The hotel management of e-wom. a cross-cultural analysis Spain-Italy

7 IJISEBC, 6(2), 2019 International Journal of Information Systems and Software Engineering for Big Companies (IJISEBC) the hotel management of e-wom. a cross-cultural analysis Spain-Italy la gestión hotelera de e-wom. un análisis intercultural España-Italia Carmen Berné-manero1, andreea v. Ciobanu1, mercedes marzonavarro1, marta pedraja-Iglesias1 1university of Zaragoza, Spain [email protected] , [email protected] , [email protected] , [email protected] aBStraCt. a lack of research in intercultural comparisons has been identified from the national and industrial perspective, and under the perceptions of those who make decisions about hotel services, particularly those related to the use of technologies. thus, the specific objective of this study is to determine if the Spanish and Italian hotel industries converge or diverge in the process of acceptance, use and management of e-wom. with this aim, the e-wIp model is tested through a multi-group analysis (mga) with partial least Squares-Structural Equation modelling (plS-SEm). then, a Hierarchical Cluster is carried out. the results announce more convergence than divergence and other collateral findings. Several useful conclusions are provided. rESumEn. Se ha observado una falta investigación en análisis comparativos interculturales desde la perspectiva nacional e industrial y desde las percepciones de los decisores sobre los servicios hoteleros, particularmente los relacionados con el uso de las tecnologías. así, el objetivo de este estudio es averiguar si las industrias hoteleras de España e Italia convergen o divergen en el proceso de aceptación, uso y gestión del boca-oído electrónico. Con este fin, se contrasta el modelo e-wIp a través de análisis multi-grupo (amg) con plS-SEm (partial least Squares-Structural Equation modelling). un análisis Cluster Jerárquico complementa el estudio. los resultados informan de una mayor convergencia que divergencia así como otra información colateral, lo que conduce a proporcionar conclusiones de utilidad. KEywordS: análisis multi-grupo, Boca-oído electrónico, Cluster jerárquico, Cultura, gestión, Hoteles. palaBraS ClavE: Culture, Electronic word-of-mouth, Hierarchical cluster, Hotels, management, multi-group analysis. Berné-manero, C.; Ciobanu, a. v.; marzo-navarro, m.; pedraja-Iglesias, m. (2019). the hotel management of e-wom. a cross-cultural analysis Spain-Italy. International Journal of Information Systems and Software Engineering for Big Companies (IJISEBC) , 6(2), 7-18. www.ijisebc.com recibido: 19-06-2019 / revisado: 21-08-2019 aceptado: 21-08-2019 / publicado: 15-12-2019 8 IJISEBC, 6(2), 2019 1. Introduction In the XXI century, cross-cultural comparison became quite a common approach to identifying the differences between national cultures or between various types of culture (Chen, Cheung & low, 2012). Culture is a collective programing of the mind that distinguishes the members of one organization from another (Hofstede, 2001). the values, beliefs, norms and behavioural patterns of a national group are an illustration of culture at the national level (yaprak, 2008). the concept of culture is used by academic researchers and industry practitioners to identify and to understand the issues that may arise in various areas of the social sciences and to develop approaches to overcome problems; at the national, industrial, occupational, corporate and organizational levels (Chen, Cheung & low, 2012). actually, the recognized industrial level of culture (pizam, pine, mok & Shin, 1997) has to consider that no single industry is homogeneous, especially hospitality industry, due to its diversity and complexity (Jones, 1999). In this sense, there is a lack of intercultural comparisons research regarding the national and industrial perspective and the perceptions of decision makers about the hotel service, particularly about the use of technologies. moreover, research using techniques such as discriminant analysis and structural equation modelling (SEm) are rarely covered in the literature as well as little information is available when probing online marketing in relation to culture (Chen, Cheung & low, 2012). authors like Cantallops and Salvi (2014), Confente (2015), Schuckert et al. (2015), and Baka (2016) calls for more research aimed at fully incorporating the e-wom, communication based on technology, into business strategy. thus, more recently, e-wIp model (Berne, Ciobanu & pedraja, 2019), which taking advantage of the Behavioural reasoned theory (Brt) (westaby, probst & lee, 2010), explains the hotel managers’ decision-making process in relation to accepting and implementing e-wom and its capacity to influence hotel performance. the e-wIp model confirms the existence of cause-effect relationships between variables such as reasons, motives, future intentions of use, ewom context, and performance, in the case of Spanish hotel industry. from these findings, the specific goal of this study is to determine if there are differences between the Spanish hotel industry and the Italian hotel industry in modelling e-wom use process and its consequences in hotel performance. with this aim, this paper presents a review of the literature about culture and e-wom hotel management and formulates a research question. then, it explains the databases used to analyse differences between Spanish and Italian hotels, and the results obtained after using multi-group analysis (mga) with plS (partial least Squares)-SEm, and Hierarchical Cluster. 2. Culture and hotel management of e-wom different cultural contexts can offer different ways to make decisions. although they are criticized as too general and simplifying, and unable to predict individual behaviours, the cultural dimensions identified by Hofstede (1980): power distance, uncertainty avoidance, individualism-collectivism, masculinity (/femininity), and later, long term pragmatism, can be considered as a good framework for cross-cultural research (Jansenverbeke, 1996). power distance is the extent to which the less powerful members of the organization within a country expect and accept an unequally distribution of power. the greater power distance, greater the inequality. uncertainty avoidance is determined by the extent to which the members of a given culture feel threatened by uncertain and unknown situations. Individualism refers to societies where ties between individuals are loose (/collectivism refers to the dependence of the group to which you belong). the dimension of masculinity points to societies where there exists a clear distinction between the social roles of men and women with the dominant male sex role pattern in the vast majority of societies. long-term pragmatism refers to the importance given in a culture to the planning of life in contrast to immediate concerns (Hosfstede, 2001). Some research focuses in validate a measurement scale of these five cultural dimensions (e.g. yoo, donthy & lenartowicz, 2011). Jansen-verbeke (1996) examined the differences and commonalities in hotel management practices of Belgium and the netherlands, which were linked to the differences in national culture. adding the pizams’ (1993) data, among the cultural dimensions in 13 countries, there were found considerable differences Berné-manero, C.; Ciobanu, a. v.; marzo-navarro, m.; pedraja-Iglesias, m. (2019). the hotel management of e-wom. a cross-cultural analysis Spain-Italy. International Journal of Information Systems and Software Engineering for Big Companies (IJISEBC) , 6(2), 7-18. www.ijisebc.com 9 IJISEBC, 6(2), 2019 between the managers in East and west. It is remarkable that Spain can be considered as a collectivist country, while Italy is six position closer to individualism; seven positions separate both countries in the case of uncertainty avoidance dimension. regarding power distance and masculinity, the positioning is similar. they are two countries that share the South European culture, an important part of their history and that are geographically located very near each other, being both are peninsulas. moreover, with the influence of globalization on people’s travel habits and preferences, and on employees’ working attitudes and behaviour, the hotel industry has been affected in many different ways, so the direction for culture research should shift toward understanding the effects of such globalization (Chen, Cheung & low, 2012). globalization could justify similarities between national and industrial cultures, as Spanish and Italian. nevertheless, as it has been observed, there are some differences between them according Hofstende’s cultural dimensions. on the other hand, since hotels are increasingly adopting the use of Information and Communication technologies, their processes are more efficient. However, problems may arise such as attitudes toward the change to high-tech systems and the requirement for more advanced skills. Jones (1999) affirms that there are structural differences between the hospitality industries of different countries. as a result, problems such as local hotel owners’ lack of knowledge of hotel operations may arise for individual hotel groups when applying managerial practices to their operations in different countries (Chen, Cheung & low, 2012; Infante-moro, Infante-moro, martínez-lópez & garcía-ordaz, 2016). the concept of e-wom includes any comments, positive or negative, made by a potential, current or former consumer about a product or company through the use of online media, whether before, during, or after the experience (Hennig-thurau, gwinner, walsh & gremler, 2004). Companies can adapt their business models taking advantage of the online reviews management in order to offer added value (Sigala 2013). most of the prior research uses secondary data, and is aimed at exploring the influence of ewom on consumers selecting a tourism destination and testing the theory of planned behaviour (e.g. Cheng, lam & Hsu, 2006; Jalilvand, Samiei, dini & manzari, 2012; Jalilvand & Samiei, 2012; di pietro, 2012; albarq, 2013; miao, 2015). However, only a few papers empirically investigate managerial responses to online reviews and the relationship between e-wom and business performance (ye, law & gu, 2009; ye, law, gu & Chen, 2011; anderson, 2012; litvin & Hoffman, 2012; park & allen, 2013; Sun, ayoun & Calhoun, 2013). all these contributions see e-wom as online interpersonal influence, informal communications directed at other consumers through Internet-based technology related to the usage or characteristics of particular goods and services and their sellers, and able to enhance the business activity. Kim et al. (2015) demonstrate that online review management leads to a performance increase. However, Xie et al. (2017) find that timely and lengthy responses enhance future financial performance. In any event, managing e-wom can improve hotel performance (phillips, Zigan, Santos Silva & Schegg, 2015; Kim, lim & Brymer, 2015; Xie, Zhang, Zhang, Singh & lee, 2016). In this line, e-wIp model provides an explanation on the way that hotel decision makers process and implement the ewom; the model even confirms the potential of this tool to influence business performance (Berne, Ciobanu & pedraja, 2019). Based on Brt (westaby, probst & lee, 2010), this model includes relationships between reasons (prior conditions favouring or impeding the use of ewom as a management tool), global motives (to develop the system), intentions to continue using the tool, and performance gains derived from changes implemented through e-wom management. It also includes a variable for the decision-making context: e-wom characteristics as seen from the decision-maker perspective (figure 1). the e-wom variable is reflected into three dimensions: quality, authority and credibility. according to rieh (2002), at an operational level, quality is identified as the extent to which users consider the information useful, good, current, and accurate; authority (cognitive) is operationalized as the extent to which users think Berné-manero, C.; Ciobanu, a. v.; marzo-navarro, m.; pedraja-Iglesias, m. (2019). the hotel management of e-wom. a cross-cultural analysis Spain-Italy. International Journal of Information Systems and Software Engineering for Big Companies (IJISEBC) , 6(2), 7-18. www.ijisebc.com 10 IJISEBC, 6(2), 2019 they can rely on the information. In the online context, the credibility of reviews influences beliefs, attitudes and intentions and is defined as an individual assessment of whether information is reliable and trustworthy, according to the receiver’s own expertise and knowledge (rieh, 2010; lu & Stepchenkova, 2012). attitude, subjective norms, and perceived control over the use of the information are three reflective criteria of global motives. while attitude represents a person’s global positive or negative evaluation toward doing the behaviour, subjective norms assess a person’s global perceived social pressure from important others to engage in the behaviour (westaby, 2005). perceived control reflects both past experiences and impediments, and anticipated consequences (ajzen, 1991; alam & Sayuti, 2011). table 1 offers the items included in the questionnaire for each variable of the model. Berné-manero, C.; Ciobanu, a. v.; marzo-navarro, m.; pedraja-Iglesias, m. (2019). the hotel management of e-wom. a cross-cultural analysis Spain-Italy. International Journal of Information Systems and Software Engineering for Big Companies (IJISEBC) , 6(2), 7-18. www.ijisebc.com figure 1. e-wIp model. Source: adaptation from Berné et al. (2019). table 1. Indicators of the dimensions in the e-wIp model. Source: Self-made. 11 IJISEBC, 6(2), 2019 from information provided by Spanish hotel decision-makers, it has been confirmed that reasons for the acceptance and use of e-wom have a direct and positive influence on hotel manager’s motives to accept and manage such information; the perceived characteristics of the e-wom have a direct, positive influence on the global motives for the use of such information by hotel management; and, global motives regarding e-wom communication have a direct and positive effect on the intentions of hotel managers as well as on hotel performance. It is important provide better understanding of the influence of e-wom on hotel managers’ behaviour depending on culture. Culture goes far beyond the country were you have born, therefore it could determine a different making-decisions behaviour and different management implications. Considering this idea, the research question is: - do the Spain and Italy hotel industries converge or diverge in the process of acceptation, use, and management of e-wom? although they are two countries that are close to each other from a cultural perspective, some differences can be expected, since the business model and the structures of the hotel industries are different. while in the Italian case, the hotel industry is composed mostly of small independent hotels dominates and is managed by its owners, in Spain the hotel industry is mainly made up of hotel chains, with a large size in number of establishments, and managed by executives (Berné, Ciobanu & pedraja, 2015; moretta, Berné & Ciobanu, 2018). whether or not significant differences are discovered, the conclusions and academic recommendations and practices should be useful. 3. methodology with the aim to provide an answer to the research question, a cross-cultural analysis is carried out; a comparative study between two independent samples. the data were collected through surveys addressed to hotel decision-makers. the Spanish data are the used in Berné et al. (2019) and the Italian data in moretta et al. (2018). the comparison was conducted applying plS-mga (with SEm) and Cluster analysis. plS Smart 3.0 is the software used. from the basic structure of e-wIp model, mga allows to identify significant differences between the structural relationships established in the model for each one of the two groups (Spanish and Italian data). the methodology is completed implementing a Hierarchical Cluster, ward method, and Euclidean distance. this analysis allows identifying different groups of individuals through levels of dimensions. 3.1. data collection the same structured survey was addressed to Spanish and Italian hotel decision-makers. prior to its distribution, the questionnaire was pre-tested on professional and academic experts. It begins with a filter question on the use or non-use of e-wom as a management tool. the first set of questions captures reasons for and against e-wom acceptance and usage, while the subsequent sets focus on the remaining theoretical model variables. the last section collects respondents’ demographic characteristics and hotel performance data. due to observed differences in the educational systems of each country, 11-points likert scale is used in Spanish case, and 5-points likert scale in Italian case, were used to measure all the variables involved in the theoretical model (see table 1), from 0/1, completely disagree, to 10/5, completely agree, with the statement1. table 2 presents the technical information of the surveys. the questionnaire was addressed to hotel decision-makers significantly involved in decisions concerning digital technologies in the company. this target populations exhibit characteristics of a “hard-to-reach” population (marpsat and razafindratsima, 2010). Berné-manero, C.; Ciobanu, a. v.; marzo-navarro, m.; pedraja-Iglesias, m. (2019). the hotel management of e-wom. a cross-cultural analysis Spain-Italy. International Journal of Information Systems and Software Engineering for Big Companies (IJISEBC) , 6(2), 7-18. www.ijisebc.com 1the scale items were based on previous literature in order to preserve content validity. the likert scale of 5 points used in the Italian case was rescaled (according dawes (2007)’ method) to 11 point to obtain a homogeneous database. 12 IJISEBC, 6(2), 2019 Sample sizes are 134 and 136 (Spain/Italy), which means response rates of 13.4%, and 27.2%; they are common in industrial market surveys (Camisón & villar, 2014), and in line with previous research in similar contexts (e.g. Chan 2013; Kim, lim & Brymer, 2015). 4. results 4.1. multi-group analysis: mga an mga (Henseler & fassott, 2010) is carried out in order to test the potential moderating influence of the culture on the relationships included in the model. accordingly, the sample was split into two groups, Spanish and Italian hotel managers. first of all, the three-step procedure to analyze the measurement invariance of composite models was applied (mICom) (Henseler, ringle & Sarstedt, 2016): (1) configural invariance, (2) compositional invariance (see table 3), and (3) an assessment of equal means and variances (see table 4). Henseler et al. (2015) recommend the use of the mICom procedure when using plS-SEm. the analysis is carried out considering just the second order variables of the model in order to simplify the relationships to be tested and diminish the number of freedom degrees. thus, the variables are reasons, ewom, motives, intentions and performance. all of them measured through summative scales. the assessment of equal means is not complied in the cases of motives and performance, and the assessment of equal variances is not valid. However, reasons, e-wom and intentions confirm partial invariance. this result is sufficient to conduct a plS-mga (Henseler, ringle & Sarstedt, 2016). next stage is to analyse the bootstrapping and permutation results. tables 5 and 6 present bootstrapping results for Spain and Italy respectively. Significant differences between groups, in this case, the two countries, emerge when p < 0,1. the structural relationships involved in the model presents differences between Spain Berné-manero, C.; Ciobanu, a. v.; marzo-navarro, m.; pedraja-Iglesias, m. (2019). the hotel management of e-wom. a cross-cultural analysis Spain-Italy. International Journal of Information Systems and Software Engineering for Big Companies (IJISEBC) , 6(2), 7-18. www.ijisebc.com table 2. technical Surveys. Source: Self-made. table 3. mICom Invariance results. Steps 1 and 2. Source: Self-made. table 4. mICom Invariance results. Step 3. Source: Self-made. 13 IJISEBC, 6(2), 2019 and Italian samples only in the relationship between e-wom and motives. this difference has to be coincident through the permutation test (table 7). However, the only relationship that could be significant is from reasons to motives. therefore, it cannot be affirmed that there are significant differences between the structural relationships by pairs of variables, presented in the model for each sample (ß estimated parameters). the e-wIp model behaves the same regardless of the country. the model confirms its predictive validity and its behaviour is very similar when applied to the Italian case. 4.2. Cluster analysis when only the practices of hotel managers in both countries are studied, differences are weakened and, therefore, cultural differences may not emerge. thus, between Spain and Italy could have larger differences than is reflected in the previous results. In other words, when some variable in a mga does not accomplish invariance, applying a cluster analysis is an alternative to identify differentiated groups (Hair, Sarstedt, matthews & ringle, 2016). thus, there may be differences in the behaviour of the Spanish and Italian sample according the characteristics of the hotel and / or the respondents. In order to detect groups of hotels differentiated through the levels of this type of variables, a Hierarchical Cluster analysis was carried out through the ward method and the Euclidean distance. to avoid the possible introduction of biases, the number of groups to be obtained was not set when executing the SpSS 21.0 application. table 8 shows the results obtained. they manifest the presence of two differentiated groups of hotels. there are significant differences, at 5% level, regarding the average levels of the variables studied (level of indicators of each first order variable –items). these differences exist in every variable between the two identified groups. the standard deviations are higher in the group 1 than in group 2, and this latter offers Berné-manero, C.; Ciobanu, a. v.; marzo-navarro, m.; pedraja-Iglesias, m. (2019). the hotel management of e-wom. a cross-cultural analysis Spain-Italy. International Journal of Information Systems and Software Engineering for Big Companies (IJISEBC) , 6(2), 7-18. www.ijisebc.com table 5. Bootstrapping Spain. Source: Self-made. table 6. Bootstrapping Italy. Source: Self-made. table 7. mga Based on permutation test. Source: Self-made. 14 IJISEBC, 6(2), 2019 higher average values in all the considered variables: reasons, quality, credibility, authority, attitude, subjective norms, perceived control, intentions and, performance (a total of 33 variables). differences in mean values goes from 1.5 to 3.2 points. In order to validate the cluster results, a t-test of independent samples was carried out through levene test (alves, 2000; Juric, mankelow & polonsky, 2000; garcía, gargallo, pastor & ramirez, 2004) (table 8). the differences between the two identified groups (group 1 and group 2) depend on some characteristics of the hotel and of the respondent. In particular: stars ranking, number of employees, charge of the respondent, gender, level of studies, and experience level. there are inexistent differences depending on the type of hotel property or on the age of the respondent. moreover, it is remarkable the country effect: group 1 is mainly composed by answers of Italian respondents and group 2 of Spanish (see table 9). these results announce differences between both countries regarding their hotel industry. It has been observed that the hotels represented in the Spanish sample, ranked with 1 or 4 stars, with 11 to 20 employees, and managed by a professional, male gender, with superior studies and no previous experience in the hotel sector, characterizing a group what gives higher importance to the variables involved in the use and management of e-wom than Italian hotels, with 3 stars, less number of employees, in property, managed by woman with high school studies and with experience in the sector. Berné-manero, C.; Ciobanu, a. v.; marzo-navarro, m.; pedraja-Iglesias, m. (2019). the hotel management of e-wom. a cross-cultural analysis Spain-Italy. International Journal of Information Systems and Software Engineering for Big Companies (IJISEBC) , 6(2), 7-18. www.ijisebc.com table 8. Cluster analysis results: mean and Standard deviation (groups 1 and 2). Source: Self-made. 15 IJISEBC, 6(2), 2019 5. discussion and conclusions research focused into national cultures and the influence of those cultures on the behaviour of managers in organizations is rather complex. there is a lack of intercultural comparison research regarding the national and industrial perspective and the perceptions of hotel decision makers, particularly about the use of technologies. this paper posed a question about the convergence or divergence of Spanish and Italian hotels regarding the process of acceptance, use, and management of e-wom. with the aim to solve the question, e-wIp model is used. this model explains the process of acceptance and management of e-wom in the hotel business. therefore, it serves as a coherent framework for predicting the effects of hotel managers' attitudes, intentions, and behaviours in their business results. managing online reviews is an opportunity for hotels both in Spain and in Italy. they can improve their performance through changes implemented as a consequence of the e-wom information management. But, what about their similarities and differences? from the previous results, obtained after apply a multi-group analysis, and a cluster, it is clear that management practices in Spain and Italy seem to converge rather than diverge. this is an inevitably process affecting the tourism business in general (Jansen-verbeke, 1996). the two studied hotel industries can be considered as one cultural area regarding the cause-effect relationships that explain the process of implementation and management of e-wom hotel. Jansen-verbeke (1996) points that the homogenizing influence of the hotel organization and the hotel industry could be a reasonable explanation for the similarities in management practices between the two countries. However, this research finds two groups of hotels that present significant differences in the mean values of the indicators involved in the model. these two groups are characterized by variables do not included in the model. one of them, very important in the current research context, is the country. the others are hotel characteristics linked to the category, size and property of the establishment, and characteristics of the decision-maker, regarding experience, studies level and gender. therefore, national culture (country) and characteristics of the research subjects are variables able to characterize behaviours of hotel decision-makers about the use and management of e-wom. particularly, in terms of reasons, the group where Spanish hotels are the most, offers more orientation to the customer, more value to the information offered by e-wom, is more confident in the customers’ reviews and they are more dependent on the group (according with pizams’ results about Spain: it is a more collectivistic country than Italy). they find more positive the use of e-wom, social pressure is higher and they consider that they have higher control over the e-wom system. regarding intentions, they manifest higher intentions to continue using the tool. In the case of performance, this group value the use and management of e-wom much more than group 1, mainly represented by Italian hotels. It means that Spanish hotels achieve that e-wom help them to achieve more co-creation with customers. this study contributes to prior knowledge building upon previous research into culture in the hotel industry. the results support Brt’s multidisciplinarity, the predictive capability of the e-wIp model, the relative utility of mga when it is applied just in behavioural variables, and the complementarity of cluster analysis to discover other hidden information in a database. moreover, this paper shows different ways to compare industrial sectors in different countries taking into account potential cultural differences. It is an opportunity to establish criteria of convergence in making decisions. this information may be useful for both organizations and industrial sectors. the main limitation of this study lies in having had to simplify the e-wIp model to the relationships Berné-manero, C.; Ciobanu, a. v.; marzo-navarro, m.; pedraja-Iglesias, m. (2019). the hotel management of e-wom. a cross-cultural analysis Spain-Italy. International Journal of Information Systems and Software Engineering for Big Companies (IJISEBC) , 6(2), 7-18. www.ijisebc.com VARIABLES AFFECTING GROUPS Group 1: 155 cases Group 2: 115 cases Country Hotel Rank Number of Employees Charge Gender Studies Experience Italy 3 stars From 2 to 10 Hotel owner Female High school From 6 to 10, from 10 to 20 Spain 1 and 4 stars From 11 to 20 Hotel manager Male University No previous experience table 9. variables determining group 1 and group 2. Source: Self-made.