Engaging customers through user-and company-generated content on CSR

Abstract

Purpose

This study aims to analyze the role of corporate social responsibility (CSR) customer perceptions, customer–company identification and customer trust on customer engagement (CE), paying special attention to the moderating effects of two types of social media communication, firm-generated content and user-generated content.

Design/methodology/approach

The study uses a mixed-methods’ approach. First, a single-factor experiment using Twitter posts as stimuli with 227 hotel guests. The structural model was analyzed using SmartPLS 3.2.7. Second, structured in-depth interviews were undertaken with three hotel industry experts to complement the conclusions of the quantitative study.

Findings

The results show that when a customer trusts a hotel and identifies with its corporate values, CSR tweets generate CE toward the hotel. CSR communications made by customers reinforce the impact of CSR tweets on customer trust more than CSR tweets posted by hotels. Hotel industry experts give insights to explain these results in different types of hotels.

Practical implications

CSR communications made through Twitter affect customers’ perceptions of a hotel’s CSR activities and customer trust in hotels, especially if they originate from a source external to the company. This result can be of use for hotel managers who have not previously given importance to active CSR communications or the interactivity of social media.

Originality/value

The authors show the moderating effect of user-generated content in the relationship between CSR customer perceptions and customer trust, thus contributing to the research into the effectiveness of social media. They use a mixed-methods’ approach to increase the validity of the results.

Propósito

Este estudio analiza el papel de las percepciones de RSC, la identificación cliente-empresa (CCI) y la confianza en el engagement del cliente (CE), prestando especial atención al efecto moderador de dos tipos de comunicación en redes sociales: Contenido Generado por la Empresa (FGC) y Contenido Generado por el Usuario (UGC).

Diseño/metodología/enfoque

Se emplean métodos mixtos: un experimento de un factor usando publicaciones de Twitter como estímulo con 227 huéspedes del hotel, cuyo modelo estructural se analizó con SmartPLS 3.2.7., y 3 entrevistas en profundidad con expertos de la industria hotelera.

Hallazgos

Los resultados muestran que, cuando un cliente confía en el hotel y se identifica con sus valores corporativos, los tweets de RSC generan engagement. La comunicación de RSC emitida por usuarios refuerza el impacto de los tweets de RSC en la confianza del cliente más que los tweets publicados por hoteles. Los expertos de la industria hotelera aportan nociones para explicar estos resultados en diferentes tipos de hoteles.

Implicaciones prácticas

La comunicación de RSC realizada a través de Twitter afecta las percepciones del cliente sobre las actividades de RSC del hotel y la confianza en el mismo, especialmente si proceden de una fuente externa a la empresa. Este resultado puede ser útil para gerentes de hoteles que no se benefician de la comunicación activa de RSC o la interactividad de las redes sociales.

Originalidad/valor

Se valida el efecto moderador del UGC en la relación entre las percepciones de RSC y la confianza de los clientes, contribuyendo así a la investigación sobre la efectividad de las redes sociales. Se emplea un diseño mixto para incrementar la validez de los resultados.

Palabras claves

Comunicación de RSC, Engagement del Consumidor, Fuente del mensaje, Contenido generado por la empresa, Contenido generado por el usuario, Twitter, Compromiso con el cliente

Tipo de artículo

Trabajo de investigación

Citation

Badenes-Rocha, A., Ruiz-Mafé, C. and Bigné, E. (2019), "Engaging customers through user-and company-generated content on CSR", Spanish Journal of Marketing - ESIC, Vol. 23 No. 3, pp. 339-372. https://doi.org/10.1108/SJME-09-2018-0043

1. Introduction

Corporate social responsibility (CSR) (for a review, see Riera and Iborra, 2017) is especially important in the hotel sector for two main reasons:

1. The hotel stay affects the tourist experience and its evaluation (Ettinger et al., 2018).

2. Hotel customers are becoming more concerned about hotels’ contributions to society and the environment (Farrington et al., 2017).

The increasing importance of social media is challenging managers and scholars to analyze the different ways to effectively communicate CSR activities (Bigné et al., 2019). Despite the calls to use social media to engage consumers through effective CSR communications (Du et al., 2010), the CSR literature has scarcely compared he effect of the different sources of CSR content. Wang and Huang (2018) compared the CSR-related posts made by CEOs and by their organizations, but not user-generated content (UGC). Majority of messages posted by tourist companies on social media (firm-generated content [FGC]) describe service attributes and neglect customers’ concerns about social issues. However, there is an increase in the volume of messages posted by tourists on this issue, and these might be the key to reaching the customer (Navío-Marco et al., 2018). Two ideas underlie this study. First, CSR is playing an increasing role in consumer decisions. Second, the implications of the social media message source for CSR communications. Therefore, we aim to compare the influence of UGC and FGC on the impact of consumers´ perceptions of companýs CSR activities on tourists’ engagement, trust and customer–company identification (CCI).

Customer engagement (CE) has been recognized as a key construct in analyzing customer responses (Kumar et al., 2010). Indeed, CE is a research priority of Marketing Science (MSI) in 2018-2020 (MSI, 2018) and CSR communication has been highlighted as a potential way to engage consumers (Du et al., 2010). This study will not analyze CSR content in social media in isolation but will include CE as a key multidimensional response. Therefore, our research integrates four constructs: CSR customer perceptions, customer trust in the hotel, CCI and CE, in a causal model of customer responses to two types of social media communications, FGC and UGC. This study focuses on message source as a determinant peripheral cue (Filieri et al., 2018) of CSR communication through Twitter, that is, FGC (Kumar et al., 2016) when made by the hotel, and UGC (Ukpabi and Karjaluoto, 2018) when initiated by customers.

The present study contributes to the literature on CSR communications about hotels in three main aspects. First, we compare two mediating factors in the relationship between CSR customer perceptions and CE: trust and CCI. Second, we show the value of the CE construct, based on its behavioral perspective, to expand knowledge of actual behaviors that have repercussions for loyalty toward firms, that is, recommendation and co-creation as a result of CSR customer perceptions. Third, we shed light on the moderating effects of two different social media message sources, FGC and UGC. Recent research (Colicev et al., 2018) posits that there is insufficient evidence as to how FGC and UGC impact on the sequential stages of the marketing funnel. We complement other studies that examine the impact of differences in tweet format, characteristics such as the valence of hotel reviews (Mauri and Minazzi, 2013), number of retweets (Bokunewicz and Shulman, 2017), type of content (Bigné et al., 2019) and graphical content (Yang et al., 2017) on different marketing outcomes.

The paper is structured as follows: Section 2 includes the literature review and the proposed hypotheses. In Sections 3 and 4, the methodology and results of the studies are presented, respectively: a single-factor experimental design with 227 hotel guests, and three structured, in-depth interviews, assess the effect of CSR customer perceptions, CCI and trust on CE. In addition, we test the moderating effect of the CSR message source on the proposed relationships. Finally, in Section 5 we discuss the theoretical contributions, in Section 6 we comment on practical implications, and on Section 7 we present potential future research lines.

2. Conceptual framework

2.1 Customer engagement

CE in this study is considered from a behavioral perspective, that is, as the voluntary contribution made by customers to the company, aside from the basic commercial transaction, which is motivated by the company’s actions (Alexander and Jaakkola, 2016). According to Harmeling et al. (2017), the majority of the existing literature confirms that CE is a behavioral response to a firm or brand rather than a psychological state of the customer. The different definitions of CE from the behavioral perspective are presented in Table I.

Our CE conceptualization is adapted from Kumar et al. (2010), who introduced four dimensions characterized by value creation by the customers for the firm, motivated by intrinsic and external factors. We use only three of the four dimensions, customer lifetime value (CLV), customer influencer value (CIV) and customer knowledge value (CKV). We consider only actions and behaviors that do not involve monetary incentives made by the hotel. Previous studies demonstrate that recommendations to members with strong ties of belonging to their social networks are unaffected by monetary incentives, rendering impotent the influence of incentivized referrals (Ryu and Feick, 2007). In addition, in the hotel industry, monetary incentives have been found to have no positive significant relationship with eWOM behavior toward hotels (Hu and Kim, 2018). Based on these arguments, we removed the customer referral value dimension.

CLV (Gupta et al., 2004) is the current value of all the expected financial benefits that customers will provide to the firm throughout their lives. In our study context, engaged customers have enjoyed the service experience, consider it as valuable and, thus, are willing to purchase the hotel’s services in the future. CIV (Kumar, 2013) includes information sharing, interactivity, WOM and assistance to other clients. This dimension reflects how engaged customers can persuade others to use the hotel, or help it to retain current clients. CKV (Kumar and Bhagwat, 2010) is the customer’s participation in the creation of new products and processes, and helps to achieve a fuller understanding of customer preferences. In keeping with the principles of service-dominant logic and value co-creation, if customers are engaged, they assist in creating customized experiences, working hand in hand with the company (Vargo and Lusch, 2017; Hernández-Ortega and Franco, 2019).

2.2 Corporate social responsibility customer perceptions and customer engagement

CSR customer perceptions refer to the status, in the customer’s eyes, of the company’s social responsibility (Alvarado-Herrera et al., 2017). Previous studies acknowledge that customers positively value CSR initiatives and attribute them to the social commitment of the company (Currás et al., 2009). CSR customer perceptions are used in this research to verify if messages sent through Twitter elicit positive views of corporate social and environmental activities, taking into account that fulfilling customer expectations about social concerns is particularly important for identification, loyalty and willingness to recommend a firm (Pérez and Rodríguez del Bosque, 2013). CLV in the hotel industry will be enhanced if the customer intends to stay at the establishment in the future, based on his/her perceptions of the hotel; and CSR communications contribute to building a positive image where they inform the consumer about actions undertaken by the hotel that support society and the environment (Martínez and Rodríguez del Bosque, 2013).

CIV is also influenced by CSR activities (Gao et al., 2016). A hotel’s social and environmental performance causes customers to post positive messages and recommend the hotel, since they feel an obligation to respond equably to its socially responsible attitude (Hu and Kim, 2018). Finally, based on service-dominant logic and social identity theory, customers that perceive a hotel to behave beyond its strict moral and legal requirements respond by undertaking pro-organizational behaviors, such as knowledge sharing and value co-creation with the organization (Luu, 2017). A significant proportion of the intention to provide feedback (i.e. CKV) about the hotel’s CSR practices is based on the beneficial results that this collaboration triggers in other stakeholders (Iglesias et al., 2018). Based on these points, CSR customer perceptions arguably positively affect CE behaviors:

H1.

CSR customer perceptions have a positive effect on CE

2.3 Corporate social responsibility customer perceptions and trust

In online environments, trust is defined as one party’s expectations about the other party’s motives and behaviors (Jarvenpaa et al., 2000). Previous research discussed three dimensions (Flavián and Guinaliu, 2006; Sanz-Blas et al., 2014):

1. honesty (the belief that the company will fulfill its promises and obligations);

2. benevolence (the belief that the company is interested in achieving joint benefits); and

3. competence (the level of training that each party must have to perform the relevant tasks).

CSR activities have two different impacts on customer trust: direct, when they immediately benefit the customer with improved goods and services, and indirect, as the impact on society of these activities also enhances the welfare of each individual customer (Fatma et al., 2018). Trust is also generated, and customer skepticism dispelled, by CSR communication through social media, if companies establish an open discourse with their stakeholders, absent of self-interested motives and unmet expectations, in which the provision of honest, understandable and appropriate information is the norm (Scherer and Palazzo, 2007; Seele and Lock, 2015). Where a hotel successfully follows social norms this legitimates its activities and makes it trustworthy in the customers’ eyes (Kim, 2017), which leads to increased positive electronic word-of-mouth (Du et al., 2010).

The positive impact of CSR communications and trust has been confirmed in tourist services, including restaurants (Swimberghe and Wooldridge, 2014) and accommodation (Jalilvand et al., 2017). CSR communication by hotels greatly improves their brand image and prestige, since the consumers’ rights are seen as being protected and their expectations fulfilled. Therefore, trust is developed toward the hotel and its products and services (Jalilvand et al., 2017). Moreover, as more customers come to realize that the hotel implements socially responsible policies, trust increases quicker, due to the trust transfer process in virtual communities (Liu et al., 2018). Based on this argument, we propose that there is a positive connection between CSR customer perceptions gained from social media and trust in the hotel:

H2.

CSR customer perceptions have a positive effect on customer trust in the hotel.

2.4 Trust and customer engagement

Trust has been found to have significant impact on customers’ propensity to continue to patronize a firm and spread positive word of mouth (Harris and Goode, 2004; Morgan and Hunt, 1994; Chumpitaz and Paparoidamis, 2005; Hazra and Srivastava, 2009; Ruiz-Mafé et al., 2014; Sanz-Blas et al., 2014). Customer trust in the company is included in our research model as it is an important factor in measuring CE with hotels that communicate CSR initiatives through social media. Responses made to users’ contributions are rewarded by a greater level of trust for messages published on the medium and greater intention to carry out positive eWOM (Eberle et al., 2013). In virtual communities, several behaviors are observed as consequences of trust in the sponsor or firm that manages the community:

• the sharing of personal information with the firm;

• working together on new product development, providing creative and innovative ideas; and

• becoming loyal to the firm (Porter and Donthu, 2008).

According to commitment-trust theory (Morgan and Hunt, 1994), trust is the basis of the creation and maintenance of positive, long-term relationships with customers. Trust is mandatory for the customer and the company to be engaged, interested in the other party’s situation and for an active and productive interaction (Pansari and Kumar, 2017). Trust, therefore, has an effect on CE due to the positive emotions evoked in the customer with respect to the brand, and the result of repeated positive experiences only strengthens this mutual interaction (Martínez-López et al., 2017). We propose that customer trust positively affects CE:

H3.

Customer trust in the hotel has a positive effect on CE.

2.5 Corporate social responsibility customer perceptions and customer-company identification

CCI is rooted in the definition of organizational identification provided by Mael and Ashforth (1992, p. 109); it is “the (cognitive) perception of oneness with or belongingness to an organization where the individual defines him or herself, at least partly, in terms of their organizational membership”. This identification fulfills the customer’s self-definitional needs (Huang et al., 2017). According to social identity theory, social connections with other groups and organizations that share similar, or at least compatible, principles and values play an important role in the process (Fatma et al., 2018). Service quality, reference groups and boundary-spanning agents are facilitating factors for the customer to identify with the hotel (Ahearne et al., 2005; Deng and Xu, 2017).

The action-based theory of persuasion perceived corporate values play a role in customer identification and perceptions of attractiveness during the consumer’s accommodation decision-making process (Funkhouser and Parker, 1999). In this sense, CSR is essential in defining corporate principles and initiating the identification process (Lichtenstein et al., 2004; Marín et al., 2009; Pérez and Rodríguez del Bosque, 2012). Customers relate to socially responsible companies as this helps satisfy their self-enhancement and self-esteem needs (Martínez and Rodríguez del Bosque, 2013). In particular, CSR actions promote an identity associated with responsiveness to society, awareness and appreciation of resources (Marín et al., 2009), and a willingness to undertake pro-social behaviors that take account of civic interests (Pérez et al., 2013).

Positive emotions arising from CSR customer perceptions are facilitators of CCI, whereas negative emotions related to perceived unfair corporate behaviors do not have the opposite effect and are not determinant in the identification process (Su et al., 2017). Nonetheless, hotels need to communicate their CSR initiatives to customers to ensure that they know they share common values; otherwise, the identification is unlikely to be as fast or lasting (Fatma et al., 2018). Therefore, we propose a positive relationship between CSR customer perceptions and CCI:

H4.

CSR customer perceptions have a positive effect on CCI.

2.6 Customer–company identification and customer engagement

Following social exchange theory, customers feel the need to help the company if they perceive it to perform well in economic, social and environmental terms (Jalilvand et al., 2017). As for CLV, when the customer identifies with the company, he/she tends to value the quality of its service and places greater trust in the hotel brand. This positive brand evaluation leads to a greater intention to stay loyal to the hotel in the future (So et al., 2013). Identification with the hotel is based on a strong, mutually beneficial relationship between the hotel and the customer. CCI dissuades the customer from switching to other hotel brands (Casidy et al., 2018).

In terms of CIV, the customer’s self-identification with the hotel increases WOM intention and behavior, since the higher the identification, the greater will be the likelihood that the customer will praise the company and recommend it to friends, relatives and acquaintances (Yang and Mattila, 2017).

The previous literature on CKV acknowledges that, once the customer has gone through a process of social identification with the hotel, co-creation is more likely and valuable for both parties (Luu, 2017). To summarize, company identification is aligned positively with CE, from disseminating positive eWOM through social media (Brown et al., 2005), to providing input to the firm as part of an active co-creation process (Bhattacharya and Sen, 2003; Romero, 2018). Based on these arguments, we posit that CCI exerts a positive effect on all the CE dimensions:

H5.

CCI has a positive effect on CE

2.7 Trust and customer–company identification

Previous research acknowledges that, for CCI to be created and developed over time, it is imperative that customers know about the company’s activities and principles and that they believe the company to be trustworthy (Palacios-Florencio et al., 2018). To satisfy their self-esteem and self-enhancement needs, customers identify with honest, competent and benevolent companies, since this is the profile that they wish to display to society (Keh and Xie, 2009). The positive relationship between trust and CCI has been examined and supported in previous literature related to the hotel industry (García de Leaniz and Rodríguez del Bosque, 2015) and online environments. All these ideas support the notion that trust is positively related to CCI:

H6.

Customer trust in a hotel has a positive effect on CCI.

2.8 Mediating effects

In addition to the direct relationships already proposed, we argue that customer trust and CCI exert a mediating effect on CSR customer perceptions and CE. We argue that trust mediates the effect of CSR customer perceptions on CE. CSR initiatives and the subsequent customer perceptions evoked create trusting and durable company-customer relationships (Park et al., 2017). Furthermore, Hollebeek (2011) points out that, in the case of existing customers, this enduring relationship of trust is an antecedent of CE. An effective CSR communication strategy, again in the hotel sector, has been found to strengthen perceptions of trust in companies (Martínez and Rodríguez del Bosque, 2013): customers are more likely to make repeat purchases from companies that behave in a socially responsible manner. In this sense, trust developed from CSR activities has been explored and confirmed as a mediating variable for the development of CLV in the hotel sector (Palacios-Florencio et al., 2018). Committed customers, who already trust the company, are usually keen on generating eWOM (i.e. CIV) (Morgan and Hunt, 1994; Serra-Cantallops and Salvi, 2014). Finally, CKV and the “win-win” approach on which it is based, is dependent on the hotel developing a trust relationship with its customers, which encourages both parties to maximize the mutual benefit (Chathoth et al., 2013). Consequently, we argue that trust acts as a mediator in the positive effect of CSR customer perceptions on CE:

H7.

Customer trust in the hotel mediates the positive relationship between CSR customer perceptions and CE.

A solid conceptual proposal is to consider the CCI generated by the company’s social responsibility messages as a key mediator variable in individual behavioral responses to CSR communications (Marín et al., 2009; Bigné et al., 2010). An analysis of CCI can provide more effective explanations for the influence of CSR on CE, because individuals can satisfy self-definitional needs through their relationship with a socially responsible hotel (Martínez and Rodríguez Del Bosque, 2013). The previous literature confirms that, if customers are particularly aware of an environmental or social cause, they positively value CSR actions taken by the hotel related to that issue and will consequently identify themselves with the hotel (García de Leaniz and Rodríguez del Bosque, 2015). This applies even if the focus is on societal issues that go beyond the customer’s own interests and that are only indirectly related to him/her (Pérez and Rodríguez del Bosque, 2015). In turn, this CCI is related positively to all the CE dimensions. Regarding CLV, affinity with a hotel’s values and beliefs shapes the customer’s social identity and, with a pleasant brand experience, results in higher revisit intention (Brakus et al., 2009; So et al., 2017). Socially responsible behavior contributes to the generation of a positive moral image for the company, which facilitates identification and leads to loyalty (Fatma et al., 2018). In terms of CIV, a high level of identification also results in positive eWOM recommendations, to reinforce one’s own sense of belonging and need for self-enhancement (Chu et al., 2018). As for CKV, CCI in social media environments encourages active participation and knowledge sharing with no expectation of specific reward through either assisting other members, taking part in activities or displaying positive behaviors toward the company (Hammedi et al., 2015). Hence, we also posit that CCI mediates the relationship between CSR customer perceptions and CE:

H8.

CCI mediates the positive relationship between CSR customer perceptions and CE.

Lee and Yoon (2018) show that trust and willingness to endorse a particular company is highly driven by its CSR profile. Customers who perceive a hotel to be truly involved in CSR have an enhanced trust relationship, have reduced skepticism of corporate hypocrisy, more strongly identify with the company and are more tolerant of corporate misbehavior (Kim et al., 2015). Furthermore, it is demonstrated that customers who firmly identify with a hotel’s corporate values develop stronger and more positive brand relationships with the company (Rather, 2018). As a consequence of CCI, customers choose to have an active relationship with the company and engage in behaviors that benefit the company (Bhattacharya and Sen, 2003), including

• positive evaluations of the company and its products;

• choosing the company over alternatives;

• positive eWOM; and

• providing product improvement suggestions (Ahearne et al., 2005).

It is argued that trust creates customer identification with the brand, which, in turn, explains CE from a behavioral perspective:

H9.

Customer trust in the hotel and CCI jointly mediate the positive relationship between CSR customer perceptions and CE.

2.9 Message source

In accordance with the integrated marketing communications paradigm, social media have become very important in terms of interaction and ubiquitous communication (Porcu et al., 2017). In contrast to other traditional sources of corporate ethics information, social media help hotels to be aware of CSR customer perceptions (Zeng and Gerritsen, 2014); and, at the same time, they influence customer behaviors toward firms, even if the CSR initiatives are not directly aimed at these particular stakeholders (Wang and Huang, 2018).

Social information processing theory and the MAIN model (van Zoonen and van der Meer, 2015) suggest that the source of a message impacts on customer behavioral intentions. In this study, we differentiate between two CSR message sources, those posted by the hotel (i.e. FGC) and those posted by the customer (i.e. UGC). FGC messages posted on social media usually come from organizational accounts, considered as “official sources”. The possibility of frequent interactions between users and official sources fosters CSR customer perceptions and purchase intention (Wang and Huang, 2018). In addition, hotels report CSR activities for two reasons: they seek

• to increase the awareness of customers and other stakeholders of their CSR activities to enhance their corporate credibility; and

• legitimacy to improve their competitive positioning and to comply with social, political and cultural norms (Du and Vieira, 2012).

In social media, the gatekeeping process or filtering of relevant information is directly transferred to the customer (Westerman et al., 2014), who determines source credibility in an individual way (van Zoonen and van der Meer, 2015).

Customers (UGC) speak about their hotel involvement in CSR, which then exerts a significant effect on purchase intentions (Sparks et al., 2013). Ettinger et al. (2018), in their study of social media communications made by hotels, note that almost half (42 per cent) of the reviews mentioned one or more CSR-related issues, especially environmental activities, diversity topics and supplier relations, showing the importance placed by customers on these aspects of a hotel’s activity. Nevertheless, the importance of the interrelation between CSR and UGC remains an open and largely unexplored area in social media and eWOM literature (Serra-Cantallops et al., 2018); the present study explores the moderating role of message source on the proposed hypotheses.

2.10 Moderating effects: user-generated content vs firm-generated content

One of the key elements that defines the importance of message source is source credibility (Petty and Cacioppo, 1986): if the message source is credible, the content of the message will be considered objective and reliable (Goldsmith et al., 2000). In this sense, messages generated by users may diminish the skepticism that arises when CSR actions are closely related to the hotel’s main business activities (van Rekom et al., 2014). In these cases, the customer may attribute intrinsic or selfish motives to the actions (Du et al., 2010), or regard differences between what the hotel says and does as a signal of incoherence in the hotel’s social and environmental discourse, which will be punished with lower perceived trustworthiness (Christensen et al., 2013).

When customers are the source of CSR information, they are perceived as proximal to other customers and, thus, as more reliable (Ukpabi and Karjaluoto, 2018); the information recipient perceives that these clients have extrinsic or disinterested reasons to highlight the hotel’s CSR initiatives and the message is, therefore, more persuasive (Kim and Kim, 2014). In fact, eWOM puts the customer into the role of an honest CSR advocate for the hotel (Du et al., 2010), and UGC is highly valued for its integrity (Dickinger, 2011). On the other hand, social media posts from organizational sources are thoroughly scrutinized, as readers attribute to them commercial purposes or ulterior motives, and they are thus considered less trustworthy (van Zoonen and van der Meer, 2015). Consequently, it is argued that UGC strengthens the relationship between CSR customer perceptions and customer trust:

H10.

UGC messages about CSR strengthen the effect of CSR customer perceptions on customer trust in the hotel more than FGC messages.

CCI is also dependent on message source. The aim of CSR communications, which are basically related to reducing skepticism and generating positive attributions of corporate motives for carrying out the initiatives, are easier to achieve when the CSR message source is not directly related to the hotel (Wang and Huang, 2018). Personal posts are seen by other users as having values more aligned with their own, and their intentions are perceived to be unrelated to profit seeking; in the case of an organizational social networks, the general image is the contrary, which makes the trust and identification processes more difficult (Wang and Huang, 2018). The fact that the message source is a hotel customer makes interaction more frequent and facilitates the development of stronger company-customer ties. In such cases, the message sender acts as a strong reference element and identification is facilitated (Phua et al., 2017). Thus, we propose that CSR communication generated by customers reinforces the impact of CSR messages and customer trust on CCI more than FGC messages:

H11.

UGC messages about CSR strengthen the effect of CSR customer perceptions on CCI more than FGC messages.

H12.

UGC messages about CSR strengthen the effect of customer trust in the hotel on CCI more than FGC messages.

Message source is also said to affect the impact of CSR communications on CE. UGC is proposed in the previous literature as increasingly more credible than organization-sourced messages (van Noort and Willemsen, 2012). In relation to CLV, source credibility impacts on eWOM communication adoption, which has an impact on customer loyalty (Aghakhani et al., 2018). Through a peripheral route, message source credibility is shown to have a significant effect on the customers’ information sharing intentions, because they are interested in building and maintaining relationships with the company (Hur et al., 2017), and thus CKV is also promoted. Based on these arguments, the effects of CSR customer perceptions, customer trust and CCI on CE can be positively moderated by UGC more than by FGC:

H13.

UGC messages about CSR strengthen the effect of CSR customer perceptions on CE more than FGC messages.

H14.

UGC messages about CSR strengthen the effect of customer trust on CE more than FGC messages.

H15.

UGC messages about CSR strengthen the effect of CCI on CE more than FGC messages.

We propose a causal model, based on these hypotheses, moderated by the source of the social media communication. The proposed conceptual model is at Figure 1.

3. Methodology

This study adopts a mixed research method approach (Molina-Azorín, 2016). Mixed methods offer a better understanding of research problems by triangulating two sets of results and thus enhancing the validity of inferences. Accordingly, a two-step study was implemented. First, an experimental design among tourists was carried out. Second, a qualitative analysis using in-depth interviews was conducted with three hotel managers.

3.1 Study 1: single-factor experimental study

3.1.1 Design of the experiment and stimuli.

The purpose of the experiment is to examine how the message source, specifically UGC, might influence CSR customer perceptions and the direct and indirect relationships with CE in a more positive way than communications from the hotel. The experiment was a single-factor (tweet source: FGC vs. UGC), between-subjects design, resulting in two scenarios, each with four tweets, meeting the following conditions:

• tweets about CSR initiatives were made by the hotel (i.e. FGC); and

• tweets about CSR initiatives were made by an independent third party, in this case former clients of the hotel or hotel chain (i.e. UGC).

Respondents were explicitly asked that they should think of the information in the tweets as if they had been posted by the actual hotel they stayed at during the previous month.

The stimuli (i.e. the tweets) were based on actual examples from hotel corporate profiles, taking into account that many customers may be entirely unaware of their chosen hotel’s CSR policies (Serra-Cantallops et al., 2018). Several design conditions were established. The tweets must actually have been published by a hotel chain on its official Twitter profile, or another profile specifically devoted to social activities and initiatives of the firm or group. The tweets were then made generic (i.e. not citing specific organizations, locations or brands) so that they could be applicable to any respondent. In the FGC scenario, we used the name of a fictitious hotel chain (@CadenaDeHoteles) for both the name of the Twitter profile and references to it in the tweets. We included a fictitious, generic hotel chain so that the experiment could be applicable to customers of different hotels, thus avoiding a bias effect from the hotel. Although the hotel chain is fictitious, the material presented to the participants was based on online posts made by real hotels. We asked the participants to accept that these posts were made by the hotels in which they had stayed during the month prior to the experiment. This allowed us to target respondents who had stayed in the specific tourist destination during the previous month, albeit in different hotels. The modifications of the scenarios involved, in the UGC case, changing the profile picture and name to create an anonymous customer account; hence, the customer accounts were fictitious but based on real accounts. To imitate the clients’ ways of expressing themselves, slight modifications were made to the UGC scenario, by adding referrals to the hotel, additional opinions, exclamations and emojis. Apart from this, the content of the message in both scenarios was the same as in the tweets published by the real hotel companies and the focus of the differences between the scenarios remained on the message source. To ensure that the stimuli and questions were appropriate and understandable, we carried out a pre-test with 27 respondents, 15 for the FGC scenario and 12 for the UGC.

To address all the relevant CSR dimensions, we included all references to the economic, social and environmental actions taken by the hotel. Moreover, to avoid cultural biases, and reach the highest possible number of respondents, we created customer profiles of different national origins. Finally, to ensure stimuli homogeneity, three formal conditions were established:

1. only text was used, with no attached contents or hyperlinks;

2. message length was three lines; and

3. the colors used in the tweets were Twitter’s official colors.

These conditions were introduced and accepted during the pre-test stage. The tweets used in each scenario are shown in Figure 2.

3.1.2 Data collection and sample.

The population used for the hypotheses testing was hotel customers who stayed in a hotel, with a rating of three stars or greater, during the previous month. The questionnaire was distributed during June 2018, to hotel clients in Valencia, a city that ranks in the top five urban tourism destinations in Spain; this allowed us to minimize the bias related to data collected in different destinations. Of the 235 hotel clients recruited, 227 completed the experiment (n = 227). The profile of the sample is given in Table II. Of the 227 questionnaires, 159 were distributed online and 68 in paper format.

The online and paper questionnaires were exactly the same. To verify that there was no method bias, a one-way ANOVA was performed between both questionnaire formats (Table III). This yielded no statistically significant differences between variables in any of the cases.

Of the 227 hotel clients, 74 per cent stayed at the hotel for leisure and 26 per cent for professional reasons; 52 per cent had a Twitter profile; 26 per cent connected to it more than once a day, 14 per cent once a day, 20 per cent 2 or 3 times a week, 10 per cent once a week and 31 per cent signed in less frequently. As to how they made their hotel bookings, 15 per cent used the hotel website, 57 per cent used a travel website (e.g. Booking.com and Trivago), 7 per cent made a phone call, 7 per cent used a travel agency and 14 per cent had a third party book the hotel for them. From the total sample, 59 per cent stayed at a hotel of a chain with an active CSR communication policy and 41 per cent at a hotel chain or an independent hotel with no CSR communication policy.

3.1.3 Measures.

All the scales were adapted from previous hotel and social media literature. The CSR customer perception data were obtained using the scale used by Bigné et al. (2010). For CCI, we used Homburg et al. (2009)’s scale adapted to hotels. To measure customer trust we used Ruiz-Mafé et al. (2014)’s 3-item adaptation of the scale validated by Flavián et al. (2006), covering the dimensions of honesty, benevolence and competence. For CE in its behavioral conceptualization, we used the measurement tool of Kumar and Pansari (2016), taking its CLV, CIV and CKV dimensions. The CLV items relate to the customer’s previous experience with the hotel and his/her future intentions based on this experience. The CIV and CKV items were adapted and recast in the conditional tense to elicit the behavioral intentions of customers following hotel stays and after having read the FGC and UGC tweets about the hotel’s CSR activities. The complete list of items included in the measurement of the variables can be seen at Table IV.

Partial least squares structural equation modeling (PLS-SEM), based on variance, with SmartPLS 3.2.7 software, was used to test the hypotheses; this is a good fit for our study (Hair et al., 2011). PLS-SEM involves a two-step process:

1. the measurement of the model; and

2. the assessment of the structural model.

3.1.4 Measurement model.

To ensure measurement validity and reliability we undertook a confirmatory factor analysis (CFA) for all constructs and their indicators. The CFA was measured by SmartPLS 3.2.7, given that an endogenous factor, namely CE, was conceptualized as a second-order formative construct. The consistent PLS algorithm was used to correct the estimations of the reflective constructs (Dijkstra and Henseler, 2015).

The results for the validity of the measurement model are very good. Content validity is assumed as all the proposed relationships between the theoretical concepts are based on a review of the literature. The results of the analyses of convergent validity and reliability are shown in Table V. Standardized loadings from indicators are all above 0.6, with average standardized loadings above 0.7 in all cases. Average variance extracted (AVE) for all factors is above the minimum level of 0.5 recommended by Fornell and Larcker (1981). The reliability results show that the measurement model is consistent, even if used repeatedly, with Cronbach’s alpha for all variables well above the minimum acceptable value of 0.7 recommended by Nunnally and Bernstein (1994). Composite reliability values are also greater than 0.6 for all latent variables (Bagozzi and Yi, 1988).

The discriminant validity results, displayed in Table VI, are also satisfactory. The HTMT ratio levels are below 0.9 in all cases (Teo et al., 2008). In addition, cross-loadings show that none of the indicators have greater loads in different factors. After applying bootstrapping with 5000 resamples, all factor loadings are found to be significant at a p < 0.001 level. As CE is included as a second-order formative construct, the model was estimated in a two-stage approach. This approach avoids collinearity problems and includes the higher-order construct as the dependent variable of the model (Ringle et al., 2012; van Riel et al., 2017). The results of the validity test show no collinearity problems (VIF <5) in the different weights. All standardized weights are found to be significant, except for CIV. Given that the intention to recommend the hotel online is an important indicator of CE, and that its corresponding loading was above 0.5 (std. loading = 0.696, p < 0.001), this dimension was kept in the model.

3.2 Study 2: in-depth interviews with practitioners in the hotel industry

A second, qualitative study was performed to give further validity to the results of Study 1. Structured in-depth interviews were undertaken with three hotel industry experts working in different sub-sectors. In particular, one was an entrepreneur and hotel manager with experience in tourism innovation, one was a manager of a small hotel in an urban area and one was the Valencian community manager of an international hotel chain. Six questions were posed in the interview, covering most of the relations between the variables in Study 1. The questions were as follows:

Q1.

What is the effect of CSR initiatives on trust felt toward the hotel?

Q2.

To what extent can the client identify with the hotel if (s)he perceives it to be socially responsible?

Q3.

What role do social media play in CSR communication? Do they contribute to generating trust and/or identification toward the hotel?

Q4.

Do you think that trust between the customer and the hotel makes the identification process with the hotel easier?

Q5.

With respect to the effects of CSR communication on trust, do you think the effects are greater when the message is sent by the hotel itself or by an independent customer who previously stayed at the hotel?

Q6.

How do trust and identification with the hotel affect CE?

The interviews were performed via skype and lasted on average between 15 and 20 min. The sessions were recorded with the prior consent of the interviewees, which allowed the contents to be transcribed. We assured the interviewees that personal data and anonymity and confidentiality would be protected in all cases. Quotes in this study are directly reproduced from the interviewees’ responses.

4. Discussion of results

4.1 Structural equations model

Having confirmed the validity and reliability of the measurement instrument, the structural model was calculated for the general model. The results of the structural model evaluation and hypotheses testing are displayed in Table VII.

The structural model evaluation shows adjusted R² values for the factors in the general and scenario models significantly above the minimum recommended level of 0.1 (Falk and Miller (1992)). The Q² statistics for all dependent latent variables are positive, confirming that the proposed model has predictive validity with respect to all its variables.

The relationship between CSR customer perceptions and CE was found to be non-significant (H1, β = 0.017, p > 0.1), therefore it cannot be stated that customers who perceive that a hotel undertakes CSR activities directly display positive engagement behaviors. Nevertheless, the effect of CSR customer perceptions on trust (H2, β = 0.665, p < 0.001) and of trust on CE (H3, β = 0.162, p < 0.05) are positive and significant, meaning that customers who perceive that the hotel undertakes social and environmental activities tend to develop feelings of trust, and that those who believe the hotel to be honest, competent and benevolent are likely to display positive CE. Similarly, the positive relationship between CSR customer perceptions and CCI (H4, β = 0.330, p < 0.001) and CCI and CE (H5, β = 0.710, p < 0.001) are confirmed: CSR customer perceptions are an important antecedent of the CCI process, which, in turn, encourages CE behaviors beneficial to the hotel. Customer trust positively influences CCI (H6, β = 0.551, p < 0.001).

4.2 Mediating effects

Table VI also shows indirect effects. The indirect effect of CSR customer perceptions on CE through trust is positive, but less powerful than via CCI. Nonetheless, this difference is found to be non-significant (Table VIII). In addition, there is a joint effect of trust and CCI, which involves a multiple mediation of these two factors between CSR customer perceptions and CE. The total indirect effect has a variance accounted for (VAF) above 90 per cent, providing further evidence of the positive mediating effect (Cepeda et al., 2017). Given that both indirect effects are significant and that the direct effect (i.e. H1) is not, both trust and CCI fully mediate the CSR customer perception-CCI relationship. The serial mediation also shows that one of the mediators (i.e. trust) has an effect on the other (i.e. CCI).

4.3 Moderation

To verify that the message source, FGC or UGC, moderates the proposed relationships, we performed a multigroup analysis, using SmartPLS 3.2.7, and used two different procedures, Henseler et al. (2009)’s MGA non-parametric significance test and permutation test. The first compares group bootstrap estimates obtained from all bootstrap samples: this shows significant differences, at the 5 per cent level, between path coefficients across the two scenarios, if the associated p-value is lower than 0.05 or higher than 0.95 (Sarstedt et al., 2011). The permutation test assesses if two data groups, in this case the FGC and UGC respondents, show statistically significant differences in their group-specific path coefficients (Hair et al., 2018). The results of this analysis are at Table IX.

Henseler’s MGA identified a significant difference between UGC and FGC regarding the effect of CSR customer perceptions on trust. As shown in Table VIII, the UGC results demonstrate a greater correlation between customers’ perceptions of the hotel’s responsiveness to society and their development of trust than in the case of FGC. Hence, H10 (p-value difference = 0.953, p > 0.95) is supported.

4.4 In-depth interviews

With respect to Q1, practitioners in the hotel industry agree that the effects of CSR on trust are yet to be completely clarified. The hotel manager who was expert in tourism innovation declared that:

“CSR initiatives undoubtedly have an impact on hotel trust, but lower than service, and depend on the segment and type of product”.

This highlights that the impact of social and environmental activities is stronger in higher-quality hotels, where the service is generally satisfactory, and where customers truly appreciate CSR contributions. Another issue raised by the interviewees was the importance of communication for CSR to have a definitive effect on trust. The international hotel chain manager confirmed that, as part of the phenomenon of the “economy of reputation”, both hotels and customers realize the importance of communicating these actions. At the same, it was highlighted that hotel managers devote little effort to this goal. In their own words:

“There is a significant gap between the orthodox definition of CSR and what companies do”.

The expert in tourism innovation and the urban hotel manager agreed that, in practice and in most cases, CSR communication is reduced to little bits of information at the “foot of the corporate webpage” and that its level of influence is low or even non-existent.

The interviewees also gave their opinions in answer to Q2 about the relationship between CSR customer perceptions and consequent CCI. The urban hotel manager stated that “CSR is a natural extension of what the company does beyond its main activity”, meaning that CSR communication has to be sufficiently effective to create identification “before the customer has formed an opinion about the company”.

For the expert in tourism innovation, CSR can have a significant effect in hotels in which “the impact of price is much lower”, and the hotel product is “more experience-based and environment-oriented”. The strong effect on CCI is explained, according to the international hotel chain manager, by the fact that:

“hotels are every time more sensitized to CSR”.

Therefore; “CSR should not be seen just a commercial matter, but as a part of the philosophy of the company”.

In the case of the international hotel chain, the company is more sensitized toward CSR and makes visible actions that are already part of the core values that define the company’s activity.

Speaking of communication of CSR initiatives (Q3) the experts agreed that social media are crucial, and often underused tools. According to the expert in tourism innovation, “generating CSR content on social networks is a challenge for hotels”, and CSR-related inputs can generate enormous customer value even for hotels less sensitive to CSR perceptions, that is, whose market segment is not particularly interested in these initiatives. The urban hotel manager regarded social media as useful, very dynamic channels that allow content generation without using intermediaries, keeping in mind that “the user has to enter into contact with the community in advance”, so that CSR communications can actually have an impact.

In terms of Q4, the responses of all three experts were unanimous: the effect of trust on CCI is positive and obvious when it comes to customers in the hotel industry, but only if it is accompanied with a positive experience at the hotel. The international hotel chain manager highlights how it contributes to the “social prestige” of the hotel.

Q5 covered the importance of the message source on the effects of CSR on trust and CCI. Here opinions differed. The expert in tourism innovation and the manager of the international hotel chain believed there is a clear tendency to regard UGC as being the more legitimate source of information:

“The impact of reviews comes mainly because the sender is different to the entity discussed in the message […] customers give much more legitimacy to external messages”.

On the other hand, the hotel manager of the urban destination argued that both sources could be considered as trustworthy as regards CSR. Specifically:

“The opinions of the clients, as long as they are not paid for and are spontaneous, are commonly regarded as more credible than FGC […]. In the case of CSR, the hotel has more scope to communicate its activities than its clients, and this can rebalance this reach and impact. The hotel is able to communicate more, and more nuanced, information than the customer, who may focus only on a small part of what (s)he has perceived”.

This insight helps to explain the lack of significance of the moderating effect of the message source in some of the proposed relationships. Although hotel guests can, and do, act as CSR message sources, and have been in the same situation as the potential customer, and hence more neutral, the information provided by them is usually less complete and can be perceived by potential customers as providing only a limited picture that does not represent the whole range of the hotel’s social and environmental activities.

Finally, in Q6 the experts were asked about the effect of customer trust and CCI on CE from a behavioral perspective. The expert in tourism innovation argued that CLV in this sense is not necessarily a consequence of trust and CCI in the case of the hotel industry, as repeat visits depend heavily on destination and other factors:

“If we depart from a more traditional concept of engagement, understood as simple repetition, it strongly depends on the segment and the destination. I may love a hotel in Japan, but I will probably travel there only once again in my life”.

This view is the contrary to that of the international hotel chain manager, who feels that repetitive hotel stays are motivated by trust and identification with the company.

Nonetheless, if we extend the concept of engagement to word of mouth or active collaboration with the hotel, then the result is positive in the view of all experts. In fact, the urban hotel manager agreed with the definition of engagement that takes into consideration the recommendation and co-creation dimensions. The urban hotel manager found that trust and CCI had the strongest effect on CIV, as:

“When I see that a hotel goes beyond with its core activities, is aware of the impact of its activities on the environment, and develops activities in which even the customer can take part, I also talk about the hotel positively”.

5. Theoretical implications

This study analyzes the role of the message source when using Twitter to report CSR activities, on the basis that UGC strengthens the effect on CE. Our findings suggest that CSR customer perceptions gained through Twitter affect CE when customers develop trust in the hotel due to its environmental and societal-supportive activities and when they identify with its corporate values and principles. Furthermore, CSR perceptions develop stronger trust when the message comes from a customer than a corporate source. This study makes several contributions. The findings contribute to the growing body of knowledge on consumer engagement in social media platforms and provide theoretical and managerial implications that may help both researchers and hotel managers understand the mechanisms underlying consumer–brand engagement with tourism services in the social media context.

The literature has recognized the emerging importance of CE as a strategic imperative for building customer–brand relationships. In testing the role of CSR perceptions, CCI and trust in developing tourism services engagement, this study has addressed calls for more research into CE to further the understanding of this emerging construct (Kumar et al., 2010; MSI, 2018). The findings provide insight into the relationships between the focal constructs that underlie the process of service engagement formation and afford a greater appreciation of how CSR perceptions interact with the key drivers of CE in creating truly committed and loyal customers.

The positive effect of hotel trust and CCI on CE can be understood in the framework of social exchange theory (Cropanzano and Mitchell, 2005). The findings of the present study suggest that, if consumers trust hotel managers, and feel identified with the hotel values, they are likely to reciprocate by increasing their engagement levels. The positive impact of trust on customers’ beyond-purchase connections with the hotel, also supports prior research proposing that behavioral dimensions of CE are a likely outcome of trust for both new and existing customers (Martínez and Rodríguez del Bosque, 2013; Yen and Tang, 2015). As argued by Martínez and Rodríguez del Bosque (2013), where the client believes that the hotel and its employees are capable of providing the necessary services, infrastructure and information, and to put themselves in the customer’s position when providing services, this increases customer revisit intention. Furthermore, the customer’s trust in the hotel’s performance encourages them to altruistically share their experiences to assist others going through the same decision-making processes (Yen and Tang, 2015).

In line with previous research, the results prove the positive impact of the CSR activities on the customers’ trust in hotels (Jalilvand et al., 2017; Martínez and Rodríguez del Bosque, 2013; Yen and Tang, 2015). This result offers new insights into the action-based theory of persuasion: as more customers come to perceive that a hotel is socially responsible and participates in activities that benefit the community, a larger number of them will trust the hotel (Jalilvand et al., 2017). This result supports service-dominant logic (Vargo and Lusch, 2017); trust is an integral part of value co-creation, since it encourages dialogic, long-term interaction between customer and company, which helps to reduce customer skepticism (Ballantyne and Varey, 2006)

The findings confirm that CCI generated on Twitter, and induced mainly due to CSR communications, is a powerful cognitive connection that can generate positive behavioral responses in relation to the brand. CCI encourages customers to spend more time and money on and in the hotel, which is a fundamental part of CE from a behavioral perspective (Pansari and Kumar, 2017). A double mediating effect has also been observed: customers trust the hotel because of its pro-environmental behavior, and identify with its corporate principles because of its support of specific causes, and display positive CE behaviors. This final contribution has not been seen in the extant literature; it provides a refinement of the CSR-CE relationship concept in the hotel industry.

As tourist companies increasingly rely on social media to engage with customers, recent studies have attempted to clarify the impact of firms’ engagement through social media (FGC) and the effects of UGC on consumer engagement. In this study, we extend and contribute to this stream of literature by examining the moderating effect of message source (FGC vs. UGC) on the relationship between customer perceptions of CSR and customer trust. Our findings are consistent with the solid framework of consumer trust and source credibility (McGinnies and Ward, 1980; van Zoonen and van der Meer, 2015).

It should be noted that the arguments in the literature are contradictory (Serra-Cantallops et al., 2018). Filieri et al. (2015) suggested that source credibility does not influence consumer trust in UGC. However, recent research showed that the effectiveness of social media content on company evaluations is more positive when the source is independent and external to the company (Dickinger, 2011; Sparks et al., 2013). Our findings are aligned with the second research stream. Thus, trust in external sources (i.e. customers) reinforces the impact of CSR perceptions on hotel trust particularly as the customer develops awareness and wants to obtain information about a company (Colicev et al., 2018).

The present paper focuses on the relationship between CSR communications made through Twitter, CSR customer perceptions and the effects on CE from a behavioral perspective. These show that customer trust and CCI are full, independent and joint mediators in this relationship. Even more important is the moderating effect caused by CSR message source, with UGC generated by third-party independent sources being considered more credible, outside corporate control and lacking in self-interest (Du and Vieira, 2012).

6. Managerial implications

The results of the study provide insights that can help hotels improve their CSR communications to produce an effective impact on CE. First, hotels should prioritize Twitter for their CSR communications. The messages should emphasize their corporate values and principles so that customers can easily find commonality with their own defining values. In this sense, hotel industry experts point out that CSR should no longer be a mere promotional tool with no relation to the hotel’s activity; CSR must be fully integrated into the hotel’s philosophy and be aligned with the company’s main activity and the impact it has on the different stakeholders.

Although customers prioritize self-centered values and attach more trust to companies that care for their personal situations (Korschun et al., 2014), growing concerns at the societal level are slowly transforming these into community-centered values, such as solidarity, welfare and environmental protection (Schmeltz, 2012). Therefore, messages should emphasize actions that take account of stakeholders’ concerns and that contribute to social and environmental issues that affect society as a whole. The hotel industry experts agree that social media are fundamental for hotels in their interaction with current and potential customers; this is especially important for small hotels, which currently do not prioritize active CSR communications in social media. Following their insights, social media are fundamental to interact with hotel current and potential customers, especially for small hotels, which are not giving priority to these means to actively communicate about CSR and give a predominant presence to this content.

The generation of customer trust and CCI results in engaged customers more likely to act toward the company in three ways:

1. to value the experience and repeat it in the near future;

2. to start conversations on social media networks about the positive aspects of the hotel; and

3. to share their experience with the company and to contribute to the development of new and improved products and services that integrate customers’ requirements (Pansari and Kumar, 2017).

From the point of view of practitioners, CSR is a main tool for generating customer trust and CCI, if well communicated and accompanied by the expected service quality. In particular, the strongest perceived effects are on CIV and CKV, which are considered as fundamental sources of customer value by the industry. As to which is the more appropriate message source, UGC is shown to be better for the generation of customer trust. Nevertheless, the question of control should be considered by the hotel: although both sources can be used to engage customers, as external sources are more difficult to control, the content of the messages might be more difficult to handle from the corporate point of view (Skard and Thorbjørnsen, 2014).

In addition, Twitter communication can have several advantages for hotels in CSR communication. CSR is an important source of social media content generation which, according to hotel managers, remains a problem in the communication strategies of companies in this sector. Moreover, CSR information can be condensed into tweets and made more accessible than reports aimed at investors and other financial stakeholders. Although hotel companies may encounter lower interaction with CSR-related tweets in their general hotel profile, this can be addressed by the configuration of a specialized CSR profile, which can then enjoy greater attention and interaction (Etter, 2013). To obtain the maximum benefit from their CSR activities, to enjoy greater communication symmetry and to build enduring relationships with their customers, companies should publish CSR-related information in a systematic way and proactively respond to customers’ inquiries and observations (Etter, 2014).

7. Limitations and future research lines

The study has some limitations. First, future investigations should examine the effect of customers’ previous knowledge of the CSR initiatives and reputation of the hotel. Second, the study focuses on the customer’s perspective, leaving unanswered questions about the economic effects on the firm and destination. The hypotheses might be tested in different destinations (e.g. a beach resort) to verify whether this shows differences in the proposed relationships, as found in other studies (Singh et al., 2008). Third, we have examined only behavioral intentions related to CE. Academicians call for more research into real behaviors to obtain insights into emotional and behavioral reactions to specific stimuli (Inman et al., 2018); future studies should measure actual CE behaviors to obtain a more realistic profile of the hotel customer with respect to CSR communications. Fourth, in future developments of the proposed model we will include individual attitudes toward CSR, such as the conceptualization proposed by López-Davis et al. (2017) in their definition of personal social responsibility. Fifth, we might examine customer satisfaction, as this might influence final behavioral intentions. Sixth, we might consider the feelings of proximity that may be developed in readers based on their cultural background/nationality.

Figures

Figure 1.

Conceptual model

Figure 2.

Experimental stimuli (FGC versus UGC)

Table I.

CE definitions following the behavioral perspective

Author(s)	Definition
Verhoef et al. (2010)	“Behavioral manifestation toward the brand or firm that goes beyond transactions” (p. 247)
Van Doorn et al. (2010)	Introduces the concept of CE Behaviors (CEB), centered in the behavioral relationship from the customer to the firm, whose different behaviors (positive or negative) are caused by motivation and can influence other existing or potential stakeholders beyond the firm/brand itself
Vivek et al. (2012)	“Beyond the purchase… events and activities engaged in by the customer that are not directly related to search, alternative evaluation and decision making involving brand choice” (p. 127). Such activities include positive WOM communication, feedback to the firm and participation in activities proposed by the company
Verleye et al. (2014)	“Voluntary, discretionary customer behaviors with a firm focus” and “customers“ interactive cocreative experiences with a firm” (p. 69). The concept includes helping other customers, complying with company's rules, cooperation and feedback
Jaakkola and Alexander (2014)	“Voluntary resource contributions that have a brand or firm focus but go beyond what is fundamental to transactions, occur in interactions between the focal object and/or other actors, and result from motivational drives” (p. 248)
Kumar and Pansari (2016)	“Customer attitude, behavior, and level of connectedness among themselves and with the firm” (p. 499)
Harmeling et al. (2017)	“Customer's voluntary resource contribution to a firm's marketing function, going beyond financial patronage” (p. 316)
Pansari and Kumar (2017)	“The mechanics of a customer's value addition to the firm, either through direct or/and indirect contribution.” (p. 4). The direct contribution consists of customer purchases, and the indirect contributions consist of incentivized referrals, the social media conversations that customers have about the brand, and the customer feedback/suggestions given to the firm

Table II.

Sample profile

Trip motivation							Leisure		Work		Total
							168	74%	59	26%	227
Twitter user							Yes		No		Total
							117	51.5%	110	48.5%	227
Twitter Frequency of Use	More than once a day		Once a day		2 or 3 times a week		Once a week		Less frequency		Total
	30	25.6%	16	13.7%	23	19.7%	12	10.3%	36	30.8%	117
Hotel booking	Hotel website		Travel website		Phone call		Travel agency		Booked by someone else		Total
	33	14.5%	129	56.8%	16	7%	17	7.5%	32	14.1%	227
Hotel chain with active CSR communication							Yes		No		Total
							134	59%	93	41%	227
Gender							Male		Female		Total
							105	46.3%	122	53.7%	227
Age			18-24		25-44		45-64		> 64		Total
			41	18.06%	107	47.1%	74	32.6%	5	2.2%	227

Table III.

ANOVA between in-person and online questionnaires

Variable	Sum of squares	df	Mean square	F	Sig.
CSRP
Between groups	3.266	1	3.266	1.900	0.169
Within groups	386.759	225	1.719
Total	390.025	226
TRU
Between groups	0.613	1	0.613	0.472	0.493
Within groups	291.929	225	1.297
Total	292.541	226
CCI
Between groups	1.745	1	1.745	0.998	0.319
Within groups	393.459	225	1.749
Total	395.204	226
CLV
Between groups	0.172	1	0.172	0.124	0.726
Within groups	313.089	225	1.392
Total	313.260	226
CIV
Between groups	0.178	1	0.178	0.062	0.803
Within groups	645.238	225	2.868
Total	645.416	226
CKV
Between groups	5.250	1	5.250	2.094	0.149
Within groups	564.170	225	2.507
Total	569.420	226
ENG
Between groups	1.087	1	1.087	0.664	0.416
Within groups	368.556	225	1.638
Total	369.643	226

Table IV.

Measurement scales

Construct	Authors	Indicators
CSR customer perceptions	Bigné et al. (2010)	CSRP1. The hotel chain is aware of environmental matters
		CSRP2. The hotel chain fulfills its social responsibilities
		CSRP3. The hotel chain puts something back into society
		CSRP4. The hotel chain acts with society's interest in mind
		CSRP5. The hotel chain acts in a socially responsible way
		CSRP6. The hotel chain integrates philanthropic contributions into its business activities
CCI	Homburg et al. (2009)	CCI1. I am well identified with this hotel chain
		CCI2. I feel good to be a customer of this hotel chain
		CCI3. I like to tell that I am a customer of this hotel chain
		CCI4. This hotel chain fits well to me
		CCI5. I feel attached to this hotel chain
Trust	Flavián et al. (2006)	TRU1. The information offered by this hotel chain is sincere and honest
		TRU2. This hotel chain is concerned with the present and future interests of its customers
		TRU3. This hotel chain has the necessary resources to successfully carry out its activities
CE	Kumar and Pansari (2016)	CLV1. I will continue staying at this hotel chain in the near future
		CLV2. My stays at this hotel chain make me content
		CLV3. I get my money's worth when I stay at this hotel chain
		CLV4. Staying at this hotel chain makes me happy
		CIV1. I would actively discuss this hotel chain in social media
		CIV2. I would love talking about my experience at this hotel chain in social media
		CIV3. I would discuss the benefits that I get from this hotel chain with others in social media
		CIV4. I am part of this hotel chain and would mention it in my conversations in social media
		CKV1. I would provide feedback about my experiences at the hotel to the hotel chain
		CKV2. I would provide suggestions for improving the performance of the hotel chain
		CKV3. I would provide suggestions or feedback about the new services of the hotel chain
		CKV4. I would provide suggestions or feedback for developing new services for this hotel chain

Table V.

Measurement model. Convergent validity and reliability

Factor	Indicator	Standardized loadings	p-value	Standardized weights	p-value	t-value	VIF	CA	rho_A	CR	AVE
CSR Customer Perceptions	csrp1	0.800	***			17.611		0.929	0.931	0.929	0.688
	csrp2	0.822	***			22.777
	csrp3	0.752	***			18.604
	csrp4	0.867	***			18.772
	csrp5	0.897	***			22.171
	csrp6	0.829	***			14.489
CCI	cci1	0.871	***			39.660		0.931	0.865	0.932	0.735
	cci2	0.899	***			37.860
	cci3	0.862	***			39.335
	cci4	0.872	***			33.781
	cci5	0.775	***			28.103
Trust	tru1	0.838	***			23.196		0.856	0.934	0.859	0.670
	tru2	0.880	***			25.332
	tru3	0.732	***			18.743
CLV	clv1	0.823	***			26.519		0.897	0.902	0.898	0.689
	clv2	0.837	***			25.139
	clv3	0.736	***			19.654
	clv4	0.887	***			25.774
CIV	civ1	0.950	***			31.745		0.954	0.956	0.954	0.838
	civ2	0.940	***			39.827
	civ3	0.936	***			42.204
	civ4	0.832	***			37.003
CKV	ckv1	0.974	***			24.142		0.952	0.953	0.951	0.829
	ckv2	0.909	***			28.389
	ckv3	0.915	***			35.906
	ckv4	0.839	***			28.157
CE	clv	0.977	***	0.814	***	14.830	1.641	N/A	1.000	N/A	N/A
	civ	0.696	***	0.089		1.459	2.042
	ckv	0.713	***	0.200	**	3.004	1.959

Notes:

VIF = Variance Inflation Factor; CA = Cronbach's Alpha; CR = Composite Reliability; AVE = Average Variance Extracted;

**p < 0.01; ***p < 0.001

Table VI.

Measurement model. Discriminant validity (HTMT ratios)

Factor	1	2	3	4	5	6
1. CSR customer perceptions
2. CCI	0.698
3. Trust	0.661	0.772
4. CLV	0.629	0.875	0.752
5. CIV	0.476	0.619	0.457	0.625
6. CKV	0.472	0.603	0.528	0.602	0.698

Table VII.

Summary of mediating effects tests

Direct effects	Coefficient	Bootstrap 95% CI
		Percentile		Bias corrected
H1: CSRP → ENG	0.017ns	−0.077	0.107	−0.084	0.103
H2: CSRP → TRU	0.665***	0.574	0.747	0.568	0.743
H3: TRU → ENG	0.162*	0.038	0.299	0.029	0.288
H4: CSRP → CCI	0.330***	0.179	0.466	0.179	0.467
H5: CCI → ENG	0.710***	0.581	0.834	0.588	0.841
H6: TRU → CCI	0.551***	0.421	0.687	0.414	0.678
Indirect effects	Point estimate	Percentile		Bias corrected		VAF (%)
H7: Via TRU (H2 × H3)	0.108	0.026	0.207	0.024	0.205	70.4
H8: Via CCI (H4 × H5)	0.234	0.116	0.358	0.124	0.364	96.7
H9: Via TRU → CCI (H4 × H6 × H3)	0.260	0.183	0.352	0.189	0.363	79.8
Total indirect effect	0.603	0.508	0.707	0.294	0.515	97.3

Notes:

***p < 0.001; **p < 0.01;

*p < 0.05;

ns

= non-significant; VAF = variance accounted for; R² (CCI) = 0.651; R² (Trust) = 0.440; R² (CE) = 0.518; Q² (CCI) = 0.409; Q² (Trust) = 0.259; Q² (CE) = 0.412

Table VIII.

Comparison of mediating effects

Differential effect	Coefficient	Bootstrap 95% CI
		Percentile		Bias corrected
CCI − TRU = (H4 × H5) − (H2 × H3)	0.126ns	−0.091	0.333	−0.086	0.337

Note:

ns

non-significant

Table IX.

Moderating effect of message source

Hypotheses	Path coefficient		Confidence interval (95%)		Path coefficients difference (FGC vs UGC)	p-value difference (one-tailed)		Status
	FGC	UGC	FGC	UGC		Henseler's MGA	Permutation test
H10: CSRP → TRU	0.520	0.672	0.394, 0.621	0.550, 0.757	0.152	0.953*	0.065**	Accepted
H11: CSRP → CCI	0.311	0.454	0.130, 0.474	0.329, 0.566	0.143	0.868	0.143	Rejected
H12: TRU → CCI	0.524	0.393	0.377, 0.663	0.252, 0.517	0.131	0.133	0.145	Rejected
H13: CSRP → ENG	0.077	0.086	−0.031, 0.173	−0.040, 0.192	0.010	0.547	0.458	Rejected
H14: TRU → ENG	0.108	0.225	−0.009, 0.242	0.077, 0.364	0.116	0.846	0.156	Rejected
H15: CCI → ENG	0.713	0.593	0.597, 0.813	0.441, 0.731	0.121	0.135	0.137	Rejected

Notes:

In Henseler's MGA method, a p-value lower than 0.05 or higher than 0.95 indicates significant differences between specific path coefficients across groups at the 5% level.

**

p < 0.05;

*

p < 0.1