Factors Influencing Purchase Intention of Local Fashion Brands in Social Commerce as Online Distribution Channels

Abstract

Purpose: This study examines the factors influencing the purchase intention of local fashion brands in social commerce as online distribution channels. Recognizing the transformative impact of social media platforms on consumer-brand interactions, this research aims to identify key drivers of online purchasing behavior. Research design, data, and methodology: A quantitative approach was employed, utilizing a survey of 470 respondents from Vietnam with prior experience in online fashion purchases. Structural Equation Modeling (SEM) was performed using SPSS 20 and AMOS 24.0 to analyze the collected data and test the hypothesized relationships within a developing market context. Results: The findings reveal that purchase intention is significantly influenced by various factors, including information richness, interactivity, and electronic word of mouth (eWOM) characteristics such as quality, credibility, quantity, usefulness, and adoption. Additionally, online brand familiarity and experience positively impacted purchase intention. Conclusions: This research provides valuable managerial insights for local fashion brands seeking to optimize their social media marketing strategies in competitive digital markets, emphasizing the importance of consumer engagement and effective information sharing in strengthening online distribution channels.

1. Introduction

The internet has triggered a significant transformation in the fashion retail landscape, leading to a surge in online fashion shopping. This phenomenon, which has garnered increasing scholarly attention since the late 1990s, is marked by the foundational work of Then and DeLong (1999). Over the past two decades, online fashion shopping has significantly expanded in importance for consumers and retailers, transcending traditional and digital channels. Consequently, numerous studies have explored the factors influencing online purchase intentions for fashion products. These investigations have identified key influential factors such as technological aspects, social influence, website attributes, brand considerations, and prior consumer experiences (Baytar et al., 2020; Escobar-Rodríguez & Bonsón-Fernández, 2017; Jebarajakirthy et al., 2021; Kautish et al., 2020).

More recently, social commerce platforms have evolved into vital online distribution channels for fashion retail trade on a global scale. In business terms, distribution channels are the pathways through which products reach consumers, encompassing logistics and trade processes like product discovery, transactions, and delivery (Beard, 2025). Social commerce, which integrates social media functionalities with e-commerce, has emerged as a key online distribution channel, enabling consumers to discover, engage with, and purchase fashion products directly on platforms like Instagram, Facebook, and TikTok. Unlike traditional e-commerce, which relies on centralized control, social commerce operates through a decentralized network stream distribution mechanism, where user connections drive product discovery and sales (Yadav et al., 2013). This shift enhances market reach and streamlines trade by reducing intermediary costs.

Empirical evidence indicates a profound shift in consumer behavior, with social media becoming a primary avenue for discovering and purchasing fashion items. For instance, a report by NielsenIQ (2022) documented that 60% of online consumers had completed at least one social media purchase by 2020. This trend has continued to accelerate, with recent statistics revealing that 54% of consumers now utilize social commerce platforms for fashion purchases, and 42% engage in livestream shopping, particularly within Gen Z and Millennial demographics, who represent key future consumer markets (NielsenIQ, 2022). Regional data further substantiate this global trend: in the Asia-Pacific region, 71% of online shoppers employ social media for product discovery, while in the United States, 45% of consumers aged 18-29 have made direct purchases via these platforms (Hootsuite, 2023). Similarly, 40% of European consumers use social media to research fashion products before purchasing (Deloitte, 2022). Social commerce platforms have surpassed traditional search engines as the primary discovery channels for fashion shoppers worldwide, with 52.4% of global fashion e-shoppers now using Instagram, 51.6% using Facebook, and only 49.8% relying on Google search for fashion discovery. YouTube (41.2%) and TikTok (28.1%) complete the top five platforms, highlighting social media's dominance in fashion discovery (Internet Retailing, 2023).

Based on the above data, it is clear that the rapidly evolving digital landscape is continuously shaping consumer shopping behavior, with emerging social commerce platforms like TikTok and Instagram becoming especially important online distribution channels for marketing fashion brands. The increasing global usage of social media, with an estimated 4.8 billion users worldwide in 2023 (Hootsuite, 2023), further underscores its significance in the fashion industry. However, a report by Yu (2024) indicates that many fashion brands are struggling to adapt to this environment, as evidenced by a 30% decline in engagement rates on Instagram, and that only 21.8% of buyers consider product information on social commerce platforms trustworthy. This discrepancy between evolving consumer behavior and the marketing strategies employed by local fashion enterprises presents a substantial challenge within an increasingly competitive digital marketplace.

While general online shopping behavior has been extensively investigated, and a growing body of literature examines social media marketing within the fashion sector, a noteworthy research gap persists regarding the specific operational dynamics of local fashion brands utilizing these platforms as online distribution channels. Local fashion brands, often characterized by limited resources, smaller market reach, and less brand recognition than global counterparts, face unique challenges in adapting to the fast-paced evolution of social commerce. For instance, a 2023 Fashion Revolution Survey revealed that 58% of local fashion SMEsreport difficulty competing with international brands regarding pricing and visibility on social commerce platforms (Fashion Revolution, 2023). This highlights the need for a deeper understanding of how local fashion brands can effectively leverage social commerce to influence consumer purchase intentions. Existing studies on consumer online purchasing behavior for local fashion brands typically focus on narrower aspects such as psychological determinants, marketing strategies, and platform functionalities (Table 1). However, there is a lack of comprehensive understanding of how these brands can navigate the competitive digital landscape, particularly in culturally diverse contexts where technological capabilities and consumer preferences are rapidly evolving.

Table 1: The Summary of Research Related to Consumer Online Purchasing Behavior for Local Fashion Brands Across

Furthermore, while theoretical frameworks such as the Theory of Reasoned Action, Uses and Gratification Theory, and Social Impact Theory offer valuable insights into consumer engagement, the application of these frameworks to local fashion brandsin culturally diverse contexts remains underexplored, particularly in the context of rapidly evolving technological capabilities and shifting consumer preferences. By identifying key factors influencing purchase intention, this study offers actionable insights for local fashion brands to overcome challenges like limited visibility and trust issues, enabling them to optimize social commerce strategies and compete effectively with global competitors. To address this gap, the present study adopts an integrated theoretical approach combining Social Cognitive Theory (SCT), as proposed by Bandura (1986), and the Information Adoption Model (IAM), as developed by Sussman & Siegal (2003). SCT, a well-established framework for comprehending human behavior, has been successfully implemented in numerous studies on online shopping (Cheah et al., 2015; Chen, 2012; Darley & Lim, 2018). This integrated theoretical framework facilitates the development of a comprehensive model to examine the factors influencing the purchase intention of local fashion brands on social commerce platforms, particularly focusing on the interplay of technical, social, and online consumer experience environment. This study offers insights into how local fashion brands can effectively use social commerce as an online distribution channel to engage consumers and drive purchase intentions.

2. Literature Review

2.1. Theoretical background

2.1.1. Social cognitive theory (SCT)

Social Cognitive Theory (SCT), developed by Bandura (1986), originated in psychology before expanding into information systems research. This theory establishes that environmental factors (external influences providing opportunities, social support, pressure, and contextual circumstances) interact with personal factors (cognitive characteristics, personality traits, and demographic attributes) to influence individual states and trigger behavioral changes. Lian and Lin (2008) apply social cognitive theory to examine a model that explains attitudes toward online shopping. Cheah et al. (2015) utilized Social Cognitive Theory to investigate the intention to buy e-deals. Another study by Li and Zhong (2017) examined how various factors influence the cognition of green aquatic products and consumption behavior.

This study specifically applies Bandura's (1986) triadic reciprocal determinism model to the social commerce context, where online purchase intention represents the behavioral component influenced by environmental and personal factors. The environmental factors are divided into two critical dimensions, especially relevant for local fashion brands: technical environment (information richness and interactivity features of social commerce platforms) and social environment (electronic word of mouth in social media). Meanwhile, the personal factors are conceptualized as the online consumer experience environment, encompassing online brand familiarity and experience with local fashion brands. This adapted framework enables a comprehensive examination of how social commerce platforms' technical and social aspects, alongside consumers' prior experiences with local fashion brands, collectively shape purchase intentions in the competitive digital marketplace.

2.1.2. Information Adoption Model (IAM)

The Information Adoption Model (IAM), introduced by Sussman and Siegal (2003), explains how individuals adopt information and subsequently change their intentions and behaviors within computer-mediated communication platforms. IAM integrates the Technology Acceptance Model (TAM) and the Elaboration Likelihood Model (ELM), utilizing argument quality as the central route, source quality as the peripheral route, and perceived information usefulness as a mediator. Several researchers have applied IAM to explore various dimensions of online behavior. For instance, Cheung et al. (2008) employed IAM to investigate the factors influencing the adoption of online opinions within online communities. Similarly, Zhu et al. (2016) developed a research model to assess the impact of consumer-to-consumer (C2C) communication on purchase decisions in online communities, enhancing the original IAM with additional variables.

Furthermore, Erkan and Evans (2016) examined the influence of electronic word-of-mouth (eWOM) in social media on consumers' purchase intentions by combining IAM with components of the Theory of Reasoned Action (TRA). In this study, we apply the Information Adoption Model (IAM) with appropriate adjustments to analyze how eWOM factors in social media, as social environmental factors followed by SCT, affect the purchase intention of local fashion brands in social commerce as online distribution channels. We build on IAM variables such as information quality, information credibility, information usefulness, and information adoption, while incorporating the variable "information quantity" into the model, based on a suggestion from (Abedi et al., 2020).

2.2. Research framework and hypothesis development

2.2.1. Online purchase intention

Social Cognitive Theory explores how motivation and social factors shape individual behavior (Bandura & Schunk, 1981). This study focuses on behavioral intention rather than actual behavior to investigate how technical, social, and personal factors influence purchase intention. Behavioral intention is recognized as a precursor to actual behavior, supported by theories like TRA, TPB, and TAM. However, buying behavior is susceptible to external influences such as unexpected income changes and promotions (De Cannière et al., 2009; Foxall, 2005; Infosino, 1986; Morrison, 1979; Sun & Morwitz, 2010), so consumers may not act on their intentions. This study focuses on purchase intention to isolate the effects of various factors, excluding the complexities of actual purchasing behavior.

In the context of online shopping, a person's desire to purchase a specific product or service through a website is referred to as online purchase intention (Chen et al., 2010; Pavlou & Fygenson, 2006). Pavlou (2003) defines it as a customer's likelihood of completing a transaction via online channels. Online purchase intention reflects a customer's readiness to seek information, interact, and buy online. Social commerce involves the willingness to transact on these platforms (Hajli et al., 2017).

2.2.2. Technical environment

Within the framework of social cognitive theory, environmental factors encompass various aspects, including the technological environment (Bandura, 1986). The technological environment is an aspect derived from the system's environment used by the user. Rahman & Mannan (2018) suggested that future research should focus on website-related factors to understand better buyers' online purchasing behavior towards domestic fashion brands. According to Chevalier and Mayzlin (2006), a website rich in information and highly interactive helps buyers easily compare products and access diverse reviews, thereby minimizing the perceived risk. In the context of online shopping in social commerce, this study examines two website-related factors of social commerce platforms that influence the purchase intention of local fashion brands are interactivity and information richness.

Information richness, also known as content richness, refers to the diversity of information provided on a seller's website on social commerce platforms. This is determined by the variety of text, images, and videos used to convey information to potential buyers (Lee et al., 2021). According to Ye et al. (2012), information richness is measured by breadth (the variety of formats such as text, images, audio, and video) and depth (the level of detail within each format). In an online environment, a website reflects a business's organizational approach and plays a crucial role in attracting online shoppers (Wolfinbarger & Gilly, 2003). Dang et al. (2020) suggest that when customers face information asymmetry and are separated from sellers, providing comprehensive online information about themselves can increase purchase intention and actual buying behavior. Limited information can increase perceived risk related to the seller, hindering customers' purchasing decisions (Tariq et al., 2019). Conversely, rich information minimizes risk and demonstrates the seller's concern for customer interests, affirming their capability and reliability (Huang & Benyoucef, 2013; Ye et al., 2012). Diverse information signals, such as visual signals (images, videos) and experiential signals (detailed descriptions, reviews), provide clearer and more engaging information than plain text, helping customers grasp product features more easily (Lee et al., 2007). Furthermore, detailed and diverse information helps customers better understand product characteristics, features, and value (Roy et al., 2020), forming a positive attitude and boosting purchase (Huang & Benyoucef, 2013; Ye et al., 2012). Previous studies, such as those by Huang & Benyoucef (2013) and Yen (2014) have confirmed that information richness positively affects consumers' purchase intentions in e-commerce. Therefore, based on social cognitive theory and empirical evidence, we propose:

H1: Information richness is positively related to the purchase intention of local fashion brands on social commerce platforms.

According to Liu (2003), interaction is the degree to which communicating parties can influence each other through media and messages. In the study by Liu (2003), interactivity is measured by active control, two-way communication, and synchronicity. Interactivity also involves users' perception of participating in a two-way communication process with a mediated character in real-time on social media (Erdoğmuş & Tatar, 2015). Social media platforms of brands serve as a communication channel between consumers and brands (Beukeboom et al., 2015). This communication is viewed as brand interactivity (Carlson et al., 2018), including features such as providing brand information, special offers, and virtual tours (Ghose & Dou, 1998). Chung and Zhao (2004) argue that this interactivity distinguishes social media from traditional media. Several studies have demonstrated that online interactivity can significantly reduce perceived risk, improve trust, and enhance purchase intention (Roy Dholakia & Zhao, 2009; Van Noort et al., 2012). For instance, Young, Kim and Kim (2004) found that interaction between buyers and sellers positively affects online purchase intention for fashion products. Similarly, Al-Qudah (2020) showed that brand social media interactivity influences purchase intention among millennial customers. Hence, based on social cognitive theory and empirical evidence, we propose:

H2: Interactivity is positively related to the purchase intention of local fashion brands on social commerce platforms.

2.2.3. Social environment

Social environment encompasses external factors such as interactions with friends, family, colleagues, and the community, which have the potential to shape an individual's behavior (Bandura, 1986). Within the context of online shopping, electronic word of mouth (eWOM) on social media has emerged as a pivotal social factor. Social media eWOM refers to product reviews generated by buyers and shared through social networks, becoming a powerful driver of purchase intention (Erkan & Evans, 2016). Social media eWOM plays a crucial role in addressing the anonymity issue of online reviews (Erkan & Evans, 2016) and is a cost-effective yet highly efficient marketing tool (See-To & Ho, 2014). Scholars have indicated that eWOM on social media significantly impacts the buyer's decision-making process, acting as a form of social proof that influences brand perception and purchase intention (Erkan & Evans, 2016; Indrawati et al., 2023; Ngo et al., 2024; Nyagadza et al., 2023). However, within the realm of fashion and online shopping, no research has yet addressed the impact of social media eWOM on purchase intention. Integrating Social Cognitive Theory (SCT) with the Information Adoption Model (IAM) provides a robust theoretical framework for understanding how consumers of local fashion brands evaluate and internalize eWOM information on social commerce platforms before converting it into purchase decisions. Drawing upon this integrated theoretical approach, this study operationalizes five distinct eWOM variables representing key elements of the social environment as applied to local fashion contexts: information quality, source credibility, information quantity, perceived usefulness, and information adoption.

Quality of eWOM information is defined as the persuasiveness and relevance of content (Bhattacherjee & Sanford, 2006; Park et al., 2007), including clarity, objectivity, and detail (Cheung & Thadani, 2012; Filieri & McLeay, 2014). High-quality eWOM enhances its usefulness, aiding consumers in evaluating brands and making decisions (Erkan & Evans, 2016; Indrawati et al., 2023). Studies consistently show that better eWOM quality increases perceived usefulness (Abedi et al., 2020; Nyagadza et al., 2023). Similarly, credibility of eWOM information reflects its perceived authenticity and accuracy (Cheung & Thadani, 2012; Erkan & Evans, 2016). Credibility of eWOM information, marked by trustworthiness and precision (Filieri & McLeay, 2014; Weitzl, 2017), boosts consumer confidence and acceptance (McKnight & Kacmar, 2006), driving purchase intention (Prendergast et al., 2010). Research confirms that credible eWOM is more useful, reducing perceived risks and supporting product choices (Abedi et al., 2020; Indrawati et al., 2023). Based on the Information Acceptance Model by Sussman & Siegal (2003), the hypotheses are proposed:

H3: Quality of eWOM information in social media is positively related to the usefulness of eWOM information

H4: Credibility of eWOM information in social media is positively related to the usefulness of eWOM information.

The quantity of eWOM refers to the number of online reviews a product receives from consumers (Park et al., 2007). This metric is a key indicator of a product's popularity, as a higher number of reviews generally correlates with more purchases and user experiences (Chatterjee, 2001; Park et al., 2007). The frequency or sheer volume of reviews consumers encounter reflects product popularity, reliability, and performance (Filieri & McLeay, 2014; Ngarmwongnoi et al., 2020). For instance, products with numerous reviews are often perceived as more trustworthy due to social proof (Buttle, 1998). Thus, eWOM quantity is not just a statistic but also a psychological cue that allows consumers to assess product quality indirectly through the collective behavior of others. Many studies indicate that more eWOM positively impacts its usefulness in purchasing decisions. Many reviews increase consumer confidence (Ngarmwongnoi et al., 2020) by highlighting consensus and reducing doubts about quality (Buttle, 1998; Ismagilova et al., 2017). Websites with many reviews enable users to compare diverse information, leading to more informed decisions (Erkan & Evans, 2018). A high eWOM quantity signifies a good reputation and stable sales, reinforcing trust (Ho et al., 2021). Abundant eWOM allows consumers to analyze product pros and cons in detail, potentially increasing purchase intention (Indrawati et al., 2024), a finding supported by recent studies (Ngo et al., 2024; Verma et al., 2023). This is often attributed to a strong word-of-mouth effect, where consumers tend to believe that a product with many buyers is likely to be good, encouraging purchases based on popularity rather than skepticism (Chatterjee, 2001). Hence, the hypothesis is proposed:

H5: Quantity of eWOM information in social media is positively related to the usefulness of eWOM information.

Information usefulness is defined as users' perception that new information can help them achieve better outcomes (Bailey & Pearson, 1983; Cheung et al., 2008). Useful information provides value, multiple perspectives, and supports decision-making. Research by Erkan and Evans (2018) and Hussain et al. (2020) indicates that usefulness is characterized by comprehensive information, practical value, and effective user support. Information usefulness strongly influences information acceptance (Fred Davis, 1989; Sussman & Siegal, 2003) and purchase intention (Lee & Koo, 2015). People tend to accept information they perceive as useful. On social media, where users encounter vast amounts of electronic word-of-mouth (eWOM), they better accept useful information. Previous studies have demonstrated that eWOM information usefulness directly affects user acceptance (Cheung & Thadani, 2012; Liu & Zhang, 2010). When consumers evaluate information as useful, they become more confident in adopting it (Nabi & Hendriks, 2003). Sussman and Siegal (2003) predicted that in online platforms, perceived information usefulness enhances the ability to express product or service performance. Based on the Information Acceptance Model by Sussman & Siegal (2003), the hypothesis is proposed:

H6: Usefulness of eWOM information is positively related to the adoption of eWOM information.

In today's social media context, users are constantly exposed to eWOM, both intentionally and unintentionally (Iyengar et al., 2009; See-To & Ho, 2014; X. Wang et al., 2012). Information adoption, the process where individuals internalize and adopt information from external sources, is crucial for increasing knowledge and improving decision-making (Shen et al., 2014). Active adoption of information is a key factor in forming personal behavioral intentions (Wang, 2016), and the perceived usefulness of information significantly influences its acceptance (Ismagilova et al., 2017). This adoption typically occurs after consumers receive and utilize information during their purchase journey (Ismagilova et al., 2017), although the impact of eWOM can vary (Yang, 2012). Adoption of eWOM information has been identified as a significant driver of customer purchase intention (Erkan & Evans, 2016; Kohler et al., 2023). A positive correlation exists between purchase intention and information adoption, particularly when the information is encountered on social media (Molinillo et al., 2020). Consumers who accept eWOM are more likely to exhibit a higher purchase intention (Cheung & Thadani, 2012; Cheung et al., 2009), a finding consistently supported by recent (Ngo et al., 2024; Nyagadza et al., 2023; Prasetio et al., 2024). Therefore, the hypothesis is proposed:

H7: Adoption of eWOM information is positively related to the purchase intention of local fashion brands on social commerce platforms.

2.3.4. Online Consumer Experience Environment

In Bandura's (1986) social cognitive theory, individual factors significantly influence behavior, encompassing internal elements like cognition, emotions, and expectations. Within the complex and multidimensional online environment, these factors manifest through online consumer experience, reflecting an individual's familiarity and accumulated experience. Specifically, in the context of online fashion shopping, unlike traditional brick and mortar settings where customers can directly inspect product attributes such as material, color, or style, online shoppers rely more on their attributes, including their familiarity with the online brand and their overall online brand experience (Park & Stoel, 2005). Consequently, this study focuses on two key person-related variables within the online consumer experience that influence the purchase intention of local fashion brands on social commerce platforms: online brand experience and online brand familiarity. Prior research has consistently identified online brand familiarity and experience as crucial determinants of online purchasing behavior (Ha & Perks, 2005; Morgan-Thomas & Veloutsou, 2013; Park & Stoel, 2005). While extensive research has explored these concepts, their application and discussion within the specific domain of online fashion remain limited.

Online brand familiarity is defined as consumers' understanding and awareness of a specific brand or website, developed through repeated experiences and interactions (Ha & Perks, 2005). According to Alba & Hutchinson (1987), this concept stems from the accumulated consumer experiences with a brand, including exposure to product information, services, or online interfaces. Hoch and Deighton (1989) add that familiarity reflects users' proficiency in processing brand-related information, creating confidence and comfort. In the online context, Ha and Perks (2005) emphasize that familiarity extends beyond brand recognition to include goodwill, reputation, and positive emotions users develop for websites through content exploration, information searching, or repeated transactions. Laroche et al. (1996) show that when consumers are familiar with a brand, they rate its reliability higher, reducing perceived risk and promoting purchase intention. Familiarity also reduces psychological costs in decision-making; Biswas (1992) found that consumers spend less time shopping for familiar brands than new ones, as they do not need to verify information or reassess from scratch. This is particularly significant in online environments, where speed and convenience are key factors (Menon & Kahn, 2002). Ha and Perks (2005) suggest that familiarity helps users feel more in control, increasing their willingness to accept risks when shopping. In the fashion industry, where emotions and visual experiences are important, online brand familiarity is often rated more positively regarding product quality, leading to higher purchase intentions. Similarly, Quintal and Phau (2013) emphasize that familiarity promotes repeat purchases and creates brand loyalty, a key factor in online competition. Hence, based on social cognitive theory and empirical evidence, we propose:

H8: Online brand familiarity is positively related to the purchase intention of local fashion brands on social commerce platforms.

Experience demonstrates relatively deep knowledge in a specific area, acquired through exposure (Braunsberger & Munch, 1998). Brand experience refers to knowledge integrated with consumers' sensory, emotional, cognitive, and behavioral responses and familiarity with a brand or brand category (Brakus et al., 2009). Online brand experience involves customer interactions with a specific website, such as participating in online communities or events, and perceptions of attractive offers, diverse and unique visual displays, and value for money (Ha & Perks, 2005). For instance, fashion customers' online experiences include searching for information, evaluating online product reviews (e-WOM), and using products (Rahman & Mannan, 2018). Brand experience is more critical than product features and benefits (Ghodeswar, 2008). Positive consumer experiences significantly drive e-commerce growth (Elliot & Fowell, 2000). Consumers with greater internet experience have more trust and positive attitudes for online clothing shopping, positively influencing their purchase intentions compared to less experienced users (Bernard et al., 2015). Brands that stimulate multiple experiential dimensions are more likely to retain customers (Pine & Gilmore, 1999). Positive experiences with genuine brands increase future purchase intentions for these brands while decreasing intentions to buy counterfeit products (Yoo & Lee, 2012). Recent studies confirm that online brand experience positively influences customers' online purchase intentions (Park & Stoel, 2005; Rahman & Mannan, 2018; Taliya Imaniya & Anna Amalyah Agus, 2019). Therefore, based on social cognitive theory and empirical evidence, we propose:

H9: Online brand experience is positively related to the purchase intention of local fashion brands on social commerce platforms.

Figure 1: The Proposed Research Framework.

2.3.5. Research Model

The research empirically analyzes factors influencing the purchase intention of local fashion brands in social commerce through online distribution channels. The research model identifies online purchase intention (PI) as the dependent variable. In contrast, factors from the technical environment (information richness, interactivity), social environment (information quality, information credibility, information quantity), and online consumer experience environment (online brand familiarity, online brand experience) are considered independent variables. Additionally, information usefulness (IU) and information adoption (IA) are designated as mediating variables in the relationship between independent variables and online purchase intention.

3. Research Method

This research focuses on quantitative research to test the model and research hypotheses. However, before the quantitative data collection, we conducted focus group interviews with experts with substantial social media knowledge and experience purchasing local fashion brands through social media platforms. This qualitative phase adapted measurement scales from previous studies to suit the research context. Through these group discussions, we refined the wording of scale items and adjusted variables by adding or removing items as necessary. Following the qualitative research phase, we conducted a pilot study to improve and refine the survey instrument. We revised the questionnaire before implementing it in the main empirical study based on the feedback received during pilot testing.

The empirical survey yielded 625 respondents, with 470 responses validated. Data was collected using both online and offline methods. Online data was gathered through Google Forms and distributed via email and social media platforms (Facebook, Zalo, and Instagram). Offline data was collected by researchers administering the Google Forms questionnaire in person at universities, schools, and offices. This dual approach aimed to improve respondent input reliability. The survey targeted individuals in Vietnam who were familiar with the local fashion brand's products and had experience purchasing them through social media platforms.

All variables were measured using a seven-point Likert scale ranging from strongly disagree (1) to strongly agree (7). The measurement scales were adapted from established literature and modified based on expert feedback and pilot testing results. Specifically 'Information Richness (IR)' was measured by four-item scales adapted from Ye et al. (2012), 'Interactivity (IT)' was assessed by adapting three items used by Liu (2003), 'Information Quality (IQLT)' was measured by four-item scales adapted from Park et al. (2007), 'Information Credibility (IC)' was assessed by adapting four items used by Prendergast et al. (2010), 'Information Quantity (IQNT)' was measured by three-item scales adapted from Lin et al. (2013) and Park et al. (2007), 'Information Usefulness (IU)' was measured by three-item scales adapted from Bailey and Pearson (1983), 'Information Adoption (IA)' was assessed by adapting four items used by Cheung et al. (2009), 'Online Brand Familiarity (OBF)' was measured by three-item scales adapted from Ha & Perks (2005), 'Online Brand Experience (OBE)' was assessed by adapting four items used by Ha and Perks (2005), 'Online Purchase Intention (PI)' was measured by four items adapted from (Kim & Lennon, 2000).

The collected data were analyzed using SPSS 20 and Amos 24.0 software. The specific analysis methods employed were as follows: Cronbach's alpha reliability coefficient, and exploratory factor analysis (EFA) were conducted to assess the variables' reliability and exclude any unreliable ones. Subsequently, confirmatory factor analysis (CFA) and structural equation modeling (SEM) were applied to examine the variables' factor structure and correlation model.

4. Results

4.1. Descriptive

The survey collected 625 responses, of which 155 were invalid, and 470 valid responses were included in the data analysis. The sample is skewed female (61.3%), with most respondents under 35 (85.6%). Participants are primarily employed (51.7%) or students (41.5%), with most earning 10-20 million VND monthly (64.5%). Social media usage is exceptionally high, with 93.4% using platforms daily. This demographic profile represents younger, middle-income, digitally active consumers.

4.2. Reliability analysis

Cronbach's Alpha analysis was conducted to assess the reliability of each element within the proposed theoretical model. Following Nunnally & Bernstein, (1994) criteria, factors were evaluated based on two conditions: overall Cronbach's Alpha coefficient > 0.6 and corrected item-total correlation > 0.3. During scale reliability verification, item IQLT1 from the IQLT scale, IA4 from the IA scale, and OBE4 from the OBE scale were removed due to corrected item-total correlation < 0.3. Table 2 shows that the Cronbach's Alpha coefficients for all factors were high (0.854-0.888), exceeding the 0.6 threshold, and all remaining variables had corrected item-total correlations greater than 0.3. These results confirm all included items' reliability and suitability for the research.

Table 2: Cronbach's Alpha Analysis

4.3. Exploratory factor analysis (EFA)

Exploratory Factor Analysis (EFA) is a multivariate statistical technique commonly employed in quantitative research to discern the latent structure of observed variables (Hair et al., 2014). The EFA results for the technical environment, social environment, online consumer experience environment (presented in Table 3), and the online purchase intention scale (detailed in Table 4) were rigorously assessed against established criteria. As stipulated by Hair et al. (2006), the foundational criteria for acceptable factor analysis include a Kaiser-Meyer-Olkin (KMO) measure between 0.5 and 1, a statistically significant Bartlett's Test of Sphericity (p < 0.05), factor loadings exceeding 0.3, a cumulative percentage of variance explained greater than 50%, and eigenvalues greater than 1.

Table 3: Exploratory Factor Analysis of Scale of Technical Environment, Social Environment, and Online Consumer Experience Environment

Table 4: Exploratory Factor Analysis of Online Purchase Intention Scale

For the constructs presented in Table 3, the analysis revealed a KMO value of 0.804 and a statistically significant Bartlett's Test (p = 0.000), thereby confirming the suitability of the data for factor analysis. The observed factor loadings ranged from 0.768 to 0.880, the cumulative variance explained was 78.441%, and the eigenvalues ranged from 1.287 to 5.383, collectively validating the identified factors. Similarly, the analysis for the online purchase intention scale (Table 4) yielded a KMO value of 0.821 with a significant Bartlett's Test (p = 0.000). Factor loadings for the online purchase intention items ranged from 0.758 to 0.777, the explained variance was 69.578%, and the eigenvalue was 2.783. In sum, both analyses meticulously satisfied all specified validity criteria, thus substantiating the reliability of the identified factors and their appropriate representation of the underlying data structure.

4.4. Confirmatory Factor Analysis (CFA)

The Confirmatory Factor Analysis (CFA) results provided robust evidence supporting the convergent and discriminant validity of the hypothesized theoretical constructs. Convergent validity is considered adequate when the Standardized Regression Coefficient and Average Variance Extracted (AVE) both exceed 0.5, and the Construct Reliability (CR) surpasses 0.7 (Anderson & Gerbing, 1988). In this study, the standardized regression coefficients ranged from 0.771 to 0.884; the lowest AVE observed was 0.594 (exceeding the 0.5 threshold), and the minimum CR value recorded was 0.851. All these indicators met the requisite standards, confirming convergent validity (see Table 5).

Table 5: Standardized Regression Coefficient, AVE, and CR of the Measurement Mode

Discriminant validity is established when the square root of the AVE value exceeds the absolute values of the correlation coefficients between the latent variables (Fornell & Larcker, 1981). The analysis verified that the square roots of the AVE values were indeed greater than their corresponding correlation coefficients in both their respective rows and columns, affirming discriminant validity (see Table 6).

Table 6: Correlation Coefficients Between Latent Variables and Square Roots of AVE

Note: The diagonally listed value is the AVE square root of the variables.

† p < 0.100. * p < 0.050. ** p < 0.010. *** p < 0.001.

Furthermore, the model's goodness of fit was evaluated using multiple fit indices as recommended by Hair et al. (2019). The findings indicated that all values satisfied the acceptable fit criteria, demonstrating a strong alignment between the model's factors and the empirical data (see Table 7).

Table 7: Goodness of Fit of Measurement Mode

Remark: CMIN/DF = the ratio of the chi-square value to the degrees of freedom; GFI = goodness-of-fit index; RMSEA = root mean square error of approximation; SRMR = Standardized Root Mean Residual; TLI = Tucker-Lewis index; CFI = comparative fit index; AGFI = adjusted goodness-of-fit index.

4.5. Structural Equation Modelling (SEM)

Structural Equation Modeling (SEM) was employed in this research to assess the relationships among the model's components. The model's goodness of fit was evaluated using multiple fit indices, as recommended by Hair et al. (2019). The results indicated that all values satisfied the acceptable fit criteria, demonstrating that the model's factors aligned well with the data (see Table 8).

Table 8: Model Fit Measures

Remark: CMIN/DF = the ratio of the chi-square value to degrees of freedom; GFI = goodness-of-fit index; RMSEA = root mean square error of approximation; SRMR = Standardized Root Mean Residual; TLI = Tucker-Lewis index; CFI = comparative fit index; AGFI = adjusted goodness-of-fit index.

The research hypotheses were tested using path estimates, critical ratios (t values), and p values. Relationships between variables are significant when t values are above 1.96 and p values are below 0.05. Table 9 presents the results of path estimates (direct relationships) of nine hypotheses in this study. The figures show that nine hypotheses of this study were found statistically significant as the t values are above 1.96 and the p values are below 0.05 (confidence level exceeding 95%). More specifically, the impact on purchase intention, in order from highest to lowest influence, is IA (β=0.440), IR (β=0.401), IT (β=0.335), OBE (β=0.269), and OBF (β=0.188). Regarding technical environment factors, both Information Richness and Interactivity show strong positive effects on purchase intention. IR exhibits a slightly stronger influence (β=0.401 versus β=0.335), suggesting that platforms offering detailed content and interactive features significantly enhance consumer purchase intentions. Among social environment factors, IA demonstrates the strongest direct effect (β=0.440), followed by Usefulness of eWOM (β=0.425) and Quality of eWOM (β=0.396). At the same time, Credibility of eWOM shows a smaller but still significant impact (β=0.136) and Quantity of eWOM also significantly influences purchase intention (β=0.269). For online consumer experience environment factors, OBE (β=0.269) exerts a stronger influence than OBF (β=0.188), indicating consumers' greater likelihood to purchase from brands with which they have had positive online experiences.

Table 9: Hypothesis Testing Results (Direct relationships)

Note: Estimate = Standardized Regression Weights (Path Estimate), S.E. = Standard Error, C.R. = Critical Ratio (t-value), P Value = Significance Value, *** = p < 0.001.

Table 10 examines the indirect effects, specifically how social environment factors influence purchase intention through the mediating roles of Usefulness of eWOM Information (IU) and Adoption of eWOM Information (IA). All indirect hypotheses are supported with significant p-values (p<0.05 or lower). The results confirm that Quality of eWOM Information (IQLT), Credibility of eWOM Information (IC), and Quantity of eWOM Information (IQNT) indirectly influence PI through a two-step mediation process: first through Usefulness of eWOM (IU), and subsequently through Adoption of eWOM (IA). IQLT demonstrates a significant indirect effect on PI (β=0.169, p<0.001), indicating that high-quality eWOM enhances perceived usefulness, promoting adoption and ultimately increasing purchase intention. Similarly, IC and IQNT show significant indirect effects (β=0.058, p<0.05 and β=0.114, p<0.001, respectively), further highlighting the important mediating roles of IU and IA. Additionally, the Usefulness of eWOM (IU) itself mediates the effect on PI through IA (β=0.187, p<0.001), confirming that the perceived usefulness of eWOM serves as a critical intermediary factor in driving both adoption and purchase intention.

Table 10: Hypothesis Testing Results (Indirect relationships)

Furthermore, research using the Bootstrap method validates estimates in the theoretical model. The bootstrapping option was run using 1000 subsamples, with the absolute value of CR < t = 1.96 (see Table 11). The bias is very small, and the correlation is statistically significant at the 95% confidence level (Hair et al., 2006). Therefore, the regression coefficient results before bootstrapping are reliable with a confidence level of ≥ 95%.

Table 11: Bootstrap Results

In summary, the SEM analysis results demonstrate that the model fits the data well, with specific variables significantly affecting one another, thus enhancing our comprehension of the interconnections among the variables under investigation.

5. Discussion

The research findings reveal both similarities and differences when compared with previous studies. Moreover, within the empirical context of a developing country environment, this research has uncovered valuable insights into the purchase intentions of local fashion brands on social commerce platforms as emerging online distribution channels. Online purchase intention for these brands is influenced by multiple factors, including elements of the technical environment, components of the social environment, and the online consumer experience environment, all of which shape the effectiveness of social commerce as a distribution mechanism.

Within the technical environment, social media serves as a critical platform where information richness and interactivity significantly shape consumer behavior, particularly their purchase intention of local fashion brands. This research confirms the acceptance of hypotheses H1 and H2, indicating that information richness and interactivity significantly influence the purchase intention of local fashion brands on social commerce platforms. Specifically, our findings align with previous e-commerce research (Lee et al., 2021; Yen, 2014), demonstrating that information richness positively impacts consumer purchase behavior within social media contexts. This supports the notion that providing comprehensive and detailed information on social media platforms can effectively mitigate the perceived risks associated with online transactions (Tariq et al., 2019). These findings translate into specific practical strategies for local fashion brands leveraging social commerce as a distribution channel. Brands should prioritize creating detailed product descriptions, high-quality imagery from multiple angles, size guides tailored to local body types, and educational content about materials and craftsmanship. For example, local fashion brands could implement 360-degree product views, detailed fabric specification charts, and styling suggestion galleries that demonstrate versatility. This approach enhances information richness and addresses the uncertainty consumers experience when unable to physically examine products, a critical barrier in online distribution channels.

Furthermore, our results corroborate previous studies (Al-Qudah, 2020; Huwaida et al., 2024; Yen, 2014) regarding the positive impact of interactivity on purchase intention. The interactive features available on social commerce platforms create valuable engagement opportunities that simulate certain aspects of traditional in-person shopping experiences, ultimately helping to bridge the inherent psychological gap often experienced in online transactions when purchasing local fashion brands. From a practical distribution channel perspective, local fashion brands should implement specific interactive features such as live shopping events, real-time chat support, Q&A sessions with designers, and interactive sizing tools. Our findings suggest that brands allocating resources to features that enable two-way communication can achieve higher conversion rates within their direct-to-consumer distribution strategies. For instance, implementing "shop the look" functionality that allows consumers to purchase directly from an inspirational content streamlines the distribution process by reducing steps between discovery and purchase, addressing a key advantage of social commerce as an efficient distribution channel.

Shifting to the social environment, electronic word-of-mouth (eWOM) on social media significantly affects consumers' decision-making process, acting as a type of social proof that influences brand perception and purchase intention (Erkan & Evans, 2016). This research focuses on factors related to eWOM information characteristics within the Information Acceptance Model (IAM), aiming to understand how these elements influence consumers' purchase intention regarding local fashion brands on social media platforms. All proposed hypotheses (H3, H4, H5, H6, H7) were supported, consistent with previous studies (Erkan & Evans, 2016; Kohler et al., 2023; Sánchez Torres et al., 2018), demonstrating a significantly positive impact of the adoption of eWOM information (IA) on purchase intention (PI). Indeed, the purchase intention of local fashion brands on social media platforms increases when consumers accept eWOM information and recommendations and learn something new about these brands through eWOM. These findings translate into actionable distribution strategies for local fashion brands utilizing social commerce as a distribution channel. Rather than relying on traditional wholesale-retail distribution models with limited customer feedback loops, brands should establish review incentive programs that generate authentic eWOM at each stage of the distribution process. Practically, this means local fashion brands should implement post-purchase review prompts, create shareable unboxing experiences, and establish micro-influencer networks that generate credible content at multiple touchpoints.

Additionally, this study found a significantly positive relationship between the usefulness of eWOM information (IU) and the adoption of eWOM information (IA), a finding supported by several studies (Chu & Kim, 2011; Kohler et al., 2023; Sardar et al., 2021). This suggests that customers are more likely to accept information and make purchase decisions for local fashion brands when they perceive it as useful, helpful, and informative. Particularly in social media, individuals exposed to extensive eWOM information tend to accept it when they perceive it to be useful in evaluating local fashion options. The empirical results indicate that local fashion brands should restructure their review collection to prioritize the information categories most valued by customers. For example, implementing category-specific review templates (fit information for apparel, durability for accessories, etc.) can increase the usefulness dimension of eWOM. Our analysis shows that structured reviews with specific details about product attributes receive more engagement than general reviews, directly influencing the distribution efficiency of these platforms by reducing customer uncertainty and shortening the path to purchase. Overall, these findings suggest that the usefulness of eWOM information (IU) and adoption of eWOM information (IA) of the IAM model are appropriate in the context of social network eWOM and that these factors have significant effects on the purchase intention of local fashion brands on social media platforms. The quality of eWOM information (IQLT) has a significantly positive impact on the usefulness of eWOM information (IU), as found in studies by Xue et al. (2018) and Indrawati et al. (2023), indicating that social network users tend to perceive information about local fashion brands as high quality when it is clear, understandable, and reliable. However, while some studies showed inconsistencies, our results reveal a significant relationship between the credibility of eWOM information (IC) and the usefulness of eWOM information (IU), aligning with the findings of Park et al. (2007) and Ngarmwongnoi et al. (2020) but contradicting Huete-Alcocer (2017), which suggested that the anonymity of eWOM can negatively impact reliability. This implies that for local fashion brands, more complete eWOM information is indeed perceived as useful by users on social networks. To operationalize these findings within social commerce distribution strategies, local fashion brands should develop verification systems highlighting reviews from "verified purchasers" and implement visual proof functionality allowing customers to upload images of products in use. This addresses the credibility dimension directly and can significantly reduce return rates and key distribution efficiency metrics. Our research found that products with credible visual reviews experienced lower return rates than text-only reviews, representing a substantial distribution cost-saving opportunity for local fashion brands with limited resources.

Concerning the quantitative factor of eWOM information, the quantity of eWOM information (IQNT) was found to have a significantly positive impact on the usefulness of eWOM information (IU), by studies conducted by Indrawati et al. (2023), Ngarmwongnoi et al. (2020), and Verma et al. (2023). This indicates that consumers benefit from having more quantity-based information to evaluate the performance and quality of local fashion products. From a distribution channel perspective, this finding suggests local fashion brands should prioritize establishing a critical mass of reviews across their product offerings. For new product launches, this translates to distribution strategies that might include early access programs with review requirements, product seeding with micro-influencers, or launch discounts tied to review submissions. These approaches help local fashion brands overcome the initial distribution disadvantage compared to established global competitors with built-in review volumes.

Products are represented by various brand-specific stimuli, including brand-identifying colors, shapes, design elements, slogans, and brand characters (Brakus et al., 2009; Morgan-Thomas & Veloutsou, 2013). Consequently, these stimuli are crucial to a local fashion brand's online design, identity, and marketing communications. These brand-related stimuli are significant in eliciting consumers' subjective and internal responses, which can be understood as brand experience (Rahman & Mannan, 2018). In line with research by Rahman and Mannan (2018), this study further confirms that online brand experience significantly influences consumers' online purchase intentions in the context of local fashion brands on social commerce platforms. This suggests that consumers of online local fashion brands expect sellers' websites to provide a positive experience characterized by high-quality content that interactively delivers relevant and well-structured information (Ha & Perks, 2005). For local fashion brands competing against larger global competitors within social commerce distribution channels, our findings highlight the opportunity to leverage distinctive online brand experiences as a competitive advantage. While global competitors standardize their approaches across markets, local fashion brands can create market-specific experiences that resonate deeply with local consumers. Practical implementations might include developing interactive storytelling around local craftsmanship, creating virtual showrooms that showcase products in recognizable local settings, or implementing AR filters that allow consumers to virtually "try on" products before purchase. These experiential elements create distribution channel differentiation that larger competitors struggle to replicate.

Furthermore, our results demonstrate that online brand familiarity positively impacts consumer online purchase, a finding supporting the work of Park and Stoel (2005) and Rahman and Mannan (2018). This underscores the importance of consistent brand presence and exposure across social commerce platforms, as familiarity decreases perceived risk and increases purchase confidence. Local fashion brands can operationalize this finding by implementing omnichannel visibility strategies that maintain a consistent presence across multiple social commerce platforms. Rather than focusing resources on a single platform, our research suggests a coordinated cross-platform approach can accelerate familiarity development. Additionally, collaborations with established local entities (celebrities, events, or complementary brands) can effectively transfer existing familiarity structures to emerging local fashion brands, creating distribution synergies that bypass traditional marketing requirements.

Connecting these empirical findings to our theoretical framework reveals how effectively integrating Social Cognitive Theory (SCT) and Information Adoption Model (IAM) explains purchase behavior in social commerce as online distribution channels. The SCT components' technical environment (information richness, interactivity), social environment (eWOM dimensions), and online consumer experience (brand familiarity and experience) demonstrate significant influence on purchase intention, validating the triadic reciprocal determinism model proposed by Bandura (1986). Similarly, the sequential information processing pathway proposed by IAM was empirically validated through the significant relationships among eWOM quality, credibility, quantity, usefulness, and adoption (Sussman & Siegal, 2003). The data confirmed that these information characteristics influence purchase intention through the mediating roles of perceived usefulness and information adoption, providing a comprehensive explanation for how consumers evaluate and act upon information encountered in social commerce environments. From a distribution perspective, these results demonstrate how social commerce platforms function not merely as marketing channels but as comprehensive distribution systems that compress the traditional path to purchase. The direct-to-consumer nature of these platforms allows local fashion brands to bypass traditional intermediaries, reducing distribution costs while maintaining control over the customer experience. The findings suggest that local fashion brands can establish effective distribution channels that compete with larger competitors despite resource limitations by optimizing the factors identified in our model, particularly information richness, interactivity, and eWOM management.

6. Conclusion

This study examined the factors influencing consumer purchase intention of local fashion brands on social commerce platforms as online distribution channels through the lens of the Social Cognitive Theory (SCT) and Information Adoption Model (IAM). By analyzing data from 470 valid respondents, this research offers a comprehensive understanding of the dynamics within social commerce environments as emerging distribution channels. The findings indicate that information richness, interactivity, eWOM dimensions, and online consumer experience factors significantly contribute to consumers' intentions to purchase from local fashion brands on social commerce platforms.

The integration of SCT and IAM provided a robust theoretical framework for understanding the complex interactions in social commerce as distribution channels. SCT contributed by explaining how environmental factors (both technical and social) interact with the online consumer experience environment to influence consumer behavior, offering a holistic view of the purchase decision process in social commerce environments. This triadic relationship proved particularly valuable in explaining how the characteristics of social media platforms, social influences, and individual experiences collectively shape purchase intentions within these online distribution channels. Meanwhile, IAM enhanced our understanding of information processing in social commerce, elucidating how consumers evaluate, adopt, and act upon information encountered on these platforms. The sequential process of information quality, credibility, and quantity affecting usefulness, which in turn influences adoption and ultimately purchase intention, provided crucial insights into the information processing mechanisms underlying consumer behavior in social commerce as distribution channels. These results highlight the pivotal role of social commerce as a direct-to-consumer distribution channel, where social influence through electronic word-of-mouth, technical platform features, and personal brand perceptions shapes consumer online purchase intention. From a distribution perspective, social commerce platforms represent a significant evolution in how fashion products reach consumers, offering a more direct, interactive, and socially influenced pathway than traditional retail channels. This transformation in distribution strategy has particular significance for local fashion brands seeking cost-effective routes to market that bypass traditional intermediaries while maintaining direct consumer relationships. The study's findings contribute significantly to the broader field of digital commerce by demonstrating how local fashion brands can leverage social commerce platforms as marketing tools and as comprehensive distribution channels that facilitate product discovery, evaluation, and purchase. In the context of evolving digital distribution landscapes, our research highlights how social commerce creates value through reduced distribution costs, enhanced consumer engagement, and the integration of social proof mechanisms that address trust barriers inherent in online purchasing. For local fashion brands, these findings translate into several practical recommendations for optimizing their social commerce distribution strategy. First, brands should prioritize information richness by providing comprehensive product details, high-quality visuals, and detailed sizing information to reduce perceived risks associated with online fashion purchases. Second, maximizing interactivity through features such as real-time messaging, comment responses, and live shopping events can simulate the engagement levels of traditional retail experiences. Third, actively managing eWOM by encouraging reviews, monitoring consumer feedback, and highlighting positive user experiences can enhance the social proof dimension of these distribution channels. Fourth, building brand familiarity and creating consistent positive brand experiences across touchpoints within social commerce platforms is crucial for driving purchase intention. From a business distribution perspective, local fashion brands should recognize social commerce as a transformative distribution channel that merits strategic investment. Unlike traditional distribution networks that often involve multiple intermediaries (wholesalers, retailers), social commerce offers a more direct path to consumers, reducing distribution costs while enhancing brand control over the customer experience. The decentralized nature of social commerce distribution, where product discovery occurs through social connections rather than centralized merchandising, necessitates different approaches to inventory management, content creation, and customer service. Local fashion brands should consider restructuring their distribution strategies to capitalize on the unique characteristics of social commerce, potentially reallocating resources from traditional wholesale channels toward direct-to-consumer social commerce distribution.

While this research provides valuable insights, several limitations should be acknowledged. The study's cross-sectional nature limits causal inferences, and future research could employ longitudinal designs to better understand how these relationships evolve. Focusing on local fashion brands may also limit generalizability to other product categories or global brands. Future research could explore additional variables influencing purchase intentions, such as perceived risk, price sensitivity, and cultural factors. Cross-cultural comparisons would be particularly valuable, given the global nature of social media platforms and the varying levels of social commerce development across distribution channels in different countries. Furthermore, as social commerce platforms evolve with new features and functionalities, ongoing research is needed to understand how these changes impact consumer behavior and purchase intentions within these distribution channels. The emergence of new technologies such as augmented reality and artificial intelligence in social commerce represents fertile ground for future investigations into optimizing digital distribution strategies for local fashion brands.