An analysis of the effects of customer satisfaction and engagement on social media on repurchase intention in the hospitality industry

Abstract The purpose of this quantitative-deductive paper is to explore the link amongst customer satisfaction and engagement on social media on repurchase intention in the hospitality industry. The study was conducted on social media because, it is the fastest growing media in history. Data was collected from hotels in the three major business hub cities (Accra, Tamale and Kumasi) in Ghana. A total of 504 valid responses were obtained from respondents in the selected cities. SmartPLS software was used to analyze the data using (PLS-SEM) method. The results show that customer satisfaction has a positive and significant relationship on the dimensions of customer engagement. The three dimensions of customer engagement (contribution, consumption and creation) were found to significantly influence repurchase intention. Finally, two dimensions of engagement (contribution and consumption) were found to mediate the relationship between customer satisfactions and repurchase intention. The study is among the few to combine the COBRA model and Social Exchange Theory to assess the nexus between customers’ engagement in an online environment and its linkages with satisfaction and repurchase intentions. Marketers should consider creating posts with photos, videos, and animation that consumers find entertaining and enjoyable, as this stimulates their desire to consume, contribute, and create content on social media pages for hotel brands.


PUBLIC INTEREST STATEMENT
The opportunities created by social media in building close relationships with consumers has gained the attention of practitioners in diverse industries across the globe. The purpose of this study was to explore the link amongst customer satisfaction and engagement on social media on repurchase intention in the hospitality industry. The study was conducted on social media because, it is the fastest growing media in history.
The engagement aspects look at the interaction between hospitality firms and its consumers by employing the social exchange theory to understand how social media becomes a forbearer of engagement and how that eventually affect consumers' repurchase intention. The study used customer engagement (COBRAs) to mediate the influence of customer satisfaction on customer repurchase. The study concludes that customer satisfaction plays a significant role in influencing buyers' repurchase intention. Social exchange theory, COBRA, and customer satisfaction literature could benefit from this study's findings. In particular, it could help shape repurchase intentions.
Currently, businesses use customer engagement strategies to build customer-brand relationships as a result of access to the internet (Santini et al., 2020). Customers are becoming increasingly aware of competing offerings and engage in active interactions with service providers and other stakeholders (Lariviere et al., 2017). As a result, service companies have strategically implemented practices to engage customers beyond repeat business and positive word-ofmouth in order to build meaningful and profitable relationships with them (Pansari & Kumar, 2017;Roy et al., 2020). Customer engagement is becoming increasingly important at the business level, according to extant empirical studies. For instance, prior research revealed that more than 80% of marketers want to win over engaged customers in order to increase advocacy and trust (Pansari & Kumar, 2017). Considering the important role of engagement in positively influencing consumer behavior, this research employs  COBRA model (Consumer Online Brand Related-Activities) to explicate the linkages between customer satisfaction, engagement and repurchase intention. Consumers can participate in in a variety of ways due to social media platforms (Harrigan et al., 2018;Schivinski, 2019).
Customer engagement has been identified as a critical factor in determining brand loyalty and repurchase behavior. However, the customer engagement mechanism in social media has rarely been thoroughly examined, particularly in the field of tourism social media (Kanje et al., 2020). As a result, the current study fills a knowledge gap by revealing the connections between customer satisfaction, customer engagement, and repurchase behavior. A growing body of literature conceptualizes or examines customer engagement in the tourism domain, in light of the aforementioned marketing benefits (Hapsari et al., 2017;Harrigan et al., 2017;So et al., 2014).
In terms of the outcomes of customer engagement, studies have shown that it has a positive impact on brand loyalty (Harrigan et al., 2017;Kumar et al., 2018). The direct effect of customer engagement on loyalty, on the other hand, was not fully supported in Steinhoff et al. (2018). As a result, the link between customer engagement and repurchase intention remains a mystery, and other variables must be investigated to clarify it. Some scholars believe that customer engagement varies depending on the situation (Kanje et al., 2020), the initiator (Beckers et al., 2018;Brodie et al., 2011), or the consumer segment (Beckers et al., 2018;Andrews et al., 2019).
Going forward, extant research has advanced conceptual connections between customer engagement and other theoretically related concepts, revealing useful information about potential and actual antecedents and consequences (Hollebeek & Macky, 2019;So et al., 2016a;Van Doorn et al., 2010) especially on repurchase intention. Many factors influence customer engagement, including involvement (Hollebeek et al., 2014;So et al., 2016a), self-brand image and value congruency and situational factors (Brodie et al., 2011;Van Doorn et al., 2010). Customer participation, interactivity, commitment (for existing customers), trust, brand attachment, and brand performance perceptions are all identified as antecedents in other studies (Van Doorn et al., 2010;Vivek et al., 2012). Few empirical studies look at customer engagement subgroups in terms of patterns of engagement. Some have also looked at COBRA and brand equity , social media and COBRAs (Majeed et al., 2020), satisfaction and engagement but no study has been combined COBRA, customer satisfaction and repurchase intention in a singly study. The study utilized the COBRA model comprising of consumption, contribution and creation, which act as mediators through which customer satisfaction predicts repurchase intention.
Moreover, the social exchange theory has been used to explain the relationship between firms and clients because it emphasizes participatory behaviors of individuals and the reasoning for resource exchanges (Ferm & Thaichon, 2021). Engagement involves social exchanges in which access to relevant information, affiliation and social status are mostly vital (Harrigan et al., 2018). Modern marketing thought suggests that customers are now partners with firms creating exchanges in a co-creation process (De Silva et al., 2019;Zhang, Gu et al., 2018). For engagement to last, both the consumer and the firm must recognize that it is equitable (Carlson et al., 2019;Moliner et al., 2018). Hence, engagement can be viewed as a social exchange. Interestingly, limited studies examining engagement behaviors in the hospitality sector have sought guidance from this theory to understand the nuances of consumers' engagement on social media. This study narrows the theoretical void by combining the COBRA model and the social exchange theory as a theoretical lens to better understand customer engagement behaviors in an online context.
The opportunities created by social media in building close relationships with consumers has gained the attention of practitioners in diverse industries across the globe (Sashi, 2012). However, academic scholarship on customer engagement in online settings has lagged practice and its theoretical foundation is comparatively underdeveloped, as such, a better understanding of the customer engagement behaviors online and its link with satisfaction and repurchase intention is essential in creating strategies for customer retention and loyalty. By combining the social exchange theory and the COBRA model, the current study intends to shed light on participatory behaviors online and their effects on consumer attitudes with data from an emerging market setting. Based on the arguments advanced, this study specifically seeks to attain the following goals: (1) To understand the effect of Customer satisfaction on customer engagement on social media; (2) To examine the effect of customer satisfaction on customer repurchase intention in the hospitality industry; and (3) To understand the effect of social media customer engagement on customer repurchase intention in the hospitality industry.
The remainder of the paper is organized as follows: first, literature is reviewed on the study's key constructs and this is followed by hypotheses development and presentation of the research model. Following that, we proceed to the research methodology, data analysis, and discussion of the findings. Finally, we discuss the study's theoretical contribution, managerial implications, limitations, and future research opportunities.

Social exchange theory
Previous tourism studies on organization-consumer relationships in non-digital utilized the social exchange theory (Rather & Hollebeek, 2019;Yuen et al., 2018). The current study examines brandconsumer relationships in an online setting using the social exchange theory. According to the social exchange theory, individuals make rational decisions to engage in a social exchange based on their perceptions of the costs and benefits (Kelley & Thibaut, 1978). An ongoing relationship exchange can be economic, social, or a mix of both social and economic benefits and costs. Customer Brand Engagement (CBE) entails social interactions in which consumer's value access to relevant information, affiliation, and social status from the consumer brand community over monetary resources and outcomes (Yuen et al., 2018). Consumers are now partners with marketers, creating exchanges through a co-creation process, according to both relationship marketing and service-dominant perspectives (Jaakkola & Alexander, 2014). Consumers exchange cognitive, emotional, social, economic, and physical resources with marketers. Both the consumer and the marketer must believe that CBE is equitable in order for it to continue (Brodie et al., 2011;Hollebeek, 2009).

Customer satisfaction
Customer satisfaction is described as the psychological state consumers experience when their expectations prior to the consumption of a product/service meets or exceeds the actual consumption experience (Patricks et al., 2020). Customer satisfaction is measured at the point when a product or service is purchased and then used, as well as throughout the process of acquiring a product or receiving service. Consumers usually decide to buy or continue buying after assessing whether their experiences with the service or product has been satisfactory or pleasurable. Extant studies have reported that purchase intention increases the likelihood of actual purchase behavior, thus, a company that seeks profits will need to make efforts to attract customers and ensure their satisfaction (Bapat, 2017;Sreejesh et al., 2018). Customer satisfaction is the sum of a customer's perceptions, evaluations, and psychological reactions to a product or service's experience, therefore, it is regarded as subjective, since consumers who use or consume a product are the only ones who can measure satisfaction with it (Benoit et al., 2020).The hospitality sector is part of the experience economy, hence, customer satisfaction is paramount if firms in this sector want to remain competitive (Cheng et al., 2019). Satisfied customers are more likely to engage in COBRA and firms can rely on their engagement on social media to differentiate their brands on the basis of positive, strong and unique experiences (Priporas et al., 2017).

Customer engagement
The fundamentals of the service provider-customer connection have changed over time particularly since the inception of the networked society (L. Hollebeek & Andreassen, 2018). Marketing academics and practitioners have long been interested in how businesses can benefit from customer engagement strategies, which can lead to value co-creation (Lemon & Verhoef, 2016). Customer engagement is the process of developing a deeper relationship with customers, with the goal of achieving customer loyalty, (Ho & Chung, 2020). Customers' roles have shifted from passive recipients to indispensable co-creators of customer values (Lemon & Verhoef, 2016), as well as key contributors to firm value, (Brodie et al., 2011), and performance (Lemon & Verhoef, 2016).
Increased customer engagement helps businesses achieve their goals, such as sales growth, cost reduction, better customer experiences, higher profitability, and customer loyalty (Hapsari et al., 2017). It can also be viewed as strategically imperative for achieving a long-term competitive advantage and a tool for establishing and maintaining effective customer relationships (Santini et al., 2020). In the service sector, engaged customers could serve as co-creators which can enhance the overall service experience. The hospitality sector needs to champion practices to promote customer engagement behaviors beyond transactions in order to sustain and increase their market share (Kanje et al., 2020). Research on the hospitality sector suggests that customer engagement facilitates the conversion of browsers to buyers and could also aid them to gain more insights into their businesses (Kanje et al., 2020).

Customer engagement (CE) on social media
CE allows businesses to engage in, participate in, and influence the conversation surrounding their brand. Effective CE increases brand loyalty and influences customer discussion and purchase behavior (Carr, 2017). Users who value big brands interact with them on social media by "liking," "sharing," and "commenting" on them (Santini et al., 2020;Wang & Gon, 2017). Every brand interaction, whether it's a purchase, a social media post, or any other exposure to the brand, builds and rebuilds CE. Consumers are increasingly gaining access to digital and social media platforms as a means of expressing opinions and interacting with companies (Santini et al., 2020). Numerous organizations have shifted promotional activities from traditional media and have begun using digital platforms to directly interact with customers. These days, businesses use social media to detect highly engaged customers for specialized marketing efforts and to ensure that they remain emotionally, profitably and sustainably connected (Wu et al., 2020).
Most of the previous studies on CE either looked into the relationship between customer engagement and brand loyalty (Harrigan et al., 2017;So et al., 2016a) or used qualitative design (Brodie et al., 2013;Harrigan et al., 2017). Also some past studies such as Dwivedi (2015) used customer engagement dimensions (vigor, dedication and absorption) which are different from what the current study is advocating (creation, contribution and consumption-COBRA) The COBRA concept is described as a behavioral construct that provides a unifying framework to think about consumer activity pertaining to brand-related content on social media platforms (Muntinga et al., 2015). However, there are limited studies applying the COBRA model to understand CE. The COBRA framework permits the assessment of a consumer's level of brand-related activities as well as the concurrent investigation and comparison of behaviors that past studies only investigated discretely (Muntinga et al., 2015).
In hospitality sector, social media platforms play an important role in travel decisionmaking and they help customers and marketers co-create value. Customers' roles have evolved to include functions such as enablers, innovators, coordinators, and differentiators (Lariviere et al., 2017). Customers in these roles participate in the development and delivery of new products or services, form communities, engage other customers and prospects, interact with noncustomers, and distinguish between competing offers in the marketplace. The emergence of social media and technological advancements support the need to reconsider current customer engagement concepts. Observing interactive experiences between multiple actors that have been facilitated via social media has led to a new perspective (Brodie et al., 2013;Lariviere et al., 2017). Social media has transformed the role of a customer by allowing customers to contribute to the creation and sharing of information, photos, reviews, and other marketing resources (Lariviere et al., 2017). Social media has a direct impact on changing perceptions and attitudes within a user when exposed to other users or groups, resulting in the formation of polarized groups (Kanje et al., 2020). The following CE dimensions will be used for this research: consumption, contribution, and creation (Majeed et al., 2020;Schivinski et al., 2019). Muntinga et al. (2015) developed the concept of COBRAs to classify social media behaviour patterns into three usage types, referring to "a set of brand-related online activities on the part of the consumer that vary in the degree to which the consumer interacts with social media and engages in brand-related online activities" (Schivinski et al., 2016, p. 66). COBRAs are a set of three usage types that span a behavior continuum of consuming, contributing, and creating, and are regarded as the lowest, medium, and highest levels of online engagement behavior respectively (Schivinski, 2019;Simon & Tossan, 2018). When consumers expect to engage and interact with the brands they consume and when brands are defined collectively, COBRA's importance is amplified (; Schivinski et al., 2020). COBRAs play a critical role because social media has altered not only how businesses communicate with their customers, but also how customers organize their decisions when planning trips (Schivinski et al., 2020). The three dimensions of COBRA namely: consumption, contribution and creation are discussed in the sub-sections below.

COBRAs
Consumption. When consumers read, view, or watch brand-related information without actively contributing or creating user-generated content, they are consuming Majeed et al., 2020). Consumers prefer reading recent trends, company news, and other important topics information about the products of interest on social networking sites (Cheung et al., 2020), which influences the number of reads and views on social media brand pages Cheung et al., 2020).
Contribution. Contribution occurs at the medium level when consumers engage in online engagement behaviors like liking, commenting, and sharing brand-related information (Majeed et al., 2020). Relationships with other customers or the brand are included in the contribution dimension (Schivinski et al., 2020). Likes and endorsements of brand-related content, as well as commenting, sharing, and reposting content, are examples of such activities . Consumers may share trending information with peers and like-minded users on social media platforms in order to make a contribution to social media brand community.
Creation. Creating at its most basic level entails uploading, dissemination, and placement content on their own and the brand's social media pages (Majeed et al., 2020). Writing and commenting on reviews, posting pictures and self-portraits featuring the brand, and starting hashtags are all examples of creation activities (Schivinski et al., , 2020. The tourism industry is especially interested in co-creating brand experiences, which is made easier by social media. Through peer-to-peer communications, storytelling campaigns, and brand narratives based on people's real stories and testimonials, social media also plays an important role in the development and growth of shared and collaborative consumption business models, allowing the co-creation of tourism experiences (Filieri & Lin, 2017). Commitment, satisfaction, brand loyalty, purchase intention, and positive referrals are all considered to be effective in generating positive business outcomes (Gensler et al., 2013). Social media is increasingly being used by brands to drive consumers' online brand-engagement activities (COBRAs), such as consuming, contributing, and creating content, in order to improve consumer-brand engagement (CBE) and loyalty intentions (Majeed et al., 2020). As a result, it improves brand awareness/ associations, and purchases  and perceived quality (Brodie et al., 2013).

Repurchase intention
Repurchase intention is the evaluative likelihood that a customer (i.e., experienced customers) will continue to buy a product from the same online seller (Wu et al., 2020;Zhang, Gu et al., 2018), or that they will buy product/services from the same company again (; Trivedi and Yadav, 2020). Previous studies have investigated possible drivers of customer repurchase intention mostly from a buyer relationship standpoint, recognizing trust and satisfaction as two important predictors (Wu et al., 2020). A positive purchase experience may lead to positive affect, which could encourage consumers to post positive reviews (Wu et al., 2020;Zhang et al., 2018). Due to the high cost of acquiring new customers and the economic value of trusted, loyal customers, repurchase is both necessary and desirable. It costs five times as much to acquire new customers and initiate transactions with them as it does to keep existing customers. A high level of e-satisfaction is required to maintain a positive customer relationship; this fosters customer trust and repurchase intent (as well as lowering switching costs (Wu et al., 2020)

Customer satisfaction and repurchase intention
There are non-standardized preconceptions that must be met from every product that is marketed, as such, fulfillment of aspects of customer satisfaction becomes an exciting dynamic for marketers (Liang et al., 2018). Consumer satisfaction is used to form an assessment of the product consumed, allowing consumers to decide whether the product is suitable for consumption in the future. Consumers who are satisfied with a product's performance are more likely to recommend it to others. When consumers want to buy a product again, the company or product provider must meet their expectations. When compared to dissatisfied customers, satisfied customers are more likely to make repeat purchases. Consumers' desire to repurchase a product in the form of goods or services for which they have previously experienced the benefits and quality is referred to as repurchase intention (Liang et al., 2018). Research in the hospitality sector has affirmed the positive effect of satisfaction on repurchase intentions (Skarlicki et al., 2019). Customer satisfaction with the overall service experience in the hospitality sector enhance the likelihood of patronizing the service again (Liang et al., 2018).

Customer satisfaction and customer engagement
Customer satisfaction and customer engagement have been linked in previous research (Carlson et al., 2017;Gopalakrishna et al., 2017;Simon & Tossan, 2018;Thakur, 2018). Customer satisfaction, according to Carlson et al. (2017), is a moderating variable in the relationship between brand experience value and customer engagement. Furthermore, according to Simon and Tossan (2018), customer satisfaction is a predictor of brand Facebook page engagement. As a result, a customer's satisfaction with a product or brand will influence his or her engagement with the brand/product. Thakur (2018) also looked at the positive and significant link between customer satisfaction with retailers and customer engagement with their mobile apps and the results depict that customer engagement will increase if the customer is pleased with the retailer's product and service.
In the context of COBRAs, several antecedents to online engagements have been identified and they include entertainment, remuneration, self-identification and empowerment (Muntinga et al., 2015). Once customers' motivations for going online have been met, they tend to actively participate in online brand related activities (Carlson et al., 2019). The consuming COBRA typology is the minimum level of online brand-related activeness, however, if customers are not satisfied with their experiences with a firm, they may show reluctance towards "consuming" (Kandampully et al., 2018;Muntinga et al., 2015). Satisfied customers may show "consuming" engagement behaviors by participating without vigorously contributing to or creating content, hence, they may view brand-related videos that businesses produce, view the product ratings and reviews that others post, and the discussions between participants of online brand forums (Ferm & Thaichon, 2021;Santini et al., 2020). Moving on, the contributing COBRA type reflects a medium level of online brand related activeness. Some satisfied customers may go beyond consumption and show contribution related engagement behaviors. It covers both user-to-content and peer-to-peer interactions. Customer satisfaction has been recognized in the literature to be linked with positive attitudes like intentions to reuse and to recommend (Cheng et al., 2019;Priporas et al., 2017). Extending this logic to customer engagement, it is argued that customer satisfaction is positively related to contribution. Thus, satisfied customers will contribute to brand related content, have conversations on brand's fan page on social media and comment on brand related posts (Lee & Kim, 2018;Muntinga et al., 2015).
Customers who enjoy an emotionally bonding relationship with an organization become engaged (Boateng, 2019;Kumar et al., 2018;Santini et al., 2020). Satisfaction and positive emotions have been identified by past studies as drivers of customer engagement on social media (Carlson et al., 2019;Jaakkola & Alexander, 2014;. The creating COBRA type is the highest level of online brand-related activeness (Muntinga et al., 2015). Based on the foregoing discussion, it is postulated that satisfied customers may express their satisfaction with the brand through the creating COBRA typology (Majeed, 2020). This group of customers will exhibit engagement behaviors by creating brand related content, post reviews, produce and upload branded videos. Based on the foregoing discussions, the following hypotheses are expected to hold true:

H1: There is a significant positive relationship between CS and COP
H2: There is a significant positive relationship between CS and CON H3: There is a significant positive relationship between CS and CRE

Social media engagement on social media and repurchase intention
While recent research has found a link between web portal cues and consumers' repurchase intent (Rather & Hollebeek, 2019), this link can be strengthened by incorporating customer engagement. Customers who are highly engaged are more likely to feel empowered as a result of their interaction, which can lead to a variety of positive transactional outcomes like repurchase intent (Lee & Kim, 2018;Rather, 2019). Relationship marketing and marketing communication are inextricably linked to customer engagement (L.D. Hollebeek & Macky, 2019;Rather, 2019;Roy et al., 2020). Repurchase intention acts as a proxy for purchase behavior (Liang et al., 2018). Through the engagement process, brand-interactivity elements on social media develop consumer attention, understanding, and positive affection, driving repurchase intention (Lee & Kim, 2018;Wu et al., 2020). Hotel customer engagement is the process of developing a cognitive, affective, and behavioral commitment to an active relationship with a hotel. As a result, customer engagement can be broken down into four key dimensions: interaction, activity, behavior, and communication, depending on the situation. The degree of closeness that customers feel toward the website is referred to as interaction engagement (Demangeot and Broderick, 2016). This phenomenon is dependent on how well the website meets and understands their expectations. The level of customer involvement in "generating" communication with online sellers is referred to as activity engagement (e.g., clicking on hyperlinks, requesting more information). Behavioral engagement is defined as "an individual's level of energy, effort, and time spent in a particular interaction" and goes beyond financial patronage (Hollebeek et al., 2014, p. 154). Finally, communication engagement is a type of commitment that occurs when someone wants to keep a relationship going (Rather & Hollebeek, 2019). According to previous research, customers who engage with a brand are more likely to develop positive beliefs about the sellers than customers who do not (Harrigan et al., 2018). The relationship also exists in e-commerce, where customer engagement helps them make better repurchase decisions online. As a result, customer engagement acts as a motivator for customers to respond in a certain way. Customer engagement is boosted by cues embedded in the virtual environment, which in turn boosts optimal consumer behaviors (Boateng, 2019). CBE has a positive impact on brand loyalty outcomes, such as repurchase intent (Cheung et al., 2020;Dwivedi, 2015;Hollebeek et al., 2014;L.D. Hollebeek & Macky, 2019;Leckie et al., 2016). So, we hypothesized that: H4: There is a significant positive relationship between COP and repurchase intention H5: There is a significant positive relationship between CON and repurchase intention H6: There is a significant positive relationship between CRE and repurchase intention 3. Methodology

Data collection and sampling
To ensure its truthfulness and consistency, a cross-sectional survey was created and pre-tested. To obtain social-demographic diversity, the survey was self-administered. Apart from that, the three regions (Accra, Tamale, and Kumasi) are Ghana's main business hubs and have the highest population density with highest number of hotels. Personal surveys of Ghanaian hotel customers were used to collect data for the study. Individuals aged 18 and above who patronizes any one of the country's hoteliers is included in the target population. Respondents were intercepted at social media and point of purchase or entry and exit of the hotels, making the sampling process nonprobabilistic dubbed convenience sampling techniques. However, three survey locations were identified in each of the low, middle, and high-income individuals in each of Ghana's three most populous cities (Accra, Tamale, and Kumasi) to ensure a fair representation of the target population. To avoid bias in the selection process, every respondent was approached for consent and participation in the study at each survey point. After deleting incomplete responses and treating outliers, a total of 504 valid responses were obtained during the months of January and February of 2021. We used both online method by employing Qualtrics (qualtrics.com) and paper-pen physical collection of data. The manual data was manually added to the downloaded online data.

Measures
To evaluate the empirical parameters of the proposed constructs, we used a five-point Likert-scale survey (1 = strongly disagree; 5 = strongly agree). For customer satisfaction, four items measuring respondents' satisfaction with the overall functionality of the services obtained from the Hotel service were adapted from previous studies (Kuo et al., 2009;Pe´rez & Rodrı´guez Del-bosque, 2015). The three main constructs of customer engagement (consumption, contribution and creation) and their 5, 6, 6 items respectively were adapted from several studies Majeed et al., 2020). We measured repurchase intention by adapting from previous studies (Chiu et al., 2009;Kim et al., 2012) and we operationalized this variable with four-item measures with regards to assessing the degree to which hotel customers are willing to repeat purchase of the hotel offerings.

Data analysis
SmartPLS software was used to analyze the data that use partial least squares (PLS-SEM) method (Ringle et al., 2015;Sarstedt and Cheah, 2019). PLS is a causal prediction method to hypothesis testing that fits the research's prediction-oriented goal Lim, et al., 2020). Furthermore, when the model includes a formative higher-order construct that can be measured, this approach is more advantageous (Hair et al., 2016). To avoid single source bias in the data, Harman's single factor test was used to check for common method bias (Fuller et al., 2016). PLS-SEM is one of the most widely used structural equation modelling (SEM) tools for data analysis  (Henseler et al., 2016). We conducted reliability/validity test (α, CR, AVE and rho_A) and path analysis (β, t-test and p-values).

Reliability and validity
Apart from Amos-SEM, another method for testing causal models is partial least squares (Sharma & Singh, 2021). Validity and reliability are two important factors to consider when evaluating a survey instrument. As a result, Cronbach's alpha is calculated for the Smart-PLS study. Table 1 shows the Cronbach's alpha values, which indicate that the scale has high reliability because the factor loadings are above 0.70. (Collis & Hussey, 2013;Greener, 2008). However, according to Dijkstra and Henseler (2015), the most important reliability test for PLSSEM is rho_A. The results for rho_A are shown in Table 1, and the values for rho_A for this study range from 0.801 to 0.936. As a result, the scores of latent variables can be assumed to be reliable and can be used in this study (Hair et al., 20172017).
Principal component analysis (PCA) with varimax rotation was used to test the dimensionality of the different instruments used to calculate the variables under investigation. Surprisingly, PCA revealed a unidimensional structure for all five variables, with a factor loading of greater than  (Trivedi & Yadav, 2020). With a bootstrapping of 5000, these results suggest a strong model fit. All of the tests had a composite reliability (CR) and Cronbach's alpha (α) value of greater than 0.7, suggesting that they were accurate. All of the items' loadings were found to be relevant, and none of them were dropped. Significant factor loading and a high CR, according to Anderson and Gerbing (1998), establish convergent validity; specifically, AVE must be greater than 0.5 but less than the CR value (Hair et al., 2012), which is true in this case (Table 1 and 2). The square root of AVE should be greater than the coefficients of correlation with other constructs for discriminant validity (Fornell & Larcker, 1981).

Direct effect
The outcome of the study supported H1, H2, and H3, demonstrating that customer satisfaction has significant positive influence on hotel social media customer contribution, creation and consumption.

Indirect effect
Two of the elements of COBRA (consumption and contribution) were also found to mediate the relationships between customer satisfactions and repurchase intentions in this study. For instance, consumption (β = 0.522; t = 12.354; and p = 0.005), and contribution (β = 0.741; t = 8.610; and p = 0.001) were both positive and significant. However, creations in COBRA could not mediate the relationships (β = −0.562; t = 0.354; and p = 0.505) has negative and insignificant connection between customer satisfaction and repurchase intentions since creation.

Coefficient of determination
R-squared (R 2 ) is a quantitative measure of the proportion of variation accounted by an independent variable or variables in a regression model for a dependent variable. R-squared explains how much the variance of one variable explains the variance of the second variable, whereas correlation explains the strength of the relationship between an independent and dependent variable. Researchers usually assess the structural model using the coefficient of determination (R 2 value) measure (Hair et al., 2013). This coefficient is defined as the squared correlation between the predicted and actual values of a specific endogenous construct (Hair et al., 2013). Furthermore, this coefficient can be used to determine the model's predictive accuracy. The coefficient is also meant to represent the combined effect of exogenous and endogenous latent variables on an endogenous latent variable. The coefficient is the squared correlation between the actual and predicted values of the variables. This coefficient also indicates the degree of  Table 3 was observed as the high value for CS and RI respectively, but the qualities in the range of 0.412, 0.375, and 0.362 for COBRA (consumption, contribution and creation) respectively which are described the moderate values as mentioned by (Liu et al., 2005). According to Table 3 and Figure 1 and 2, the model had a mediating predictive power, with nearly 41.2 percent, 37.5 percent, 36.2 percent, and 40.8 percent of the variance in consumption, contribution, and creation respectively, supported by the model. Furthermore, the R 2 value of customer satisfaction was found to explain 85.2% of the variance, indicating that this construct has ahigh predictive power. Muntinga et al. (2015) model of consumers' brand related activities on online media identifies three types of brand related activities: passive consumption (e.g., viewing videos and pictures, reading product reviews), active content contributing (e.g., actually replying to posts and other SNS comments, posting one's own customer reviews, uploading user-created videos) and creation.

Discussion
The hypotheses of the first direct effect were all supported. Hence, H1, H2, and H3 were supported, depicting that customer satisfaction has both positive and significant relationships with the three dimensions of customer engagement. This is coherent with previous findings which indicate that customer satisfaction and customer engagement are positively related (Carlson et al., 2017;Gopalakrishna et al., 2017;Simon & Tossan, 2018;Thakur, 2018). Of the three dimensions of customer engagement, the relationship between customer satisfaction and contribution was the strongest, followed by consumption. This result suggests that the more satisfied customers are, the more likely it is that they would contribute to brand related activities on social media. Satisfied customers would engage in behaviors like product rating, joining a brand profile on social media, engaging in branded conversations and commenting on brand related social media posts to express their satisfaction with the brand (Muntinga et al., 2015). H5, H6 and H7 were all supported. Implying that customer engagement elements significantly predict repurchase intention. with an increasing number of hotel services disrupting traditional business models through web 2.0 technologies, SM-CE trends have a significant impact on customer repurchase intention in the hospitality industry, (Guttentag, 2019). Our results depict that creation, which is the highest level of brand related activity on social media, has the greatest effect on customer repurchase intention. This suggests that customers who publish brand related weblog, upload brand related content, write articles and reviews about the brand will most likely have intentions to repurchase. The interactive feature of social media with its potential to create a platform for conversations among hospitality firms and clients has allowed customers to co-create services with firms. In line with Sashi (2012) our results indicate that customers who are involved in content generation and value creation are more likely to keep patronizing a service firm. Consumption and contribution were also found to significantly influence repurchase intention, however, consumption had a comparatively higher influence than contribution. In effect, engaged customers who consume a firm's social media content can serve as potential loyal customers, though consumption is the lowest form of engagement behavior according to the COBRA model (Muntinga et al., 2015).
The results of the fourth hypothesis (H4), suggests that customer satisfaction influences Repurchase Intention. The desire to make a repurchase is a reflection that expectations are met, also, past studies in the hospitality sector links revisit intentions and intentions to recommend to satisfaction with previous purchase (Liang et al., 2018). The results of mediation tests depict that contribution and consumption partially mediated the relationship between satisfaction and repurchase intention. This suggests that for contributing and consuming typologies of COBRA, repurchase intentions are to some extent dependent on the level of brand activity. However, creation did not mediate the relationship between customer satisfaction and repurchase intention. Thus, customer engagement partially mediates the relationship between satisfaction and repurchase intention (Boateng, 2019;Wang & Gon, 2017), however, this mediation effect is dependent on the type of engagement behavior.

Conclusion
The study sought to analyse the relationship between customer satisfaction, customer engagement and repurchase intentions. All hypothesized relationships were supported. Our results demonstrated that indeed customers' satisfactory experiences in the hospitality industry influenced their engagement behaviors. Customers who are satisfied with their experiences with a firm will likely involve in medium to low engagement behaviors like contribution and consumption respectively. Moreover, the results of the study also suggests that consumers' level of brand activeness and satisfaction is positively related to their repurchase intentions. Customers who show high level of brand activeness on social media are more likely to repurchase the brand, as our results demonstrated that customers who exhibit engagement behaviors like creation will likely repurchase from the firm. Also, contribution and consumption are mediate the effect of satisfaction on repurchase intentions

Theoretical contribution
Social media and COBRAs are emerging as important research domains in service consumer research, according to the service marketing. This study adds to the marketing literature by demonstrating how COBRAs influence consumers' intent to repurchase and engage in an ongoing search for a highinvolvement hotel service/product via social media. For starters, whereas previous research has looked at customer engagement as a set of three constructs known as COBRAs (Majeed et al., 2020;Schivinski et al., 2019), our research delves deeper into specific user experiences and their roles in the relationship between satisfactions and repurchase intention. This is a novel approach in which we learn that some experiences, such as consumption, contribution, and creation (Calder et al., 2009), may not be equally relevant in all circumstances. With the growing body of literature on customer engagement constructs, it's more important than ever to identify appropriate engagement experiences and investigate their role in various contexts. This study adds to the body of knowledge on customer satisfaction and repurchase intention by identifying customer engagement experiences that are relevant to hotel clients. Consumption, contribution, and creation all play important roles in the satisfaction-repurchase intention relationships, according to our findings. Customers with higher levels of social media engagement have a stronger effect of customer satisfaction on repurchase intention, according to our findings. Customer engagement (COBRAs), on the other hand, may not always play a consistent role in influencing the satisfaction-repeat purchase relationship. In line with Poyry et al. (2013) recognition of entertainment as a driver in strengthening consumers' intent to browse and participate on social media brand pages, our findings support the use of entertaining and interactive content in driving consumers' intention to consume, contribute, and create brand-related content. As a result, our findings support the importance of interaction in driving consumer engagement and value co-creation behaviors (Cheung et al., 2020a)in order to encourage repeat purchases.
The level of consumer involvement with the focal hotel service could be one reason for the discrepancy between our findings and the existing literature. As previously stated, Ghanaian consumers are very interested in hotel services and have developed a culture around them. As a result, they are motivated to look for information about hoteliers and stockpiling patterns on a regular basis. In other words, as highly engaged consumers who prefer to access more comprehensive information, they are more interested in learning from like-minded peers on social media platforms through interaction (Cheung et al., 2020) than in creating content. Furthermore, we believe that in encouraging consumers to read, share, and create online hotel brand-related content, consumption and contribution content is far more important than creation, especially for high-engage products.

Practical contribution
This study looks into the numerous aspects of customer engagement as motivators for consumers using social media. Furthermore, from a managerial standpoint, this study suggests that hotel marketers in Ghana should employ SM strategies. Marketers should also consider communicating brandrelated news and current hot topics on social media platforms to encourage consumers to create content. It emphasizes the value of using social media to strengthen consumers' intentions to participate in COBRAs by creating, contributing, and consuming SMM content. Individuals' concerns about engagement may have been predicted by recent social media trends that included contributing, consuming, and creating hotel content. The current study demonstrates and validates the significance of customer satisfaction concerns and COBRA attributes in driving hotel repurchase intent in Ghana. Despite the fact that CE and CS concerns were given higher priority in the current sample than other concerns, the correlation between CE and CS concerns suggests that vendors should treat the consumption, creation, and consumption features of their social media as a joint priority.
As a result, marketers should consider creating posts with photos, videos, and animation that consumers find entertaining and enjoyable, as this stimulates their desire to consume, contribute, and create content on social media pages for hotel brands. Marketers should consider including interactive content on SM pages, such as product usage discussions, makeup styles, and product comparisons, to encourage consumer-brand and peer-peer interaction. According to the theory, such interactive posts will pique customers' interest, prompting them to read, comment on, and share the posts, prompting them to upload content describing their hotel experience

Limitations and suggestions for future research
The testable model used in this study showed that the independent variables chosen (customer satisfaction) have a high level of explanatory power for repurchase intention. Other determinants of repurchase intention could be usefully added to the model in future research so that repurchase intention can be predicted more accurately. Furthermore, this study only looked at the mediation roles of COBRA dimensions; other mediators could be investigated and tested. This study was carried out in Ghana's northern region, but because Ghanaians have different cultural backgrounds depending on where they live, it would be beneficial to test samples from other parts of the country in the same experimental context. It is also suggested that future research be based on a larger sample size in order to obtain more precise results. In addition to service quality and satisfaction, future studies should include other factors and dimensions that can help predict consumer behavior, such as value, loyalty, and purchasing motivations.