Customer identification leads to behaviors that are beneficial for firms. This paper aims to analyze the effect of firm identification and community identification on content creation, which indirectly may affect offline word of mouth and online word of mouth.
This paper proposes a research model that is tested using data from 491 users of online travel agencies. To do so, partial least squares method is used.
The results show a positive relationship between firm identification and community identification. Moreover, both variables exert a positive effect on content creation. Furthermore, content creation positively influences offline and online word of mouth. This influence is moderated by self-enhancement in the case of online word of mouth.
Firm managers must enhance customer identification, as it can turn in behaviors that are beneficial for the company. Moreover, firms that own online communities must apply segmentation strategies based on identification and self-enhancement to encourage positive behaviors from customers.
This research tests the relationship between firm identification and community identification. Additionally, this study jointly analyzes the impact of these variables on several beneficial behaviors.
La identificación del consumidor genera comportamientos que son beneficiosos para las empresas. Esta investigación analiza el efecto directo de la identificación con la compañía y la identificación con la comunidad sobre la creación de contenido, así como el efecto indirecto de estas variables de identificación sobre el boca- oído offline y online.
Esta investigación propone un modelo teórico, el cual es estimado mediante Partial Least Squares a partir de información procedente de 491 usuarios de agencias de viajes online.
Los resultados muestran una relación positiva entre la identificación con la firma y la identificación con la comunidad. Además, ambas variables ejercen un efecto positivo en la creación de contenido. Asimismo, la creación de contenido influye positivamente sobre el boca-oído offline y online. Esta influencia es moderada por la necesidad de reconocimiento de los consumidores.
Los resultados del trabajo recomiendan potenciar la identificación del cliente con la empresa, dado esta identificación conlleva comportamientos beneficiosos para la compañía. Además, las compañías que poseen comunidades online deberían aplicar estrategias de segmentación basadas en la identificación y la necesidad de reconocimiento de cara a potenciar que sus clientes llevan a cabo comportamientos positivos para la empresa.
Esta investigación examina la relación entre la identificación con la compañía y la identificación con la comunidad. Adicionalmente, este estudio analiza conjuntamente el impacto de estas variables en comportamientos que son beneficiosos para la empresa
Tipo de trabajo
Trabajo de investigación
Romero, J. and Ruiz-Equihua, D. (2020), "Be a part of it: promoting WOM, eWOM, and content creation through customer identification", Spanish Journal of Marketing - ESIC, Vol. 24 No. 1, pp. 55-72. https://doi.org/10.1108/SJME-11-2019-0092
Emerald Publishing Limited
Copyright © 2020, Jaime Romero and Daniel Ruiz-Equihua.
Published in Spanish Journal of Marketing – ESIC. Published by Emerald Publishing Limited. This article is published under the Creative Commons Attribution (CC BY 4.0) licence. Anyone may reproduce, distribute, translate and create derivative works of this article (for both commercial and non-commercial purposes), subject to full attribution to the original publication and authors. The full terms of this licence may be seen at http://creativecommons.org/licences/by/4.0/legalcode
Customer identification refers to the sense of belongingness to a group (Ashforth and Mael, 1989). It positively affects firms, as it provokes valuable customer behaviors (Casaló et al., 2010). Therefore, firm managers should try to enhance customer identification. Because of the benefits obtained from being part of the group, the individuals develop a moral responsibility that makes them help other members of the group (Muniz and O’Guinn, 2001). Focusing on the relationship with a company, when customers obtain several benefits from the company, they develop a commitment to the firm, which in turn motivates positive customer behaviors (Sääksjärvi et al., 2007). That is, because of these benefits, customers try to maintain a long-term relationship by performing positive behaviors for the firm.
Previous research studies, as far as we know, have mainly focused on the impact of identification on specific customer behaviors, such as intentions to participate in online firm communities (Casaló et al., 2010), resilience to negative information (Augusto et al., 2019) or co-creation (Romero, 2017), among many others. However, these works have analyzed the relationship between customer identification and some behaviors separately. That is, the extant literature does not consider the connection between behaviors elicited by identification. This approach could be biased and lead to misinterpretation of the phenomenon. For example, if customers feel that they have helped the company too much through providing suggestions to improve service delivery, they could stop spreading positive word of mouth. Similarly, if the firm publicly acknowledges customers’ reviews of company services, customers could feel grateful to the firm and try to make suggestions to maintain the relationship with the company. Thus, this research aims to deepen our comprehension about customer identification by analyzing the relationship between behaviors produced by customer identification in the context of online communities.
Our research distinguishes between firm identification and community identification. The former refers to the degree to which a customer feels that a company shares the same attributes as the customer (Dutton et al., 1994). The latter refers to the degree to which people see themselves as part of the group, and being part of it has significant value for the group (Casaló et al., 2010).
This research considers three different behaviors that customer identification could provoke:
offline word of mouth (hereafter WOM); and
online word of mouth (hereafter eWOM).
Content creation refers to the media-generated content by users to share information or opinions with other users (Tang et al., 2014). Companies that own online communities provide customers with an environment where they can generate value for firms through content creation. We propose that content creation could also influence other valuable behaviors, namely, eWOM and WOM. Word of mouth refers to any brief statement made by customers about a product or service (Sundaram et al., 1998) and it is relevant for companies, as it can elicit positive behaviors such as revisiting intentions (Abubakar et al., 2017), booking intentions (Sparks and Browning, 2011) or sales increases (Chevalier and Mayzlin, 2006). Depending on where it takes place, we distinguish between WOM and eWOM. Both behaviors are very relevant for firms (Casaló et al., 2009, 2015; Ruiz-Equihua et al., 2019).
Our research proposes that the relationships between content creation, WOM and eWOM could be influenced by self-enhancement, that is, customers’ need of positive recognition (Hennig-Thurau et al., 2004). Additionally, as a control variable, we include relationship quality, that is, customer evaluation of the customer–firm relationship, which has been traditionally related to trust, commitment or satisfaction, among others (Hennig-Thurau et al., 2002).
We conducted our research in the hospitality industry whose importance in modern economies is unquestionable. For example, in Europe in 2016, a total of 1,998,320 hospitality firms employed 11,900,00 people (Eurostat, 2019).
This study thus contributes to previous research in the following ways. First, we extend social identity theory by considering the relationship that firm identification and community identification might have. Second, we enhance our comprehension of the effects of customer identification by showing how the resultant behaviors interact between them – in this case, content creation, eWOM and WOM. Specifically, we study the relationship between content creation, eWOM and WOM. Third, we analyze the moderating effect, if any, of self-enhancement in the relationship between content creation, eWOM and WOM. Finally, we develop managerial insights from our research, allowing firms a more profitable management of customer identification.
The structure of this research is as follows. In the next section, we present the literature review of our study and the research hypotheses. Subsequently, we explain the methodology of our research and its main results. Finally, we discuss the study findings and their implications.
2. Literature review
This research proposes that customer identification leads to content creation and that the higher the content creation behavior, the more likely it is to elicit other positive behaviors, namely, eWOM and WOM. Furthermore, we analyze the moderating effect of self-enhancement in the relationship between content creation and eWOM and WOM.
Previous research analyzes the effect of identification on several industries, e.g. food service (Bergami and Bagozzi, 2000), software (Bagozzi and Dholakia, 2006) or financial services (Cardador and Pratt, 2018). Identification leads to positive behaviors among the customers. For example, identification is positively related to the willingness to pay (Augusto and Torres, 2018), engagement (Romero, 2018) or content creation (Dholakia et al., 2004). However, companies that own online communities need to pay attention to customer identification with such communities. In this research, we propose that firm identification and community identification are related. Furthermore, we propose that firm identification and community identification could influence content creation in the firm online community.
Content creation involves information sharing in virtual environments, such as customer blogs, online forums, communities and social networks. This information is shared by customers trying to help peer customers (Romero, 2018). As posited above, a relationship might exist between the customer behaviors elicited by identification. We propose that content creation, WOM and eWOM are linked. WOM includes personal and one-to-one conversations. In contrast, eWOM is impersonal and reaches several people because of its ubiquitous nature. For the purpose of this research, we focus on eWOM (and, of course, WOM) that takes places outside the firm online community.
Our model includes two more variables, self-enhancement and relationship quality. Self-enhancement reflects customers’ need to enhance their awareness of themselves as experts and smart shoppers (Sundaram et al., 1998). Our study analyzes whether self-enhancement moderates the influence of content creation on eWOM and WOM behaviors. Relationship quality is a control variable in our model. Because relationship quality captures customers’ positive feelings about a company (Alves et al., 2019), it may also be related to other variables such as customer identification, content creation and word of mouth (Ng et al., 2011; Romero, 2017, 2018). Thus, we include it as a control variable.
3. Research model and hypotheses formulation
Figure 1 shows our research model. First, we analyze the effect of firm identification on community identification and on content creation. We also consider that community identification might influence content creation. Second, we analyze the link between customer behaviors elicited by identification, specifically between content creation, eWOM and WOM. Finally, we analyze the moderating role of self-enhancement in the relationship between content creation, eWOM and WOM.
Next, we present the hypotheses for where we ground our research model.
3.1 The effect of identification on content creation and community identification
Social identity theory (Tajfel and Turner, 1986) provides the background to understand the customer identification phenomenon. The social identity theory refers to “the perception of oneness or belongingness to some human groups” (Ashforth and Mael, 1989, p. 21). Firms are human groups and therefore can generate identification (Ahearne et al., 2005). Previous research shows that customer identification arises with several aspects of firms. For example, Ahearne et al. (2005) find that customer–firm identification is related to the firm’s salespersons. In addition, an existing identification with the brand, which means valuing it for the functional and symbolic benefits it provides, motivates customers to be integrated and identified with the brand community (Algesheimer et al., 2005).
Hence, according to previous studies, customer identification can focus on a range of firm-related elements, such its staff, and we propose that this identification could spill over beyond these elements and reach other entities such as online communities, given that customers can establish links between these entities and the firm. As Algesheimer et al. (2005, p. 23) note, “a harmonious relationship with the brand can lead customers to seek out and interact with like-minded customers who share their enthusiasm.” Thus, we propose the following hypothesis:
Firm identification exerts a positive effect on community identification.
In addition, Bhattacharya and Sen (2003) suggest that identification elicits positive customer behaviors for the firm. In this regard, previous research examines the relationship between firm identification and positive customer behaviors. For instance, firm identification leads to co-creation (Romero, 2018), resilience to negative information (Bhattacharya and Sen, 2003), loyalty (Rather et al., 2018) or engagement (Badenes-Rocha et al., 2019). Thus, consistent with previous research and the social identity theory, we expect that firm identification exerts a positive effect on content creation. Furthermore, and in line with the social identity theory and previous research (Bhattacharya and Sen, 2003), we believe that community identification may also exert a positive effect on content creation. Previous research has found that community identification elicits positive customer behavior, such as intention to participate in the community (Bagozzi and Dholakia, 2006) or knowledge sharing (Qu and Lee, 2011). Well-identified members are more prone to participate in the online community exchanging information (Bagozzi and Dholakia, 2006; Casaló et al., 2010; Qu and Lee, 2011). Therefore, customers with high community identification might be prone to create content in the online community. Thus, we posit the next hypotheses:
Firm identification exerts a positive effect on content creation.
Community identification exerts a positive effect on content creation.
3.2 The relationship between content creation and online/offline word of mouth
Next, we consider the influence of content creation on eWOM and WOM behaviors. Customers who actively participate in online communities and create content may obtain benefits from the company beyond the purchases. For example, customers interact with related peers in the online firm communities (Wellman, 2001). Through such interactions, customers satisfy some basic socializing needs (Casaló et al., 2010). Arising from such benefits, customers might develop a level of commitment with the firm (Sääksjärvi et al., 2007). Commitment is the desire to maintain a valued relationship with a specific entity (Brown et al., 2005). Previous research has found that commitment elicits a range of positive customer behaviors, such as loyalty, willingness to pay (Hur and Kang, 2012), membership continuance, community recommendation, community participation (Algesheimer et al., 2005) or customer identification (Rather et al., 2018). Reasonably, because of the greater value obtained by participating and creating content in the community, committed customers might be expected to perform other behaviors that are beneficial for firms to restore equity in the relationship. WOM and eWOM are positive for companies and thus they are among such behaviors. Indeed, customers acknowledge that they help firms through word of mouth (Hennig-Thurau et al., 2004). In addition, when creating new content, it is possible that customers are already talking positively about the company, as most conversations in the community are around the brand products (Casaló et al., 2010). Hence, we expect that customers who create content in online communities are more prone to help companies through eWOM and WOM. Therefore, we develop the following hypotheses:
Content creation is positively related to eWOM behavior.
Content creation is positively related to WOM behavior.
3.3 The moderating role of self-enhancement in the relationship between content creation and online/offline word of mouth
Self-enhancement may strengthen the relationship between content creation, eWOM and WOM, thus exerting a moderating role. Specifically, customers with high self-enhancement levels need to gain attention (Hennig-Thurau et al., 2004) by enhancing their self-image among other customers (Sundaram et al., 1998). For example, previous research showed that customers with self-enhancement needs are more prone to develop positive behaviors for companies. For example, Nambisan and Baron (2010) find that customers with self-enhancement support the firms’ product in the online community. Moreover, Hennig-Thurau et al. (2004) find that customers with self-enhancement are more prone to post their experiences on online platforms. Therefore, customers with high self-enhancement might be more prone to benefit the company through eWOM and WOM. In addition, customer behaviors such as eWOM and WOM can imply extra recognition from more people. Therefore, self-enhancement might positively moderate the aforementioned relationships – that is, the effect of content creation is stronger as self-enhancement increases – and so we propose the following hypotheses:
Self-enhancement positively moderates the relationship between content creation and eWOM.
Self-enhancement positively moderates the relationship between content creation and WOM.
For the sake of completeness, relationship quality is included as a control variable affecting customer identification (with both the firm and the community), content creation, WOM and EWOM.
4.1 Data collection
Data to test our hypotheses were collected from users of hospitality services, specifically, online travel agencies. Particularly, we focused on two agencies, namely, Logitravel and Atrápalo, which hold accounts in social network sites such as Facebook and Twitter.
A market research firm collected our data through a Web-based survey. The results of online surveys are comparable with the ones of other data collection methods if the respondents are used to online environments (Deutskens et al., 2006). We requested that our respondents answer our questions with regard to the travel agency they use the most. Specifically, participants answered a questionnaire including multiple-item reflective measurement scales inspired by previous literature (see Appendix). These scales used seven-point Likert-type response formats, which respondents rated from 1 = strongly disagree to 7 = “strongly agree, except one of the items used to measure customer identification with the online agency (taken from study by Bergami and Bagozzi, 2000). This item is a visual scale in which respondents express the overlap between their own identity and the identity of the travel agency. To ensure content validity of scales and to confirm that the adaptations made to our research context were right, we followed Zaichkowsky’s (1985) method and asked a panel of ten experts in tourism, e-commerce and marketing to classify each item in terms of its relevance and representativeness of the construct to which it was assigned. Only items that produced a high level of consensus among the experts were retained and are part of the final scales (Lichtenstein et al., 1990).
Finally, this process allowed us to collect data from 491 users of online travel agencies in Spain. The demographics of the sample are representative of the internet users in Spain who browse the internet for travel and accommodation purposes (Table I).
4.2 Common method bias assessment
Using a single survey to collect data may cause problems related to common method bias. To avoid this potential problem, procedural recommendations were first taken into account to minimize common method bias through study design (Podsakoff et al., 2003). Particularly, we assured participants that their responses would be confidential and specifically stated that no right or wrong answers existed. In addition, we also guaranteed participants’ anonymity to reduce the possibility of answering in a dishonest manner (Fraj et al., 2015). In addition, although the data come from the same sources with no temporal separation, the design of the questionnaire tries to avoid any direct connection between dependent variables (DVs) and independent variables (IVs) by inducing a psychological disconnection because of the inclusion of other questions not related to the research objective (García Rodríguez et al., 2008). Specifically, individuals described some aspects related to their use of social network sites. Thus, participants could not infer any cause–effect link between variables (Fraj et al., 2015). Second, evidence of common method bias may have also resulted in extremely high correlations, greater than 0.90 (Pavlou et al., 2007). As we show later in this paper (Table III), correlations do not exceed the critical threshold. Hence, common method bias does not seem to be a problem in this research. Finally, we also assessed common method bias statistically following Liang et al.’s (2007) test, based on Podsakoff et al. (2003). This method is not novel in research involving tourism firms, as in our study (Lo et al., 2016; Mayr and Zins, 2012; Steinhoff and Palmatier, 2016), and has been extensively used in other areas too. However, we take its results with caution, as an extra assessment of common method bias, given that the method has been sometimes criticized (Chin et al., 2012). To be precise, we included a common method factor in our model and calculated the variance of each item explained by its corresponding substantive construct and by the common method factor. On average, the method factor determines less than a 2 per cent of the indicators’ variance. Together with other evidences, we concluded that common method bias is not a major problem in this research.
5.1 Measures validation
We used partial least square (PLS) method to estimate our model because it is especially useful in situations with low theoretical information, or when the phenomenon under research is relatively new (Roldan and Sanchez-Franco, 2012), as it is in our case. In addition, PLS can handle data without multivariate normality better than covariance-based structural equation models, which assume data normality. More specifically, we used SmartPLS 3.0 (Ringle et al., 2015) and used consistent PLS, as it solves inconsistency problems that can arise in traditional PLS (Dijkstra and Henseler, 2015a, 2015b). These inconsistency problems lead to wrong path estimates and construct measures. In addition, our sample consists of 491 participants, which meets the sample size requirements when using PLS; that is, ten observations multiplied by the maximum between: the construct with the highest number of indicators; or the endogenous construct with the largest number of exogenous constructs (Davcik, 2014). In our case, WOM has six indicators, which requires a minimum sample size of 60. The sample was balanced in terms of age and gender (Table I).
We evaluated the reliability, convergent validity and discriminant validity of our constructs. Except for company identification (0.63), the Cronbach’s alpha values of these latent variables are all above the cut-off value of 0.7 proposed by Nunnally (1978). Our research includes a visual item for measuring company identification, which we eliminated for the sake of reliability, hence measuring this variable using only a verbal item. This verbal item provides a general measure of the variable and uses a seven-point Likert-type scale, as with all the remaining variables in the model. Additionally, some of the indicators for community identification and relationship quality are lower than 0.7, thus recommending their elimination. After depurating our scales, the entire alpha values range between 0.87 and 0.96 (see Appendix for further details). The loadings of all indicators are above 0.70. Composite reliability varies between 0.87 and 0.95. The average variance extracted (AVE) varies between 0.62 and 0.78, above the 0.50 cut-off value suggested by Fornell and Larcker (1981).
To evaluate discriminant validity, we used two criteria. First, we applied the Fornell and Larcker’s (1981) criterion, by which the square root of the AVE of a variable needs to be higher than its correlations with other variables. Second, we assessed the heterotrait–monotrait ratio (HTMT) of the correlations. The HTMT is more sensitive to a lack of discriminant validity than the other criterion (Henseler et al., 2015). To indicate discriminant validity, the HTMT between two constructs has to be lower than 0.85. Both criteria support the discriminant validity of all our variables (Tables II and III).
5.2 Hypotheses testing
Once measures were validated, we developed a structural equation model (Figure 2) that gives support to our hypotheses and achieves acceptable values of model fit. All path estimates in our model are significant at 99 per cent according to a nonparametric bootstrapping procedure with 10,000 subsamples (no sign change). We find that firm identification is positively associated with community identification (β = 0.46, p < 0.01) and content creation (β = 0.25, p < 0.01); therefore, H1 and H2 were supported. Moreover, community identification (β = 0.57, p < 0.01) is positively associated with content creation; as a consequence, H3 was supported. Similarly, content creation is positively associated with eWOM (β = 0.69, p < 0.01) and WOM (β = 0.20, p < 0.01); therefore, H4 and H5 were supported. The relationships between content creation and eWOM/WOM are moderated by the self-enhancement in the case of eWOM (β = 0.06, p < 0.05), but not for WOM (β = 0.03, p > 0.10); consequently, H6 was supported, while H7 was not. Finally, relationship quality, which we include in our model as a control variable, has a significant impact on firm identification (β = 0.75, p < 0.01), community identification (β = 0.40, p < 0.01), content creation (β = −0.15, p < 0.01) and WOM (β = 0.35, p < 0.01), but not on eWOM.
Because of these relationships, this model allows us to partially explain our DVs. The adjusted R2 of firm identification, community identification, content creation, WOM and eWOM are 0.56, 0.64, 0.44, 0.73 and 0.64, respectively. These values indicate a substantial fit for WOM and a moderate for the remaining variables (Chin, 1998).
Figure 2 also shows the effect sizes. According to Cohen (1988), the effect size of content creation on electronic eWOM is strong (0.95). The effect sizes of firm identification on community identification (0.26) and of community identification on content creation (0.21) are moderate. Finally, the effect sizes of firm identification on content creation (0.04) and content creation on WOM (0.06), as well as the moderation of self-enhancement in the relationship between content creation and eWOM (0.01), are weak.
In the next section, we focus on possible indirect or mediated effects to gain extra insights regarding the impact of firm identification and community identification on customers’ eWOM and WOM behaviors.
5.3 Post hoc analyses of the indirect effects
According to our results, firm and community identification might have an indirect or mediated effect on eWOM and WOM. The effects of both types of identification could be mediated by content creation. Subsequently, we analyze these potentially mediated relationships. To do so, we follow Vinzi et al. (2010), Williams and Mackinnon (2008) and Zhao et al. (2010), who suggest calculating bias corrected and accelerated confidence intervals of such effects. This method has been used in marketing research (Song et al., 2013). The indirect effect of an IV on a DV in each sample is measured as the product of the effect on the IV on the mediating variable (IV →MV) and the effect of the mediating variable on the DV (MV→DV), that is, IV→MV × MV→DV. The indirect effects in each sample are used to build bias corrected and accelerated confidence intervals. If the intervals do not contain the value zero, we can conclude that the indirect effects are significant. In our study, we used 10,000 subsamples with no sign change. We show the results of such analyses in Table IV.
This estimation indicates that firm identification indirectly influences eWOM (confidence interval: 0.25-0.46) and WOM (0.05-0.17). Similarly, community identification also indirectly influences these variables (confidence intervals are 0.26-0.54 and 0.05-0.20, respectively). According to Zhao et al. (2010), the existence of these indirect effects constitutes sufficient evidence of mediated relationships. Therefore, we can conclude that content creation mediates the impact of firm identification and community identification on eWOM and WOM.
Following Matzler et al. (2016), we also examine the direct effects of firm identification and community identification on eWOM and WOM. For this purpose, we estimate an extended version of our research model, which also includes such direct effects. This further analysis provides extra insights about these mediated relationships. According to Zhao et al. (2010), if the direct effect of a variable (IV→DV) is not significant, the mediated relationship between both variables is classified as indirect only. If the direct effect is significant, the sign of the product of the direct effect by the indirect effect has to be evaluated. If this sign is positive, the mediation is classified as a complementary mediation. In contrast, if the sign is negative, the mediation is classified as a competitive mediation.
We show the estimation results for our extended research model in Table IV. These results reveal that one type of mediation relationship exists in our model. None of the direct effects is significant, thus indicating indirect-only effects.
This study analyzes the effects of customer identification, with both the company and the brand community developed around it, on content creation, eWOM and WOM, taking into account the moderating effect of self-enhancement. This section contains the main theoretical and managerial contributions of our research, presents its main limitations and suggests avenues for further research.
6.1 Summary of results and theoretical contributions
Customer identification influences positive behaviors that create value for companies (Bhattacharya and Sen, 2003). In this sense, our study contributes to the extant theory in the following ways. First, we find that firm identification exerts a positive effect on community identification. This is in line with the social identity theory (Ashforth and Mael, 1989). Social identity theory posits that customers are identified when they feel that they belong to some group (Ashforth and Mael, 1989). In this regard, previous research studies the effect of firm identification and community identification on customer behaviors separately. Despite research advances, previous studies do not test a potential relationship between firm identification and community identification. This gap might create a misunderstanding of the customer identification process. Thus, this research advances the identification theory, noticing the relationship between firm identification and community identification. Moreover, in line with previous research, we find that identification with both the company and the community elicits positive behaviors, such as content creation, WOM and eWOM, subsequently.
Second, our results also indicate that the benefits obtained from identification elicit other positive behaviors. In this regard, we find that content creation elicits positive eWOM and WOM behaviors. Thus, customers get involved in “helping” behaviors such as eWOM and WOM to maintain the relationship with the firm. This is in line with previous research, as it demonstrates that these “helping behaviors” serve to restore equity in the relationship (Casaló and Romero, 2019) and maintain the long-term relationship between customers and companies (Brown et al., 2005).
Finally, we find that self-enhancement reinforces the relationship between content creation and eWOM, but not between content creation and WOM. Self-enhancement reflects the need of awareness as experts and smart shoppers (Sundaram et al., 1998). Online environments allow customers to reach numberless customers. Therefore, it is reasonable to think that only in the online environment can customers fulfill their self-enhancement needs.
Additionally, this study includes the role of relationship quality as a control variable. Research results indicate that relationship quality has a positive impact on firm identification, community identification and WOM, but not on content creation or eWOM. These latter results might indicate that the impact of relationship quality found in previous research (Hennig-Thurau et al., 2002; Ng et al., 2011; Alves et al., 2019) might be because of a mediating effect of identification on these behaviors.
6.2 Managerial implications
From a managerial perspective, this research offers interesting implications. First, in general, firm managers must enhance firm and community identification. To enhance the identification, firms must clearly communicate their identity dimensions through their marketing channels. For example, they need to communicate values or behaviors that customers may perceive as similar to their own (Bhattacharya and Sen, 2003).
Second, firms without an online firm community should develop it. Our research results indicate that community identification elicits positive behaviors for the company, such as content creation. Moreover, firms can gain extra positive customer behaviors, as customer commitment might arise from such behaviors. Therefore, committed customers involve themselves in eWOM/WOM behaviors, allowing firms to reduce marketing costs.
Finally, the managers of firms that already own online communities should try to detect their most identified customers. Research results show that these identified customers are more prone to be involved in commitment behaviors. Managers could develop a free community membership to take advantage of this identification. Managers might create a rewards program for customers with high content creation behaviors. Additionally, it would be interesting for managers to somehow evaluate customer self-enhancement and incentivize the customers with higher levels of self-enhancement. Research results demonstrate that customers with high self-enhancement are even more prone to involve themselves in valuable behaviors for the company.
6.3 Limitations and further research
This research is not exempt from limitations, which can motivate further research. First, we focused our study on online travel agencies. For further research, it might be interesting to test research results in different contexts. This would enhance the validity of the results. Second, the use of a Spanish sample recommends caution about generalizing the research results. Further research might replicate the study in other cultural contexts. Third, we focused on firm identification and community identification. Further research might test customer identification with different aspects of the firm, such as staff, specific products, etc. Furthermore, because behaviors elicited by identification might trigger other behaviors (such as WOM or eWOM considered in this research), future research might take into account other behaviors to benefit companies (i.e. buying intentions, engagement, etc.). Finally, it seems that some benefits might mediate the connection between content creation in online firm communities and helpful customer behaviors. This potential effect could be relevant for further research about customer identification.
Note: All figures are expressed in percentages
Source: INE (2012)
Discriminant validity analysis: Fornell and Larcker criterion
|Company identification (1)||1.00|
|Community identification (2)||0.76||0.81|
|Content creation (3)||0.57||0.65||0.86|
|Electronic word of mouth (4)||0.52||0.64||0.84||0.88|
|Relationship quality (7)||0.75||0.74||0.46||0.48||0.67||0.57||0.79|
Italic numbers on the diagonal show the square root of the AVE; numbers below the diagonal represent construct correlations
Discriminant validity analysis: HTMT
|Company identification (1)||–|
|Community identification (2)||0.76||–|
|Content creation (3)||0.56||0.64||–|
|Electronic word of mouth (4)||0.52||0.64||0.84||–|
|Relationship quality (7)||0.74||0.74||0.45||0.47||0.67||0.56||–|
Indirect and direct effects
|Estimate||t-statistic||95% Bias corrected and accelerated confidence interval|
|Research model: firm identification|
|Indirect effect firm identification → electronic WOM||0.35||6.46||***||(0.25, 0.46)|
|Indirect effect firm identification → WOM||0.10||3.19||***||(0.05, 0.17)|
|Research model: community identification|
|Indirect effect community identification → electronic WOM||0.39||5.71||***||(0.26, 0.54)|
|Indirect effect community identification → WOM||0.12||3.19||***||(0.05, 0.20)|
|Extended research model: firm identification|
|Direct effect firm identification → electronic WOM||−0.09||1.15||n.s.||(−0.24, 0.06)|
|Indirect effect firm identification → electronic WOM||0.39||5.80||***||(0.27, 0.53)|
|Direct effect firm identification → WOM||0.08||1.28||n.s.||(−0.04, 0.21)|
|Indirect effect firm identification → WOM||0.13||2.81||***||(0.05, 0.24)|
|Extended research model: community identification|
|Direct effect community identification → electronic WOM||0.10||1.08||n.s.||(−0.08, 0.30)|
|Indirect effect community identification → electronic WOM||0.39||5.44||***||(0.26, 0.54)|
|Direct effect community identification → WOM||0.12||1.32||n.s.||(−0.05, 0.29)|
|Indirect effect community identification → WOM||0.09||2.68||***||(0.03, 0.16)|
***p < 0.01; n.s. = nonsignificant
|Company identification (α = NA; CR = NA; AVE = NA)
Adapted from Bergami and Bagozzi (2000)
|I identify with this OTA||4.53||1.25||0.01||−0.20|
|Community identification (α = 0.88; CR = 0.88; AVE = 0.65). Adapted from Nambisan and Baron (2010)|
|This OTA’ community thinks like me||4.80||1.21||−0.06||−0.22|
|This OTA’ community is like me||4.59||1.29||−0.36||−0.16|
|This OTA’ community could be a friend||4.63||1.31||0.35||−0.12|
|This OTA’ community means a lot to me||3.88||1.56||−0.72||0.10|
|Content creation (α = 0.90; CR = 0.90; AVE = 0.74)
Adapted from Casaló et al. (2009)
|I assess and share with other users my opinions and experiences about the products and services of this OTA on the company website||4.21||1.59||−0.63||−0.24|
|I write comments in the blog and/or in the profile of this OTA in social networking sites (e.g., Facebook, Twitter, etc.)||3.64||1.69||−0.97||0.12|
|I write comments in the forums on this OTA||3.59||1.73||−1.04||0.17|
|Electronic word of mouth (α = 0.87; CR = 0.87; AVE = 0.77)
Adapted from Casaló et al. (2009)
|When I find an interesting content at this OTA, I say it at social networking sites||4.04||1.71||−0.97||0.01|
|I follow this OTA at social networking sites (Facebook, Twitter, Tuenti, etc.)||4.02||1.79||−1.00||−0.05|
|WOM (α = 0.95; CR = 0.95; AVE = 0.78)
Adapted from Brown et al. (2005)
|I mention to others that I do business with this OTA||4.62||1.45||−0.41||−0.29|
|I make sure that others know that I do business with this OTA||4.20||1.54||−0.61||−0.10|
|I recommended this OTA to family members||4.84||1.35||−0.07||−0.39|
|I speak positively of this OTA to others||4.89||1.36||−0.07||−0.50|
|I recommend this OTA to acquaintances||4.91||1.38||−0.13||−0.49|
|I recommended this OTA to close personal friends||5.01||1.39||−0.05||−0.56|
|Self-enhancement (α = 0.92; CR = 0.92; AVE = 0.73)
Adapted from Hennig-Thurau et al. (2004)
I regularly visit this online travel agency (hereafter, OTA) (its website, social networks, etc.), because …
|I like telling others that I have made a good choice||4.44||1.39||−0.37||−0.2|
|I like when I can tell others about my buying success||4.68||1.39||−0.18||−0.28|
|I like telling others about a good experience||4.81||1.38||−0.26||−0.36|
|My contributions show others that I am a clever customer||4.30||1.47||−0.57||−0.19|
|Relationship quality (α = 0.87; CR = 0.87; AVE = 0.62)
Adapted from Ng et al. (2011)
|Overall, I am satisfied with this OTA||5.30||1.03||0.31||−0.52|
|I am very happy with this OTA||5.26||1.06||0.08||−0.46|
|I am very committed to this OTA||4.34||1.38||−0.42||−0.15|
|I believe that I will continue to use this OTA frequently in the future||5.26||1.17||−0.03||−0.53|
All items are measured with a seven-point Likert scale anchoring (1) = strongly disagree and (7) = strongly agree. α = Cronbach’s alpha, CR = CR, AVE = average variance extracted, SD = standard deviation, OTA = online travel agency
Abubakar, A.M., et al. (2017), “eWOM, revisit intention, destination trust and gender”, Journal of Hospitality and Tourism Management, Vol. 31, pp. 220-227.
Ahearne, M., Bhattacharya, C.B. and Gruen, T. (2005), “Antecedents and consequences of customer – company identification: expanding the role of relationship marketing”, Journal of Applied Psychology, Vol. 90 No. 3, pp. 574-585.
Algesheimer, R., Dholakia, U.M. and Herrmann, A. (2005), “The social influence of brand community: evidence from European car clubs”, Journal of Marketing, Vol. 69 No. 3, pp. 19-34.
Alves, H., Campón-Cerro, A.M. and Hernández-Mogollón, J.M. (2019), “Enhancing rural destinations’ loyalty through relationship quality”, Spanish Journal of Marketing – ESIC, Vol. 23 No. 2, pp. 185-204.
Ashforth, B.E. and Mael, F. (1989), “Social identity theory and the organization”, Academy of Management Review, Vol. 14 No. 1, pp. 20-39.
Augusto, M. and Torres, P. (2018), “Effects of brand attitude and eWOM on consumers’ willingness to pay in the banking industry: mediating role of consumer-brand identification and brand equity”, Journal of Retailing and Consumer Services, Vol. 42, pp. 1-10.
Augusto, M., Godinho, P. and Torres, P. (2019), “Building customers’ resilience to negative information in the airline industry”, Journal of Retailing and Consumer Services, Vol. 50, pp. 235-248.
Badenes-Rocha, A., Ruiz-Mafé, C. and Bigné, E. (2019), “Engaging customers through user-and company-generated content on CSR engagement del cliente mediante contenido de RSC generado por el usuario y la empresa engaging customers”, Spanish Journal of Marketing – ESIC, Vol. 23 No. 3, pp. 339-371.
Bagozzi, R.P. and Dholakia, U.M. (2006), “Open source software user communities: a study of participation in Linux user groups”, Management Science, Vol. 52 No. 7, pp. 1099-1115.
Bergami, M. and Bagozzi, R.P. (2000), “Self-categorization, affective commitment and group self-esteem as distinct aspects of social identity in the organization”, British Journal of Social Psychology, Vol. 39, pp. 555-577.
Bhattacharya, C.B. and Sen, S. (2003), “Consumer-company identification: a framework for understanding consumers ‘relationships with companies”, Journal of Marketing, Vol. 67 No. 2, pp. 76-88.
Brown, T.J., et al. (2005), “Spreading the word: investigating antecedents of consumers ‘positive word-of-mouth intentions and behaviors in a retailing context”, Journal of the Academy of Marketing Science, Vol. 2, pp. 123-138.
Cardador, M.T. and Pratt, M.G. (2018), “Becoming who serve: a model of multi-layered employee-customer identification”, Academy of Management Journal, Vol. 61 No. 6, pp. 2053-2080.
Casaló, L.V., et al. (2009), “Determinants of success in open source software networks”, Industrial Management and Data Systems, Vol. 109 No. 4, pp. 532-549.
Casaló, L.V., et al. (2015), “Do online hotel rating schemes influence booking behaviors?”, International Journal of Hospitality Management, Vol. 49, pp. 28-36.
Casaló, L.V. and Romero, J. (2019), “Social media promotions and travelers ‘value-creating behaviors: the role of perceived support”, International Journal of Contemporary Hospitality Management, Vol. 31 No. 2, pp. 633-650.
Casaló, L.V., Flavián, C. and Guinalíu, M. (2010), “Determinants of the intention to participate in firm-hosted online travel communities and effects on consumer behavioral intentions”, Tourism Management, Vol. 31 No. 6, pp. 898-911.
Chevalier, J.A. and Mayzlin, D. (2006), “The effect of word of mouth on sales: online book reviews”, Journal of Marketing Research, Vol. 43 No. 3, pp. 345-354.
Chin, W.W. (1998), “The partial least squares for structural equation modeling”, in Marcoulides, G.A. (Ed.), Modern Methods for Business Research, Lawrence Erlbaum, Mahwah, NJ.
Chin, W.W., Thatcher, J.B. and Wright, R.T. (2012), “Assessing common method bias: problems with the ULMC technique”, MIS Quarterly, Vol. 36 No. 3, pp. 1003-1019.
Cohen, J. (1988), Statistical Power Analysis for the Behavioral Sciences, 2nd ed., Lawrence Erlbaum Associates, Hillsdale, NJ.
Davcik, N.S. (2014), “The use and misuse of structural equation modeling in management research a review and critique”, Journal of Advances in Management, Vol. 11 No. 1, pp. 47-81.
Deutskens, E., et al. (2006), “Comparing the generalizability of online and mail surveys in cross-national service quality research”, Marketing Letters, Vol. 17 No. 2, pp. 119-136.
Dholakia, U.M., Bagozzi, R.P. and Pearo, L.K. (2004), “A social influence model of consumer participation in network- and small-group-based virtual communities”, International Journal of Research in Marketing, Vol. 21 No. 3, pp. 241-263.
Dijkstra, T.K. and Henseler, J. (2015a), “Consistent partial least squares path modeling”, MIS Quarterly, Vol. 39 No. 2, pp. 297-316.
Dijkstra, T.K. and Henseler, J. (2015b), “Consistent and asymptotically normal PLS estimators for linear structural equations”, Computational Statistics and Data Analysis, Vol. 81, pp. 10-23.
Dutton, J.E., Dukerich, J.M. and Harquail, C.V. (1994), “Organizational images and member identification”, Administrative Science Quarterly, Vol. 39 No. 2, pp. 239-263.
Eurostat (2019), “Annual enterprise statics for special aggregates of activities”, available at: https://appsso.eurostat.ec.europa.eu/nui/submitViewTableAction.do (accessed 17 September 2019).
Fornell, C. and Larcker, D.F. (1981), “Evaluating structural equation models with unobservable variables and measurement error”, Journal of Marketing Research, Vol. 18 No. 3.
Fraj, E., Matute, J. and Melero, I. (2015), “Environmental strategies and organizational competitiveness in the hotel industry: the role of learning and innovation as determinants of environmental success”, Tourism Management, Vol. 46, pp. 30-42.
García Rodríguez, N., Sanzo Pérez, M.J. and Trespalacios Gutiérrez, J.A. (2008), “Can a good organizational climate compensate for a lack of top management commitment to new product development?”, Journal of Business Research, Vol. 61 No. 2, pp. 118-131.
Hennig-Thurau, T., et al. (2004), “Electronic word-of-mouth via consumer-opinion platforms: what motivates consumers to articulate themselves on the internet?”, Journal of Interactive Marketing, Vol. 18 No. 1, pp. 38-52.
Hennig-Thurau, T., Gwinner, K.P. and Gremler, D.D. (2002), “Understanding relationship marketing outcomes: an integration of relational benefits and relationship quality”, Journal of Service Research, Vol. 4 No. 3, p. 230.
Henseler, J., Ringle, C.M. and Sarstedt, M. (2015), “A new criterion for assessing discriminant validity in variance-based structural equation modeling”, Journal of the Academy of Marketing Science, Vol. 43 No. 1, pp. 115-135.
Hur, W. and Kang, S. (2012), “Interaction effects of the three commitment components on customer loyalty behaviors”, Social Behavior and Personality: An International Journal, Vol. 40 No. 9, pp. 1537-1542.
Liang, H., et al. (2007), “Assimilation of enterprise systems: the effect of institutional pressures and mediating role of top management”, MIS Quarterly, Vol. 31 No. 1, pp. 59-87.
Lichtenstein, D.R., Netemeyer, R.G., Burton, S. (1990), “Distinguishing coupon proneness from value consciousness: an acquisition-transaction utility theory perspective”, Journal of Marketing, Vol. 54 No. 3, pp. 54-67.
Lo, A.S., et al. (2016), “Building brand relationship quality among hotel loyalty program members”, International Journal of Contemporary Hospitality Management, pp. 458-488.
Matzler, K., et al. (2016), “Brand personality and culture: the role of cultural differences on the impact of brand personality perceptions on tourists ‘visit intentions”, Tourism Management, Vol. 52, pp. 507-520.
Mayr, T. and Zins, A.H. (2012), “Extensions on the conceptualization of customer perceived value: insights from the airline industry”, International Journal of Culture, Tourism and Hospitality Research, Vol. 6 No. 4, pp. 356-376.
Muniz, A.M. and O’Guinn, T.C. (2001), “Brand community”, Journal of Consumer Research, Vol. 27 No. 4, pp. 412-432.
Nambisan, S. and Baron, R.A. (2010), “Different roles, different strokes: organizing virtual customer environments to promote two types of customer contributions”, Organization Science, Vol. 21 No. 2, pp. 554-572.
Ng, S., Meredith E, D. and Dagger, T.S. (2011), “Generating positive word-of-mouth in the service experience”, Managing Service Quality: An International Journal, Vol. 21 No. 2, pp. 133-151.
Nunnally, J.C. (1978), Psychometric Theory, 2nd ed., Mc Graw-Hill, New York, NY.
Pavlou, P.A., Liang, H. and Xue, Y. (2007), “Understanding and mitigating uncertainty in online exchange relationships: a principal-agent perspective”, MIS Quarterly, Vol. 31 No. 1, pp. 105-136.
Podsakoff, P.M., et al. (2003), “Common method biases in behavioral research: a critical review of the literature and recommended remedies”, Journal of Applied Psychology, Vol. 88 No. 5, pp. 879-903.
Qu, H. and Lee, H. (2011), “Travelers ‘social identification and membership behaviors in online travel community”, Tourism Management, Vol. 32 No. 6, pp. 1262-1270.
Rather, R.A., Tehseen, S. and Parrey, S.H. (2018), “Promoting customer brand engagement and brand loyalty through customer brand identification and value congruity”, Spanish Journal of Marketing – ESIC, Vol. 22 No. 3, pp. 321-341.
Ringle, C.M. Wende, S. and Becker, J.M. (2015), “SmartPLS 3”, available at: www.smartpls.com (accessed 23 September 2019).
Roldan, J.L. and Sanchez-Franco, M.J. (2012), “Variance-based structural equation modeling: guidelines for using partial least squares in information systems research”, Research Methodologies, Innovations and Philosophies in Software Systems Engineering and Information Systems, pp. 193-221.
Romero, J. (2017), “Customer engagement behaviors in hospitality: customer-based antecedents”, Journal of Hospitality Marketing and Management, Vol. 26 No. 6, pp. 565-584.
Romero, J. (2018), “Exploring customer engagement in tourism”, Journal of Vacation Marketing, Vol. 24 No. 4, pp. 293-306.
Ruiz-Equihua, D., Romero, J. and Casaló, L.V. (2019), “Better the devil you know? The moderating role of brand familiarity and indulgence vs. restraint cultural dimension on eWOM influence in the hospitality industry”, Journal of Hospitality Marketing and Management, doi: 10.1080/19368623.2019.1630698.
Sääksjärvi, M., et al. (2007), “Love at first sight or a long-term affair? Different relationship levels as predictors of customer commitment”, Journal of Relationship Marketing, Vol. 6 No. 1.
Song, Z., Su, X. and Li, L. (2013), “The indirect effects of destination image on destination loyalty intention through tourist satisfaction and perceived value: the bootstrap approach”, Journal of Travel and Tourism Marketing, Vol. 30, pp. 386-409.
Sparks, B.A. and Browning, V. (2011), “The impact of online reviews on hotel booking intentions and perception of trust”, Tourism Management, Vol. 32 No. 6, pp. 1310-1323.
Steinhoff, L. and Palmatier, R.W. (2016), “Understanding loyalty program effectiveness: managing target and bystander effects”, Journal of the Academy of Marketing Science, Vol. 44 No. 1, pp. 88-107.
Sundaram, D.S., Mitra, K. and Webster, C. (1998), “Word-of-mouth communications: a motivational analysis”, Advances in Consumer Research, Vol. 25, pp. 527-532.
Tajfel, H. and Turner, J.C. (1986), “The social identity theory of intergroup behavior”, The Psychology of Intergroup Relations, in Worchel, S. and Austin, W.G. (Eds), Nelson-Hall, Chicago.
Tang, T.(Y.)., Fang, E.(E.). and Wang, F. (2014), “Is neutral really neutral? The effects of neutral user-generated content on product sales”, Journal of Marketing, Vol. 78 No. 4, pp. 41-58.
Vinzi, V.E., et al. (2010), How to Write up and Report PLS Analyses, Springer, Berlin Heidelberg.
Wellman, B. (2001), “Physical place and cyberplace: the rise of personalized networking”, International Journal of Urban and Regional Research, Vol. 25 No. 2, pp. 227-252.
Williams, J. and Mackinnon, D.P. (2008), “Resampling and distribution of the product methods for testing indirect effects in complex models resampling and distribution of the product methods for testing indirect effects in complex models”, Structural Equation Modeling: A Multidisciplinary Journal, Vol. 15 No. 1, pp. 23-51.
Zaichkowsky, J.L. (1985), “Measuring the involvement construct”, Journal of Consumer Research, Vol. 12 No. 3, pp. 341-352.
Zhao, X., Lynch, J.G., Jr. and Chen, Q. (2010), “Reconsidering Baron and Kenny: myths and truths about mediation analysis”, Journal of Consumer Research, Vol. 37 No. 2, pp. 197-206.