How social CRM and customer satisfaction affect customer loyalty Open Access

Recently, an increasing number of companies have developed social customer relationship management (CRM) strategies to integrate the opportunities offered by Web 2.0 into their marketing approaches. According to Greenberg (2009), social CRM is a business strategy that leads to customer engagement through social media (SM) with the objective of sustaining customer trust and brand loyalty. This innovation allows firms to center their strategy on customers, reach and attract them with user-generated content, enhance their engagement using online social interactions and retain them by generating new relationships with other customers (Rodriguez and Trainor, 2016). On the other hand, social CRM has given the ability to the customer who can express himself through blogs and forums and give his opinion quickly, freely and easily – about the products and services he has purchased – to other customers or advertisers.

Despite its different advantages, the use and effect of social CRM on customer satisfaction (CS) and loyalty (CL) differ from one sector to another and between developed and developing countries. Indeed, while organizations in developed economies use smart technologies, those in developing countries care about providing basic information systems (IS), and the majority of them are still trying to digitize their services (Stokić et al., 2019). As a consequence, investments in social CRM solutions in developing countries still lag behind those in developed countries, and their payoff and impact on CS and CL are still unassessed (Jaber and Simkin, 2017).

Although widely discussed, the influence of social CRM on CS and CL remains one of the most challenging issues faced by marketing researchers (Nyadzayo and Khajehzadeh, 2016). First, the mediating role of CS between social CRM and CL has received restricted scientific interest (Khan et al., 2022). Indeed, Cakici et al. (2019) underlined that CL can be increased through CS, which is a potential mediator between CRM and loyalty. Second, some specific sectors where customers are implicated in long-term and deep relations with companies need more investigation to inform managers about how to benefit from social CRM systems and how to generate higher levels of satisfaction and loyalty (Chen et al., 2021). This is the case for the automotive sector, where the minimum price of a car is over US$15,000, and unlike other products, the ownership cycle exceeds five years. In a context marked by digitalization and enhanced competition, many automotive firms still fail to reach CS and CL due to deficiencies in resources and a lack of strategies (Khan et al., 2021). Hence, firms operating in such an environment need to understand all social CRM factors that affect CL to develop sustainable relationships with customers, which can be maintained and strengthened beyond the transaction (Nyadzayo and Khajehzadeh, 2016). Third, most of the studies in this area were undertaken in developed countries (Tsokota et al., 2021). Nevertheless, firm practices, market conditions and product life cycles are different in developing countries compared to developed countries (Wang and Chen, 2004). Cultural values, customer behaviors, incomes and price sensitivity are too divergent; the technological infrastructure needed for social CRM remains less efficient, and the usage and practices of internet users are quite different (Medjani and Barnes, 2021). Therefore, the study of social CRM in economies with different markets and contexts in general and developing countries in particular can help us to generalize the social CRM scale and to expand our global understanding of how it impacts CS and CL (Akroush et al., 2011).

In summary, the scarcity of studies addressing social CRM in the environment of developing countries’ organizations and specific sectors, such as the automotive industry, is still a major issue (Kebede and Tegegne, 2018). Hence, it is necessary to conduct more studies to understand the mediating effect of CS between social CRM and CL from a customer angle (Aldaihani and Ali, 2018). From this perspective, this study revolves around the influence of social CRM on CL by answering the following research question: Does CS play a mediating role in the relationship between social CRM and CL?

To address this question, the structure of this paper is divided into six sections. Section 2 presents the literature review and the proposed hypotheses. The research methodology is described in the Section 3; follow by the results obtained in the Section 4. Finally, in Sections 5 and 6, we provide theoretical and managerial implications as well as the limitations of the study.

The framework of this research has been developed based on relationship marketing theory (RMT). Arose in the early 90s, this theory tends to be long-term oriented; it originated from the idea of being focused on building value networks and valuable relationships instead of transactions (Caliskan and Esmer, 2020). Hence, RMT can be considered an approach to attract, establish, maintain and enhance long-term relationships with customers and other stakeholders by mutual exchange and fulfillment of promises (Koiranen, 1995).

Relationship marketing ties many seemingly unrelated marketing ideas, such as business-to-customer (B2C) marketing, CRM, promotional strategy and database marketing. According to this theory, it costs three to five times less to preserve a customer than to attract a new customer (Farber and Wycoff, 1991), and the firms’ profits can be boosted from 25% to 85% by decreasing customer defections by 5% (Hanley and Leahy, 2008). Indeed, when firms act to reinforce relationship policies and practices, commitment, trust and cooperative activities are likely to be improved, and marketing costs and perceived risk should be reduced (Palmatier, 2008). Consequently, customers’ perceptions and behavior will be shaped, their satisfaction and loyalty will increase, and firm performance will be enhanced (Halimi et al., 2011).

Therefore, relationship marketing aims to strengthen the firm–customer relationship and to progressively turn them into “clients” (who make regular purchases and then), “supporters” and eventually “advocates” for the company (Herington et al., 2006). Achieving this goal in today’s digital age requires a special solution that includes personalized messages and services and responsive communication such as social CRM.

Social CRM includes the principles of CRM using social networks. Therefore, this managerial innovation should be considered an enrichment of traditional CRM (TCRM), which follows the evolution of firms moving from purely transactional logic to interactional logic with their customers (Greenberg, 2009).

The goals of social CRM are to extract the highest value from customers over the lifetime of the SM relationships and to turn them into loyal customers. Once companies adopt social CRM, they can facilitate communication and collaborative experiences with their customers. Furthermore, through social CRM, firms become able to build a dynamic, interactive and multichannel process of mutual value creation and cocreation product development with customers, which allows them to create market-based products and to reach CS and CL (Hidayanti et al., 2018).

Recently, social CRM has become one of the most debated subjects in digital marketing (Yasiukovich and Haddara, 2021), and many researchers have proven that using CRM not only establishes superior relations with customers but can also boost CS and CL. For instance, Nadeem (2012) indicated that engagement through SM makes the customer more loyal and pushes him to spend more than other customers. Moreover, Dutot (2013) estimated that social CRM use and presence on social networks boost satisfaction and loyalty from a long-term view. In other studies, Siti Hasnah et al. (2019) and Zaim et al. (2020) found that social CRM adoption could increase CS. Finally, Arora et al. (2021) detected a significant impact of social CRM on customers’ engagement, satisfaction, retention and loyalty.

Based on the studies of Yawised et al. (2018) and Suhasini and Kumar (2018), social CRM was operationalized in this study through two variables: TCRM and SM.

TCRM can be defined as the process of managing the whole relationship between the company and its customers, with all their various contacts, interactive processes and communication elements (Grönroos, 2007). CRM is an iterative process in which companies review their goals regularly to affect customers’ behavior and consequently decrease costs, enhance CS and CL, attract new customers and increase sales and profitability.

Several researchers, such as Seify et al. (2020), have identified a positive relationship between CRM and CS. Furthermore, CL can also be achieved via a successful CRM application.

On the other hand, recent studies, such as Kim (2012), pointed out that CS plays the role of a mediator variable that connects CRM and CL. From this perspective, CRM influences CS, which in turn impacts CL (Nyadzayo and Khajehzadeh, 2016). In the context of developing countries, although firms are increasingly adopting TCRM systems at considerable cost, there is limited research examining the impact of CRM on other variables, such as CS and CL (Askool and Nakata, 2010), particularly in North African countries. Therefore, the following hypotheses were suggested:

SM are the technology used for interactions by the interactive Web (Yawised et al., 2018). This technology is changing customer behavior and marketing practices, especially in emerging markets (Enikolopov et al., 2018).

According to Potra et al. (2016), SM can immensely assist companies in delivering on the promise of the advertising concept, market orientation and relationship marketing by providing tools to improve CS and boost customer engagement that leads to CL. In this line of idea, Huang et al. (2018) and Pinto (2015) indicated that companies that have better communication with their customers on SM gain not only the knowledge of customers but also their satisfaction and loyalty. Furthermore, Yee et al. (2021) underlined CS as a mediating variable that affects the relationship between SM and CL. Thus, the following hypotheses were assumed:

First, CS is considered customers’ judgment of the perceived performance of a product or service in line with their expectations, as well as the level of pleasure acquired from consumption-related fulfillment (Kotler and Keller, 2016). In the online environment, satisfaction arises as one of the most important indicators of CL and firm success (Jacka and Keller, 2013). Indeed, satisfied online customers are likely to make purchases again and recommend online retailers to others (Pereira et al., 2017).

On the other hand, CL is a concept that has been widely adopted and used in the field of customer behavior for many years, and original studies have defined it as a behavioral manifestation that includes rebuying company products or services (Tellis, 1988). CL is considered a multifaceted construct that includes attitudinal and behavioral components (Fournier and Yao, 1997). Jacoby and Chestnut (1978) interpret that the behavioral dimension of CL is a form of repeat purchase behavior for a particular product or service. This dimension involves some degree of positive attitude toward the distinctive value associated with the product or service. In the current study, we measure loyalty from a behavioral perspective as the degree to which a customer exhibits repetitive purchasing behavior and shows a positive attitudinal disposition toward the products and/or services provider (Nyadzayo and Khajehzadeh, 2016).

CL depends to a greater extent on cumulative satisfying experiences with product/service attributes. This means that satisfied customers are more likely to be loyal (Minta, 2018). In this line of idea, Kotler and Armstrong (2006) argue that highly satisfied customers will repeat purchases and share their experiences with others. Therefore, companies need to better understand the relationship between CS and CL to allocate their marketing efforts between satisfaction initiatives and loyalty strategies (Shankar et al., 2003).

Empirical evidence from previous studies conducted in different contexts and countries has indicated that CS is a significant determinant of CL (Hu et al., 2009). This is the case for Liat et al. (2017) and Nobar and Rostamzadeh (2018), who found that CS induced CL in Malaysia and Iran. Moreover, Şenel (2011) mentioned that CS is a prerequisite of CL in the automotive sector in Turkey. Based on these results, the following hypothesis is proposed:

In light of this discussion, the theoretical model shown in Figure 1 was formed.

Like almost all developing countries, most Algerian firms struggle to provide Web connections and basic computer equipment (Stokić et al., 2019), and few sectors have seen a wide diffusion of social CRM systems (Jaber and Simkin, 2017). In this context, the Algerian automotive sector has emerged as an appropriate research field to investigate the influence of social CRM on CS and CL. As seen from the statistics of the Algerian National Office of Statistics (NOS), the Algerian automotive market recorded during the period of this study had one of the highest growth rates in Africa. Between 2012 and 2019, the Algerian automotive fleet grew from 4812555 to 6577188 vehicles, with average sales of 265507 vehicles per year (NOS, 2013, 2019). This period also saw the establishment of several multinational firms and the development of mounting activities, which encouraged the adoption of social CRM practices, especially by French and German leader companies.

From these considerations, an empirical study was conducted on customers of the Algerian automotive sector.

The questionnaire construction was based on previous studies. As shown in Table 2, each variable was measured using a scale tested and validated in several studies. To ensure that survey participants understood and answered the questionnaire in the way that our study intended, the questions were tested before the main survey. Three customers and four researchers specializing in the area performed the pretest. Both the structure of the survey (e.g. order of questions) and some unclear wordings were adjusted.

To test the hypotheses, an electronic questionnaire was administered – from January to May 2020 – to Algerian customers who have at least one car from any automotive brand and who use SM and were exposed to a social CRM campaign.

Items of the second section of the questionnaire ensure that the respondents have experience in social CRM. The SM platforms most frequently used by respondents to interact with automotive brands are Facebook and Instagram. In contrast, other social networks, such as Twitter, LinkedIn or YouTube, are never or rarely used. Moreover, the most valuable SM functions for the respondents are successively instant messaging, video hosting/sharing, news/live feed and photo sharing. Finally, these customers seem to be exposed to different CRM actions.

Similar to other e-marketing and CRM studies (Kumar et al., 2023), we used convenience and snowball sampling techniques. A total of 340 questionnaires were collected. Then, actual cleaning and data screening led us to eliminate 26 inconsistent responses and to retain 314 usable questionnaires for data analysis.

Table 1 shows that the respondents consisted of 268 males (85.4%) and 46 females (14.6%), which is representative of Arab societies. Despite progress in gender equality and even if an increasing number of women drive and own cars, male drivers remain the majority in these countries (Aissaoui, 2022). Furthermore, the vast majority of respondents were within the age range of 20–30 (61.1%) and 31–40 (28%) years old, which corresponds to the characteristics of the young Algerian population (70% are under 30 years old) (Boukhelkhal, 2022), of which the majority of people who are able to drive or own a car is between 20 and 40 years old. In addition, Millennials (born from 1982 to the mid-1990s) and Generation Z (born from the mid-1990s onward) who grew up in the age of ICT, internet connectivity and SM (Agrawal, 2022) are more likely to interact with brands through social CRM tools and share their consumption-related experiences across digital platforms (Khan, 2022). Finally, most of the respondents had cars from the leading brand in Algeria (25.2%), while the second brand reached 14.7%, the third brand reached 8% and the rest of the percentage was divided among the other brands.

This study used the covariance-based structural equation modeling approach and the bootstrapping approach (AMOS version 23.0) for data analysis. A three-stage data analysis was performed to validate the proposed framework. First, reliability and validity tests were performed. Second, the direct relationships were tested. Then, the mediation analysis was tested.

To ensure that the data were normally distributed, skewness and kurtosis tests were performed. As recommended by Hair et al. (2010), the obtained values in this study are between −2 and 2 for skewness and between 7 and 7 for kurtosis, which support the assumption that the data follow a normal distribution (Table 2).

To measure the reliability and convergent validity of the constructs, Cronbach’s alpha, factor loading, composite reliability (CR) and the average variance extracted (AVE) were calculated for each construct. Regarding factor loading, all item loadings in the model should be 0.50 or higher (Hair et al., 2010). Furthermore, Cronbach’s alpha and CR values should exceed 0.70 (Hair et al., 2010), and the AVE should exceed 0.50 (Kline, 2011). As shown in Table 2, all values of factor loading, CR, Cronbach’s alpha and AVE exceed the recommended thresholds. Thus, the reliability and convergent validity of the model were established.

To evaluate discriminant validity, the Fornell–Larcker criterion (square roots of the AVEs) was used. The recommended standard of the Fornell and Larcker techniques is that they should not obtain the same variance as any other variables, which is more than the AVE (Fornell and Larcker, 1981). The values in Table 3 ensured that the observed variables in the model in each construct show the given latent variable, authenticating the discriminant validity of the model.

In addition, all good fit indices of the measurement model were within the recommended level, indicating that the model is acceptable and has sufficient goodness of fit to the observed data. Relative chi-square = 2.885, incremental fit index = 0.939, normed fit index = 0.91, Tucker–Lewis index = 0.924, comparative fit index = 0.939, root mean square error of approximation = 0.078.

Finally, the explanatory power of the model – estimated by the coefficient of determination (R²) – is 25.6% for CS and 45.4% for CL (Figure 2).

The main causal paths were tested using structural equation modeling. As seen in Figure 2, all hypotheses (H1, H2, H4, H6 and H7) were supported except for H5. Therefore, the results indicate that TCRM (β = 0.606, p < 0.001) and SM (β = 0.135, p = 0.013) positively and significantly influence CS. In addition, the findings showed that CS (β = 0.692, p < 0.001) and TCRM (β = 0.224, p = 0.027) have a positive and significant effect on CL (0.224, p = 0.027). However, there was no direct relationship between SM and CL (β = 0.080, p = 0.157). In fact, the relationship between SM technology use and CL seems to be indirect rather than direct. As seen below, CS has a significant mediating effect on this relationship.

To examine the mediation relationship, we used the bootstrapping method. The bootstrapping method allows us to provide the upper and lower bounds for the confidence interval with a 95% confidence level and to confirm that the estimate of the indirect effect is inside this confidence interval (Arbuckle, 2009). As shown in Table 4, the estimate of the indirect effect between TCRM and CL (0.259) is significant (p < 0.01) and inside the confidence interval (0.251, 0.640). Thus, the CS variable is considered a mediator between TCRM and CL, and H3 has been supported. In addition, the indirect effect between SM and CL (0.085) is also significant (p < 0.01) and inside the confidence interval (0.024, 0.187). Therefore, the CS variable is considered a mediator between SM and CL, and H6 has been supported.

The competitiveness of the automotive sector and increasing virtual interactions that occur today among customers through SM make it critical for managers to be present and available on Web 2.0 and to develop a digital long-term connection with their current and potential customers. Indeed, ensuring CS and CL consolidates relationships with companies and can be decisive in a maturing sector characterized by an accelerated rhythm of innovation and strong interest shown by customers in after-sales services (YuSheng and Ibrahim, 2019).

This research contributes to the literature by the following:

• analyzing the influence of social CRM on automotive CL; and

• examining how CS affects the precedent effect.

The findings have provided empirical support for the contention that social CRM (TCRM and SM) has a direct and positive impact on CS in developing countries’ automotive sector. This result is consistent with the findings of previous studies conducted in other sectors and countries, such as Aldaihani and Ali (2018) and Seify et al. (2020). Furthermore, CS has a positive effect on CL, which is consistent with the results of Şenel’s (2011) research in Turkey in the same sector and in other sectors, such as Nyadzayo and Khajehzadeh (2016) and Nobar and Rostamzadeh (2018). Therefore, in addition to social CRM efforts, CS creates CL in developing countries’ automotive sector, leading to future behavioral intentions, which means that loyal customers can recommend the automotive brand to their relatives and friends and post positive comments on that brand on SM. In this context, companies can use a social CRM strategy to upgrade the quality standards of products/services, which are produced or delivered in the end to gain CS that leads to CL.

Moreover, the study results also revealed that CL to automotive brands in developing countries is directly affected by TCRM. This result supports the results of previous studies, such as Shaon and Rahman's (2015) research.

On the other hand, even if the research failed to detect any direct effect of SM on CL, the mediating role of consumer satisfaction between social CRM (TCRM and SM) and CL was confirmed. These results are consistent with those underlined by Kim (2012) and Yee et al. (2021) and attest that CL is affected indirectly by social CRM through CS. Thus, the relationship between social CRM and CL appears more complex. In addition to the mediating effect played by CS, other potential mediators could explain the nonsignificant direct effect of SM technology use on CL: CRM capabilities (Trainor et al., 2014), customer engagement (Arora et al., 2021) and customer empowerment (Aldaihani and Ali, 2018). In the context of this study, even though the respondents were exposed to social CRM campaigns, their level of engagement seems to be low. Referring to the collected data, they use, on average, Facebook and Instagram to interact with automotive companies from time to time and rarely or never use other SM. A final explanation is related to customer relationship quality and strength. In developing countries, where customers have a relatively lower income (compared to developed countries), automotive brands tend to offer standardized cars with little possibility of customization. In addition, despite the quality of the after-sales services provided by these brands, the counterfeit spare parts and the services offered by independent repair remain very cheap, which can affect social CRM efficiency. Therefore, it is worth noting that our findings provide more evidence supporting the results in the marketing literature, developing country context (Medjani and Barnes, 2021) and automotive sector (Nyadzayo and Khajehzadeh, 2016).

While several studies conducted in different sectors focus solely on either CS or CL (Nadeem, 2012; Aldaihani and Ali, 2018), this research reduces the knowledge gap in the fields of RMT and social CRM literature by exploring the outcomes of social CRM on CS and CL.

First, our model was tested in the Algerian automotive sector, as Yasiukovich and Haddara (2021) recommended expanding to new sectors and countries. Therefore, the scales of social CRM (Iriana and Buttle, 2008; Trainor et al., 2014), CS (Chang, 2015) and CL (Nyadzayo and Khajehzadeh, 2016) are generalizable to the Algerian automotive sector as an example of developing countries’ sectors. Such empirical validation is needed to provide sufficient evidence on how to properly translate the social CRM concept into a comprehensive set of specific organizational activities conducive to CRM success (Akroush et al., 2011). Second, this study tests and validates the direct relationship among important constructs of social CRM (TCRM and SM) and CS (Aldaihani and Ali, 2018). Third, this research not only empirically examines the direct relationships but also validates the mediating effect of CS, which plays the role of a mediator variable in various previous studies (Kim, 2012; Yee et al., 2021). Finally, even if the results validate the direct relationship between TCRM and CL, they failed – in turn – to detect any significant and direct effect of SM technology use on the dependent variable (CL). This prompts us to ask about the moderating and mediating variables–inherent to developing countries–likely to explain such results.

Practically, our study provides some precious implications for practitioners and firms activated in developing economies, particularly those in the automotive sector.

First, companies should develop strong and sustainable relationships with their customers to improve CS and CL and consequently enhance their competitiveness and market share. This goal can be reached by implementing social CRM best practices.

Second, companies should focus their social CRM campaign on the most SM platform used by customers.

Third, in a growing market, where automotive offers are standardized, companies in developing countries should offer more choices to customers and allow them to personalize their requests.

Fourth, it is important to promote and encourage communication with current and potential customers to increase the purchase and repurchase of products and to improve value cocreation. Marketing managers should develop strategies that make customers able to share their experiences and opinions, which help the company explain and understand needs.

Fifth, our study revealed that young people use SM and the internet more than older generations. Consequently, companies, especially B2C businesses and automotive firms, should reconsider SM as an integral element of their communication strategy to maintain connections with them.

Beyond the automotive sector, other industries where customers are involved in deep and long-term relationships with companies could also benefit from the study findings, especially in developing countries.

From an international business and marketing perspective, our research offers empirical insights for international organizations that are considering entering the North African and Middle Eastern automotive market by providing a deeper empirical understanding of social CRM – CS and CL relationships. This could reveal new business opportunities or provide clues in understanding automotive markets in neighboring countries that share similar characteristics with those of Algeria (Akroush et al., 2011).

Table 5 summarizes the research conclusions and implications:

Despite its important contribution, this study is not without limitations. First, the generalizability of our results is limited, the data were collected from a self-report survey, and the sample size is relatively small compared to the total number of automotive customers present in the area. The limited size of the sample covered in this study may negatively reflect the generalizability of the results to the whole population. In addition, the majority of respondents were men and young. This may not be generalizable for women or customers over age 40. Further studies should consider collecting real social CRM use data from a larger sample. This will lead to an increase in the limited degree of freedom in the model.

Second, although the conceptual model explains 45.4% of the variance of CL, we can improve it by adding other variables such as customer engagement or customer empowerment. Furthermore, we used data from a B2C relationship within a single sector, suggesting that the results cannot be immediately applied to business-to-business (B2B) contexts. Therefore, future studies should investigate the antecedents of CRM and its impact on brand loyalty in other B2C or B2B markets.

Finally, only a quantitative analysis was undertaken in this research. For future studies, a mixed method using both quantitative and qualitative techniques can be used to test whether the supported hypotheses are still valid.

These authors contributed equally to this work for Djihane Malki & Mohammed Bellahcene