Introduction

The banking sector is among the fastest growing sectors across the globe (Rinaldi and Spadavecchia 2021). Nevertheless, this growth in banking sector has its own negative consequences such as non-compliance of regulatory framework, declining values and financial crisis (Aslam and Haron 2020) that create a negative image of the banks in consumers’ minds. Banks are trying to negate these impacts through a responsible behaviour toward society and environment (Nosratabadi et al. 2020) and by introducing corporate social responsibility (CSR) in their strategy.

In order to demonstrate their commitment to CSR, an increasing number of Indian banks have initiated numerous socially responsible practices (Fatma et al. 2020). The Indian banks have been focusing on CSR due to its significant influence on electronic word of mouth (EWOM) and consumers’ identification with a company (C–C identification) (Gupta et al. 2021). These concepts are increasingly gaining importance since many Indian public sector banks are facing challenges in improving their market shares due to global competition and slow economic growth (Garg and Gupta 2020). Apart from competition, the recent advances in technology have made significant impact on global banking industry that has forced retail banks to rethink their marketing strategies (Martino 2019). For example, in services sector, social media is providing new means to brands for strengthening long-term relationships with their existing and potential consumers, through continuous interactions and information sharing (Angelini et al. 2017). Despite social media’s popularity and its ability to create enormous business opportunities, many organizations have not shown extra enthusiasm in using social media (Cao et al. 2018). For instance, Indian public sector retail banks are yet to catch up other financial intermediaries operating in India in terms of social media usage (Fatma et al. 2020).

Due to an increasing significance of social media, the academia and practitioners are evaluating different ways of effectively communicating firm’s CSR practices to its consumers (Bigné et al. 2019). For example, the impact of user-generated content (UGC) on the relationship between a company's perceived CSR initiatives on social media and EWOM needs more attention from research scholars (Schivinski and Dabrowski 2016). Research in the recent past (Colicev et al. 2018) also suggests that there is a need to investigate the moderating effects of user-generated content on the relationship between consumer perceived CSR initiatives and EWOM.

Despite the fact that there have been studies which integrated the three variables CSR, C–C identification and EWOM in the recent past, there have been gaps in these studies. For example, few of these studies have focused on single social media platform such as WeChat (e.g., Wu and Zhu 2021) and Facebook (Fatma et al. 2020), thus restricting the generalizability of their findings. The study by Castro et al., (2022) in Philippines was also restricted to two TV channels only. The study by Fatma et al. (2020) in Indian context covered both private and public sector banks. This study specifically failed to address the consumer perceptions of CSR communication by Indian public sector banks which form the backbone of Indian economy. Our study will enrich the extant literature by incorporating different social media platforms and focusing on public sector banks for an in depth understanding of the mechanism to build EWOM. Another study was conducted in Pakistan healthcare sector (Ma et al. 2021) but had geographical limitations such as data being collected from only one city. We intend to overcome this limitation by collecting data from different Indian cities.

Moreover, most studies in past have been conducted in traditional environment, such as corporate websites (Chu and Chen 2019). A study by Wu and Zhu (2021) also suffered from another drawback that the study was conducted during Covid-19 period, again limiting the generalizability of their findings. Although many studies in past have discussed the positive impact of CSR engagement and C–C identification on consumers’ behavioral intention such as intention to spread positive EWOM, the social media context in banking sector has been an under-researched topic (Chu et al. 2018; Puriwat and Tripopsakul 2021).

Thus, the current study will explore the role of social media in spreading a positive EWOM in Indian banking sector, particularly in public sector banking context. Also, the study will investigate the effectiveness of bank’s CSR communication through social media. The study will examine whether UGC strengthens the effect of CSR on EWOM and the mechanism of communicating bank’s CSR initiatives to consumers directly and indirectly. This study will also improve the generalizability of the findings since many studies have been conducted across countries during Covid-19 period. In fact, this will be among the few studies during post-Covid-19 period. The study further improves the generalizability of previous studies by including multiple social media platforms as opposed to single platform used in earlier studies.

Our study thus proposes a model that explicates C–C identification and favorable EWOM in terms of consumer perceived banks’ CSR initiatives on social media in Indian banking context. Specifically, our study will address the following research questions in context of Indian public sector banks:

  1. 1.

    What are the direct effects of consumer perceived bank’s corporate social responsibility communication in social media (hereafter, CPBCSRSM) on C–C identification and EWOM?

  2. 2.

    What are the direct effects of C–C identification on EWOM and indirect effects of C–C identification between CPBCSRSM and EWOM?

  3. 3.

    What are the moderating effects of user-generated content (UGC) between CPBCSRSM and EWOM?

From a practical perspective, the findings of this study will provide banks’ policymakers directions to develop their strategies by comprehending the perceptions of consumers about banks’ CSR communications in social media, and its consequences such as consumers’ identification with bank, their intention to generate UGC and spread favorable EWOM.

To address various research issues discussed in preceding sections, this paper has been structured as follows: We discuss relevant literature and theoretical background of this study in the next section. The subsequent section explains the constructs and scales, reliability and validity. We report the hypothesis testing in the consecutive section followed by discussions, theoretical and managerial implications. In the last section of the paper, we discuss our study’s limitations and provide directions to future scholars.

Theoretical background

The current study uses attribution theory and social identity theory as grounding theories. Consumer behavior research has frequently used attribution theory for explaining the behaviors of individuals (Jackson 2019; Deb et al. 2021). Attribution theory is highly significant as it facilitates marketers in understanding the factors influencing the process of consumer decision making and the subsequent impact of these decisions on EWOM (Folkes 1988). In the current study, we argue that CSR perceptions of consumers are critical in shaping the attribution process by influencing their evaluation of any brand. The attribution theory is relevant for current study due to the fact that the attribution process helps consumers in brand evaluation. For example, when consumers perceive a particular brand to be actively engaged in societal causes, they attribute this brand to be socially responsible and hence evaluate it favorably (Ginder et al. 2021). Attribution theory also specifies how consumers infer causality by considering intrinsic motives (personal internal characteristics) and extrinsic motives (situational characteristics) (Norberg et al. 2007). There is a greater likelihood that consumers will create online content in the expectation that other members will use social media platforms to share their experiences about the brand without any extrinsic profit motive. Hence, we can expect that consumer-created online content will have positive impact on their willingness to spread EWOM through their intrinsic motives.

This study also seeks help from social identity theory Tajfel (1979) in formulating hypothesis. This theory suggests that the individuals tend to identify with a group due to their emotional attachment and identification with that group. Scholars in past (Li et al. 2017; Nason et al. 2018; Wang 2017) have used this theory to explain how the firms engaged in CSR practices can develop consumers’ identities and sense of belongingness. Consumers can easily identify themselves with such socially responsible firms since these firms satisfy consumers’ need for self-definition (Walsh and Bartikowski 2013; Lii and Lee 2012). Scholars have been arguing that C–C identification can improve the positive evaluation of a company and its brands by consumers (Rather and Sharma 2016) due to fact that consumers identify easily with the company or the brand that matches their self-concept (Steffens et al. 2021; Stokburger-Sauer et al. 2012).

Social identity theory is relevant in our context since the self-concept of consumers is expected to change positively when they observe that their bank serves the society by taking different CSR initiatives. The consumers start identifying themselves with such CSR-oriented bank and endorse the bank through different media which includes spreading positive EWOM and content creation. The social media platforms provide social identity to those consumers who generate content (Fujita et al. 2018). Since the providers and receivers of this content share similar interests and opinions, consumers value these ideas and thoughts. These consumers even associate their self-concept with content creators on social media (Bergami and Bagozzi 2000) and support those brands that are congruent with their social identity (Todd and Melancon 2018). Hence, these consumers will support socially responsible companies and spread a positive EWOM. Based on this discussion related to social identity theory, we argue that UGC will moderate the relationship between CSR and EWOM.

Literature review and hypothesis development

Research scholars have employed C–C identification to examine the path connecting CSR and consumer response. C–C identification has been referred to as a shared identity between a firm and its consumers (Ashraf and Merunka 2013), who prefer to identify with a firm or a brand that matches their self-definition. By drawing on earlier research (Bhattacharya and Sen 2003) we define consumers’ identification with a company (C–C identification) as “an active, selective, and volitional act motivated by the satisfaction of one or more self-definitional needs.” This definition by Bhattacharya and Sen (2003) characterizes C–C identification as a consumers’ self-categorization state that includes consumers’ active support for a company which has similar values as that of its consumers. Thus, socially responsible consumers who believe in particular values, are likely to identify with a brand or a firm whose CSR practices are aligned with that of its consumers. The C–C identification generated through firm’s CSR messages can be a key mediator in consumer responses to firm’s CSR communications on social media (Marín et al. 2009; Bigné et al. 2019). As C–C identification can satisfy consumers' self-definitional requirements because of their identification with a socially responsible company, an analysis of C–C identification can better explain the effect of consumer perceived CSR actions on EWOM. (Martínez and Rodríguez Del Bosque 2013).

The current study uses the definition of CSR as proposed by Mascarenhas (2011) which states that CSR refers to “commitment and accountability of a firm to improve the society’s well-being through judicious use of firm’s resources.” The current study examines consumers’ perceived CSR communication by firm on social media and this communication can in turn be of two types-firm created and user generated (Rim and Song 2013). Our study is primarily devoted to user-generated content due to its reliability, non-commercial intentions and trustworthy nature (Christodoulides 2012). Thus, UGC is information about businesses and/or their brands that is created by users and may take the form of user-generated blogs, online product evaluations, online communities and so on (Lee and Hong 2016). In the banking context, we refer UGC as the content created by current consumers about the bank on various social media platforms.

The EWOM communication also has great impact on the information exchange among consumers and their reaction to the information (Brown et al. 2007). In this study, EWOM has been defined as “any positive or negative statement made by potential, actual, or former consumers about a product or company, which is made available to a multitude of people and institutions via the Internet” (Hennig-Thurau et al. 2004).

CPBCSRSM and C–C identification

A growing body of research suggests examination of the linkages between CSR and consumer behavior, with a focus on consumer identification with a company. In fact, scholars in past have shown the positive impact of CSR on C–C identification (He and Li 2011; Ma and Kaplanidou 2021; Marín et al. 2009). According to research on organisational identification, a firm’s CSR initiatives reflect a corporate character and identity that enable consumers to identify with the firm based on an assessment of the degree to which their own identities and that of the firm overlap (Lichtenstein et al. 2004; Maignan and Ferrell 2004). C–C identification literature (Bhattacharya and Sen 2003; Marín et al. 2009) also reveals that CSR initiatives are critical for building corporate identity that encourage consumers to identify with a company. Thus, those consumers who know about the firm’s socially responsible efforts will strongly identify with the firm than those consumers who lack awareness about such initiatives. Moreover, it has been suggested that consumers will strongly identify with a socially responsible firm so as to project their ethical image and improve self-esteem. (Aquino and Reed II 2002). This type of feeling also reconfigures the group identity of these consumers (Pettigrew 1998) and their intense involvement in firm’s CSR activities will create greater economic value for the business (Mahmood et al. 2021). Hence, the consumers who are knowledgeable about the company’s socially responsible initiatives will easily identify with the company. Based on this discussion, we propose that:

H1

There exists a significant and direct relationship between CPBCSRSM and C–C identification in retail commercial banking sector.

C–C identification and EWOM

A socially responsible firm can create C–C identification by engaging consumers through CSR communication and hence building its image as one that adheres to and believes in ethical values (Meehan et al. 2006). When consumers find congruence between a firm’s socially responsible image and their self-identity, they will tend to identify with the firm and form positive emotional bonds (Balmer 2001). Such emotional bonds establish strong and meaningful relationships between the firm and the consumers who become brand advocates or ambassadors (Bhattacharya and Sen 2003). Moreover, it can be inferred from social identity theory that C–C identification will lead to consumer’s discretionary role behavior, for example purchasing sustainable products (Noh and Johnson 2019) and favorable product recommendations on different social media (Romero and Ruiz-Equihua 2020). Similarly, when consumers identify themselves with any organization, they are likely to promote that organization as they feel to be a part of the community (the organization). An organization’s CSR initiatives motivate its consumers to provide favorable ratings and actively engage in social media conversations leading to positive EWOM (Fatma et al. 2020).

Hence, the following set of hypotheses is proposed.

H2

There exists a significant and direct relationship between C–C identification and EWOM in social media in retail commercial banking sector.

CPBCSRSM and EWOM

Firms’ engagement in CSR provides them several benefits, including an improvement in EWOM. The consumers of such businesses are more likely to tell their peers, friend and family, about the firm’s socially responsible practices (Bhattacharya and Sen 2004). This communication creates a socially responsible image of the firm and thus increasing consumers’ interest in it and its brands (Chu et al. 2020). When consumers find that a firm is actively engaged in CSR and contributing to the environment and society, they respond positively through positive EWOM about that firm (Schaefer et al. 2020). Cheng et al. (2021) have shown that a bank's CSR engagement aids in generating a favourable EWOM from its consumers on social media. Similar findings have been shown by Khan et al. (2015) in Pakistan’s banking sector. Other researchers have also verified a connection between bank’s CSR activities and a favorable EWOM (Abbas et al. 2018). Such consumers spread a favorable EWOM and personally recommend a service through informal networks (Bansal and Voyer 2000; van Doorn et al. 2010). In a study of Hong-Kong retail commercial banks, researchers (e.g., Tong et al. 2012) have demonstrated positive impact of CSR implementation on EWOM. Hence, we posit the following hypotheses:

H3

There exists a significant and direct relationship between CPBCSRSM and EWOM in social media in retail commercial banking sector.

Mediating effect of C–C identification between companies’ CSR initiatives and EWOM

The research is still inadequate about different paths connecting CSR with

EWOM (Marín et al. 2009). C–C identification provides one path linking CSR with

EWOM (Deng and Xu 2017). Consumers prefer to be identified with a socially responsible company (Bhattacharya and Sen 2004) due to its ability to improve their self-esteem and social image (Porter and Donthu 2008). Consumers with strong C–C identification levels frequently engage in behaviors such as spreading a positive word about company on social media (Hur et al. 2018). Scholars (e.g., Lichtenstein et al. 2004) have found significant effects of a firm’s CSR initiatives on C–C identification. Studies reveal that whenever consumers identify with a brand or a company, they feel connected to it and share common interests. These emotional connections encourage consumers to put extra efforts in realizing the goals of the firm voluntarily. Thus, C–C identification can further affect consumers’ response to firm’s CSR initiatives. Hence, we posit that C–C identification will show mediating effects in the relationship of organizational CSR initiatives and a favorable EWOM in social media.

H4

C–C identification shows mediating effects in the relationship between consumer perceived bank’s CSR  communication and EWOM in social media.

Moderating effects of user-generated content

When consumers themselves are source, the other consumers perceive CSR-related information as more reliable and persuasive (Ukpabi and Karjaluoto 2018) since the recipients perceive that the consumers have no intrinsic motive to highlight or promote the company’s CSR activities (Kim and Kim 2014). In fact, any favorable EWOM by consumers makes them an honest CSR proponent of the company, and therefore the fellow consumers value integrity of UGC (Dickinger 2011). On the contrary, any social media posts by the company or its sources is thoroughly scrutinized by readers, who attribute such posts to commercial motives of company, and not worthy of a favorable EWOM (van Noort and Willemsen 2012; van Zoonen and vander Meer 2015). Hence, we argue that the relationship between consumer perceived bank’s CSR communication and favorable EWOM is strengthened by UGC.

H5

UGC messages about consumer perceived bank’s CSR communication in social media strengthen the direct effect of EWOM.

The summary of all these relationships is provided in Fig. 1.

Fig. 1
figure 1

Hypothesized model. Note: CPBCSRSM: Consumer perceived Bank’s CSR communication in  social media; CCI: consumer company identification; UGC: user-generated content; EWOM: electronic word of mouth; dotted lines represent indirect effects; ***p < 001, **p < 0.01, *p < 0.05

Full size image

Methodology

The hypothesized model in the current study was tested in Indian public sector banks through a cross-sectional study conducted from April to August 2022. The author selected those public sector banks that were communicating their CSR initiatives on different social media platforms. Based upon a review of different websites, the five banks namely State Bank of India (SBI), Punjab National Bank (PNB), Bank of Baroda (BoB), Vijaya bank and Bank of India (BoI) were found to be most active on social media. These five banks have presence across India, and are actively engaged in socially responsible activities. Hence, these banks have high brand recognition among consumers. The author collected data only from those individuals who had at least one bank account in these banks and were active on any social media platform for past one year. This step would ensure the selection of only those consumers who are active on social media and are familiar with the above-mentioned banks’ CSR initiatives on social media.

We personally contacted 700 consumers in national capital region of India including the cities of New Delhi, Gurgaon, Faridabad and Noida. who voluntarily answered the questions based on their perceived communication of CSR initiatives of banks on social media. Despite an increase in usage of internet-administered surveys, concerns have been raised about their overall quality, reliability and validity of the responses obtained through internet (Birnbaum 2004). Therefore, to overcome this limitation, pen-and-paper questionnaire was distributed to the consumers leaving a bank branch or an ATM counter. We assured the respondents complete anonymity and finally retained 648 valid surveys for further analysis.

The author used a 7-point Likert scale (where 1 = strongly disagree and 7 = strongly agree) which was pre-tested on five bank managers, ten marketing professors and 25 bank consumers that provided suggestions about item comprehension and wording (Hair et al. 2014). The pre-testing ensured that respondents have clearly understood the measurement items which are relevant for the target population. For better clarity of measurements items, the respondents were free to seek any further clarification. This step helped in ensuring scale’s content validity. For testing the model and examining the relationships among the constructs, we used CFA and SEM to test the hypothesis, in two steps. First, the author used structural model for testing the direct effects without any mediator. Then the author used the bootstrapping option (AMOS) for testing the output of mediation effects.

Sample

The population in this study includes all retail consumers who evaluate CSR practices of Indian public sector banks which have presence on social media. For deciding the sample size, the author followed the guidelines provided by researchers in past (Sekaran and Bougie 2016; Kline 2015). As per the recommendations of these researchers, the preferred sample size should be more than ten times the number of items. Hence, a sample size of 648 valid responses in this study is justified. We used convenience sampling technique, one of the most popularly used technique for sample selection (Zhao 2021). Researchers in past (Peterson 2001; Simonson et al. 2001) have reported that most of the samples in convenience sampling have been college students. Researchers (e.g., Lucas 2003) have also suggested the use of non-college students as sample for external validity. Therefore, this study has used convenience sampling but with non-college students.

Measures

The current study has adopted previously established scales (“Appendix”). These scales have already been tested for their reliability and validity in other countries and industries. As discussed in the next few sections, these scales were found to be reliable and valid for Indian banking sector in social media context.

Results

Common method bias

By using three different techniques, the current study examined common method bias (CMB). By using Harman’s test, our findings show that any one factor could explain a maximum of only 27.8% of variance, that is, less than the recommended value of 50% (Podsakoff et al. 2003). Moreover, the VIF values in current study were not more than the recommended threshold value of 3 (Hair et al. 2014). Finally, all correlations among constructs in this study are significantly less than 0.9 (Table 2), indicating that there is an absence of common method bias (Bagozzi et al. 1991).

Demographics

The respondents’ descriptive profile (Table 1) reveals that there were 52.16% male and 47.83% female respondents. Among the respondents, 16.98%,33.81%, 20.68% and 14.66% were in the age group of 18 to 30 years, 31 to 40 years, 41 to 50 years and 51 to 60 years, respectively. The remaining 13.89% were in the age group of above 60 years. These respondents were high school pass outs (18.98%), graduates (41.36%), postgraduates (21.14%) and with miscellaneous qualifications (18.52%).

Table 1 Sample demographics
Full size table

Reliability and validity

The first step involved determining whether items were sufficiently loaded on their corresponding factors. All factor loadings in this study are in the range 0.61 to 0.89 (Table 2) and the loadings were more than 0.5, which satisfy the recommendations of Fornell and Larcker (1981). The output of correlation analysis (Table 2) shows a positive correlation with each other. Moreover, a moderate range of these values indicates absence of multi-collinearity among variables. The skewness and kurtosis values in our study were in the range (+ 1.16 to + 1.94), indicating the normal distribution of data used in this study (Brown and Dacin 1997).

Table 2 Scale reliability and validity
Full size table

The author employed confirmatory factor analysis to ensure that the data fits in the proposed model. The results show good fit as indicated by appropriate indices (χ2/df = 2.16, RMSEA = 0.041, CFI = 0.916, GFI = 0.959, AGFI = 0.918, NFI = 0.933). The author then tested the scale for its convergent and discriminant validity. To confirm the scale’s convergent validity, the value of average variance extracted (AVE) for a construct should be more than 0.5, and for better internal consistency of items composite reliability (CR) value has to be more than 0.6 (Fornell and Larcker 1981). The AVE values in our study were in the range 0.57 to 0.62, and CR values were in the range of 0.71 to 0.78 (Table 2) thus confirming the convergent validity and internal consistency. Moreover, all the correlation values of different constructs were less than the squared root of AVE of each construct (Table 2). Thus, the criterion of discriminant validity assessment is satisfied. Cronbach alpha (α) values in our study are higher than recommended values of 0.7, indicating that the scale is reliable (Nunnally 1978).

Results

The hypotheses in this study were tested in different steps. In the first step, the author examined the direct path effects. The results (Table 3) revealed that all three hypotheses are validated since all the direct paths were found to be significantly positive. The direct path (H1) from CPBCSRSM to C–C identification was supported in our study (β = 0.381, t-value = 26.748, p < 0.001). The direct path from C–C identification to EWOM was also supported in our study (β = 0.312, t-value = 22.547, p < 0.001), supporting H2. The findings in this study show the significance of CPBCSRSM and their role in influencing EWOM, supporting H3 (β = 0.227, t-value = 18.619, p < 0.001).

Table 3 Statistics of hypotheses testing
Full size table

Although the inference of bootstrapping method is based on the estimates of indirect effects, this method is preferred over Sobel test (Hayes 2015) since bootstrapping method is free from any assumption about the sampling distribution of indirect effects and its shape. Hence, we used bootstrapping method (n = 2000) in AMOS 26 for testing the mediation effects. The results (Table 3) indicate full mediation of C–C identification since the β value (= 0.026) becomes insignificant when the mediating variable is introduced as compared to a highly significant β value of 0.381 (p < 0.001) without any mediator. Thus, the bootstrapping results in this study show mediating role of C–C identification between CPBCSR and EWOM supporting hypothesis H4.

We tested the moderating effects of UGC through interaction effects (Table 3). First, we examined the direct effect of CSR on EWOM and then interaction effects. Our findings indicate that CSR has significantly positive effect on EWOM (β = 0.227, t = 18.619, p < 0.001). Similarly, UGC has significant positive impact on EWOM (β = 0.103, t = 2.014, p < 0.01). Moreover, interaction effects of CSR and UGC (CSRXUGC) on EWOM are also significantly positive (β = 0.117, t = 2.318, p < 0.01) and suggest that UGC strengthens the positive relationship of CSR with EWOM. Therefore, the relationship between CSR and EWOM is positively moderated by UGC, supporting H5.

Discussion

The direct path (H1) from CPBCSRSM to C–C identification was supported in our study. By using social identity theory, we can infer that a bank’s CSR communication is appreciated by its consumers who willingly identify with a socially responsible bank. Further, a bank’s socially responsible behavior is an organization-related attribute. While making a purchase decision, the digital age consumers consider product-related attribute such as product’s quality and also organizational attributes such as its CSR initiatives. The consumers’ identification with the bank and its attributes leads to a sustainable relationship between the bank and its consumers. This relationship has also been supported in extant research (Baskentli et al. 2019; Cheng et al.2021; Choi et al.2019).

The direct path from C–C identification to EWOM was supported in our study supporting H2. A key consequence of consumers’ identification is the favorable EWOM about the identified-with brand or organization, on different social networking platforms. Such EWOM includes the recommendation of a firm’s brands to relevant others or even defending the firm in adverse conditions (Schouten et al. 2020).

The findings in this study show the significance of CPBCSRSM and their role in influencing EWOM, supporting H3. The CSR communication by a bank on social media creates the bank’s image as an honest and transparent as its consumers feel that a socially responsible bank will be fair in its transactions. Hence, they evaluate such bank favorably and recommend these banks to others by spreading a favorable EWOM. Further, consumer purchase risk is high in case of services (Sweeney and Swait 2008) such as banking, due to experiential nature of services. In such situations, risk can be reduced by consumers through recommendations or reviews of other users or buyers (Ng et al. 2011).

Our study has confirmed the mediating role of C–C identification between CPBCSR and EWOM supporting hypothesis H4 (Table 3). These findings can be explained through social identity theory. The social identity theory suggests that a company’s communication of socially responsible practices on social media increases the likelihood of consumers’ identification with that company who spread a favorable EWOM about the bank. Our findings have also been supported by research scholars in past (Ahmad et al. 2021; Bialkova and Te Paske 2020; Contini et al. 2020). Hence, in line with social identity theory, the current study has established that bank’s CSR communication influences EWOM both directly and indirectly. These findings have been supported by previous studies (Arıkan and Güner 2013; Fatma et al. 2018; Vo et al. 2019).

The results confirmed that a bank’s CSR communication on social media provides the consumers a base to identify with a social group (bank) and the consumers take pride in being associated with such group. This identification with the bank encourages the consumers to share their opinions and feelings about the bank with their peers, friends and family members on different social media platforms such as Facebook, Twitter and Instagram. Based on social identity theory, this study has contended that a bank’s socially responsible behavior is an organizational attribute and consumers consider it to be a discretionary attribute for the benefit of the society and its environment. While making any purchase decision, these consumers look for the product attributes such as a product’s quality, as well organization-related attributes such as its CSR practices. Hence, by using social identity theory, our study has established the direct and indirect influence of CSR on EWOM in Indian banking sector. These findings have been supported in previous studies (Deng and Xu 2017; Huang et al. 2017; Vo et al.2019).

Our findings also indicate that the relationship between CSR and EWOM is positively moderated by UGC, supporting H5. Although few scholars (e.g., Filieri et al. 2015) suggest that source credibility is not significant when communicating CSR through social media, other scholars (e.g., Dickinger 2011; Sparks et al. 2013) have shown that effectiveness of CSR communication from an external (to the company) and independent source on social media is more favorable. Our findings concur with the second school of thought.

Theoretical and practical contribution

Our findings will provide a significant forward step in banking sector research regarding the factors influencing EWOM and the role of CSR communication in social media. First, our study demonstrates the important role of social media in dissemination of a firm’s CSR-related content. Our results reveal that consumers willingly identify with the bank that they perceive to be socially responsible and active on social media. The consumers also share their opinions and feelings about such banks. Second, this study advances the literature on CSR, C–C identification, EWOM and UGC by providing an integrative framework for a comprehensive study of the factors that directly or indirectly influence consumers’ intention to spread a favorable EWOM through social media. Although a number of studies have been conducted in EWOM domain, the studies are limited about CSR communication in online banking context and have examined these variables in isolation (Augusto and Torres 2018; Choi et al. 2019; Vo et al. 2019). This study provides an integrated model to incorporate UGC as a moderating variable. The study also demonstrates that UGC strengthens the effects of CSR on EWOM. Third, our study also adds to contemporary CSR literature in a developing country where the topic of CSR engagement through social media is still in nascent stage. In fact, CSR practices in banking sector of emerging economies have received limited attention of scholars from consumers’ viewpoint (Ahmad et al. 2021; Fatma et al. 2020). Fourth, this study has also expanded earlier studies which have suggested that EWOM communication can play an active role in services industry (Harris and Goode 2004) through mediating variables such as C–C identification (He et al. 2012). Fifth, this study has also improved the generalizability of the findings since many studies have been conducted across countries during Covid-19 period. In fact, this study will be among the few studies during post Covid-19 period. Sixth, this study has further improved the generalizability since it includes multiple social media platforms as opposed to single platform used in earlier studies. Finally, researchers in past (e.g., Fatma and Khan 2023) have conducted studies in Indian banking context, their studies included only a limited number of banks and retail banks in general. By incorporating five public sector banks, our study has increased the number of banks and acknowledged the contribution of public sector banks in India. Our study has demonstrated the mediating role of identification similar to other studies in past (e.g., Fatma and Khan 2023; Fatma et al. 2022).

The current study’s practical implications are also important for banking sector. The current study has highlighted the significance of banks’ CSR communication on social media in generating consumers’ favorable word-of-mouth and positive content online. The consumers give more weightage to the opinion of their close acquaintances including family members and friends, when compared with commercial advertising. A possible explanation can be that consumers attribute advertisement as an external motive which attempts to manipulate profits.

Moreover, when a bank’s consumers spread a positive EWOM, the results create more impact than any advertising campaign of banks since the consumer feedback to potential consumers is viewed as more honest and transparent. Hence, the consumers find the opinions and online content of other consumers more trustworthy than the advertisements. In other words, consumers trust UGC and EWOM. Therefore, banking sector professionals can eliminate or reduce the focus on paid advertisements and focus on CSR engagement for favorable UGC and EWOM. The consumers will share their association with a socially responsible bank on digital platforms with existing and potential consumers. Banks can translate this association to consumers’ identification with the bank. Thus, the policymakers in banking sector can use these findings in designing their strategies around the CSR communication and consumer engagement in social media.

Conclusion

To address various gaps in existing research, our study proposed a model that incorporates CPBCSRSM, C–C identification, UGC and EWOM in Indian public sector banking context. These banks play a major role in supporting Indian economy and are facing problems in defending their market share due to competitive market. This study has used attribution theory and social identity theory as grounding theories. The current study explored the role of social media in spreading a positive EWOM. The findings indicate that CPBCSRSM on social media have direct effects on EWOM. Banks therefore need to encourage consumers’ participation and engagement to develop consumers’ identification with the banks, which will further enhance favorable marketing outcomes. CSR communication through social media can also help banks in getting their messages or information being communicated more effectively and efficiently to a larger but meaningful audience. The consumers’ interaction through CSR dialogues is also likely to strengthen their identification with the bank and hence improve their willingness to spread a favorable EWOM. Our study thus provides mechanism of communicating bank’s CSR initiatives to consumers both directly and indirectly. The findings also indicate the mediating role of C–C identification and moderating role of UGC between CPBCSRSM and EWOM. Our findings indicate that the social media is beneficial for Indian public sector banks for garnering a strong and favorable EWOM communication. The current study has affirmed that when consumers find their bank actively involved in CSR activities, they easily identify themselves with their bank. This identification in turn spreads a favorable EWOM. The study also examined the strengthening effect of UGC between CSR and EWOM. Thus, in extremely competitive Indian banking sector, public sector banks need to focus on C–C identification, UGC and EWOM communication. The banks can use the insights gained from our study to integrate the CSR initiatives in their business strategy.

Limitations and future research

In this study, we have considered only one construct that mediates the relationship of consumer perceived CSR communication in social media and EWOM. In fact, C–C identification is not the only mediator but other alternative possible paths could also explain the linkage of CPBCSRSM with EWOM. Also, there are specific CSR activities that determine consumer perceived CSR value. For example, CSR activities related to product improve self-oriented value. Such values mediate firm’s CSR communication and favorable outcomes. Scholars in future may like to explore more such mediating variables.

Moreover, this study has considered only positive aspects of EWOM. Any EWOM communication among consumers will have repercussions as well. Research scholars in future may examine the role of banks’ CSR activities, which may alleviate the harmful effects of consumers’ negative EWOM. Scholars in future may also like to broaden this study by analysing the impact of CSR communication on cognitive states, such as consumer disidentification with a company. This situation may arise if the consumers find any difference in actual CSR practices of a bank and its communication on social media. Scholars in future may like to test the hypotheses in different banks (e.g., private and foreign) to verify the hypothesized relationships. They may even compare the consumer perceived CSR practices and the outcomes in these banks. Finally, our study has investigated only consumers’ behavioral intentions related to EWOM. Scholars in future may examine real behaviors with respect to CSR communications.