Social media marketing and brand loyalty: exploring interrelationships through symmetrical and asymmetrical modeling

Abstract

Purpose

This study aims to examine the interrelationships between social media marketing activities, self-brand connections, brand equity, trust and loyalty.

Design/methodology/approach

A total of 402 valid responses were collected from Amazon MTurk, and the data were subjected to partial least squares structural equation modeling and fuzzy-set qualitative comparative analysis (fsQCA).

Findings

Findings indicate that social media marketing activities strongly and positively influence self-brand connection, brand equity and brand trust. Moreover, brand loyalty was strongly and positively influenced by self-brand connection, brand equity and brand trust. Moreover, the findings from fsQCA indicate that three causal paths lead to a high level of brand loyalty, and one causal path determines a low level of brand loyalty.

Originality/value

This research extends current knowledge by bridging the literature between social media marketing activities and branding using self-brand connections. Additionally, this study uses the strength of two complimentary methods – symmetrical and asymmetrical modeling – to uncover how social media marketing activities bridge customer-brand relationships.

Introduction

Social media has become essential to modern marketing strategies, providing customers with instant and direct means of engaging with companies (Hafez, 2021). Social media ads are 55% more effective than conventional ads (Seo and Park, 2018). Over 91.9% of businesses with over 100 workers use social media for marketing, making it a crucial brand platform (Chaffey, 2022). Social media marketing activities (SMMA) are marketing practices applied through social media mediums and created by brands to impact consumers’ behavior (Chen and Lin, 2019; Khan, 2022). SMMAs enable consumers to interact with brands at different stages of their purchasing journey (Chu et al., 2020). While these activities are conducted in different virtual consumer environments, many scholars have emphasized the essential role of brand communities formed and maintained through social media (Laroche et al., 2013). SMMAs such as online brand communities have improved customer-customer and business-customer interactions (Zollo et al., 2020), making them an integral customer relationship management tool (Seo and Park, 2018).

Interestingly, some believe brands are unwanted intruders on social media (Fournier and Avery, 2011), making the research on SMMAs and branding inconclusive (Laroche et al., 2013), suggesting additional research in this domain (Zollo et al., 2020). As per Guo and Zhou (2021), the link between service providers’ usage of brand communities and brand equity is understudied in marketing literature. Studies on SMMAs have discussed their impact on purchase intentions (Dolega et al., 2021), revenue generation (Phan et al., 2011), consumer attitude (Jin, 2012) and inclination to premium rates (Godey et al., 2016). While Ebrahim (2020) investigated the impact of SMMAs on brand equity and loyalty, the role of self-brand connection has been overlooked. Notwithstanding the importance of self-brand connections and trust to establish the relationship between customers and brands, it is unclear how consumers develop trust toward the brand through SMMAs nor how the connection between their self and the brand is enhanced (Ebrahim, 2020).

The current study aims to depart from previous approaches by emphasizing the importance of using qualitative comparative analysis (Pappas and Woodside, 2021; Rasoolimanesh et al., 2021). Previous research examining examines the interrelationships between SMMA, self-brand connections, brand equity, trust and loyalty extensively relied on symmetric inferential statistical tests, such as correctional or multiple regression analysis, assuming symmetrical associations between variables (Olya, 2023). The symmetrical modeling approach implies that high values of one variable are associated with high values of another, and vice versa (Ali et al., 2016, 2019; Woodside, 2016). However, recent advancements, particularly in hospitality and tourism research, advocate for the use of asymmetrical modeling, such as causal asymmetrical modeling to understand complex causal relationships among variables (Olya, 2023; Rasoolimanesh et al., 2021). Unlike symmetric models, asymmetrical modeling considers scenarios where high values of one variable are either necessary, sufficient or both for high values of another (Ali et al., 2023). By incorporating complexity and configural theories, the study examines how the four antecedent conditions (i.e. SMMA, self-brand connections, brand equity and brand trust) collectively contribute to high as well as low scores in the outcome (i.e. brand loyalty). In addition, it highlights that these four antecedent conditions leading to high scores might not necessarily be the exact opposite of those leading to low scores (Ali et al., 2019; Fiss, 2011; Woodside, 2016). Neglecting causal asymmetrical modeling might result in incomplete findings, hindering a comprehensive understanding of the issue at hand (Ruiz-Equihua et al., 2023). Thus, the asymmetrical modeling approach provides a deeper understanding of complex causal associations and configurational combinations of outcome conditions, allowing researchers to grasp the intricate realities better than symmetric models. Accordingly, this study adopts a symmetrical and asymmetrical approach to explore the impact of SMMAs on customer brand loyalty. The study investigates multiple causal paths of four causal antecedent conditions resulting in high and low levels of brand loyalty by using self-brand connections, brand equity and trust. It aims to provide deeper insights via comparison of conventional regression-based analyses with fuzzy-set qualitative comparative analysis (fsQCA), which identifies alternative combinations of four causal antecedent conditions determining various levels of brand loyalty. Figures 1 and 2 illustrate symmetric and asymmetric models, respectively.

Literature review

Social media marketing activities

Social media is “a group of internet-based applications that build on the ideological and technological foundations of Web 2.0, and that allow the creation and exchange of user-generated content” (Kaplan and Haenlein, 2010, p. 61). Social media allow for connectivity and interactivity between brands and consumers (Chi and Lieberma, 2011). It provides marketers with incredible prospects to reach consumers and build long-lasting relationships with them (Kelly et al., 2010). Due to its interactive features that allow for participation, collaboration and knowledge sharing, social media is considered an essential medium for creating and distributing brand-related information (Knoll, 2016), with greater ability to reach consumers as compared to traditional media, including radio, television or print. Therefore, companies will find it helpful to incorporate social media in their marketing promotional mix. Although several definitions have been offered for SMMA, a useful one is shared by Yadav and Rahman (2018):

A process by which companies create, communicate, and deliver online marketing offerings via social media platforms to build and maintain stakeholder relationships that enhance stakeholders’ value by facilitating interaction, information sharing, offering personalized purchase recommendations, and word of mouth creation among stakeholders about existing and trending products and services (p. 1296).

This definition of SMMAs implies that social media is used in brands’ marketing as a two-way communication tool (Ebrahim, 2020). They essentially allow for content sharing and diffusion (Chang et al., 2015). Therefore, companies extensively use social media to reach consumers through SMMAs, including YouTube or Facebook advertising, blogger endorsements and consumer-generated content (Yu and Yuan, 2019). Researchers have used different dimensions for investigating SMMAs, including entertainment, interaction, trendiness, customization and word of mouth (Chen and Lin, 2019). It can also be noted that Yadav and Rahman (2018) used the dimensions of interactivity, informativeness, word of mouth, trendiness and personalization. Sano (2015) used four dimensions:

• Interaction;

• Trendiness;

• Customization; and

• Perceived risk.

Ebrahim (2020) considered trendiness, customization and word of mouth as components of SMMAs. For this study, entertainment, interaction, trendiness, customization and word of mouth as components of SMMAs.

Self-brand connection

Consumers form social and emotional connections with brands to fulfill their self-identity desires (Arghashi and Yuksel, 2022). Developing such brand-consumer relationships requires compatibility between the consumer and brand identity, leading to a self-brand connection characterized by the perceived brand meaning and its benefits (Nguyen et al., 2022). Self-brand connection is the extent to which individuals have incorporated brands into their self-concept (Escalas and Bettman, 2017).

Consumers sense more connection to brands that reflect their self-identities (Japutra et al., 2019; Nyadzayo et al., 2018). For instance, perceived customization or personalization of advertisements on social media like Facebook was conjectured to enhance self-brand connections (Heirman and Walrave, 2012). Such personal relevance of brand advertising messages fits with consumers’ self-concepts, enhancing self-brand connections (De Keyzer et al., 2022). In addition, influencer advertising on social media enhances the relationship between brands and consumers, as influencers supplement meaningfulness and relevance to brand messages from a consumer perspective (Schimmelpfennig and Hunt, 2020). In the light of the above argumentation, the following hypothesis is proposed:

H1. Social media marketing activities positively affect the self-brand connection.

Brand equity

Brand equity can be defined as:

Brand assets and liabilities linked to a brand, its name and symbol that add to or subtract from the value provided by a product or service to a firm and/or to that firm’s customers (Aaker, 1991, p. 15).

From a consumer’s perspective, brand equity is a unique value that differentiates one brand from other and encompassing various brand properties (Seo and Park, 2018). Brands’ marketing efforts on social media entice several benefits, such as enabling interaction between stakeholders, delivering personalized suggestions and improving consumer relationships (Chen and Qasim, 2021). As such, the success of social media advertising can be evaluated through its capability to enhance brand equity, especially in terms of developing brand awareness and brand image (Adetunji et al., 2018). As a result, service providers use SMMA to build value, foster relationships with consumers and develop brand equity (Yu and Yuan, 2019).

This study contends that SMMAs involving the aspects of word of mouth, customization, trendiness, interaction and entertainment for portraying brand elements are prone to raise awareness and favor more positive associations in consumers’ minds, thus enhancing brand equity. Through customization, social media allows brands to disseminate content tailored to users’ preferences. Consumers’ perceptions of customization influenced brand equity, especially for younger consumers (Kim and Lee, 2019). SMMAs embody the aspect of word of mouth, as consumers can publicly share their experiences, which tends to positively influence image, awareness and associations with the brand (Chen and Qasim, 2021). Kim and Lee (2019) also found a positive influence of e-word of mouth on consumers’ brand equity. In addition, SMMA can elicit positive emotions in consumers through the diffusion of entertainment driven contents (Godey et al., 2016), which also impacts brand equity. Jayasuriya et al. (2017) noted that most studies on social media brand communication used entertainment, interaction, trendiness, customization and word of mouth as components of social media, confirming the positive effects of SMMAs on brand equity. More recent studies in the marketing literature (e.g. Ebrahim, 2020; Felix et al., 2017; Zollo et al., 2020) have investigated the role of social media marketing in improving brand equity and found a positive relationship. In the context of e-commerce, Yadav and Rahman (2018) and Zarei et al. (2022) found a positive impact of SMMAs on brand equity. For luxury brands, Husain et al. (2022) found that advertising campaigns on social media boosted brand equity. In the light of this discussion, the following hypothesis is posited:

H2. Social media marketing activities are positively related to brand equity.

Brand trust

Trust is a psychological predisposition of a consumer toward a brand (Lin and Lee, 2012). It depicts consumers’ positive expectations regarding how brands meet their promises. Brand-related information on social media may consist of stories or product usage patterns, which meet the needs of lowering the risks of information failure and uncertainty (Mammadli, 2021). According to Puspaningrum (2020), authentic information about brands disseminated on social media contributes strongly to perceived trustworthiness toward the brand. Marketers can use brand communities to stress several components of brand elements that echo consumers’ perspectives, which is also important for brand trust (Laroche et al., 2013). As a result, social media platforms have become an avenue for brands to highly communicate and interact with their customers (ElAydi, 2018).

Social media interactions increase perceived brand trustworthiness through two-way communication and instant feedback, promoting an objective perception of brand elements (Sohail et al., 2020). Social media interactions can also reinforce consumers’ trust with the brand as they can mitigate uncertainties that might deter their intentions to purchase from the brand (Haudi et al., 2022). By cultivating emotional connections across their brand presence on social media, companies can enact more intimate relationships in which consumers perceive them as more reliable (Dwivedi et al., 2021). Yu and Yuan (2019) also confirmed that positive experiences with brands through social media are a crucial anchor for securing trust in brand-consumer relationships. Furthermore, Sanny et al. (2020) found a positive influence of social media marketing on brand trust in male skincare brand marketing on Instagram, YouTube, Facebook and Twitter. Similarly, Khoa (2020) focused on the fashion business and found SMMAs positively impacted brand trust. Finally, Ibrahim et al. (2021) focused on Northern Cyprus coffee shop brands’ marketing activities on Facebook and found positive effects on brand trust. In consideration of these findings and drawing the above rationale, it is hypothesized that:

H3. Social media marketing activities are positively related to brand trust.

Brand loyalty

Loyalty is manifested in the repeat purchase of a service or product from a brand (Oliver, 1999). Brand loyalty is “repeated purchases of particular products or services during a certain period of time” (Yi and Jeon, 2003, p. 231). Several studies have assessed consumers’ bonds with brands (Cooper et al., 2010). Strong consumer-brand bonds are likely to bring about more loyalty toward the brand, with probrand behaviors including positive word of mouth (Batra et al., 2012). Previous studies have pointed out that the more a consumer identifies with a brand, the stronger the loyalty toward that brand (Lin et al., 2017). In other words, self-brand connection, the extent to which a brand is integrated into a consumer’s self-concept, can be a predictor of brand loyalty (Papista and Krystallis, 2013). As consumers develop connections with brands as an extension of their self-concept, such linkage positively influences their responses toward the brand (De Keyzer et al., 2022). Eelen et al. (2017) explained that self-brand connection plays a vital role in strengthening the relationship with word of mouth. Similarly, Hur et al. (2013) also emphasizes the power of self-brand connection positively affecting brand loyalty. Escalas and Bettman (2017) explained that self-brand connections were powerful in eliciting long-term positive attitudes and loyalty to a brand. Hemsley‐Brown (2022) explained that strong connections between brands and consumers led to attachment, which was influential in terms of repurchase behavior and loyalty to the brand. Based on the above discussion, the following hypothesis is proposed:

H4. Self-brand connection has a positive influence on brand loyalty.

Prior literature assessed the positive influence of brand equity on brand preference and loyalty (Keller and Lehmann, 2006). Chen and Qasim (2021) discussed that brand equity had been recognized as a critical element of overall companies’ branding, as it allows for predicting brand success and customers’ loyalty. Therefore, social media has become a vital component of the overall communication strategy of brands. They provide an extraordinary venue for brands to deliver accurate information related to their products or services, which can sway consumers’ perceptions and positively affect their loyalty behaviors (Schivinski and Dabrowski, 2019). Brakus et al. (2009) discuss in their study that brand equity influences brand loyalty when there are direct experiences with the brand, in the current case, social media interactions. Brand equity resulting from marketing efforts and brand communication on social media positively influences consumers’ future purchase intentions (Husain et al., 2022). Arya et al. (2022) explained that brand equity played a nonnegligible role in consumers’ willingness and intention to purchase from a brand. Considering the above discussion, the following hypothesis is proposed:

H5. Brand equity has a positive influence on brand loyalty.

Finally, a plethora of previous research has investigated the relationship between brand trust and brand loyalty (Ebrahim, 2020). Because trust leads to a highly valued exchange relationship, loyalty becomes an outcome of that relationship (Morgan and Hunt, 1994). Previous studies have assessed the construct of trust as one of the main antecedents of loyalty (Hong and Cho, 2011). For instance, Delgado-Ballester et al. (2003) also found a direct and positive effect of brand trust on brand loyalty. Hartmann and Ibáñez (2007) found that consumer trust influences brand loyalty. Hajli (2014) also found that consumers’ social interaction enabled by social media contributed to brand trust, which, in turn, had a positive effect on buying intentions. Additionally, interactions on social media that are entrenched around trust bring about higher consumer purchasing intentions (Liu et al., 2018) and Gamboa and Gonçalves (2014) found a positive effect of brand trust on customers’ loyalty through Facebook fan pages. In addition, managing relationships with consumers on an individual basis might reveal an arduous and demanding task for brands (Laroche et al., 2013). As such, using communities on social media has notable advantages in enhancing trust through mainly facilitating information sharing and support provided to customers, which tend to positively affect brand loyalty (Sohail et al., 2020). Based on these previous findings, the following hypothesis is postulated:

H6. Brand trust has a positive influence on brand loyalty.

The proposed connections between the factors indicate a sequential relationship starting with SMMA, then moving through self-brand connections, brand equity, brand trust and finally culminating in brand loyalty. Therefore, to gain a more profound understanding of how these variables are interconnected and the intricate relationships among them, this study uses complexity theory and configural theory. Its aim is to analyze how different combinations of causal factors contribute to either a high or low level of brand loyalty. In the context of this study, the application of asymmetrical modeling finds its rationale in the theory of complexity. Complexity theory has gained recognition across various fields, including hospitality and tourism, for advancing theoretical frameworks (Ali et al., 2023; Loureiro et al., 2024; Ruiz-Equihua et al., 2023; Olya and Altinay, 2016). Prior research has extensively used complexity and configural theories. For example, drawing on complexity and configurational theories, Ali et al. (2023) used fsQCA and predicts that several causal associations through a set of hospitality, authenticity, destination quality and destination love lead to high and low levels of destination advocacy. Loureiro et al. (2024) used complexity theory with fsQCA to predict configurational paths for high and low levels of chatbot advocacy. Leveraging complexity and configural theories, Ruiz-Equihua et al. (2023) contends that the asymmetrical configuration of antecedent conditions (e.g. social cognition, psychological ownership, robot anthropomorphism, gender, age, past experience and restaurant setting) influencing two outcome conditions (attitude and revisiting intentions) provides deeper insights into the intricate interdependencies among these constructs. Olya and Altinay (2016) used complexity theory with fsQCA to forecast paths for tourist weather insurance purchase intention and destination loyalty.

Building on prior research (Ali et al., 2023; Loureiro et al., 2024; Ruiz-Equihua et al., 2023), this study uses complexity theory to delve into a pattern phenomenon that offers a more comprehensive understanding of the relationships among SMMA, self-brand connections, brand equity, brand trust serving as causal factors and brand loyalty as the resulting outcome. Scholarly community in hospitality and tourism (Olya and Altinay, 2016; Olya and Mehran, 2017) recognizes that using complexity theory will yield deeper insights into how SMMA, self-brand connections, brand equity, brand trust in various combinations, stimulate high or levels of brand loyalty (Fiss, 2011; Olya and Altinay, 2016; Olya and Mehran, 2017). This study proposes that there are intricate configurations of SMMA, self-brand connections, brand equity, brand trust, linked to both low and high levels of brand loyalty. Consequently, the focus of this study is on achieving a high level of brand loyalty, aiming to achieve this outcome through diverse combinations of SMMA, self-brand connections, brand equity, brand trust. The anticipated relationships imply a sequential progression from SMMA, self-brand connections, brand equity, brand trust toward a high level of brand loyalty. Specifically, this study aims to explore various causal associations among SMMA, self-brand connections, brand equity, brand trust that lead to high/low levels of brand loyalty (Figure 2). Based on these premises, the study sets forth the following hypothesis:

H7. Varied combinations of a social media marketing activities, self-brand connections, brand equity and brand trust associate with brand loyalty.

Research methodology

This study focused on users over the age of 18 who use social media platforms to connect with various brands in the USA. Using self-selection sampling method, data was gathered from Amazon MTurk from respondents with a minimum of 95% human intelligence to ensure data quality (Ali et al., 2021). Moreover, questionnaire also included two attention check questions. Respondents who failed the attention-check questions were eliminated from the data set. Consequently, 402 valid responses were selected for further examination. Of the total respondents, more than half of the respondents were male (53.7%), whereas 40.2% were aged between 18 and 34. About half of the respondents were full-time employees (48.1%), and around 60% used more than three social media channels of their favorite brands. As shown in Table 1, scales were adopted from previous studies. All the scales were measured using a five-point Likert scale, whereas loyalty was measured using a seven-point Likert scale. Multiple methodological and statistical tools were used to account for the common method variance – first, different cover stories for each measurement scale to achieve psychological separation among respondents. Second, the questionnaire only included 27 items. Therefore, it was short enough to avoid tiredness and confusion, which can negatively impact the cognitive effort of the respondents to answer the items accurately. Third, different measurement scales were used to measure independent and dependent variables. Harman’s single-factor test was also conducted, and results revealed that one factor (36.72%) did not account for most of the variance. Hence, common method bias was not likely to have affected this study in a significant way.

Findings and analysis

Partial least squares structural equation modeling

First the data set was subject to multivariate data normality using Mardia’s coefficients. Results indicated that data was not normally distributed. As such, partial least squares structural equation modeling (PLS-SEM) was chosen as the preferred method for this study. Ali et al. (2018) suggested that reflective constructs were validated through composite reliability (>0.70), Cronbach’s alpha (>0.70) and average variance extracted (AVE > 0.50). All the values passed the recommended thresholds, and each construct’s reliability and validity were established (see Table 1).

Furthermore, the constructs’ discriminant validity was assessed using the Fornell–Larcker criterion. Table 2 shows that the square roots of the AVE (values in bold, off-diagonal) are all greater than the correlations in the respective columns and rows. As a result, the measurement model exhibited adequate discriminant validity. The Heterotrait-monotraite (HTMT) method was also used to assess discriminant validity (Henseler et al., 2015). Table 2 also shows that all HTMT values are less than 0.90, satisfying the condition of HTMT.90 (Kline, 2015) and confirming the satisfactory discriminant validity for all constructs in this study.

Additionally, for the model fit assessment, the SRMR value was used. A value of less than 0.08 is considered a good fit. The SRMR value for both the saturated and estimated models in this study was 0.051, indicating that the proposed model has an excellent fit to the data. After the overall measurement model was acceptable, the structural model was tested. Initially, all variance inflation factor values were calculated and found to be less than 5; thus, no multicollinearity issues were found in the structural model. Next, R-square, path estimates and corresponding t-values were calculated using a bootstrapping procedure with 5,000 subsamples. The results are shown in Table 3.

SMMA was found to have a strong and positive influence on the self-brand connection (β = 0.131, p < 0.05), brand equity (β = 0.782, p < 0.05) and brand trust (β = 0.800, p < 0.05). Thus, H1 to H3 were supported. Moreover, brand loyalty was found to be strongly and positively influenced by self-brand connection (β = 0.561, p < 0.05), brand equity (β = 0.169, p < 0.05) and brand trust (β = 0.101, p < 0.05). Thus, the hypotheses H4 to H6 were supported. Finally, SMMA explains 21.3%, 61.2% and 64.0% variance in self-brand connection, brand equity and brand trust, respectively, whereas all three further explain a 42.1% variance in loyalty.

Fuzzy-set qualitative comparative analysis

To address H7, this study applies fsQCA – an asymmetric modeling approach used to investigate causal antecedents that explain complex conditions for achieving high and low levels of brand loyalty. The analysis of fsQCA is performed by using guidelines in Oana et al. (2021), Pappas and Woodside (2021), Rasoolimanesh et al. (2021) and procedure in Rihoux and Ragin (2008). Using Boolean algebra logic, fsQCA enables generating sets of causal conditions (i.e. social marketing activities, self-brand connection, brand equity, brand trust,) which lead to both high and low (negation) levels of the outcome (i.e. brand loyalty). The analysis of fsQCA is performed in several steps: calibration, construction of a truth table and counterfactual analysis.

The first step is data calibration, where raw data (survey data) is transformed into scores of set memberships. Data calibration refers to transforming the survey data, such as Likert scales, into fuzzy sets of values ranging from “0” to “1.” For calibration, this study followed the direct method suggested by Rihoux and Ragin (2008). For this purpose, the averages of all four antecedents (or conditional) and single outcome variables in the survey data are used to complete the calibration process. The three commonly used qualitative anchors are assigned, where 1 = full set membership, 0.50 = crossover point and 0 = no set membership. Based on the guidelines provided in Pappas and Woodside (2021), the variable having value greater than the 95th percentile is calibrated into 1 – full set membership of the corresponding fuzzy set. The variable has a value smaller than 5% and is calibrated into 0 – no set membership of the fuzzy set. If a variable has a value of 50%, it is the crossover point – the data are neither in nor out of the fuzzy set. Calibrated data > 0.5 represent “presence” in set theory; those < 0.5 represent “absence”; while = 0.5 represent “do not care” (Ren et al., 2016). The second step is constructing the truth table – a tool to analyze causal antecedents’ conditions, and the minimization of the truth table by applying the Quine–McCluskey algorithm to produce three solutions (complex, parsimonious and intermediate). The calibrated data from the first step is used to build a truth table. A truth table refers to a data matrix that provides empirical information captured in the sets formed from the raw data. A truth table is an important tool in fsQCA as it allows the unraveling of set relationships and views observations as “configurations” of conditions. We used coverage and consistency – the two commonly used measures to refine the solutions for condition outcome – brand loyalty. Finally, the same procedure in the second step is performed for the low scores of conditional outcomes – brand loyalty-referred to as negation analysis.

Table 4 highlights the causal antecedents’ conditions conducive to a higher or a lower level of brand loyalty. The analysis of fsQCA provides evidence of the combinations of three different model configurations for predicting a high level of brand loyalty (coverage: 0.729; consistency: 0.809). The first configurational model (M1: sbc*∼beq) suggests a high-level presence of self-brand connection and a low-level absence of brand equity provide a condition that leads to achieving a high level of brand loyalty (coverage: 0.499; consistency: 0.859). The second model (M2: sbc*btr) suggests that both a high-level presence of self-brand connection and brand trust provide a condition that leads to a high level of brand loyalty (coverage: 0.531; consistency: 0.852). Alternatively, the final model (M3: smma*∼sbc*beq*∼btr) provides evidence that a high level of brand loyalty can also be attained by the presence of a high level of SMMAs and self-brand connection, brand equity and brand trust (coverage: 0.285 and consistency: 0.810).

The negation analysis of fsQCA provides evidence for predicting a low level of loyalty (coverage: 0.832 and consistency: 0.735), as shown in Table 4(B). The fsQCA analysis generates a single configurational model (M1: ∼sbc) that suggests that the low level of self-brand connection is enough to lead to a low level of brand loyalty (coverage: 0.603 and consistency: 0.850). The results in Table 4 confirm that a high level of brand loyalty can be achieved through a varied combination of SMMAs, self-brand connections, brand equity and brand trust, which consists of three configurational models [Table 4(A)]. While a single configurational model leads to the absence of brand loyalty, as shown in Table 4(B). These results suggest that several combinations of a SMMA, self-brand connections, brand equity, brand trust associate with brand loyalty. Therefore, H7 was supported.

Discussion and implications

Recent statistics show a steady increase in social media users worldwide, uncovering newer trends in consumers’ needs to interact with brands. Service providers offer direct and real-time methods to connect with customers via virtual consumer environments, including social media-based brand communities, to cater to this need. However, the existing literature is inconclusive about how SMMAs can develop into brand equity. Moreover, it is also unclear how consumers develop trust toward the brand and increase the intensity of connection with the user through SMMAs. To deal with the disparities mentioned above in the literature, this study implements a symmetrical and asymmetrical approach to explore the impacts of SMMAs on customers’ brand trust, self-brand connections and brand equity, leading to brand loyalty. This study responds to recent calls from Panigyrakis et al. (2020) and VanMeter et al. (2018) to explore the ins and outs of the relationship between marketing activities carried through social media and consumer-brand relationships.

On the one hand, this study hypothesized a positive impact of SMMAs on the self-brand Connection (H1), brand equity (H2) and brand trust (H3). Findings from this study supportH1, H2 and H3. Per H1, SMMAs positively influence self-brand connections, like findings from previous literature. Panigyrakis et al. (2020) confirmed that SMMAs by a brand positively influences self-brand connections, inducing brand attachment to consumers. Per H2, SMMAs positively influence brand equity, confirming results from previous studies (e.g. Felix et al., 2017; Kim and Ko, 2012; Çifci et al., 2016; Zollo et al., 2020). Although those studies found a significant relationship, some interestingly found the relationship between SMMAs-brand equity insignificant (see Ebrahim, 2020; Hafez, 2021). Per H3, SMMAs positively influence brand trust, a relationship verified in previous literature (Ebrahim, 2020; Habibi et al., 2014; Tatar and Eren-Erdoğmuş, 2016).

The present study hypothesized a positive and strong influence of self-brand connection (H4), brand equity (H5) and brand trust (H6) on brand loyalty. As per the current findings, self-brand connection, brand equity and brand trust were found to have a strong and positive influence on brand loyalty. Thus, H4 to H6 were supported. To the best of our knowledge, it is worth noting that the relationship between self-brand connection and brand loyalty (H4) has not been tested before. However, in a conceptual paper, Escalas and Bettman (2017) suggested the role of self-brand connections in eliciting long-term positive attitudes and loyalty to a brand. Similarly, Panigyrakis et al. (2020) also suggested that self-brand connections developed through SMMA lead to brand attachment. Furthermore, regarding the effect of brand equity on brand loyalty (H5), this study’s findings attest to the positive impact on brand loyalty, consistent with previous studies (Aaker et al., 2007; Ebrahim, 2020; Kim and Ko, 2012). Finally, the current analysis found a positive impact on the effect of brand trust on brand loyalty, like previous related literature (Chiu et al., 2010; Ebrahim, 2020; Kang et al., 2011; Kim et al., 2011). Finally, results from the asymmetrical analysis show that a high level of brand loyalty is achieved through a combination of SMMAs, self-brand connections, brand equity and brand trust, which consists of three configurational models. In contrast, a single configurational model leads to the absence of brand loyalty (H7). Significant theoretical and practical contributions are drawn from the present examination of the hospitality industry, brand and social media marketing managers and responding directly to Kandampully et al. (2015) call for more research into customer loyalty dynamics.

Theoretical implications

This present investigation provides several contributions to the literature. First, the past literature is inconclusive about the role of SMMA in developing consumers’ self-brand connections, brand equity and brand trust, leading to brand loyalty. While previous studies have confirmed the effect of SMMA on brand equity and other related constructs (Godey et al., 2016; Khan, 2022); recent examinations have highlighted the ineffective role of SMMA in developing customer participation in virtual environments, leading to discontinued use of social media platforms (Tang et al., 2019). This research proposed and empirically verified a holistic model to explain the role of SMMAs in developing brand loyalty through consumer self-brand connections, brand equity and brand trust. This study’s findings contribute to the growing social media marketing literature in improving connections to the customers and branding goals, including brand trust, equity and loyalty (Khan, 2022). While several studies have examined the individual relationships among SMMA, self-brand connections, brand equity, brand trust and brand loyalty, this is the first study to consider all these relationships in one model, which adds value to the existing social media marketing and branding literature. Finally, the existing literature is based on symmetrical relationships among these constructs; this study uses symmetrical and asymmetrical analytical tools to examine them. The application of fsQCA complements results from PLS-SEM by exploring the causal combination of SMMAs, self-brand connections, brand equity and brand trust that led to high and low levels of brand loyalty. To this end, this examination provides a theoretical contribution to the existing social media marketing literature. It also extends the discussion on brand loyalty by studying combinations of causal configurations (i.e. combinational of antecedents) that lead to high or low levels of brand loyalty. Most marketing-related studies that investigated the relationship between brand trust and brand loyalty from a social media perspective did not necessarily integrate brand equity within their theoretical models (Upadhyay and Tripathi, 2023). Very few studies also evaluated the role of self-brand connections as antecedents to brand loyalty in the context of brands’ marketing activities on social media. As a result, they fall short of providing a thorough understanding of the interrelationships between SMMAs and their components on the one hand and brand trust, brand equity and self-brand connections on the other hand. For instance, results from the asymmetrical approach used in this study showed that both high levels of self-brand connection and brand trust are necessary to cause high levels of loyalty; however, even in the presence of brand trust, a low level of self-brand connection alone can suffice to have negative impacts in loyalty behaviors. It implies that consumers trust products or services from a brand; however, if it does not reflect their self-concept, they may decide to purchase from other brands.

Finally, approaching the analysis from both symmetric and asymmetric approaches, this study responds to the advocated combination of methodologies (Rasoolimanesh et al., 2021; Ruiz-Equihua et al., 2023), showcasing the practicality of using PLS-SEM and fsQCA. Using these methods in tandem complements and enhances the outcomes derived from each individual technique. PLS-SEM allows the research community to effectively model intricate structural relationships among observed and latent variables. PLS-SEM predicts direct, indirect and interactive causal paths among SMMA, self-brand connections, brand equity, brand trust and brand loyalty. FsQCA, on the other hand, serves as a bridge between quantitative and qualitative approaches, identifying causal configurations that sufficiently generate high and low levels of brand loyalty (Ruiz-Equihua et al., 2023). As a result, future studies might benefit from using these various analytical techniques (PLS-SEM and fsQCA) in a single study more comprehensive empirical insight.

Managerial implications

Industry practitioners can strategically invest in social media marketing to effectively reach their target audience and cultivate stronger connections with their brand. The attributes of SMMA identified in this study can offer valuable insights for brands aiming to enhance their online presence. For instance, these attributes can guide practitioners in creating and sharing trendy, entertaining and interactive content on various social media channels. Starbucks, for example, leverages Instagram and TikTok to share visually appealing content, engaging customers with behind-the-scenes glimpses, entertaining videos and interactive challenges, thereby actively motivating customers to use and engage with their social media channels.

As the brands are facing high competition and the market is filled with increased customer incredulity toward traditional advertisement, social media mediums and online brand communities are the best vessels to connect and forge a brand connection with customers. Brands can develop a consistent and authentic brand voice that resonates with their target audience to build stronger self-brand connections. For example, a restaurant can develop a brand voice that reflects their unique cuisine and dining experience. They can use this voice in their social media posts, website content and advertising to build a stronger connection with their audience. Airbnb has also successfully built a community of hosts and guests on its platform. By facilitating direct communication and user-generated content, Airbnb strengthens its self-brand connections. Similarly, Wendy’s, a fast-food restaurant chain, has developed a witty and humorous brand voice on Twitter, resonating with its target audience and creating a strong online connection. However, it is essential to understand that the intensity of consumers’ connection to the brand ultimately influences the effect of SMMAs on brand loyalty. It implies that brands providers should consistently cultivate and develop their customer’s self-brand connections, acknowledging the vital influence on behavioral outcomes.

Industry practitioners can build trust with their audience by being transparent in their social media marketing efforts. For example, an illustrative example is Southwest Airlines, known for its transparent communication on social media platforms regarding flight updates, policies and customer service interactions. Such transparency contributes to building trust among its customer base. They also respond promptly to customer inquiries and complaints to demonstrate their commitment to customers. Brands can segment customers based on their interaction with the brand presence across social media platforms. This information can provide specific content to different groups of people. However, this content must be reliable and accurate, as those are determinants of customers’ trust in the brand. It is noteworthy that the effectiveness of SMMAs in generating brand loyalty is contingent upon customers’ degree of trust in the SMMAs. Therefore, building an intimate relationship with customers would benefit marketers and brands.

Moreover, brands can foster brand loyalty by engaging with their audience through social media. A prime example includes Sephora, a cosmetics retailer, encouraging customers to share product reviews and beauty tips on its social media platforms. By responding to these reviews with personalized messages and creating a sense of community, Sephora strengthens its brand loyalty. Another example is how Marriott Hotels uses social media monitoring tools like Hootsuite to track customer sentiment, engagement and reach. This data-driven approach helps Marriott make informed decisions, optimize its social media marketing strategy and continually enhance its brand’s online presence.

Applying the fsQCA results, marketers may attempt to satisfy conditions matching with causal processes of high levels of brand loyalty. However, the causal conditions leading to high levels of brand loyalty are not an opposite mirror to the causal processes leading to low levels of brand loyalty. Consequently, marketers need to be cautious in promoting the practices of conditions for which a desired combination of antecedents remains consistent with the causal receipt, leading to an effect of a high level of brand loyalty along with a regulation of conditions for which the sufficient antecedents are not following causal processes to brand loyalty negation.

Limitations and future research suggestions

Limitations of the current study include its cross-sectional design, which precludes capturing changes in constructs over time. Future studies should consider a longitudinal design and compare alternative postpurchase constructs. Moreover, social media platforms were treated as one despite their heterogeneity in usage and purpose; future research should investigate specific platforms to provide a more nuanced understanding of SMMA. Other potential avenues for future research include considering the moderating effects of marketing activity, services and platform use on brand loyalty, exploring the multidimensional nature of brand equity, examining the impact of demographic and cultural characteristics and investigating the nature of e-electronic word of mouth and customer engagement in social media. See Table 5.

Competing interests: Authors do not have any financial or nonfinancial interests directly or indirectly related to the work submitted for publication.

Figure 1.

Theoretical model of the study

Figure 2.

Configurational model

Table 1.

Validity and reliability

Constructs	Items		Loadings	Cronbach’salpha	rho_A	CR	AVE
Entertainment(Kim and Ko, 2012)	ENT1	The content found in brand X’s social media seems interesting	0.914	0.703	0.732	0.856	0.749
	ENT2	It is fun to collect information on products through brand X’s social media	0.814
Customization(Kim and Ko, 2012)	CUS1	It is possible to search for customized information on brand X’s social media	0.835	0.704	0.714	0.825	0.701
	CUS2	Brand X’s social media provides customized services	0.840
Interaction(Kim and Ko, 2012)	INT1	It is easy to convey my opinion through brand X’s social media	0.876	0.823	0.825	0.895	0.739
	INT2	It is easy to convey my opinions or conversations with other users through brand X’s social media	0.854
	INT3	It is possible to have two-way interaction through brand X’s social media	0.848
Word of mouth(Kim and Ko, 2012)	WOM1	I would like to pass information from brand X’s social media to my friends	0.946	0.871	0.876	0.939	0.885
	WOM2	I would like to upload content from brand X’s social media on my own social media	0.935
Trendiness(Kim and Ko, 2012)	TRE1	Using brand X’s social media is very trendy	0.931	0.851	0.851	0.931	0.870
	TRE2	The content on brand X’s social media is the newest information	0.934
Brand equity(Ebrahim, 2020)	BE1	Even if another brand has the same offerings, I prefer brand X	0.896	0.908	0.916	0.942	0.844
	BE2	If there is another brand as good as this one, I prefer brand X	0.924
	BE3	If the services of another brand are not different from brand X in any way, it seems smarter to purchase brand X	0.937
Brand trust(Tran and Strutton, 2020)	TR1	Brand X social media advertising seems genuinely committed to my satisfaction	0.823	0.899	0.900	0.930	0.768
	TR2	Overall, I trust brand X’s social media advertising	0.920
	TR3	In terms of usability, I know what to expect from brand X’s social media advertising	0.896
	TR4	If brand X makes a claim about in their social media advertising, it is probably true	0.863
Self-brand connections (Thomas and Jewell, 2019)	SBC1	Brand X reflects who I am	0.865	0.913	0.915	0.932	0.697
	SBC2	I can identify with brand X	0.853
	SBC3	Liking brand X communicates who I am to other people	0.822
	SBC4	I think brand X is consistent with the type of person I want to be	0.797
	SBC5	I consider brand X to be reflective of “me.”	0.842
	SBC6	Brand X suits my personality well	0.829
Brand loyalty(Tran and Strutton, 2020)	LOY1	I prefer brand X’s products/services over other alternatives	0.914	0.867	0.880	0.918	0.790
	LOY2	I am loyal to brand X	0.916
	LOY3	Brand X is my first choice for the product/service category they offer	0.835

Table 2.

Discriminant validity

Constructs	1	2	3	4	5
F&L criterion
Brand equity	0.919
Brand loyalty	0.446	0.943
Brand trust	0.879	0.574	0.876
Social media marketing activities	0.782	0.523	0.800	0.837
Self-brand connections	0.048	0.564	0.561	0.631	0.835
HTMT criterion
Brand equity
Brand loyalty	0.459
Brand trust	0.797	0.784
Social media marketing activities	0.739	0.541	0.794
Self-brand connections	0.558	0.628	0.673	0.748

Note:F&L = Fornell and Larcker

Table 3.

Hypotheses testing

Hypotheses		Beta	T-value	p-values	Decision
H1	Social media marketing activities → Self-brand connection	0.131	2.471	0.050	Supported
H2	Social media marketing activities → Brand equity	0.782	31.40	0.000	Supported
H3	Social media marketing activities → Brand trust	0.800	34.06	0.000	Supported
H4	Self-brand connection → Brand loyalty	0.561	15.81	0.000	Supported
H5	Brand equity → Brand loyalty	0.169	2.696	0.050	Supported
H6	Brand trust → Brand loyalty	0.101	2.010	0.050	Supported

Table 4.

Sufficient causal configurations

	Raw coverage	Unique coverage	Consistency
(A) Models for predicting a high level of brand loyalty
Model: loy = f (smma, sbc, beq, btr)
M1: sbc*∼beq	0.499	0.139	0.859
M2: sbc*btr	0.531	0.160	0.852
M3: smma*∼sbc*beq*∼btr	0.285	0.056	0.810
Solution coverage: 0.729
Solution consistency: 0.809
(B) Models for predicting a low level of brand loyalty(negation)
Model: ∼ loy = f (smma, sbc, beq, btr)
M1: ∼sbc	0.603	0.424	0.850
Solution coverage: 0.832
Solution consistency: 0.735

Notes:loy = Brand loyalty; smma= social media marketing activities; sbc= self-brand connection; beq= brand equity; btr= brand trust; ∼ indicates negation

Table 5.

Conclusions and theoretical and managerial implications

Conclusions	Theoretical and managerial contributions
This study empirically verifies a holistic model showing SMMA’s role in developing brand loyalty through consumer self-brand connections, brand equity and brand trust	Enhances understanding of the complex relationships between SMMA and key branding constructs, proposing a comprehensive model that integrates these elements for the first time
Findings indicate that both strong self-brand connections and brand trust are crucial for achieving high brand loyalty. Conversely, a lack of self-brand connection can undermine loyalty, even in the presence of brand trust	Emphasizes the importance of creating content that resonates with the consumer’s self-image and maintaining transparency and engagement to build trust
The research underscores the value of a holistic approach to analyzing the impact of SMMA on brand loyalty, suggesting that certain combinations of factors are more effective than others	Encourages a nuanced examination of the factors contributing to brand loyalty, highlighting the importance of considering multiple brand-related constructs in relation to each other
Effective social media strategies are identified as key drivers of brand loyalty, with a focus on developing authentic connections and trust with consumers	Recommends leveraging social media to develop a consistent brand voice and engage with consumers in a meaningful way. Highlights the role of interactive and engaging content in strengthening brand loyalty