The impact of positive emotional experiences on eWOM generation and loyalty Impacto de las experiencias emocionales positivas sobre la lealtad y la generaci on de eWOM

Purpose – The purpose of this paper is to examine the role that positive emotional experiences play in the generation of electronic word-of-mouth (eWOM) and the influence both variables have, together with customer satisfaction and brand reputation, on customer loyalty in the context of hospitality services. Design/methodology/approach – An online survey has been conducted including customers of an important hotel chain in the holiday up-market segment. The study sample is composed of 878 customers from Germany and the UK. The model has been tested using partial least squares technique. Findings – Results indicate that positive emotional experiences have a positive effect on satisfaction, eWOM generation and company reputation. This study identifies the provision of emotional experiences as a powerful generator of positive eWOM in the hospitality context and it states that customer satisfaction, in itself, does not guarantee positive eWOMgeneration. Practical implications – Designing unique, tailor-made, memorable experiences can become a key element to increase loyalty and, particularly, to foster the generation of recommendations, that is positive eWOM, in the hotel industry. Originality/value – The study makes a contribution on the antecedents of customer satisfaction, eWOM engagement and loyalty in the specific context of hotel services.


Introduction
The role of emotions has gained the interest of both academics and practitioners as a critical element in understanding consumption experiences and behavior (Han and Jeong, 2013;Mattila and Enz, 2002). Emotions are markers, mediators and moderators of consumer responses (Bagozzi et al., 1999). These influence information processing, choices, satisfaction, perceptions of service failure, word-of-mouth (WOM), complaining behavior and loyalty (Bagozzi et al., 1999;Han and Jeong, 2013;Joireman et al., 2013).
When we travel for leisure, we pursue basically experiences and emotions. And once we are back at home, we basically retain experiences and emotions in our minds. As Brunner-Sperdin et al. (2012) and Dong and Siu (2013) reported in previous research, cognitive models are limited in their ability to explain service encounter assessment. Service experience evaluation is both cognitive and emotional (Edvardsson, 2005;Han and Jeong, 2013), but particularly emotional in tourism.
This paper tries to make a contribution to the rich line of research about emotions in tourism by analyzing the role of positive emotional experiences on customer behavioral intentions, customer satisfaction and company reputation in the specific field of hotel services. When traveling, emotive experiences usually come from enjoying tourism destination attractions, either natural or built (natural scenarios and beautiful landscapes, historic and cultural heritage, urban atmosphere, etc.), and the interactions that take place with the local resident community and the travel peer group during the stay. But with competition becoming more and more intense, some hotel industry companies are exploring their ability and capability to provide emotional experiences and tailor-made products to their customers and using it as a clear differentiation element that could contribute to increasing customers' loyalty and, thus, becoming a source of competitive advantage. They are increasingly trying to sell unique experiences instead of simple commodities. Experiences are more difficult to replicate.
All these mentioned studies show that tourist emotional reactions are fundamental determinants of post-consumption behaviors. For example, Bigné et al. (2005), Grappi and Montanari (2011) and Yuksel and Yuksel (2007) examined emotions as an antecedent of satisfaction and behavioral intentions. Although tourism research acknowledges the importance of studying emotions, satisfaction and behavioral intentions, mixed support exists regarding the relationship among these three constructs (Bigné et al., 2005). For example, while most of the studies confirm a positive relationship between satisfactory tourist experiences and intention to recommend or positive WOM (Han and Ryu, 2012;Lii and Lee, 2012;Prayag et al., 2017;Tsao and Hsieh, 2012), other studies (Dolnicar et al., 2015;Swanson and Hsu, 2009;Yang, 2017) raise some concerns regarding the link between satisfaction and intention to recommend.
In addition, with the new century, new platforms have emerged conveying opinions and emotions felt during service encounters. TripAdvisor, Booking.com and Holidaycheck are some specific examples but, in general terms, all online travel agencies (Expedia, Trivago, etc.) include costumer opinions about their past experiences. It is the electronic word-of-mouth (eWOM), consumer-to-consumer information in form of opinions and reviews. Within tourism services in general, particularly in the hospitality industry, eWOM is of fundamental importance.
Due to the impact of digital technologies, traditional WOM is evolving toward eWOM. The emergence of eWOM raises new research questions. In particular, can WOM and eWOM be treated as equivalent concepts? In other words, are the links between some constructs and WOM, proven by previous research, also applicable to eWOM? When analyzing antecedents of intention to recommend, many researchers do not distinguish between traditional WOM and eWOM, whose fundamentals could be different.
Given the mixed and non-conclusive results mentioned before and the widespread of eWOM, this paper tries to shed light on the relationships among these constructs in the context of the hotel industry. As pointed out by some researchers (Filieri and McLeay, 2013;Litvin et al., 2008;Vermeulen and Seegers, 2009), there is a need for more research that explores the behavioral implications of eWOM on travelers. Tanford et al. (2016) also highlighted the need for research about the antecedents of loyalty through eWOM. In addition, there is also a need for a better understanding of the impact of positive emotional experiences on customer satisfaction and brand reputation in the hotel industry context. Finally, we think it is worth analyzing the role emotional experiences could play as an antecedent of positive eWOM generation.
Thus, the main objective of this paper is to test a conceptual framework, a holistic model, to measure the impact of positive emotional experiences provided in hotel service contexts on customer satisfaction and brand reputation and the role emotional experiences play as a review-generating factor. In addition, the model tries to measure the influence of all these aforementioned constructs on "customer loyalty." Finally, the role customer satisfaction can play as an antecedent of eWOM generation and its impact on brand reputation are also analyzed, as well as the potential relationship between eWOM and brand reputation.
To provide additional information useful for practitioners, the model tries to check whether nationality moderates these relationships between the provision of "emotional experiences" in the hotel service context and customer satisfaction, brand reputation and eWOM generation, considering the two most important tourism outbound markets in Europe (Germany and the UK).

Theoretical background and research hypothesis
Among the academic and professional fields, interest is growing in identifying the factors that influence customer loyalty, with the aim of developing the most appropriate marketing strategies. Customer loyalty has been defined in many different ways. Although the costumer's intention to repeat purchase selection is central to all definitions, as Kabiraj and Shanmugan (2011) pointed out, loyalty is a very complex multidimensional concept, influenced by a large set of variables: attitudinal, emotional, psychological. . . The literature examines loyalty from different perspectives, but the most relevant ones are behavioral and attitudinal (Dick and Basu, 1994;Gounaris and Stathakopoulos, 2004;Pic on et al., 2014). The behavioral perspective refers to a customer's behavior on repeat purchases, indicating a preference for a brand or a service over time (Bowen and Shoemaker, 1998;Wilkins et al., 2010). The attitudinal perspective refers to a customer's intention to repurchase and recommend (Caruana, 2004;Getty and Thompson, 1995;Kandampully and Suhartanto, 2000;Oliver, 1999).
The attitudinal approach depicts loyalty as a psychological state that the customer may attain as a result of the relationship with the firm. The customer may voluntarily maintain that relationship on the basis of the benefits he receives during the relationship (Caruana, 2004). This approach advocates the idea that the foundation of true loyalty is a positive attitude toward the firm (Pic on et al., 2014). A positive attitude toward the provider is precisely what ensures a repeat purchasing behavior in the future. The importance of customer's loyalty to long-term company's sustainability is widely acknowledged. As Callarisa et al. (2012) emphasized, building and maintaining customer brand loyalty contribute to the establishment of a sustainable competitive advantage. For the purpose of this study, the behavioral approach is assumed: the customer intends to repurchase from the same firm whenever possible in the future, recommends to other people and maintains a positive attitude towards the firm.
Customer satisfaction is a comparison between the expectation of the level of service (before purchase) and the perceived service (after the purchase or consumption) (Parasuraman, 1997;Ravald and Grönroos, 1996). Satisfaction is seen as an antecedent of loyalty.
On the other hand, image plays a critical role in tourism. Grönroos (1990) conceptualized corporate image as the result of how the customer perceives the firm; in other words, corporate image is the customers' mental representation of the firm, their vision of the company. Corporate and brand reputation is one dimension of this image. Reputation is judged within the context of competitive offerings (Fombrun and Van Riel, 2002;Schultz et al, 2006). A company with a good reputation is likely to be perceived by customers as being more trustworthy and credible as opposed to one with a poor reputation. Companies and brands with a good reputation are likely to attract more customers and retain their current ones. Reputation can reduce perceived risk, which is shown to be particularly high in the tourism decision-making process (Loureiro and Kastenholz, 2011). Hence, reputation is a core intangible asset of the firm and creates barriers to competitive threats. Considering that the sample used in this research are customers from a large international hotel chain operating in the holiday segment using single brand strategy (the company name is also the brand name), reputation means, for the purpose of this study, "brand reputation." Brand reputation is a "customer-centric" concept and refers to how a particular brand is viewed and valued by consumers. A favorable brand reputation means consumers trust your company and feel good about purchasing your goods or services over time.
An impressive amount of research supports the link between customer satisfaction, brand reputation and loyalty. In summary, research demonstrates that customer satisfaction has a direct and positive impact on customer loyalty (Callarisa et al., 2012;Da Silva and Alwi, 2008;Loureiro and Kastenholz, 2011;Tu et al., 2012;Wilkins et al., 2010); customer satisfaction is linked with brand reputation and the later has a positive effect on customer's loyalty (Da Silva and Alwi, 2008;Loureiro and Kastenholz, 2011;Wilkins et al., 2010;Wu, 2011). In the specific context of the hospitality industry, Wang (2010) identified corporate and brand image as potential determinants of customer loyalty and Martínez and Rodríguez del Bosque (2015) identified customer trust as a key antecedent of hotel customer loyalty.
Thus, to explore the relative weights of these variables in comparison with the rest of variables analyzed in the model, the following hypotheses are considered: H1. Customer satisfaction has a positive direct effect on reputation.
H2. Customer satisfaction has a positive direct effect on loyalty.
H3. Reputation has a positive direct effect on loyalty.
Tourist experiences include satisfying and pleasurable emotions (Aho, 2001;Coghlan et al., 2012). Positive emotions are associated with memorable experiences (Tung and Ritchie, 2011). However, the impact of emotional content potentially derived from hotel guest experience remains under-researched.
As Xiang et al. (2015) stated, the guest experience is, undoubtedly, an extremely complex construct. Depending on the research design and methods, researchers could get very different pictures of what constitutes customer or guest experience in tourism services. Customer experience can be defined as the entirety of the interactions a customer has with a company and its products. The overall experience reflects how the customer feels about the company and its offerings. Throughout these processes, customer experiences a lot of perceptions about the hotel company but, obviously, most of them cannot be categorized as "emotional experiences." Positive "emotional experiences" in the hotel context could be defined as service encounters that far exceed the customer's expectations (i.e. what one experienced traveler can expect to find in a four-star hotel for example), and are flawlessly executed; the design and delivery of customized, innovative and unique experiences are difficult to find "on average"; surprising and/or unexpected "levels" of service, a touch of additional differentiation. In summary, "emotional experiences" refer to emotionally engaging offerings in hotel service encounters with the ability to generate strong positive emotions and memories; "emotional experiences" imply exceeding expectations and being different "to the average" in terms of global offering, entertainment, pleasure, stimulation, excitement, surprise, curiosity, passionate, romantic, energizing, trust generating, etc. Offerings one was not expecting that make, just for themselves, the stay in a particular hotel worthwhile. It is important to highlight, to clarify the results, that only positive feelings are considered in this work.
As emotions have gained academic interest, a remarkable amount of research supports the relationship between emotions displayed during service encounters, service evaluation and future behavioral intentions. Prior studies in marketing (Ladhari, 2007;Walsh et al., 2011) and tourism (Bigné et al., 2005;Grappi and Montanari, 2011;Han and Jeong, 2013;Yuksel and Yuksel, 2007) establish a positive path between positive emotions, satisfaction and behavioral intentions. More specifically, the link between positive emotions and customer satisfaction is confirmed in several research papers (Ali et al., 2016;Bigné et al., 2008;Hosany et al., 2017;Rodríguez del Bosque and San Martín, 2008). However, in a recent study, Prayag et al. (2017) obtained mixed results when measuring the link between positive emotions and satisfaction in a tourism destination context. Interestingly, the positive surprise was the only emotional aspect showing a positive link to customer satisfaction.
When Berry (2000) examined branding in consumer service settings, and what was found to be more important was the brand's "meaning" that the customers derived from the service experiences they had. Therefore, it can be hypothesized that the delivering of emotionally engaging offerings in hotel service encounters with the ability to generate strong positive emotions and memories should have a positive impact on reputation.
The above discussions frame the following hypotheses: H4. Positive emotional experiences have a positive direct effect on customer satisfaction.
H5. Positive emotional experiences have a positive direct effect on reputation.
eWOM, also often referred to as online reviews, online recommendations or online opinions, has gained importance with the emergence of new technology tools. Litvin et al. (2008) defined eWOM as "all informal communications directed at consumers through Internetbased technology related to the usage or characteristics of particular goods and services, or their sellers." Traditional WOM has evolved into eWOM in the new digital era. The main differences between WOM and eWOM can be identified in the reach of the reviews' impact (number of people who can be influenced) and the speed of interaction. With regard to this comparison, Sun et al. (2006) concluded that "compared to traditional WOM, online WOM is more influential due to its speed, convenience, one-to-many reach, and its absence of face-toface human pressure." This breadth of eWOM scope and ease in accessing reviews can deeply affect a company's performance. Therefore, companies are increasingly seeking to understand the factors that influence the use of eWOM, as well as the impacts resulting from its use. The tourism industry is strongly affected by eWOM and, within the tourism industry, hotels are probably the most affected. Serra and Salvi (2014) classified articles analyzing eWOM and hotels into two general lines of research: on the one hand, the factors related to the generation of comments; and on the other hand, the impacts these comments have on consumers and on company perspective. In relation to review-generating factors, most of the analyzed studies highlight aspects such as "service quality and satisfaction," "failure and recovery," "customer dissatisfaction" and "sense of community belonging" as consumers' main motivations for writing reviews (Kim et al., 2009;Nusair et al., 2011;Sánchez and Currás, 2011;Qu, 2011, Swanson andHsu, 2009). Therefore, most of the previous studies confirm a positive relationship between satisfactory tourist experiences and intention to recommend or positive WOM (Lii and Lee, 2012;Prayag et al., 2017;Tsao and Hsieh, 2012). But results are not conclusive, yet some studies seem to indicate that customer satisfaction does not necessarily lead to eWOM generation. For instance, Swanson and Hsu (2009) reached to the conclusion that customers who experience satisfactory incidents are not necessarily more likely to recommend the service provider or to convince others to use the service provider's offerings than their dissatisfied counterparts. In the context of restaurant services, Yang (2017) did not find support for the link between satisfaction and positive eWOM intentions. Dolnicar et al. (2015), in their meta-analysis of methodological aspects of guest surveys in the context of tourism destinations, also raised some concerns regarding the link between satisfaction and intention to recommend. Therefore, the relationship between customer satisfaction and eWOM generation deserves additional research effort and, particularly, in the hotel service context where no research has been carried out. In addition, when analyzing antecedents of intention to recommend, researchers normally do not distinguish between traditional WOM and eWOM, whose fundamentals could be different from traditional WOM. Given the previous reasoning, the following hypothesis is suggested: H6. Customer satisfaction has a positive direct effect on electronic word-of-mouth generation.

Positive emotional experiences
When considering the impacts of eWOM from the consumer perspective, it has been observed that, to a greater or lesser extent, all the studies consider the influence of reviews (WOM or eWOM) in the decision-making process (Serra and Salvi, 2014). Xie et al. (2011) argued that eWOM is prevalent in today's lodging market and it has potential to influence consumers' decision-making. Litvin et al. (2008), in the same line, pointed out that interpersonal influence and WOM are ranked as the most important information source when a consumer is making a purchase decision. These influences are especially important in the hospitality and tourism industry, whose intangible products are difficult to be evaluated prior to their consumption. Ye et al. (2009) in their study showed that positive online reviews can significantly increase the number of bookings in a hotel. The studies indicated that eWOM can influence loyalty. According to Worthington et al. (2010), loyal customers have higher propensity to express positive eWOM and brand advocacy. But the reverse direction on this relationship has not been analyzed yet: that is, could eWOM be considered as an antecedent of consumer brand loyalty? In other words: could the fact that a customer is willing to invest time and effort in spreading positive opinions online be considered as a precursor of a higher propensity to loyalty in the future? On the other hand, it is acknowledged that positive comments have a positive impact on brand reputation (Yacouel and Fleischer, 2012), thus aiding the company to capture new customers and increase the propensity of these new customers to become loyal. But, it could also be argued that a customer's intention to spread positive eWOM should have a positive impact on his/her perception of the reputation of the brand.
Thus, based on the previous discussion, the following hypotheses are proposed: H7. Positive electronic word-of-mouth has a positive direct effect on loyalty.
H8. Positive electronic word-of-mouth has a positive direct effect on reputation.
Some previous research confirms the expected relationship between positive emotions and willingness to recommend (Jang and Namkung, 2009;Lee et al., 2008) and WOM (Ladhari, 2007), though not in the hospitality context. Hospitality companies acknowledge that their existence and growth depend on their ability to create unique, memorable and positive experiences for customers (Walls et al., 2011). In the hotel service context, emotions derived from the provision of unique, tailored and unforgettable experiences could be a means to create brand ambassadors and co-creators of value through positive recommendation. Therefore, the provision of emotional experiences to hotel guests may constitute a powerful review-generating factor in hospitality services. While the topic of eWOM has drawn much attention in the academy and industry, little research has been published addressing the factors influencing eWOM intentions (Yang, 2017). In an attempt to cover this lack of research, the next hypothesis is proposed: H9. Positive emotional experiences have a positive direct effect on electronic word-ofmouth generation.
On the basis of the increasing importance of online reputation derived from eWOM and the impact both elements can have on customer's loyalty, in this paper, we design a model that considers these new variables that are strongly affecting consumers' buying behavior in the hospitality market ( Figure 1). But consumer responses to the same stimulus can be different from one geographic market to the other (Crouch, 1994;De Mello et al., 2002). It can be useful for future academic research as well as for practitioners to know whether results can be generalized. Therefore, nationality is also introduced in the model to check whether this variable introduces any significant behavioral differences. The German and British markets have been selected as they are the most important tourism-generating markets in Europe (World Tourism Organization, 2017): H10. Nationality moderates the effects of positive emotional experiences.
H10.1. Nationality moderates the relationship between positive emotional experiences and customer satisfaction.
H10.2. Nationality moderates the relationship between positive emotional experiences and electronic word-of-mouth generation.
H10.3. Nationality moderates the relationship between positive emotional experiences and reputation.

Methodology
To test the hypotheses, a questionnaire has been designed containing some items already used in previous studies and others from our own elaboration. The measurement items of constructs for the variables included in the causal model have been adapted from previous research: emotional experience (Bigné et al., 2008;Brunner-Sperdin et al., 2012;Han and Jeong, 2013), satisfaction (Bigné et al., 2008;Loureiro and Kastenholz, 2011;Wilkins et al., 2010), eWOM (Kim et al., 2009;Sun et al., 2006), brand reputation (Loureiro and Kastenholz, 2011) and loyalty (Bigné et al., 2008;Loureiro and Kastenholz, 2011;Wilkins et al., 2010). To measure the items, a five-point Likert scale has been used. The questionnaire was completed with questions dealing with demographic aspects (sex, age, country of residence, etc.) and Positive emotional experiences trip information (travel period, length of stay, information sources used, booking method, etc.). Before engaging in full field research, the questionnaire was previously tested to detect the potential weaknesses and problems of understanding. The questionnaire's languages were English and German.
British and German customers have been selected from an important Spanish hotel chain, which operates over 100 hotels, four and five stars, in Europe and America, and targeted to the holiday segment of the tourism market. The reason for selecting these two nationalities is because they constitute the two most important tourism-generating markets in Europe. Interviews have taken place from April to June 2014. Customers who have stayed in any of the hotel chain's establishments all over the world in the previous three months (January, February and March 2014) were selected. Data have been obtained through an online survey (Table I). A total sample of 878 valid questionnaires was obtained, 403 from British customers (sampling error of 4.98 per cent) and 475 from German customers (sampling error of 4.59 per cent). The research model has been tested using the partial least squares (PLS) technique, a variance-based structural equation modeling (SEM) method and complemented using SPSS 20. The PLS technique is especially suitable for predictive research (Loureiro and Kastenholz, 2011) and theoretical developments (exploratory analysis). More precisely, this study uses SmartPLS 2.0 software for the PLS analysis (Ringle et al., 2005) to process two causal analyses (one for each nationality), composed by five reflective constructs and the multigroup analysis necessary to contrast the hypotheses dealing with the customer's nationality moderator effect.

Results
Before proceeding to the analysis of the proposed structural models, it is necessary to previously analyze the measurement model. Items that do not meet criteria in both models should be removed because the same scale of measurement must be used in both models to avoid errors when making comparisons. Furthermore, the usual analysis applied to the scale of measurement must be complemented with the invariance analysis. First, the reliability and convergent validity should be analyzed. A factor loading minor than 0.707 on its respective construct (Carmines and Zeller, 1979) forced to eliminate the variable "I have received positive recommendations concerning the hotel," which was part of the "reputation" construct. Construct reliability is usually assessed using Cronbach's alpha (Cronbach, 1970) and composite reliability (Anderson and Gerbing, 1988;Bagozzi and Yi, 1988;Werts et al., 1974). When both values are greater than 0.7, it is usually considered as a valid reliability criteria (Nunnally, 1987;Santesmases, 2009). It can be observed in Table II that values for Cronbach's alpha and composite reliability are systematically greater than 0.78 in both models, which indicates the construct reliability. In addition, the average variance extracted (AVE) values should be greater than 0.50 (Baggozi and Yi, 1988;Fornell and Larcker, 1981), and AVE values for all latent variables are greater than 0.67 in both cases (Table II).
There are two approaches to assess discriminant validity (Anderson and Gerbing, 1988) in PLS: no item should load higher on another construct than it does on the construct it intends to measure and each construct should load higher with its assigned indicators than other items; the square root of the AVE of each latent variable should be greater than its correlations with any other latent variable in the assessment (Chin, 1998). Problems with Positive emotional experiences discriminant validity forced to remove from the final measurement model the following items: "I am excited about the choice of this hotel," from the "experience" construct, and "This hotel inspires trust," from the "reputation" construct. Analysis of the discriminant validity of the final measurement model appears in Table III.
Finally, an invariance analysis must be applied to the constructs to be sure that weights in both models do not show significant differences. Invariance problems forced to remove the following items: "I consider this brand my first choice when selecting a hotel" and "I will probably return to this hotel chain" which were part of the "loyalty" construct. Table IV illustrates the final measurement model proposed.
After testing the measurement model, we proceeded to the analysis of the structural model and the proposed causal relationships (Figure 2). PLS-SEM does not assume that the data are normally distributed, which implies that parametric significance tests cannot be applied to test whether coefficients such as outer loadings and path coefficients are significant. Instead, PLS-SEM relies on a nonparametric bootstrap procedure (Davison and Hinkley, 1997;Efron and Tibshirani, 1993) to test the significance of estimated path coefficients. As the proposed moderator variable is a categorical one, it has been necessary to carry out a multigroup analysis to test H10 to determine whether differences between path coefficients of both samples are statistically significant. The moderating effect is examined using a t-test with pooled standard errors (Table V).
Results demonstrate that positive emotional guest experiences have a strong and very powerful positive effect on customer satisfaction (H4); these also have a positive effect on intention to recommend, that is, eWOM generation (H9) and on company's reputation (H5). Their impact on reputation, although positive, is not so substantial. These impacts are higher in the UK market when compared with the German market, but the differences are not significantly different. Thus, the impact of emotional experiences on customers' postpurchase behavior is, without a doubt, significant.
Customer satisfaction has a positive impact on reputation (H1) and its positive impact on customer loyalty is really strong (H2). This can be considered an expected outcome as it is in line with results of previous research (Callarisa et al., 2012;Da Silva Loureiro and Kastenholz, 2011;Tu et al., 2012;Wilkins et al., 2010;Wu, 2011). Results also prove the mediating effects of customer satisfaction on the relationship between positive emotional experiences and loyalty (Table VI). This mediating effect is fairly strong.
But, interestingly, there is no significant effect of customer satisfaction on intention to recommend online (eWOM). Therefore, the results do not support H6. This is, in our opinion, an important finding. On the other hand, intention to recommend or make comments online (eWOM generation) is positively linked with Positive emotional experiences reputation (H8) and loyalty (H7). Results also support the hypotheses that reputation has a positive effect on customers' loyalty (H3). Better hotel chain's reputation leads to higher levels of customer loyalty. Therefore, both results confirm findings from previous research.
Finally, regarding potential differences in consumer behavior between the German and British tourism markets, no significant differences between them can be found in the structural proposed model (Table V), particularly regarding the emotional guest experience (H10). Therefore, the structural model analyzed could be of application to both nationalities and, possibly, to the rest of western European tourism markets.

Conclusions and managerial implications
In recent years, academics have shown an increased interest in studying and understanding the affective dimensions of tourists' experiences (Morgan et al., 2010;Ritchie and Hudson, 2009;Tung and Ritchie, 2011;Wang and Pizam, 2011). Intensifying competition requires marketers to appreciate the value of emotional experiences potentially derived from their tourism offerings. Tourists seek pleasurable and memorable experiences through the consumption of their vacations. Despite becoming a recurrent issue in hospitality professional debates, relatively few studies investigate tourists' responses toward emotional experiences lived in the hotel context. This study has proven that positive emotional experiences are a powerful determinant of behavioral intentions (eWOM and loyalty) in the hotel service context. The positive impact on loyalty through the mediating effect of satisfaction is particularly remarkable.
Results confirm the positive link between customer satisfaction, company reputation and positive eWOM generation on customer loyalty being customer satisfaction, by far, the most influential element on loyalty. However, it is worth mentioning that the intensity of the  Positive emotional experiences relationship between positive eWOM generation and loyalty is the weakest one among the three aforementioned. Therefore, it can be argued that customer satisfaction (in turn strongly affected by positive emotional experiences felt during hotel service encounters) really has a strong positive impact on loyalty, whereas the intention to spread positive comments (also strongly affected by positive emotional experiences) is not exercising such a strong influence on loyalty.
Our model supports previous research findings regarding the positive link existing between customer satisfaction and customer loyalty and brand reputation; it also supports the positive relationship between reputation and loyalty and the link between eWOM and brand reputation and loyalty. Our research shows that providing emotional experiences to customers during their stay is a key factor for customer satisfaction in hospitality services. In this sense, results are consistent with prior studies in marketing (Dubé and Menon, 2000;Nyer, 1997;Walsh et al., 2011) andtourism (Faullant et al., 2011;Grappi and Montanari, 2011;Lee et al., 2008;Rodríguez del Bosque and San Martín, 2008;Yuksel and Yuksel, 2007), proving that emotions are strong predictors of satisfaction, and are in line with the results obtained by Prayag et al. (2017) in a tourism destination context.
The provision of emotional experiences also has a positive impact on the company's reputation, eWOM generation and, finally, on customer loyalty. Thus, this paper highlights the importance of offering unique, unexpected and tailor-made "emotional experiences" to customers in hotel services due to their proven power as an eWOM generator and their strong positive impact on customer satisfaction, not to mention the very positive impact on brand reputation.
Customer satisfaction, in itself, does not guarantee eWOM generation. This result confirms previous research findings of Swanson andHsu (2009) andYang (2017). Interestingly, results suggest that customer satisfaction is a necessary but not a sufficient condition to guaranty the generation of positive eWOM in hospitality. What really has a powerful impact or influence on intention to recommend online (eWOM) is the provision of unique emotional experiences to hotel customers. These are, probably, the most interesting findings of the present study. Although it is widely accepted that customer satisfaction leads to positive traditional WOM, our results seem to indicate that consumer behavior could be different in online settings.
The generation of positive eWOM is extremely important in hotel services, not only by the positive effect on the company's reputation but also because it facilitates to attract new customers in a context in which the vast majority of consumers express that they do not book a hotel without having consulted the reviews comments from previous hotel Notes: *p < 0.05; **p < 0.01; ***p < 0.001; ns , not significant Source: Own elaboration customers. Therefore, hotel managers must explore and look for ways to foster the generation of positive comments. As customer satisfaction, in itself, does not guaranty the generation of positive eWOM, this study indicates that the provision of emotional experiences to customers during their stay is one method to increase the likelihood of positive comments through the different online platforms.
In terms of service design, the greater the number of reviews from customers, the greater the amount of relevant information at the company's disposal to continuously improve and innovate its product/service offering. If reviews are positive, they positively affect the company's reputation, which favors customers' loyalty and affects positively on brand equity. Thus, a "virtuous circle" is created. If marketing departments have the ability to properly manage the large amount of data and information available coming from different sources like social networks, review websites, booking profiles, etc., they can use this relevant information to design tailor-made services better adapted to the specific needs and customer's requirements and, thus, improve the customer experience: from the process of information search, through the stay at the hotel till customer follow-up. Some hotel industry managers highlight a growing trend toward tailor-made offerings and the information generated by the client facilitates this process. Yacouel and Fleischer (2012) and Anderson (2012) proved that positive eWOM reviews have an impact on price and hotel occupancy levels. If hotel management has the ability to properly manage emotional guest experiences, and positive reviews are generated placing the hotel at the top positions in online reputation rankings, revenue managers could set premium prices without affecting occupancy rates. Furthermore, online reputation is a key factor to succeed in hotel meta-searchers, as most of them show customers reviews. In practical terms, this means that although many resources are invested in other marketing aspects, such as content, corporative image and sales promotions, any of these elements can be distorted or have little credibility if the hotel fails to complement them with positive comments from customers. This study shows that the best way companies have at their disposal to generate positive eWOM is the provision of emotional experiences to their customers.
From the results of this study, hotel managers also know that there is no significant difference between the two most important European tourism-generating markets in terms of the impacts of emotional guest experiences on eWOM generation, customer satisfaction and company reputation. Results could probably be generalized to eastern European markets.

Limitations of the study and future research
The research model has been tested using SmartPLS 2.0 instead of the latest available version of this software (SmartPLS 3.0). This may be considered as a limitation of this study as SmartPLS 3.0 provides more current indices and procedures than the 2.0 version.
The sample used in this study is quite large. However, to generalize the results, it has to be considered that respondents were customers of upscale hotels and most of them were of age over 50 years. Thus, respondents were experienced travelers. This fact could partially explain their reluctance to get involved in spreading positive opinions online unless emotional elements of positive surprise were delivered during their stay. It would be interesting to analyze whether these effects remain the same when comparing upscale with low-middle scale hotels. On the other hand, it could be worth checking the moderating effect of nationality on these relationships in other business settings.
Further research should be devoted to improve the measurement scales both in regard to emotional experiences and eWOM. In addition, it would be of significant interest to include Positive emotional experiences the concept of "quality of service" in the model to assess simultaneously the relative importance of quality in comparison with positive emotional experiences and the impact of both concepts on satisfaction, eWOM, reputation and loyalty. Finally, in future research, it would be interesting to test the direct relationship between positive emotional experiences and loyalty in the hotel sector.