Does Gender Matter for Electronic Word-of-Mouth Interactions in Social Media Marketing Strategies? An Empirical Multi-Sample Approach

Abstract

Considering social media’s expansion worldwide, marketing academics and marketers emphasize the need to consider electronic word of mouth (eWOM) for strategic marketing decisions. However, there is limited research regarding the ways in which male and female consumers engage in eWOM behaviors. This research aims to explore the gender-specific dynamics of eWOM drivers in social media marketing, by validating a proposed model of key predictors for two samples (of female and male respondents). Data were gathered from two samples of social media users. For this empirical research, we integrated structural equation modeling and an artificial neural network (PLS-SEM-ANN) for a comprehensive approach intended to generate practical and theoretical insights for eWOM. Hypothesis testing reflected contrasting results—for female respondents, the key eWOM drivers were customer participation, involvement, loyalty, and customer satisfaction; whereas, for male respondents, the key predictors were brand familiarity, loyalty, and satisfaction. The significant variables supported by SEM were included in ANN models as input neurons, showcasing nonlinear relationships among constructs for both samples. Thus, this research provides theoretical contributions regarding eWOM, gender assessments, and the social media marketing literature. From a practical perspective, this study advances targeted social media marketing strategies to enhance consumer–brand interactions.

1. Introduction

Due to its exponential growth, social media has acquired a broad spectrum of applications and uses for final consumers and organizations. In fact, while there are over 5.3 billion internet users on a global scale, there are 4.95 billion social media users [1]. Indeed, social media is influencing the way consumers interact with other consumers, but also how they interact with brands. Moreover, social media marketing enables access to key brand information for consumers, influencing their behaviors based on company-generated content, as well as their behaviors regarding the development of user-generated content and electronic word of mouth (eWOM). Due to social media’s value, worldwide marketing expenditure has increased significantly and is estimated to reach 276.72 billion USD in 2025 [2]. Among the social media platforms available worldwide, the most popular one is Facebook, with more than three billion monthly active users worldwide [3].

Extrapolating from this practical perspective, social media has a distinctive role in people’s lives, becoming an essential computer-intervened communication platform [4]. For organizations and brands, social media offers many opportunities and serves as an important outlet for eWOM. Nonetheless, these opportunities may entail certain challenges. A prominent challenge is related to managing and understanding eWOM while considering the many options of user content creation on social media. Indeed, marketing academics and marketing managers acknowledge the value of word of mouth (WOM) and eWOM in terms of affecting consumer attitudes and behavior, due to their associated credibility and trustworthiness [5].

Therefore, WOM communications reflect the “act of exchanging marketing information among consumers” [6] (p. 691) with significant potential to influence consumers’ perceptions and decision-making processes. Considering the extension of WOM with eWOM to encompass electronic mediums, eWOM represents positive or negative brand-related statements via the internet [7]. Compared to traditional marketing tactics, consumers rely on eWOM to make brand-related decisions and acquisitions, considering it to be more reliable [8]. In fact, Moisescu et al. [5] noted that brands attempt to generate positive eWOM. Due to these aspects, eWOM has a meaningful role in marketing strategies and enhanced overall business performance [9]. Based on its value, marketers and academics regard eWOM as an essential tool for marketing strategies [10,11] and a powerful influencer of consumers’ buying decisions and brand perceptions [7].

Consequently, grasping the key drivers of eWOM behavior has increased in importance for academic and practical frameworks due to the exponential rise in social media marketing. Prior studies have addressed multiple eWOM predictors. Thus, brand familiarity, involvement, and customer participation in social media have been established as predictors of eWOM [12,13]. As consumers display familiarity and increased interest along with participation in brand-related social media activities, they have a higher level of probability of engaging in eWOM behaviors. Also, previous studies have emphasized the role fostering loyalty and satisfaction plays in eWOM [13,14,15]. Loyal and satisfied customers are likely to act as ‘brand ambassadors’ or brand advocates in social media settings, sharing their recommendations and brand-related knowledge.

Even though eWOM denotes a popular topic in the marketing literature, there is a research gap related to gender assessments for exploring eWOM on social media. Prior research [4,5,6,7,8,9,10,11] has investigated various antecedents of eWOM; however, empirical studies tend to disregard the contrast of these drivers among male and female consumers in validating models of consumer behavior related to eWOM.

Gender has a substantial impact on consumers in terms of information processing and behavioral impact [16,17,18,19]. In fact, prior research has exhibited that male and female consumers consider different attributes when engaging with and/or considering marketing offerings [14]. However, there is limited research in terms of discovering gender specifics that drive eWOM behaviors in social media marketing strategies.

Given this gap in the research, this study aims to investigate the gender specifics of eWOM drivers in social media marketing based on the multi-sample model validation of female and male consumers. Compared to prior research that explores gender differences based on descriptive or multi-group analysis, this study tests and validates a model considering two separate samples of male and female respondents. The separate validation of the respondents helps reinforce the robustness of the models based on all the steps required for justification. For the proposed eWOM model, this research focuses on key predictors, such as brand familiarity, customer participation, involvement, loyalty, and customer satisfaction. In terms of practical implications, this research aims to advance targeted social media marketing strategies, resulting in enhanced consumer–brand interactions in an overly competitive marketing environment.

Also, this paper focuses on offering theoretical contributions to understanding gender’s role in behaviors related to eWOM in a social media setting. Specifically, this research anticipates contributions to the literature pertaining to eWOM, gender assessments and dynamics, and social media marketing. First, considering the theoretical framework of social capital theory [6], investigations of eWOM drivers showcase its important value for strategic marketing decisions. Responding to calls for new empirical studies on eWOM [6,18,20,21], this research focuses on exploring key drivers of electronic word of mouth. Baykal and Hesapci Karaca [6] focused their empirical study on exploring the mechanisms through which social capital factors affect eWOM, considering the different dimensions of social capital theory. The study by Haro-Sosa et al. [18] established a key distinction between eWOM consultation and eWOM adoption in an empirical investigation that included the moderating effect of gender, with no effect on the results. With a focus on Millennials and Gen Z consumers, Dorie et al. [21] developed a mixed-methods approach to examine critical incidents that motivate consumers in eWOM behavior. Based on a multi-sample analysis of young adults and adults, San-Martín et al. [22] explored the key role of satisfaction in impacting eWOM for mobile shopping.

Methodologically, eWOM empirical studies have mainly focused on SEM [22] or regression [23] to investigate linear relationships between proposed constructs. This focus on linear approaches represents a research gap in empirical investigations of eWOM [19]. Prior studies have not focused on validating a model that includes both linear and nonlinear relationships in establishing these proposed drivers of eWOM across genders [4,5,6,7]. Correspondingly, Alnoor et al. [19] emphasized the importance of focusing on nonlinear relationships of constructs and the extension of studies with artificial neural networks (ANNs), which offers an “accurate measurement of each variable’s relative power and increases the validity of PLS-SEM results” [19] (p. 811). As such, this empirical investigation complements prior research by integrating PLS-SEM with an ANN to develop linear and nonlinear analyses. For this study, this multi-analytical effort aims to address this research gap and establish additional insights with practical and theoretical value for eWOM frameworks. Other studies have employed this hybrid PLS-SEM–artificial neural network (ANN) modeling approach to examine the purchasing of virtual goods [24], online grocery shopping acceptance [25], impulsive buying behavior in metaverse commerce [26], and consumer preference for domestic products [27]. In exploring green apparel consumption, Armutcu et al. [28] included three different techniques—bibliometrics, PLS-SEM, and ANN.

Second, this study aims to inspect predictors of eWOM considering the perspective of gender differences. Preceding studies have investigated gender differences in social media interactions [14] and consumer information search behaviors [17]; however, the topic of eWOM across genders represents a topic that has not been thoroughly investigated in multi-sample studies. As such, other authors have emphasized the importance of focusing on the role of gender [18,19] in impacting the “consumer-brand relationship and consumers’ perceptions in a social-media-mediated environment” [14] (p. 147). Intending to fill this gap, this research sets out to improve the gender literature considering eWOM’s propagation on social media, focusing on various predictors.

Third, this study sets out to amplify the social media marketing literature by expanding the theoretical backgrounds of consumer–brand interactions in a highly competitive environment for consumers’ attention. Prior studies [4,20] have emphasized the prominence of social media marketing due to its broad spectrum of value creation opportunities for consumers, businesses, brands, and marketers. As social media marketing entails a resounding practical dimension, this manuscript focuses on enhancing perspectives for marketers in fostering and leveraging eWOM behaviors for strategic implementations.

This article is divided into multiple sections. Section 2 encompasses the literature review and the hypotheses incorporated in this study, pertaining to the main aim of this study and the aforementioned research gaps. Section 3 provides the methodology. Section 4 shows the analyses for the measurement models, hypothesis testing, and artificial neural networks for the two samples of female and male respondents. Section 5 discusses the results and compares the findings with the existing literature. Finally, Section 6 comprises the conclusions, along with theoretical and practical perspectives, while also presenting the study’s limitations and additional opportunities to expand this research.

2. Literature Review and Hypotheses Development

2.1. Social Capital Theory

Initiated by Nahapiet and Ghoshal [29], social capital theory (SCT) provides a multi-dimensional and unified perspective on social capital, that covers a multitude of aspects related to consumers’ acquisition of knowledge. Baykal and Hesapci-Karaca [6] (p. 693) defined social capital as “the set of all kinds of resources such as values, relationships, thoughts, ideas, vision, trust, culture, representations that are embedded within social networks accessed and used by the network actors”. Thus, SCT explains how the benefits received by individuals participating in a specific network (such as social media) can reflect sources of social capital that can subsequently enhance the motivation to create new knowledge. Thus, this knowledge can be reflected in eWOM which will influence other people’s decisions and intentions in social media outlets [6,29,30]. According to Nahapiet and Ghoshal [29], SCT explores three key dimensions. First, relational capital encompasses the emotional elements within a social network [29], highlighting values such as “friendship, trust, expectation and respect” [6] (p. 693). Related to this study, the relational dimension is reflected in the emotions encompassed in the social media recommendations or feedback associated with eWOM. Second, structural capital is linked to the connectivity within the network. Based on the structural dimension, eWOM is enhanced in various settings based on the network available to each consumer. Based on the interconnectedness of consumers, eWOM is widely spread across social media. Third, cognitive capital reflects resources highlighting shared representations, systems, and interpretations [29,30]. Based on shared common interests and experiences, consumers are more likely to engage in online discussions and eWOM, thus shaping other people’s perspectives. Based on these aspects, SCT offers a substantial basis for exploring eWOM on social media platforms and reflects the theoretical background of this study.

2.2. eWOM’s Strategic Importance in Social Media Settings

Dichter (1966) defined traditional WOM communication as distributing information between “a non-commercial communicator and a receiver” [31] regarding a marketing object (brand or product). The technological development of the internet has advanced a new dimension to WOM, namely eWOM. When it comes to exchanging product information, consumers rely on various platforms (e.g., search engines, blogs, and social media). This creates networks of eWOM that reshape the manner in which customers interact and relate with each other and with brands [21,32,33]. Unlike traditional word of mouth, eWOM is unrestricted by time or location due to the availability of the internet [34]. Consequently, eWOM spreads faster compared to traditional word of mouth. Indeed, online settings offer access to a broad spectrum of social media platforms and communities to enhance eWOM. These eWOM enhancement settings help other consumers discover or recommend products by providing sources of brand information. In fact, certain authors associated eWOM with online reviews, online recommendations, or online opinions, emphasizing its advantages of “speed, convenience, one-to-many reach, and its absence of face-to face human pressure” [35] (p. 6).

Especially in today’s expansive social media marketing environment, eWOM is increasing in strategic importance as consumers have access to an increased range of platforms and outlets to express their opinions about the brands they interact with. Indeed, “consumers are increasingly relying on social media to acquire knowledge about unfamiliar brands” [36] (p. 318). As social media has grown in popularity, consumer online content has developed into a popular eWOM behavior [4]. Technological advancements and social media platforms offer consumers the ability to create eWOM and consumer online content in various interactive forms, ranging from images, gifs, videos, text, animations, stories, and using emojis to highlight their brand experiences. Sharing brand experiences can impact other consumers’ perceptions and, ultimately, behaviors [15,37]. Prior research has shown that engagement with a social media brand page has a positive effect on eWOM [5,38]. Notably, the effect on consumers’ decisions is significant regardless of the eWOM valence (positive, negative, or neutral [39]). Thus, for strategic marketing investments in social media, brands, organizations, and advertisers need to comprehend how social media interactions and brand decisions impact eWOM.

Social media represents a catalyst for connection beyond geographical frontiers, enabling communication and connection worldwide. Billions of users who come from different backgrounds and countries can interact with other consumers and with brands on social media. As such, social media represents integral channels for eWOM, with a worldwide perspective, especially for international brands. Individuals from all over the world can engage in online discussions, share experiences, and exchange thoughts about products and brands on social media. On social media, eWOM receptiveness and eWOM behavior can impact consumers from different countries. To this point exploring the causes of eWOM in worldwide contexts can help provide an understanding of effective social media marketing. Studies have discovered that eWOM effectiveness needs to be addressed based on gender [16,17,40], especially since gender differences “are predominately driven by biological, cognitive, behavioral, or social influencing factors” [16] (p. 1). Thus, this research aims to provide new strategic frameworks for fostering eWOM, considering gender specifics reflected in different predictors.

2.3. Hypotheses Development

To advance the topic of eWOM in social media, this research sets out to address key predictors. Consequently, brand familiarity represents a key concept that has an impact on consumers’ storage of brand information [41], creating multiple cognitive depictions [42]. Extrapolating these cognitive dimensions of brand familiarity, other authors have linked this concept with the time spent processing brand information [43]. Thus, once a consumer engages with a brand, they start to develop familiarity with that brand. As a result, brand familiarity reflects the accumulation of brand-related knowledge based on interactions over time. Customers’ familiarity with a brand often stems from their repeated exposure, connection, experience, and understanding of that brand [44]. When consumers are familiar with a product or brand, this familiarity reduces their perception of risk associated with an acquisition [42]. Thus, a higher level of brand familiarity can lead to increased eWOM in online and social media settings. Prior research has established brand familiarity as an antecedent of eWOM [41,43,44,45], leading us to propose the following:

Hypothesis 1 (H1): 

Brand Familiarity has a direct and positive impact on eWOM for social media-liked brands for female (H1a) and male consumers (H1b).

Olaniran [46] described customers’ participation from the perspective of consumers’ contributions to the communication process of a brand. Thus, customer participation represents “the degree to which customers engage to create interactions” [47]. Based on this level of participation in social media-based brand communities, Fernandes and Castro [12] posit that active or passive participation is a key driver of eWOM, showcasing direct and indirect impact based on a ‘self-brand connection’. Passive participants monitor brand-related communications online, while active participants contribute to the amplification of the brand in various manners (sharing, discussing, or communicating) [47]. On social media, customer participation implies reviewing and sharing brand experiences, leading to the eWOM of consumers who feel compelled to share product- or brand-related benefits [10,47]. Thus, we propose the following:

Hypothesis 2 (H2): 

Customer participation has a direct and positive impact on eWOM for social media-liked brands for female (H2a) and male consumers (H2b).

Involvement is reflected in the “personal relevance” of a marketing object considering consumers’ “needs, values and interests” [48]. Consumers who feel a high level of involvement with a brand are inclined to devote substantial time and effort to processing information about the brand, cognitively considering the brand’s features, image, and characteristics [49,50]. Involvement can take many forms, such as the cognitive, affective, and motivational; however, this concept does not definitely lead to a specific behavior. Furthermore, involvement is regarded as the motivation with which consumers seek information and knowledge that can be utilized to manage and moderate any potential risk inherent in online settings to facilitate a buying decision [51]. Involved customers tend to concentrate their attention on brands [13]. Thus, consumers’ increased interest, along with their allocated resources of time and energy in the decision-making process, suggests that involvement is an antecedent of eWOM, as consumers have a higher predisposition to want to share their thoughts and opinions in online settings and social media platforms [13,52]. Hence the following:

Hypothesis 3 (H3): 

Involvement has a direct and positive impact on eWOM for social media-liked brands for female (H3a) and male consumers (H3b).

Loyalty is recognized as “a deeply held commitment to re-buy or re-patronize a preferred product/service consistently in the future, thereby causing repetitive same brand or same brand-set purchasing, despite situational influences and marketing efforts having the potential to cause switching behavior” [53] (p. 34). In the marketing literature, loyalty reflects a ‘biased behavior’ conveyed over time. Thus, brands understand the fact that loyalty enhances their chances of engaging in eWOM behaviors [54], leading to the strategic importance of eWOM in social media marketing. Loyal customers act as ‘brand ambassadors’ or ‘brand advocates’ for other consumers, especially in social media outlets. In fact, studies have shown that social media members with pronounced levels of brand loyalty are more prone to foster positive eWOM [14]. Prior research has emphasized various important effects of brand loyalty, including recommendations to others and eWOM [13,55,56,57]. As explained by Eelen et al. [58], loyal consumers tend to display stronger mental associations and memories with specific brands from their direct experiences, making those brands more accessible in their minds. Thus, ideas about the brand are accessible and easier to recall for loyal consumers, compared to those who are not loyal. This level of brand accessibility for loyal customers leads to increased levels of word of mouth. To expand this hypothesis and advance knowledge on gender’s role in marketing strategy, Rialti et al. [14] explored and confirmed the impact of loyalty on eWOM for male and female samples, discovering that female consumers are more prone, compared to men, to recommend a brand they are loyal to. Thus, we propose the following:

Hypothesis 4 (H4): 

Loyalty has a direct and positive impact on eWOM for social media-liked brands for female (H4a) and male consumers (H4b).

According to Oliver [53], customer satisfaction describes consumers’ perceptions of “fulfillment” with a brand, product, or service. This level of fulfillment is established based on consumers’ evaluations of the product’s performance considering initial expectations. Thus, in the marketing literature, satisfaction represents a highly subjective concept that involves an emotional process from the consumers’ perspectives [15]. Satisfaction plays a key role in any commercial setting. Studies have examined the relationship between customer satisfaction and eWOM, highlighting a positive link between these concepts [55,59]. Thus, consumers might feel compelled to engage in eWOM and share their experiences when their expectations are met or surpassed. Specifically, studies have considered satisfaction as a driver of eWOM, as satisfied customers have a stronger predisposition to share their experiences in social media settings [13,15]. Indeed, prior research has confirmed the positive impact of satisfactory experiences on consumers’ eWOM and overall intention to recommend a brand [35]. Thus, we aim to examine the following:

Hypothesis 5 (H5): 

Customer satisfaction has a direct and positive impact on eWOM for social media-liked brands for female (H5a) and male consumers (H5b).

To expand the study, this research aims to focus on the moderating roles of age and time spent on social media considering the relationship between satisfaction and eWOM. Satisfaction is a key construct of paramount strategic value in the marketing literature. Prior studies have recognized customer satisfaction as one of the most important antecedents of eWOM, because satisfied consumers are more likely to share their experiences in online (and offline) settings [13,15]. While examining the role of satisfaction as a driver of eWOM, it is important to note that its effects can vary based on consumers’ characteristics.

The previous literature has examined multiple consumer characteristics as moderators. Specifically, there is an increasing interest in exploring consumers’ age in empirical studies, notably in aging societies in which the adult population reflects a substantial segment with purchasing power [60]. Age was considered in various settings, such as the moderating role of age on the influence of entertainment and subjective norms on m-shopper satisfaction and WOM [22] and on social media marketing activities and customer brand engagement [61]. Another study found that while eWOM is significantly linked to customer satisfaction, there is a moderation effect based on consumers’ altruism and age [62]. These aspects are notably important considering the significance of consumer satisfaction in today’s overcrowded space of offerings and the effect of satisfaction in generating eWOM to enhance brands’ reach in social media marketing.

Moreover, the role of social media usage in eWOM has been explored based on contexts [63] and platform type [64]; however, this study sets out to uncover the moderating role of ‘time spent on social media’, considering the relationship between satisfaction and eWOM. While time spent on social media has been examined in different psychological studies [65,66], to the knowledge of the authors, this dimension has not been the subject of an eWOM social media marketing study. As such, this proposition reflects a novel perspective in the existing social media marketing literature. Nonetheless, we note the importance of exploring the habit of using social media as a moderator in this satisfaction–eWOM relationship. Consumers who spend more time on social media tend to be more engaged in digital environments and interactions. Thus, these consumers are more prone to translating their satisfaction into eWOM behaviors. Therefore, we propose the following:

Hypothesis 6 (H6): 

Age is a significant moderator for the relationship between satisfaction and eWOM for female (H6a) and male consumers (H6b).

Hypothesis 7 (H7): 

Time spent on social media is a significant moderator for the relationship between satisfaction and eWOM for female (H7a) and male consumers (H7b).

Figure 1 presents the proposed model.

3. Method

A web survey was developed for data collection. Respondents were recruited randomly from social media community groups (on Facebook) and contact lists. The web survey was accompanied by a short survey introduction and a link to the questionnaire. The respondents were invited to indicate their preferred brand on Facebook. Facebook brand pages have “become a commonly-used marketing channel and their importance as a sales channel is enhancing” [36], highlighting the relevance of examining brand pages on Facebook for the context of this study.

Overall, 324 responses were collected from female respondents and 248 responses from male respondents. These responses were evaluated for accuracy, outliers, and incomplete responses. The final sample included 272 usable responses for female and 191 for male respondents, showcasing acceptable accurate response rates (of 83.9% for female respondents and 77% for male respondents). All participants reported using Facebook on a regular basis. Male respondents tend to spend an average of 6.79 h/week on Facebook, whereas female respondents spend 7.82 h/week on Facebook.

For both samples, generation Y were the most prominent respondents (of 49.6% for female respondents and 55.5% for male respondents). The education level of most of the participants was a bachelor’s degree (89 percent for female respondents, 86.6 percent for male respondents), with an average or above-average income (64.7 percent of male respondents reported an average monthly income of 500–3000 EUR/month, whereas 62.5 percent of female respondents reported the same categories of monthly income).

To verify the purpose of this research, a quantitative research framework was applied via an online questionnaire. In the beginning, the respondents were asked for their consent to complete the questionnaires, they were informed of the anonymized answers given in the survey and the fact that it would not be possible to identify each individual respondent. The online questionnaire consisted of demographic questions and purpose-specific questions. The demographic details encompassed generation, education level, and Facebook usage. The questions related to the purpose of the research encompassed scale items from renowned studies to attest to the validation of the research instrument, focusing on the research constructs. Specifically, to ensure the validity of the proposed model, all items pertaining to latent constructs were adapted for this study and adopted from prior studies as follows (Appendix A): the brand familiarity scales were adapted from Acharya [67]; the items for customer participation were adapted from various sources, such as Casaló et al. [68], Kamboj et al. [69], and Phan Tan [47]; the involvement scales were adjusted from Chen [70]; the Vinerean and Opreana [71] scales were adjusted from Zeithaml et al. [72], Rialti et al. [14], and Vinerean and Opreana [71] to assess customer loyalty; customer satisfaction was explored based on items extracted from Ruiz-Alba et al. [15] and Rialti et al. [14]; and the eWOM items were adjusted from Moisescu et al. [5], Phan Tan [47], and Choi et al. [73]. All scale items were measured using five-point Likert scales. Throughout the questionnaire, randomization was used to avoid common method bias.

In total, 272 of the completed surveys were gathered and examined for female respondents, and 191 for male respondents. These two samples were analyzed based on structural equation modeling (SEM) and artificial neural network (ANN) data analysis strategies. Specifically, partial least squares structural equation modeling (PLS-SEM) is a key and popular statistical technique that aids in exploring compensatory and “linear relationships between constructs” [5,74,75]. While PLS-SEM is valuable in confirming relationships, its main limitation lies in its ability to capture nonlinear interactions between constructs [19]. These nonlinear interactions are becoming progressively more important in understanding consumer behavior in digital settings, including eWOM. To account for this limitation, we aim to extend the empirical study with a complementary analysis—artificial neural networks. ANNs aid in discovering non-compensatory and nonlinear connections between constructs [19]. Thus, the integration of PLS-SEM and ANN offers a hybrid procedure with an advanced analytical perspective and amplifies insights into eWOM and social media marketing. SPSS v.25 and SmartPLS v.4 were used to develop the data analyses and to assess two models for female and male respondents. The empirical investigations are presented in the following section.

4. Results

For this study, hypothesis testing encompassed the PLS-SEM approach. PLS-SEM provides the opportunity to explore relationships among complex variables based on exploratory research and conceptual development [74,75], which highlights this study’s aim. Additionally, considering the multi-sample approach of this study, PLS-SEM showcases effective performance regardless of sample size and normal data distribution [74]. PLS-SEM involves two stages, i.e., the measurement model and the structural model (detailed in the next sub-sections). This procedure was applied for the male sample and the female sample to assess gender in a multi-sample framework, according to the study’s aim.

Prior to model testing, several quality tests were established. Each sample involved data screening procedures to eliminate irrelevant and incomplete observations. Moreover, the empirical analysis also addressed the issue of common method bias. First, the data collection process considered a priori processes to reduce potential predispositions and biases in the responses [64]. Second, the analysis for both samples involved the assessment of the multicollinearity condition based on the variance inflation factor (VIF) [76]. As the VIF-registered results (Appendix A) did not exceed the recommended 5.0 level [74,77,78], multicollinearity did not represent an issue for the female or the male samples of respondents [74].

4.1. Analysis of Measurement Models for Female and Male Samples of Respondents

The proposed research model (Figure 1) was explored in a structural equations modeling (SEM) analysis using SmartPLS software v.4 [79,80], which enabled the examination of the dependency connections among the model’s reflective constructs. For the measurement models, the constructs were assessed based on convergent validity (loadings, internal consistency, average variance extracted—AVE and CR) and discriminant validity (the heterotrait–monotrait criterion) according to well-established thresholds [78,79,80,81]. Appendix A presents the loadings for all items which surpassed the recommended threshold of 0.7 for both samples [78], as the lowest loadings were registered for LOY5, highlighting a score of 0.783 for the female sample and a score of 0.789 for the male sample. Further, composite reliabilities (CR [78,81] and Cronbach’s alpha [74]) values met the 0.7 threshold for all latent variables examined in both samples, as exhibited in

Table 1. Convergent Validity.

Sample	Construct	Cronbach’s Alpha (>0.7)	Composite Reliability (CR > 0.7)	Average Variance Extracted (AVE > 0.5)
Female sample	BF	0.892	0.895	0.754
	CP	0.787	0.788	0.824
	INV	0.869	0.874	0.792
	LOY	0.872	0.878	0.660
	SAT	0.865	0.867	0.788
	eWOM	0.885	0.887	0.743
Male sample	BF	0.893	0.896	0.759
	CP	0.731	0.731	0.788
	INV	0.865	0.865	0.787
	LOY	0.890	0.903	0.694
	SAT	0.864	0.880	0.785
	eWOM	0.874	0.878	0.726

Notes: BF = brand familiarity; CP = customer participation; INV = involvement; LOY = loyalty, SAT = customer satisfaction; eWOM = electronic word of mouth.

. Moreover, both samples reflected an average variance extracted (AVE) higher than 0.5 (Table 1), as the lowest value of 0.660 > 0.5 was recorded for loyalty (female sample). All the criteria for convergent validity were met [78].

Next, both samples of respondents were assessed based on discriminant validity testing which implied measuring the degree to which the analyzed constructs are distinct. Thus, we used the heterotrait–monotrait (HTMT) test (

Table 2. Discriminant validity based on HTMT testing.

Sample		BF	CP	INV	LOY	SAT	eWOM
Female sample	BF
	CP	0.614
	INV	0.649	0.591
	LOY	0.64	0.629	0.807
	SAT	0.692	0.612	0.74	0.788
	eWOM	0.658	0.654	0.747	0.76	0.799
Male sample	BF
	CP	0.647
	INV	0.542	0.599
	LOY	0.515	0.697	0.850
	SAT	0.612	0.73	0.698	0.698
	eWOM	0.72	0.683	0.69	0.762	0.787

Notes: BF = brand familiarity; CP = customer participation; INV = involvement; LOY = loyalty, SAT = customer satisfaction; eWOM = electronic word of mouth.

). The HTMT values are required to be less than 0.9 [74,82] and, for both samples, this condition was met, indicating the fact that none of the concepts are similar. The calculated results showed that measurement models for the female and male samples conformed with the discriminant validity criteria.

4.2. Hypothesis Testing of Direct Effects

Considering the validation of the measurement models for both explored samples, of female and male social media users, hypothesis testing entailed the assessment of the structural model. Using SmartPLS4, the structural model involved the estimation of the models’ quality (for both models, considering the two examined samples) and hypothesis analysis. Thus, a bootstrap procedure was implemented to test the hypotheses and their associated relationships. Considering the proposed data analysis approach, this empirical investigation aimed to apply and validate the proposed hypotheses set (see Figure 1) in two models in a multi-analytical effort based on multi-sample analysis, considering samples of female and male respondents. Each sample was investigated, modeled, and confirmed separately. For the structural models of the female and male samples, we considered the following criteria: R² (explained variance), f² (effect size), Q² (predictive relevance), and the size and statistical significance of the structural path coefficients.

Showcasing ‘in-sample predictive power’ [75,83], R² indicated the ability of the antecedents to explain eWOM. The results for R² showcased substantial values of 0.614 or 61.4% for the female sample, and 0.654 or 65.4% for the male sample. Thus, the outcome variable is determined by the predictor variables. In terms of examining the effect sizes, highlighted by the f² values of latent variables, we considered the widely recognized classification of small, medium, and large effects, proposed by Cohen [84], for values of 0.02, 0.15, and 0.35, respectively. Overall, the resulting effect size values were deemed to be of small and medium effects.

Furthermore, a blindfolding procedure was developed to establish the model’s predictive relevance based on Q² values. As Q² values were higher than zero for the main dependent construct of eWOM, the path models’ predictive relevance was confirmed for both samples. Specifically, Q2 values were 0.596 and 0.617 for the female and male samples, respectively. Additionally, the goodness of fit value explored through the standardized root mean squared residual (SRMR) was lower than the recommended threshold of 0.080 since for the female sample SRMR was 0.05, and for the male sample it was 0.066.

Considering the size and statistical significance of the structural path coefficients, for the female sample, four out of five hypotheses were accepted (

Table 3. Structural model analysis results.

Sample	Hypothesis	Path Coefficients	2.50% 1	97.50% 1	St. dev.	T-Statistics	f-Square	p-Values	Result
Female sample	H1a: BF -> eWOM	0.109	0.003	0.223	0.056	1.954	0.017	0.051	Reject
	H2a: CP -> eWOM	0.140	0.031	0.247	0.055	2.521	0.032	0.012	Accept
	H3a: INV -> eWOM	0.193	0.077	0.314	0.061	3.18	0.042	0.001	Accept
	H4a: LOY -> eWOM	0.196	0.052	0.327	0.070	2.801	0.039	0.005	Accept
	H5a: SAT -> eWOM	0.305	0.171	0.438	0.069	4.445	0.103	0.000	Accept
Male sample	H1b: BF -> eWOM	0.303	0.155	0.428	0.069	4.395	0.165	0.000	Accept
	H2b: CP -> eWOM	0.018	−0.119	0.15	0.069	0.255	0.000	0.799	Reject
	H3b: INV -> eWOM	0.002	−0.166	0.215	0.098	0.022	0.000	0.982	Reject
	H4b: LOY -> eWOM	0.346	0.116	0.52	0.102	3.377	0.128	0.001	Accept
	H5b: SAT -> eWOM	0.304	0.159	0.463	0.077	3.935	0.126	0.000	Accept

Notes: BF = brand familiarity; CP = customer participation; INV = involvement; LOY = loyalty, SAT = customer satisfaction; eWOM = electronic word of mouth; ¹ CI = confidence interval [2.5–97.5%].

, Figure 2). For the male sample, three out of five hypotheses were accepted based on significant relationships considering the t-statistics calculations. These relationships are discussed and compared to prior studies in Section 5.

4.3. Hypothesis Testing of Interaction Effects

The analysis was expanded based on an analysis of interactions [85] (

Table 4. Interaction results.

Sample	Hypothesis	Path Coefficients	2.50%	97.50%	St. dev.	T-Statistics	p-Values	Result
Female sample	TimeSpent × SAT -> eWOM	0.007	−0.057	0.06	0.03	0.223	0.823	Reject
	Age × SAT -> eWOM	−0.075	−0.145	−0.008	0.035	2.151	0.032	Accept
Male sample	TimeSpent × SAT -> eWOM	−0.013	−0.152	0.135	0.073	0.183	0.855	Reject
	Age × SAT -> eWOM	0.069	−0.051	0.206	0.065	1.055	0.292	Reject

). Considering the calculations, only the interaction between age and satisfaction is significant. Overall, this result shows that age exerts a significant and negative effect on the connection between satisfaction and eWOM, showing that the higher the age of the social media users, the weaker the association between customer satisfaction and eWOM.

4.4. Artificial Neural Networks

Considering previous recommendations [85,86,87], an ANN was applied only to the model’s significant drivers. As a result, two analyses using artificial neural networks were applied for the dependent variable of eWOM. For the female sample, ANN modeling included four predictors—customer participation, involvement, satisfaction, and loyalty. For the male sample, ANN modeling included three predictors—brand familiarity, satisfaction, and loyalty. For the application of ANNs, the multilayer perceptron with a feedforward backpropagation algorithm was considered. Based on a comprehensive study of empirical ANN investigations in marketing settings, Kalinić et al. [86] found that Sigmoid was the activation function most widely used for hidden and output layers, with one hidden layer. Therefore, the same function was applied in this empirical investigation (Figure 3). Furthermore, the ANNs reflected the ten-fold cross-validation framework (90% data for training and 10% for testing), consistent with previous implementations of this technique [88,89,90]. Moreover, the ANN models were examined for relevancy and accuracy based on low calculated scores for the root mean square error (RMSE) and the average RMSE. The findings calculated in the table show that both ANN models (for the female and the male samples) reflect acceptable levels of predictive precision [85,87,88].

To measure the performance of the analysis, ANN models for both samples were investigated considering the calculated R² coefficient [87,91]. For the female sample, the R² value of 0.6054 reflected that the ANN model explained 60.54% of eWOM’s variance. Similarly, for the male sample, the ANN model explained 67.21% of the variance of the eWOM. Comparing these R² values calculated in ANN modeling to their corresponding PLS-SEM investigations, we noted that the R²-ANN value was slightly higher for the female sample, and considerably higher for the male sample. Thus, these results (

Table 5. Artificial neural network results.

ANN	Female Sample		Male Sample
	R-Square Average:	0.6054	R-Square Average:	0.6721
	Training RMSE	Testing RMSE	Training RMSE	Testing RMSE
1	0.094	0.091	0.093	0.094
2	0.094	0.093	0.107	0.090
3	0.098	0.083	0.093	0.083
4	0.098	0.118	0.093	0.064
5	0.099	0.087	0.087	0.068
6	0.103	0.074	0.090	0.102
7	0.095	0.084	0.092	0.071
8	0.092	0.122	0.094	0.115
9	0.096	0.094	0.086	0.045
10	0.092	0.088	0.098	0.064
Average	0.096	0.093	0.093	0.080
St. deviation	0.003	0.015	0.006	0.021

) indicate the fact that eWOM, as an endogenous variable, was better explained in a neural network analysis [88].

Additionally, the relative importance of the predictors’ contributions was ranked using ANN modeling [88,92]. Compared to PLS-SEM and the resulting coefficients, ANN modeling showcases normalized importance (

Table 6. Sensitivity analysis for ANNs.

ANN	Female Sample				Male Sample
	CP	INV	LOY	SAT	BF	LOY	SAT
1	0.203	0.229	0.213	0.355	0.275	0.474	0.251
2	0.210	0.269	0.212	0.309	0.379	0.322	0.299
3	0.199	0.258	0.266	0.276	0.160	0.572	0.268
4	0.207	0.211	0.229	0.353	0.294	0.363	0.343
5	0.180	0.226	0.318	0.277	0.401	0.365	0.234
6	0.206	0.170	0.350	0.275	0.229	0.461	0.310
7	0.162	0.235	0.262	0.341	0.241	0.397	0.362
8	0.163	0.237	0.251	0.349	0.421	0.303	0.276
9	0.199	0.278	0.196	0.327	0.392	0.356	0.253
10	0.200	0.193	0.270	0.336	0.307	0.396	0.298
Average	0.193	0.231	0.257	0.320	0.310	0.401	0.289
Normalized importance	58.57%	70.37%	77.87%	96.56%	75.47%	93.87%	69.55%

). Thus, the next step for ANN implied determining the relative significance of each predictor. Normalized importance was calculated as the “ratio of relative importance of each predictor over the highest relative importance” [85]. As such, for the female sample, the most prominent predictor of eWOM was satisfaction; whereas, for the male sample, the most important determinant was loyalty.

Table 7. Comparison and ranking of PLS-SEM and ANN modeling results.

Sample	Examined Driver of eWOM	Path Coefficient	ANN Result—Normalized Importance	PLS-SEM Ranking	ANN Ranking	Remark
Female sample	CP	0.140	58.57%	4	4	Matched
	INV	0.193	70.37%	3	3
	LOY	0.196	77.87%	2	2
	SAT	0.305	96.56%	1	1
Male sample	BF	0.303	75.47%	3	2	Partially matched
	LOY	0.346	93.87%	1	1
	SAT	0.304	69.55%	2	3

Notes: BF = brand familiarity; CP = customer participation; INV = involvement; LOY = loyalty, SAT = customer satisfaction; eWOM = electronic word of mouth.

provides a summary of the ANN and PLS-SEM results [93], based on a ranking comparison of eWOM’s significant predictors, considering the two samples of female and male participants.

5. Discussion of Results

The testing of the hypotheses (showcasing direct effects) was explored before the interaction analysis. As shown in Table 3, H1 explored the relationship between brand familiarity and eWOM for female and male samples. This hypothesis showcased contrasting results, as it was rejected for the female sample (H1a); however, it was confirmed and supported for the male sample (H1b) based on a positive and significant result of 0.303 (t = 4.395, p-value < 0.001, 95% CI = [0.155, 0.428]). Thus, this hypothesis is only accepted for the male sample, consistent with previously reported outcomes [41,44,45]. Based on these results, male consumers are more prone to becoming involved in eWOM behaviors in online and social media settings if they are familiar with a specific brand. For female respondents, this hypothesis should be explored in further studies.

Similarly to H1, H2 reflected contrasting results. In exploring the positive impact of customer participation on eWOM, H2 was confirmed only for the female sample based on the significant result of 0.140 (H2a, t = 2.521, p-value = 0.012, 95% CI = [0.031, 0.247]). This finding is consistent with prior studies [10,47]. As customer participation involves reviewing and sharing brand experiences in social media settings, female consumers might be compelled to engage in eWOM behavior to share product or brand-related benefits based on their brand experiences. While this finding is important and relevant for the expanding literature on customer participation studies, the relationship was not confirmed for the male sample. For male respondents, this hypothesis should be examined further.

H3 was aimed at exploring the role of consumers’ involvement in eWOM behavior. For female respondents, this hypothesis was confirmed (H3a, β = 0.193, t = 3.180, p-value = 0.001, 95% CI = [0.077, 0.314]), showcasing congruence with prior studies in highlighting involvement as an antecedent of eWOM [13,53]. In this case, female consumers who are involved with a particular brand are inclined to reflect a higher predisposition to share their insights and feedback via eWOM on social media platforms, as the new communication medium. With regard to the sample of male respondents, this hypothesis was rejected based on a t-value of 0.022 (H3b, p-value = 0.982).

For both samples, the results demonstrate that customer loyalty has a strong and very powerful positive effect on electronic word of mouth (H4). For female respondents, H4 was confirmed based on a positive and significant result of β = 0.196 (H4a, t = 2.801, p-value = 0.005, 95% CI = [0.052, 0.327), and for male respondents the significant result was β = 0.346 (H4b, t = 3.377, p-value = 0.001, 95% CI = [0.116, 0.520]). This is an important finding that corroborates the previous literature [13,14,54,55], focusing on enhancing theoretical frameworks that highlight loyalty and eWOM. Thus, engaging in eWOM is highly determined by the loyalty both male and female consumers feel toward a brand or company. As other studies have shown, loyal customers are more prone to develop positive eWOM. As such, this study reconfirms the importance of customer loyalty for marketing strategies, including driving positive recommendations from other consumers.

Considering the two examined models of female and male consumers, the results also reinforce the hypothesis that customer satisfaction reflects a positive and significant impact on eWOM. For female respondents, H5 was confirmed based on a positive and significant result of β = 0.305 (H5a, t = 4.445, p-value < 0.001, 95% CI = [0.171, 0.438), and for male respondents the significant result was β = 0.304 (H5b, t = 3.935, p-value < 0.001, 95% CI = [0.159, 0.463]). As expected, satisfaction and loyalty are influential for both samples, echoing the pertinent literature [13,14,55,59]. As such, this study provides additional theoretical support for this literature framework. As a key marketing construct, satisfaction has a strategic role in any organization’s success. Based on these results, we can state that satisfied consumers might feel more compelled and inclined to develop eWOM and share feedback on their brand-related experiences. This study showcases the important role of satisfaction in driving eWOM for male and female social media users, especially considering the growth of social media marketing.

Regarding the ANN results, for the female sample, the ranking reflects a match and congruence of the results considering the two analyses. For the male sample, only the main predictor (loyalty) reflected the same ranking in both analyses, thus a partial match is showcased in this case.

6. Conclusions

6.1. Theoretical and Practical Implications

From a theoretical perspective, this research primarily provides contributions to the gender differences in the marketing literature, showcasing the significance of these contrasts in examining predictors of e-WOM in a social media-mediated environment. As such, this research is valuable in emphasizing the measure in which gender has a significant impact on consumers’ social media interactions and eWOM behavior.

By integrating insights from social capital theory, this study’s findings emphasize several contrasts between male and female consumers. From a psychological perspective, gender differences can be explained based on selectivity theory and the socialization perspectives of each group [94]. Male and female consumers exhibit different perspectives in terms of acquiring, providing, and processing information [95]. Considering the selectivity theory, female consumers tend to process information in a more comprehensive manner, with an intent to consider as much available information as possible [96]. In contrast, men rely on heuristics and select key information in a manner that minimizes their time and effort [94,97]. Moreover, male consumers reflect a key focus on goals and efficiency in decision-making, based on their cognition-driven perspective [96]. With a focus on heuristic processing [97], male consumers prioritize familiarity with brands in their eWOM behaviors.

Prior studies [94,97] (Bae, 2011; Hyde, 2014) found that females are more prone to trust and be receptive to views from other people than men. Moreover, female consumers have a higher predisposition for social connections, which leads to enhanced consumer participation and WOM [94]. As such, female consumers are more prone to participate in brand-related initiatives, in collaborative settings, and in online discussions. Therefore, female consumers prioritize active involvement and participation in relation to eWOM, as shown in the findings of this study.

Considering the contributions to gender-based studies in eWOM, this study has generated certain interesting and contrasting findings. On the one hand, female consumers’ engagement in eWOM behavior is determined by customer involvement, interaction with the brand, brand loyalty, and satisfaction (which is the most prominent predictor). On the other hand, male consumers engage in eWOM based on brand familiarity, satisfaction, and loyalty (which is the most important predictor). As a contrasting result, males are more prone to eWOM behaviors in online and social media settings if they are familiar with the brand. Considering gender assessments, females need additional levels of involvement and participation with the brand to develop eWOM, highlighting the complexity of eWOM behaviors for female consumers.

Moreover, this empirical study provides a key contribution to the eWOM literature. By empirically investigating and validating key eWOM influences (i.e., brand familiarity, customer participation, involvement, loyalty, and customer satisfaction), this manuscript offers an extensive understanding of the processes that drive eWOM on social media. These findings underline the essential roles of loyalty and satisfaction in generating eWOM and consumers’ associated intentions for this communication format, providing supplementary insights into consumers’ interactions and their processes related to information sharing.

Also, this research offers contributions to the social media marketing literature. Specifically, this research enhances theoretical frameworks by emphasizing the key role of eWOM in social media, along with the dynamics of user-generated content and brand advocacy in online social communication frameworks. This research accentuates the value of customer–brand relationships in social media marketing. Thus, these findings inform and amplify the theoretical background for future studies that aim to focus on consumer–brand interactions in social media marketing.

This study also enhances the literature on social capital theory (SCT) for exploring eWOM. Social capital reflects an “intangible asset that arises from networks of relationships characterized by norms of voluntary participation, reciprocal exchanges, and trust among individuals” [98]. From this perspective, the elements of SCT are embedded in eWOM and consumers’ predisposition to share their brand-related opinions and perspectives online. In the context of this paper, the dimensions of SCT are reflected in the results of the manuscript. This study’s findings confirm the hypotheses for loyalty and satisfaction in relation to eWOM, for both female and male consumers, aligning this finding with the relational dimension of SCT in terms of fostering long-term and strong relationships with consumers’ favorite brands. Further, structural capital refers to connectivity [98]. In this study, we explored brand familiarity and customer participation in brand-related activities that foster this connectivity and networks. Brand familiarity significantly influences male consumers’ eWOM behavior, aligning this finding with structural capital. Customer participation was deemed to be a key driver only for the female sample, extending the structural capital to this perspective. Moreover, the cognitive dimension of SCT is observed in this study based on the concept of involvement which is regarded as the motivation with which consumers seek information and knowledge.

From a practical perspective, the findings of this research point out the need for marketers and managers to develop gender-segmented strategies and campaigns to enhance eWOM in a social media-mediated environment.

Regarding male consumers, social media strategies should have a key role in driving brand familiarity. Thus, the focus of social media marketing strategies should be to engage male consumers in online settings, to increase the level of brand visibility and familiarity in social media settings.

Concerning female consumers, eWOM behavior is driven by customer participation in brand-related activities and involvement. As such, social media activities should be strategically used by managers to foster the participation of female consumers. Also, social media enables dialog between brands and consumers. Managers should leverage this dialog to communicate directly with female consumers and create premises for involvement in social media conversations [14]. Thus, marketing strategies centered on social media should converge on participation and involvement.

Male and female consumers are more inclined to recommend a brand in social media settings if they are loyal to that brand. Thus, loyalty needs to be nurtured by companies to drive positive eWOM and enhance brand advocacy. Additionally, satisfaction must be monitored and addressed due to its favorable impact on eWOM for female and male respondents. While there are certain similarities between female and male respondents, especially related to the importance associated with satisfaction and loyalty, marketers need to address and consider gender in their (general) marketing and social media marketing strategies. Marketers can use these findings to leverage user-generated content for promotional strategies and boost their social media brand advocacy levels. In social media settings, consumer interactions are instantaneous and highly influential for future purchasing decisions. Thus, this research emphasizes the importance of targeted campaigns based on segmentation to drive eWOM based on gender assessments.

Specifically, marketers should delineate the predictors of eWOM, and its associated behavior and design tailored campaigns for male and female social media users. These campaigns should focus on different messaging and content types to showcase alignment for male and female audiences, and consequently, maximize the effectiveness of social media marketing. The findings support the strategic role of eWOM. This strategic role should be leveraged to foster consumer–brand interactions and brand enhancement in social media frameworks, while also considering gender differences. Brands that aim to attract male consumers should emphasize brand recognition and brand credibility, considering the fact that men are more likely to be influenced by familiarity and factual content. Whereas brands targeting female consumers should focus on interactivity and participatory marketing strategies to engage this group.

Based on this research, marketers can develop various tactics to advance their social media marketing strategies, while leveraging and fostering eWOM. First, for both male and female consumers, marketers should constantly engage in social media listening and the assessment of share of voice to monitor and engage in relevant brand-related social media discussions. These processes can help companies address negative issues in a swift manner and capitalize on positive developments regarding brand interactions. As such, marketers should identify and engage consumers who mention the brand in a positive way, to further enhance positive eWOM prospects. Second, for female consumers, marketers should implement initiatives that focus on developing a brand community on social media, based on user-generated content. Building a brand community would foster a feeling of belonging among customers, which would further enhance their satisfaction and loyalty. For female consumers, social media marketing strategies should also emphasize brand storytelling. Third, especially for female consumers, marketers should develop interactive campaigns (i.e., treasure hunts or contests) that would enable consumers to participate, become involved with the brand, share their brand experience, encourage brand interactions, and advance the creation of user-generated content (UGC). Fourth, as brand familiarity is a key factor for male consumers, brands should focus their marketing efforts on establishing more online content with educational and technical dimensions for this group. Fifth, to appeal to male consumers, brands could develop incentivized review mechanisms based on gamification.

6.2. Limitations and Future Research Directions

There are certain theoretical and methodological limitations associated with this study that need to be addressed for research transparency. First, this study considered a key demographic characteristic of online shoppers (i.e., gender); however, other demographic or psychographic characteristics should be considered. Consumers with different demographic characteristics may have different motivations to create eWOM.

Second, this study was developed considering the Facebook context, as it represents the most popular social media platform [3]. Hence, the results may differ across distinct platforms. To consider the findings’ generalizability, future empirical investigations can reapply this model to other social media outlets for investigating eWOM behavior, under the gender assessment. As such, this study could be extended to different social media platforms (Instagram, TikTok, or LinkedIn) to address the issue of generalizability. Moreover, a study could also be expanded to include cultural premises for eWOM behavior.

Third, this research disregarded a predetermined classification of products/services. While this aspect helped minimize biases and common method variance [64], subsequent research may emphasize a specific category of brands, e.g., eWOM on social media pertaining to hotels or luxury goods.

Fourth, this study included age as a moderator. However, research could be extended to include generational differences (e.g., Millennials vs. Gen Z vs. Gen Alpha) in terms of influencing eWOM behavior. Future studies could address this research direction and provide a deeper exploration of age effects in terms of driving eWOM.

Future studies could enhance this research by discovering other avenues related to eWOM. While this study focused on primary data, additional studies could utilize secondary data based on social media listening and share-of-voice tools to track eWOM trends, along with UGC brand developments in social media. Moreover, future studies could explore the effect of influencer marketing on consumers’ eWOM to advance insights into the effectiveness of these brand collaborations from users’ perspectives. Also, considering the expansion of technologies in virtual and augmented realities, new studies could focus on eWOM receptiveness in these new digital channels. Additional studies could focus on including control variables that would enhance eWOM understanding. For example, studies could include the frequency of content creation on social media. This control variable could provide new insights as consumers who are more actively posting and creating content on social media might have a higher predisposition to engage in eWOM behavior.