***	0.073	0.172
PES → PSF -> CT → CE → CL	0.092	0.020 ***	4.555	0.000 ***	0.057	0.135

Note: * p < 0.05; *** p < 0.001; SE = Standard error; LLCI = Lower-level confident interval; ULCI = Upper-level confidence interval.

shows the indirect and serial mediation effects, estimated using bootstrapping. All indirect effects are statistically significant, as indicated by t-values greater than 1.96, p-values less than 0.05, and 95% confidence intervals that do not include zero. This supports the recommended practices for mediation analysis in SEM [67,68,69,70]. Regarding single-mediator and parallel mediation paths, ethical perceptions of sellers (PES) significantly increase customer engagement through customer trust (PES → CT → CE: β = 0.116, p < 0.001) and through perceived sustainable fashion products (PES → PSF → CE: β = 0.127, p < 0.05). Therefore, the data suggests that the perceived ethics of sellers enhances engagement by increasing both trust and perceptions of sustainability [24,29,31,40]. PES also indirectly strengthens customer loyalty through both customer trust (β = 0.064, p < 0.05) and customer engagement (β = 0.186, p < 0.001). This suggests that ethical sellers primarily build loyalty by fostering trust and encouraging active engagement, as demonstrated by [21,24,29,32,40]. Similarly, perceived sustainable fashion products demonstrably affect customer engagement indirectly through customer trust (PSF → CT → CE: β = 0.209, p < 0.001) and customer loyalty via both customer trust (β = 0.115, p < 0.001) and customer engagement (β = 0.092, p < 0.05). This underscores the significance of trust and engagement as essential mechanisms by which sustainability perceptions influence loyalty, a finding corroborated by [24,31,40,71]. Consequently, CT indirectly promotes customer loyalty via customer engagement (CT → CE → CL: β = 0.239, p < 0.001), thus confirming engagement as a direct factor in loyalty, a finding corroborated by [31,40].

The mediation analysis results support a process model. This model shows that perceived ethics of sellers and perceived sustainable fashion products mainly affect loyalty indirectly, through trust and engagement. This finding aligns with current recommendations that emphasize the importance of understanding indirect relationships.

5. Discussion

The results of this investigation offer compelling support for the assertion that perceived ethics of sellers and perceived sustainable fashion product apply a significant impact on customer trust, engagement, and, consequently, customer loyalty within the sustainable fashion domain. The substantial direct effect of perceived ethics of sellers on perceived sustainable fashion product suggests that consumers place considerable reliance on ethical signals from sellers when evaluating the sustainability of fashion products, thereby corroborating previous studies that have highlighted the significance of ethical branding in shaping sustainability perceptions and trust [25,35,39]. The notable mediating roles of customer trust and customer engagement between perceived ethics of sellers and perceived sustainable fashion product and customer loyalty emphasize a multi-stage process wherein ethical perceptions cultivate trust, which subsequently drives engagement, ultimately resulting in loyalty. This result corroborates earlier studies that have identified trust as a critical determinant of engagement and loyalty in the realm of sustainable consumption [24,38,42]. Furthermore, the serial mediation analyses provide evidence that sustainability perceptions do not exert a direct influence on loyalty; rather, they function through relational mechanisms, thereby reinforcing the stimulus–organism–response model utilized in sustainable fashion marketing [24].

This study contributes to the existing literature by incorporating ethics, sustainability perception, trust, engagement, and loyalty into a holistic model, which was subsequently tested using robust SEM analysis. In contrast to previous studies that have primarily concentrated on direct effects or purchase intention [8,35]. These findings illuminate the intricate indirect pathways through which ethical and sustainability cues foster enduring customer relationships.

The results further build upon prior investigations by showing that customer engagement functions not merely as a mediating factor but also as a more potent predictor of loyalty compared to trust alone, thereby emphasizing the active role consumers assume in co-creating brand value through their participation and interaction [3,24]. This finding corroborates the need for more dynamic models of sustainable consumption that integrate psychological states, such as engagement, alongside cognitive assessments like trust [3,18].

This study adds to the existing research on sustainable marketing by providing empirical evidence for a multi-step process. The relationship between a seller’s ethical behavior and customer loyalty is strengthened by how customers view sustainability, their level of trust, and their engagement. This study highlights the critical role of trust as a basic relationship element in sustainable fashion consumption. It also shows that customer engagement is a key psychological factor in building loyalty. This integrated methodology enhances established frameworks, including the value-belief-norm theory and the stimulus–organism–response model, by explicitly incorporating ethical dimensions as determinants of sustainability beliefs and the cultivation of relationships [24,51]. Additionally, the research highlights the significance of using serial mediation analyses to better understand the complex ways consumers make decisions, which are often neglected in simpler models [35]. Therefore, this study connects ethical branding research with consumer behavior theories, particularly in areas related to sustainability. These findings offer useful insights for fashion brands aiming to improve their sustainability efforts and build customer loyalty. Given the significant influence of PES on how consumers view sustainability and the resulting development of trust, transparent ethical practices are crucial.

To address skepticism and apprehensions related to greenwashing, as highlighted in prior studies [8], brands should explicitly articulate their ethical principles. Nevertheless, trust-building is insufficient on its own; brands must actively engage customers through interactive platforms, such as partnerships with social media influencers or participatory campaigns aimed at fostering emotional bonds with sustainability values [24]. Engagement strategies might involve storytelling focused on eco-friendly production processes or co-creation initiatives that position consumers as active participants in sustainability efforts. Given the significant impact of customer engagement (CE) on brand loyalty, the establishment of enduring, reciprocal communication channels is likely to enhance customer retention.

Although the relationship between perceived sustainable fashion products and customer engagement was statistically significant, the effect size was relatively weaker compared to the effects of customer trust and seller perceived ethics [7,20,29]. Sustainability perceptions alone may not be sufficient to generate active consumer engagement unless consumers also develop emotional trust and relational attachment toward the brand or seller [31,53]. In the social media environment, engagement behaviors such as sharing, commenting, or participating in brand activities may depend more strongly on relational and emotional factors than on sustainability evaluations alone [18]. Therefore, sustainability perceptions may function more indirectly through trust-building mechanisms rather than acting as a direct driver of engagement.

In addition, the very strong relationship between seller perceived ethics and perceived sustainable fashion products suggests that consumers may closely associate ethical seller behavior with the sustainability credibility of fashion products [22,30]. While this finding supports prior literature emphasizing the importance of ethical signaling in sustainable consumption, it may also indicate partial conceptual proximity between the two constructs in consumers’ perceptions [29].

From a sustainability standpoint, this research highlights how the integration of ethical considerations into marketing strategies can promote responsible consumption patterns by fortifying consumer–brand relationships. Sustainable fashion brands can leverage these insights to align their business goals with environmental stewardship by fostering authentic relationships grounded in shared values, rather than depending exclusively on transactional exchanges [3,25]. This strategy is consistent with circular economy tenets, as it encourages enduring brand loyalty. Consequently, this could mitigate the swift turnover characteristic of fast fashion, thereby diminishing its detrimental environmental impacts. Furthermore, the incorporation of these relational mechanisms aligns with the global Sustainable Development Goals (SDGs), specifically SDG 12, which addresses responsible consumption and production [42]. Consequently, the practical application of these findings can significantly contribute to both corporate success and wider ecological aims.

5.1. Theoretical Contributions

This study extends the Stimulus–Organism–Response (SOR) theory by empirically validating a multi-stage model where perceived ethics of sellers (PES) and perceived sustainable fashion (PSF) serve as stimuli that activate internal psychological states—customer trust (CT) and customer engagement (CE)—which then lead to the behavioral response of customer loyalty (CL). Unlike traditional SOR applications that often treat engagement as a passive outcome, this research positions engagement as an active psychological mechanism that amplifies trust’s effect on loyalty, highlighting a more dynamic organism phase in sustainable fashion consumption [24]. The serial mediation through PSF → CT → CE pathways reveals complex indirect effects, encouraging future research to adopt sophisticated modeling techniques to capture these nuanced consumer decision processes in high-involvement sustainability contexts [33,72]. Regarding the Theory of Planned Behavior (TPB), this study enriches the framework by integrating ethical perceptions and relational constructs such as trust and engagement into the attitude-intention-behavior sequence. While TPB traditionally emphasizes attitudes, subjective norms, and perceived behavioral control as predictors of intention, this research demonstrates that trust acts as a foundational attitudinal construct mediating between ethics/sustainability perceptions and downstream outcomes like engagement and loyalty. Engagement further extends TPB by embodying perceived behavioral control through consumers’ participatory involvement in co-creating sustainable brand value, thus bridging cognitive intentions with affective and behavioral commitment [26,27,30]. This integration addresses limitations in prior TPB applications by capturing how ethical considerations translate into durable consumer–brand relationships beyond mere purchase intentions.

The SOR framework serves as the primary theoretical foundation explaining the process through which external sustainability-related stimuli (seller perceived ethics and perceived sustainable fashion) influence internal psychological states (customer trust and customer engagement), which subsequently lead to behavioral responses (customer loyalty) [28]. Meanwhile, TPB is used as a complementary perspective to explain how ethical and sustainability perceptions shape consumers’ attitudinal and behavioral decision-making processes in the social media environment [30]. Specifically, trust reflects consumers’ attitudinal evaluation toward sustainable fashion brands, whereas engagement represents consumers’ active behavioral involvement and participatory commitment toward the brand [21,36]. Therefore, the integration of SOR and TPB provides a more comprehensive explanation of how ethical and sustainability perceptions translate into long-term customer loyalty.

Furthermore, these theoretical extensions provide a comprehensive understanding of sustainable fashion consumption. The SOR framework captures the stimulus-driven activation of trust and engagement leading to loyalty, while TPB explains how ethical attitudes and social norms shape these internal states [73]. By situating trust and engagement within both models, the study highlights their critical roles in fostering long-term loyalty grounded in shared ethical values. This dual-theoretical approach advances sustainable marketing literature by emphasizing multi-dimensional psychological pathways essential for cultivating authentic consumer–brand relationships in sustainability-driven markets [25,28,41].

5.2. Practical Contributions

The practical implications derived from this study are highly relevant for managers seeking to enhance both business performance and environmental responsibility within the fashion industry. First, emphasizing ethical transparency is critical because perceived ethics strongly shape consumers’ assessments of product sustainability—a prerequisite for building trust essential for long-term relationships [35,39]. Brands should invest in clear communication about sourcing practices, labor conditions, environmental impact reduction efforts, and certifications to mitigate skepticism or greenwashing concerns prevalent among conscious consumers [8]. Second, fostering customer trust requires consistent demonstration of genuine commitment beyond superficial claims; this includes third-party verifications or storytelling highlighting authentic efforts toward social responsibility [25,39]. Trust-building initiatives should be integrated into all touchpoints including online platforms where consumers increasingly seek credible information before purchase decisions. Third, brands must prioritize customer engagement strategies that encourage active participation rather than passive consumption. Social media influencer partnerships aligned with brand values can create emotional resonance around eco-conscious narratives, facilitating deeper psychological connections with customers [24]. Interactive campaigns inviting user-generated content or co-creation opportunities empower consumers as collaborators in advancing sustainability goals—strengthening both engagement levels and brand advocacy. Fourth, given customer engagement’s strong influence on loyalty, companies should develop ongoing relationship management programs leveraging digital tools such as apps or communities focused on education about circular economy practices or product lifecycle transparency. These initiatives not only enhance retention but also promote responsible consumption patterns aligned with Sustainable Development Goal 12: Responsible Consumption and Production [42]. Finally, embedding these insights into corporate strategy helps reconcile profitability with ecological stewardship by shifting focus from short-term transactions toward cultivating enduring brand relationships grounded in shared ethical values. Such alignment supports broader industry transformation toward circularity while meeting growing consumer demand for authenticity in sustainable fashion offerings.

6. Conclusions

This study examined the factors associated with customer loyalty toward sustainable fashion brands in the social media context. The findings indicate that seller perceived ethics and perceived sustainable fashion are positively related to customer trust and customer engagement, which are in turn associated with customer loyalty among Thai consumers purchasing sustainable fashion products through social media platforms. The results also highlight the important roles of trust and engagement as relational mechanisms linking ethical and sustainability perceptions with loyalty outcomes.

The study contributes to the sustainable fashion and consumer behavior literature by integrating ethics, sustainability perception, trust, engagement, and loyalty within a single empirical framework grounded in the SOR and TPB perspectives. From a practical perspective, the findings suggest that sustainable fashion brands should emphasize ethical transparency and meaningful customer engagement strategies to strengthen long-term consumer relationships.

However, the findings should be interpreted within the context of Thai consumers and the social media environment examined in this study. Future research may further validate the proposed relationships across different cultural and market settings.

7. Limitation and Future Research

This study has several limitations that should be acknowledged. First, the data were collected using a cross-sectional survey design, which restricts the ability to draw causal inferences among the proposed relationships. Longitudinal or experimental designs are recommended in future studies to better capture the dynamic nature of customer perceptions and loyalty formation over time. Second, the sample was limited to consumers in Thailand who purchase sustainable fashion products via social media platforms. Therefore, the generalizability of the findings to other cultural contexts, countries, or online retail environments may be limited. Third, this research relied on self-reported measures, which may be subject to common method bias and social desirability bias, particularly regarding perceptions of ethics and sustainability.

Future research could explore comparative studies across different countries or regions to examine potential cultural differences in the determinants of loyalty toward sustainable fashion businesses. Furthermore, future studies may incorporate objective behavioral data (e.g., purchase histories, engagement metrics) and qualitative methods to gain deeper insights into consumers’ decision-making processes. Researchers could also extend the proposed model by including additional antecedents (e.g., environmental concern, social influence) or outcomes (e.g., advocacy, willingness to pay a premium) to provide a more comprehensive understanding of loyalty in sustainable fashion contexts. Additionally, this study employed purposive sampling and recruited respondents from online social media communities focused on sustainable fashion consumption. Although this approach was appropriate for targeting relevant consumers, it may introduce self-selection bias, as participants with stronger interests in sustainability and ethical consumption may have been more likely to participate in the survey. Therefore, the findings should be interpreted with caution regarding the representativeness of the broader consumer population. Furthermore, although the discriminant validity and multicollinearity assessments were within acceptable thresholds, several constructs showed relatively high correlations due to their conceptual proximity within the sustainable consumption context. Future studies may further validate the model using alternative analytical approaches and broader samples to confirm the robustness of the proposed relationships.

Future research could also examine potential moderating variables such as cultural orientation, environmental awareness, social influence, or platform type to better understand variations in sustainable fashion loyalty across contexts. In addition, incorporating objective behavioral indicators, such as purchase frequency, clickstream data, or actual engagement metrics, may provide deeper insights into consumer behavior beyond self-reported perceptions.