The current competitive climate paves the way for a change in the management of the customer portfolio by attempting to improve the efficiency and profitability of the relational approach. Accordingly, the study of the company–customer relationship incorporates the concept of customer engagement. This new concept suggests that transactional criteria (repurchasing, cross-selling, level of use) are insufficient to assess the profitability of each customer. Therefore, it is necessary to consider customers’ non-transactional behaviour, including word of mouth and blogging, in order to gain a more accurate idea of the current and future worth of each customer.
In line with the previous ideas, this research analyses the role of frontline employees, who, the majority of the time, are regarded as spokespeople for the company in the company–customer interaction, as well as being influential in the level of satisfaction and engagement. This interaction is vital in the case of a service failure or complaint management.
The contribution of this study is justified by the originality of the concept analyzed and the lack of any previous works dealing specifically with the possible relationship between the actions of employees and customer engagement. Furthermore, it analyses whether the fact that a customer has made a complaint or not has any effect on the causal relationships proposed. The theoretical and practical implications are included in the final part of the paper.
El clima competitivo actual allana el camino hacia el cambio en la gestión de la cartera de clientes tratando de mejorar la eficacia y la rentabilidad del enfoque relacional. En consonancia, el estudio de la relación empresa-cliente incorpora el concepto de «compromiso del cliente». Este nuevo concepto sugiere que los criterios transaccionales (readquisición, venta cruzada, nivel de uso) son insuficientes para evaluar la rentabilidad de cada cliente. Por lo tanto, hay que considerar el comportamiento no transaccional de los clientes, incluido el boca a boca y la publicación en blogs, para obtener una idea más exacta del valor actual y futuro de cada cliente.
De acuerdo con las ideas expuestas, en esta investigación se analiza el papel de los empleados de atención al cliente, que la mayor parte del tiempo se consideran portavoces de la empresa en la interacción empresa-cliente y, asimismo, influyentes en el nivel de satisfacción y compromiso. Esta interacción es fundamental en el caso de errores en el servicio o de gestión de reclamaciones.
La contribución de este estudio se justifica por la originalidad del concepto analizado y la falta de obras previas dedicadas específicamente a la posible relación entre las acciones de los empleados y el compromiso de los clientes. Además, se analiza si el hecho de que un cliente haya interpuesto una reclamación o no tiene algún efecto sobre las relaciones causales propuestas. En la última parte del artículo se incluyen las implicaciones teóricas y prácticas.
Companies have traditionally paid much attention to the transactional behaviour of consumers (repurchasing, cross-selling, level of use, length of the relationship, etc.) as these actions had an immediate effect on the sales volumes and results. However, the current competitive climate favours a new approach to managing the company–customer relationship which incorporates non-transactional behaviour (word-of-mouth, blogging or referrals, co-creation, amongst others), which can have a strong impact on company results, even if these are not registered with immediate effect. This has led to the inclusion of a new concept in specialized literature (customer engagement) which has become a prominent issue in the areas of both relational marketing and consumer behaviour. Research on customer engagement has focused its attention on analyzing customer profitability, using both their transactional and non-transactional behaviour (Kumar et al., 2010; Van Doorn et al., 2010).
This concept, considered to be one of the top research priorities by the Marketing Science Institute, emerged in marketing literature as an evolution of the relational paradigm. It is based on the continual interactions that firms maintain with their customers, and those between customers and other consumers, which affect non-transactional behaviour (Vivek, 2009). The peculiarity of these behaviours is that they are voluntary (Mollen & Wilson, 2010), and although they do not generate immediate revenue for the company, they help reinforce the company image and can influence future purchase decisions of other consumers, which could boost profitability and the value of the business over the long-term.
Engagement can be seen as an indicator that reflects the level of customer involvement and connection with the products, services and activities of the firm (Verhoef, Reinartz, & Krafft, 2010). Therefore, an engaged customer is the result of feeling sufficiently satisfied and from the company–customer interaction itself (Hollebek, 2011).
This interaction is of particular interest in services marketing. The literature highlights that in this interactive process those employees who are in direct contact with customers play a vital role (Berry, 1981; Gounaris, 2008; Paswan, Pelton, & True, 2005). This is due to the fact that employees often act as the spokespeople of the company and take an active role in delivering the service, providing information, showing customers how to make better use of the service acquired, process complaints or claims, apologize, propose solutions or alternative uses, and generate feedback, all of which is of great importance for the present and future value of the company. Aspects such as training, actions and motivation of employees can therefore help explain why certain customers are willing to become engaged customers.
However, the novelty of the concept of customer engagement and the relative shortfall of specific literature explain the lack of previous research on the subject and provide justification for this study. Our work, which is of exploratory nature, seeks to analyze the role of frontline employees in the level of customer engagement. This study is of interest to both academics and practitioners given that, as Seltzer, Gardner, Bichard, and Callinson (2012) indicate, frontline employees often act as public relations, by being the ambassadors of the company when they communicate and interact with external audiences (e.g., customers). The actions taken by these employees and their responses to customer enquiries are one of the most effective communication tools. In fact, consumers sometimes trust the messages conveyed by these employees more than any other source of communication (Van Laer & De Ruyter, 2010).
In this sense, our research analyzes the role of frontline employees to increase the perceived level of customer satisfaction when using a service (in this case, mobile phone operators). This study represents the first empirical research to tackle this matter. In a second step, we examine whether those customers who are dissatisfied have made a complaint or not, and whether making a complaint moderates the customer–employee interaction. In order to do this we consider engagement a specific construct that simultaneously includes non-transactional attributes such as co-creation or word-of-mouth (Bolton, 2011; Van Doorn, 2011).
Specifically, we wish to examine
- -
Whether employees, in their roles as spokespeople of the company, affect (i) the level of customer satisfaction and (ii) the level of engagement of these customers.
- -
Whether customers can become engaged customers on the basis of their feeling of satisfaction alone, or whether employees play a part in this role.
- -
Whether relationships where no complaint has been made are comparable to those where a complaint has been registered.
In order to achieve these objectives the second section describes the concept of customer engagement and analyzes the possible relationship with frontline employees. The third section presents the reference model and hypotheses that form the basis of the empirical study. The fourth section details the empirical study, whilst the subsequent section shows the results. To conclude we provide a section on discussions, implications, possible limitations and propose future lines of investigation.
Customer engagement and frontline employeesEngagement recently emerged in customer management literature and in the field of relational marketing (Verhoef et al., 2010). To date, international consulting firms have published numerous studies to analyze the concept by trying to assess its influence on business results (e.g., Smith & Wallace, 2010; Voyles, 2007). Academic investigations have added to the stream of professional studies, with authors such as Brodie, Hollebeek, Juric, and Ilic (2011) demonstrating their multidisciplinary approach through links with sociology, psychology, and of course, marketing. The concept is defined as behavioural manifestations of the customer towards the company that surpass purchases and repurchases of a product and service, resulting from different motivations of the individual (Van Doorn et al., 2010). In this sense, the non-transactional behaviour that customers display in a disengaged manner, which have a delayed effect on company results, should also be added to the economic transaction that is derived from sales (Kumar et al., 2010).
Engagement refers to the strength of the behavioural tie the customer has with the company (Van Doorn et al., 2010). This tie not only serves to retain present customers – by consolidating the relationship with the company – but also to attract new ones (Bowden, 2009). Engagement requires continuity and continual two-way communications as a result of the customer–company connection and the interactivity of these relationships (Druckenmiller, 2009). Terms such as connection, participation, involvement or absorption are used (Higgins & Scholer, 2009). The interactive experience and co-creativity of the customer are also highlighted in the relationship with the company (Brodie et al., 2011).
For the purpose of this study we will use the definition given by Van Doorn et al. (2010) which contemplates the behavioural profile of the construct, depending on the different motivations of the individuals. These authors believe that customer satisfaction is decisive in order to understand their level of engagement. This definition highlights the importance of certain behaviours of customers, manifested in a disengaged manner, which could hold future value for the company. This is understood to be the case as authors such as Kumar et al. (2010) state that a customer with a high level of engagement is more valuable for the company and provides increased profitability. For example, in the case of services which are characterized by their intangibility, there is a high level of uncertainty prior to signing a contract or service delivery, so the fact that a company already has these connected customers could strengthen the trust of other possible users, improve its brand image and even attract other customers. In this sense, our investigation seeks to analyze the importance of employee actions in determining the level of customer engagement.
The research by Gounaris (2008) is particularly interesting as it highlights that, within the services context, the interactions between frontline employees and customers influence the overall perception of the service quality and ultimately, customer satisfaction. This is due, in part, to the fact that services are characterized by inseparability and intangibility, and it is often the employees themselves who deliver the service and act as a communication channel (Berry, 1981; Cadwaller et al., 2010; Paswan et al., 2005).
Therefore, given that frontline employees play a significant role in the overall service rating, companies should take great care in managing their performance. Consequently, it makes sense to promote positive attitudes among these employees. Saari and Judge (2004) define employee attitude as an emotional state based on their own experiences in the workplace. This attitude is based on both objective and emotional issues such as salary, training received, work conditions, the socio-cultural profile of the employee, feelings towards the company or compatibility with company values, and ultimately determines whether employees are satisfied or unsatisfied (Gregory, Harris, Armenakis, & Shook, 2009). Along these lines, Allen and Grisaffe (2001) and Tharenou, Shaks, and Moore (2007) posit that employees with a positive attitude who feel satisfied are able to perform better, which in turn favours company results. Therefore, it is vital for a company to understand the needs and expectations of their employees to increase their level of satisfaction, as satisfied employees are likely to treat customers better and offer an improved service (Ahmed & Rafiq, 2003; Berry, Hensel, & Burke, 1976; Paswan et al., 2005; Wieseke, Ahearne, Lam, & Van Dick, 2009).
Furthermore, despite the great advances made in technology, telecommunications and virtual services, the interaction between employees and customers continues to be relevant and significantly influences the experience the customers receive (Ballantyne, 2003; Zeithaml, Parasuraman, & Malhotra, 2002). This point is highlighted by the fact that many consumers still prefer human contact to the virtual environments, which many prefer to use as an information channel. Authors such as Gummensson (1987), Ahmed and Rafiq (2003) or Ballantyne (2003) believe that employees form part of the value chain and that companies should strive to underline both the importance of the interdependencies that exist between all employees and the interfunctional coordination to maximize efficiency and final customer satisfaction. Specifically, the interaction between frontline employees and customers provides an excellent marketing opportunity for the organization. This manifests itself in various ways, not only in terms of the provision and delivery of services but also as an opportunity to understand customers better, their changing needs and expectations, level of satisfaction or dissatisfaction, the possibility of developing a relationship, etc. by means of feedback which is of great value for the future of the company (Ballantyne, 2003; Cadwaller et al., 2010).
Therefore, from a management point of view, aspects such as attracting, selecting and retaining the right staff (Rafiq & Ahmed, 2000), their development, training, motivation and rewards (Babakus, Yavas, Karatepe, & Avci, 2003; Cadwaller et al., 2010; Zhang & Bartol, 2010), understanding and trust (Ahmed & Rafiq, 2003), communication, work environment, procedural justice and sincerity (Paswan et al., 2005) and transparency (Ballantyne, 2003) help determine the level of employee satisfaction and are of great interest to organizational performance.
Based on the previous arguments we believe that, within the services context, the actions of the employees are important in determining both customer satisfaction and achieving the right levels of engagement for the company.
The influence of employees in customer engagementThe aim of this study is to examine the influence that frontline employees possibly exert on the level of customer satisfaction and engagement, within the service sector. We therefore propose a model by means of hypotheses that is based on customer management literature and is shown in Fig. 1. This model allows us to assess whether satisfaction is a necessary condition to generate customer engagement or whether the action of employees is sufficient to evoke this feeling and future behaviour.
The literature highlights that it is important to first satisfy the needs and expectations of employees before being able to satisfy the needs of consumers (Gounaris, 2008). This fact is relevant because the actions of all employees, particularly those who are frontline, influence the perceived experience of the customer (Zeithaml et al., 2002) and the assessment of the service quality received (Berry, 1981; George, 1990; Paswan et al., 2005; Wangenheim, Evanschitzky, & Wunderlich, 2007).
The overall sense of satisfaction is influenced by aspects such as correct and courteous treatment, the effort and interest shown, product knowledge, the ability to transmit clear and concise information, empathy, or solving problems, are highly valued by customers (Ahmed and Rafiq, 2003; Babakus et al., 2003; Gounaris, 2008). In order for customers to feel satisfied they should receive a level of service that meets or exceeds their expectations (Anderson, Fornell, & Mazvancheryl, 2004; Torres & Tribo, 2011). From an academic standpoint, unanimity appears to exist by postulating that there is a positive relationship between the quality of employee actions and the level of perceived customer satisfaction (Gummenon, 1999; Ballantyne, 2003), which allows us to present our first hypothesis:H1 An appropriate attitude of frontline employees positively influences customer satisfaction.
In line with the previous arguments we can assume that the attitude of the employees, in so much as they are sometimes considered the true spokespeople of the company (Seltzer et al., 2012), can lead to customers talking positively about the company, are willing to contract additional services or simply act as referrals (Cambra, Melero, & Sese, 2012). It is important to note that in some instances the service contract runs for a sustained period of time and customer satisfaction is not determined until the contract ends, or a certain level of initial dissatisfaction exists. In these instances, the employee–customer interaction can generate certain predisposition of the customer towards engagement. However, we are forced to tentatively pose this assumption due to the lack of previous studies that have explicitly examined this relationship.H2 An appropriate attitude of the frontline employees positively influences in the degree of customer engagement.
One of the fundamental pillars of relational marketing is building, maintaining and developing long-lasting and profitable relationships between a company and its customers (Morgan & Hunt, 1994). Furthermore, the theory postulates that satisfaction must exist for a customer to be willing to maintain a relationship, show commitment or speak favourably about the company and recommend it to other potential customers.
However, customer engagement is considered a global construct, which makes it necessary to determine the extent to which satisfaction affects the resulting level of engagement. In this sense, the work by Higgins and Scholer (2009) considers customer satisfaction a key antecedent in determining the extent of engagement, highlighting that it is a necessary condition. However, the work of Van Doorn et al. (2010) provides the strongest arguments to support this fact. These authors believe that engagement is a satisfaction-driven construct, meaning that satisfaction is a fundamental prerequisite.
Therefore, based on the previous arguments, we propose that:H3 Customer satisfaction positively influences their degree of engagement.
However, failures may occur in any relationship. Often, these failures are inevitable and can affect the level of customer satisfaction. These situations are known as “service failures” (Cambra, Berbel, Ruiz, & Vázquez, 2011; Maxham, 2001; Michel & Meunter, 2008) and have been defined as a real or perceived problems by the customer during the interaction between themselves and the company. In the case of services, these interactions determine the level of satisfaction of the users but also provide the company with the possibility of fixing the problem and regaining satisfaction. The actions which are taken subsequent to the service failure, aimed at fixing the problem, are denominated “service failure processes” (Bitner, Booms, & Tetrault, 1990; Grönroos, 1998; Varela, Vázquez, & Iglesias, 2009). In this sense the actions of the frontline employees play a decisive role (De Matos, Henrique, & Vargas, 2007), not only by processing the complaint but also in their role as spokespeople of the company they can offer an apology and propose possible solutions (Guenzi & Pelloni, 2004). For many customers the speed, energy and apparent honesty of these employees play a pivotal role in determining their level of satisfaction, even when a perfect solution has not been found (Mohr & Bitner, 1995).
Intuitively, we recognize that these service recovery processes are more complex than those situations characterized by an initial level of satisfaction. We therefore stress the interest in examining whether a complaint has been made or not, and whether this determines the degree to which frontline employees play a role in customer engagement. The attitude of the frontline employees influences both the satisfaction of customers and their level of engagement, and this influence can be stronger for those customers who do not show a state of initial dissatisfaction. Furthermore, given that the relationship is more intense during service recovery processes the link between satisfaction and engagement will also be more intense. Therefore, we propose:H1A An appropriate attitude of frontline employees positively influences customer satisfaction to a stronger degree during a service failure than in situations of initial satisfaction. An appropriate attitude of frontline employees positively influences the level of customer engagement more intensely during a service failure than in situations of initial satisfaction. The satisfaction of customers positively influences their level of engagement more intensely during a service failure than in situations of initial satisfaction.
In order to test the hypothesis posed in the model we developed a questionnaire aimed at mobile phone users. During 2012, the sector recorded more than 58 million lines in Spain, according to data of the National Commission of Telecommunications Market (CMT, 2012). However, in January 2013, the sector registered 52.7 million lines in Spain. This figure is certainly revealing given that more than 5 million lines were lost in one year. Furthermore, in January of the same year a record figure of mobile phone portabilities was reached. Specifically, 633,616 customers change operators every month. Since the emergence of virtual telephone companies the choices available have multiplied and competition has increased in the sector. In January, 170,000 mobile phone portabilities went to virtual telephone operators. We are therefore witnessing a sector with constant growth which is characterized by intense competitive pressure. This is mainly due to the fact that it is difficult to develop innovations that are difficult to imitate, which leads operators to try and gain market share through aggressive tactics to attract and retain customers (Polo & Sese, 2009).
For this reason, the employee–customer interaction can help stop the flow of customers leaving – and could even become a focal point of interest for new customers – and justifies the suitability for this study.
The scales used in the questionnaire were modified to suit the precise needs of this investigation, and were adjusted from previously validated scales in the literature (see Table 2). Additionally, to verify the validity and comprehension of the items we carried out a pre-test among marketing researchers from several Spanish universities and telephone service users. Subsequently we amended the scales. In order to measure the employee attitude variable we used a three item scale proposed by Karatepe (2006). This scale reflects the fundamentals of human resource management with regard to employee attitude (e.g., Saari & Judge, 2004; Tharenou et al., 2007) such as an assessment of their behaviour, interest and energy. The variable satisfaction was measured using two indicators – based on the works of Chin and Gursoy (2009) and Geyskens, Steenkamp, and Kumar (1999) – that express the overall rating by the customer after the service provision. This overall perception of satisfaction intuitively compares the service received with the prior expectations, which is presented in the works of Anderson et al. (2004) and more recently Torres and Tribo (2011). We considered the work of Sprott, Czellar, and Spangenberg (2009) as a starting point to develop the scale for the variable customer engagement. This scale is comprised of three indicators and makes reference to a series of disengaged behavioural manifestations by customers and adheres to the recommendations of renowned authors such as Bolton (2011), Brodie et al. (2011) and Van Doorn (2011), in so much as customer engagement is seen as an aggregation of constructs that contemplate the foundations of the global concept. For example, Van Doorn et al. (2010) recognize the role of word-of-mouth to help understand customer engagement. Furthermore, Kumar et al. (2010) considers that it is vital to also include concepts related to co-creation, word-of-mouth and referrals. The work of Bijmolt et al. (2010) also refers to word-of-mouth and co-creation. Given the exploratory nature of this study it seems appropriate to use a simple scale that encompasses the main behaviours related to engagement: recommendations and co-creation.
The inclusion criteria for individuals in the sample were: (i) adults of legal age, (ii) phone mobile service users and (iii) the current contract had been held with the same operator for at least two years. This last prerequisite is important because during this time it is possible that several employee–customer contacts had taken place which gives consumers an overall impression of the employees of the company, thus eliminating the possible bias of isolated employee actions. Furthermore, respondents were asked whether they had made a complaint or not. The average of the satisfaction variable for those users who had made a complaint was 1.96, whilst the figure for those who had not registered a complaint rose to 4.9. Here it is important to highlight, as Cambra, Melero, Sese, and Vázquez (2013) note that the mobile phone market heads the list for the poorest customer service. For this reason, a significant number of complaints and claims are made every year, which results in the sector residing first place in the classification of sectors for the highest number of complaints and unsatisfactory situations. According to data from the Telecommunications User Support Office, the number of claims registered for services in the mobile phone market reached 20.085 in 2011, which represents a 50% increase on the previous year, 2010.
The technical data for our study can be found in Table 1.
Technical data.
| Universe | Mobile phone users (of legal age) who had maintained their current contract with the same company for at least 2 years. |
| Geographical scope | National. Spain |
| Sample | 185 respondents |
| Sample method | Stratified by quotas |
| Type of survey | Telephone |
| Profile of the respondents | Sex: Male: 97 (52.43%); female: 88 (47.57%)Age: 18–25 years old: 66 (35.67%); 26–35 years old: 56 (30.27%); 35–50 years old: 41 (22.16%); >50 years old: 22 (11.89%)Complaint: 80 (43.24%); No complaint: 105 (56.76%) |
| Analysis technique | PLS and SPSS |
In order to assess the quality of the data obtained we analyzed individual and composite reliability, convergent validity and discriminant validity (using cross-loading and overview techniques). We should highlight that all the constructs were considered as first-order. First, we carried out an individual reliability analysis where the values corresponding to each item exceeded the threshold stipulated by Carmines and Zeller (1979). The same is true when assessing the composite reliability of the variables using Cronbach's Alpha and the Composite Reliability Index (see Table 2).
Measurement scales (reliability). (The scales run from 1=totally disagree to 7=totally agree).
| Variables/indicators | Results | ||||
| Average | Std. deviation | Cross loadings | Composite reliability | AVE | |
| Attitude of the employees (Karatepe, 2006) | |||||
| The employees use all their energy when advising me which service to choose | 3.70 | 1.671 | 0.8894 | 0.9313 | 0.8188 |
| The employees show interest in understanding my needs | 3.44 | 1.707 | 0.9056 | ||
| The overall behaviour of the employees is adequate | 3.48 | 1.768 | 0.9193 | ||
| Customer satisfaction (Chin & Gursoy, 2009; Geyskens et al., 1999) | |||||
| I am highly satisfied with the service of the company | 3.16 | 1.810 | 0.9419 | 0.9448 | 0.8954 |
| I have a good overall opinion of the service provided by the company | 3.21 | 1.752 | 0.9506 | ||
| Customer engagement (Sprott et al., 2009) | |||||
| I like to share my experiences with other consumers | 3.44 | 1.464 | 0.7767 | 0.8466 | 0.7491 |
| I am willing to continue doing business with the company in the future | 3.57 | 1.961 | 0.7483 | ||
| I would recommend the services of this company to friends or family | 3.15 | 1.699 | 0.8856 | ||
The convergent and discriminant validity of the model was also confirmed, as the average variance extracted (AVE) is higher than 0.5 (Fornell & Larcker, 1981) and the comparison of the square root of the AVE of each construct exceeded the correlations between the variables (see Table 3).
Discriminant validity of the variables in the structural model.
| Variables | Attitude of the employees | Customer satisfaction | Customer engagement |
| Attitude of the employees | 0.9048 | 0 | 0 |
| Customer satisfaction | 0.6729 | 0.9468 | 0 |
| Customer engagement | 0.6576 | 0.7905 | 0.8655 |
Data appearing on the main diagonal are the square roots for the AVE (average variance extracted) of the variables. The rest of the data represent the correlations between constructs. All correlations are significant p<0.01 (Fornell & Larcker, 1981).
Once the quality of the data was confirmed, we carried out a structural equation analysis Partial Least Squares (PLS) using the software program SmartPLS 2.0M3 version to test the robustness of the model and hypotheses posed. This methodology has recently been defended in the area of marketing (Chung, 2009; Hair, Ssrstedt, Ringle, & Mena, 2012; Lindgreen, Palmer, Wetzels, & Antioco, 2009; Reinartz, Haenlein, & Henseller, 2009) and is appropriate to meet the objectives of this study. We chose this method as our study is based on the prediction and relationships of an exploratory nature (Roldán & Sánchez-Franco, 2012). Furthermore, we consider our model to be relatively incremental as it is partly based on previous models (attitude–satisfaction) but also includes new measurements (attitude of the employees-engagement; satisfaction-engagement) (Chin, 2010). In addition, PLS is robust for moderately sized samples (Cassel, Hackl, & Westlund, 2000; Reinartz et al., 2009).
ResultsGeneral modelIn this section we analyzed the structural model using the SmartPLS software. In order to do this we calculated the path coefficients and t values of the parameters obtained using the Bootstrap technique. These measurements confirm the precision and stability of the estimations. Table 4 shows the significance of the structural paths and acceptance or rejection of the three hypotheses posed in the model.
Results of the structural model.
These results show the significance of the three hypotheses posed. Furthermore, R2 of the dependent variables are acceptable (R2 SAT=0.4531; R2 ENG=0.7942). The results of the estimation show that the parameters associated with the three contrasted relationships are positive and significant. We can therefore confirm that a positive attitude of the employees can generate customer satisfaction in the employee–customer interaction, within the service industry context. This satisfaction positively and significantly influences the level of customer engagement, which can be of great value to the company, in non-transactional terms. Lastly, we examined the possibility of the existence of a direct relationship between the attitude of the employees and the level of customer engagement. This relationship is significant and suggests that their interaction with customers has a direct effect on the level of engagement.
To assess the predictive relevance of the model we used the Stone–Geisser test. The resulting Q2 value of the two dependent variables was satisfactory (Q2-SAT=0.3996; Q2-ENG=0.4909). Thus, we can infer that the dependent variables can be predicted via the independent variables. Additionally, we calculated the goodness of fit proposed by Tenenhaus, Esposito, Chatelin, and Lauro (2005) that shows a value of 0.7008 and can be considered high based on the references proposed by Cohen (1988).
Indirect effectAs well as calculating the relationships proposed in the model, owing to the fact that the third hypothesis – which establishes a link between the attitude of the employees and the level of customer engagement – is significant, we needed to calculate the mediating effect of satisfaction. Table 5 shows the data corresponding to the calculation of the total effect (direct and indirect) of the attitude of the employees on the level of customer engagement. To verify the significance of the indirect effect we used the Sobel (1982) test and obtained the “z” statistic.
As Table 5 shows, the mediator effect of satisfaction is confirmed with the “z” statistic in the relationship that relates the attitude of the employees (ACT) to the level of customer engagement (ENG), with a p-value <0.001. The size of the indirect effect on the total effect is given by the VAF (variance accounted for) proposed by Iacobucci and Duhachek (2003). In this sense, almost 84% of the total impact of the variable ACT on ENG is due to the indirect effect. The total effect of ACT over ENG, including the indirect effect that satisfaction exerts, substantially exceeds the direct effect (0.6577 versus 0.1068). Therefore, the mediator effect of customer satisfaction in the relationship between the attitude of the employees and customer engagement is demonstrated. Therefore, despite the fact that the results indicate that the actions of frontline employees can be a sufficient condition to generate engagement, satisfaction logically reinforces this effect.
Analysis of the moderator effectAs we outlined in previous sections, this investigation aims to assess whether the degree to which frontline employees can affect customer engagement is conditioned by the existence or absence of a complaint. In order to do this we firstly produced a multi-sample, following the guidelines of Chin and Frye (2003), which consists of comparing the β coefficients for each of the sub-samples. This first analysis provides an overall vision which should be subsequently corroborated with the moderator effect. The results of this analysis are shown in Table 6.
Results of the structural multi-sample.
| Impact on the endogenous variables | Complaint (n=80) | No complaint(n=105) |
| Path coefficients (β)T value (bootstrap) | Path coefficients (β)T value (bootstrap) | |
| H1: attitude of the employees→satisfaction | 0.6945*** (10.911) | 0.4818*** (11.9738) |
| H2: attitude of the employees→customer engagement | 0.3488** (2.7346) | 0.0874* (2.0338) |
| H3: satisfaction→customer engagement | 0.5437*** (7.6653) | 0.3275*** (6.5078) |
The data show that the proposed relationships are significant in both sub-samples; that is to say that the actions of frontline employees are decisive in determining the level of customer engagement, irrespective of whether customers have made a complaint or not. Furthermore, from analyzing the path coefficients we can deduce that the values are higher for the sub-sample “complaint” meaning that the actions of employees are more important where a complaint has been made. This all lends weight to the argument that in the case of a service failure the attitude of the employees possibly has more importance in terms of generating customer engagement.
However, to assess whether these differences are significant it is necessary to carry out an analysis based on the T-test proposed by authors such as Chin and Frye (2003) or Keil et al. (2000). The results of this test are included in Table 7. We can conclude that the existence or absence of a complaint moderates all the relationships proposed in the causal model. Therefore, we can confirm that, in the case of sector analyzed in this study, the fact that a complaint has been made can influence the perception the customers have of the activities of frontline employees. Where no complaint has been made customers value a favourable attitude and satisfaction with the service received, but when a complaint exists customers tend to value more the performance of the employees who deal with them.
Results of the analysis of the moderator effect.
| T-test | Complaint (β) (n=80) | No complaint (β) (n=105) | SE | SP | T-value | |
| Complaint | No complaint | |||||
| H1: attitude of the employees→satisfaction | 0.6945 | 0.4818 | 0.0637 | 0.0653 | 0.608 | 2.287* |
| H2: attitude of the employees→customer engagement | 0.3488 | 0.0874 | 0.0858 | 0.0942 | 0.839 | 2.036* |
| H3: satisfaction→customer engagement | 0.5437 | 0.3275 | 0.0696 | 0.0811 | 0.697 | 2.028* |
*p<0.05 (t=1.96). SE: standard error. SP: Separate variance estimate.
The importance of customer engagement lies in the impact this concept can have on the future of the company. As Van Doorn et al. (2010) commented, an engaged customer can improve the reputation and financial value of companies. Firstly, customer loyalty generates a certain flow of future sales and revenue, whilst non-transactional activities such as blogging, referrals, co-creation and positive word-of-mouth can serve to attract new customers and also generate future sales. Secondly, these behaviours help strengthen the reputation of the company over the long-term as consumers tend to trust the recommendations of other buyers. Companies should therefore dedicate resources to identifying the most profitable customers and those who could potentially become so based on their non-transactional behaviour.
These ideas are particularly relevant to the service sector, where companies should strive to highlight the tangible aspects of their services and demonstrate their ability to meet customer expectations. In this sense, our investigation aimed to analyze the role that frontline employees could potentially play in determining the degree of customer engagement. Furthermore, we wished to examine whether managing the employee–customer interaction was a sufficient condition to achieve an engaged customer. In this regard, as we proposed in our model, the role of employees is relevant. Data suggest that the employee–customer interaction is sufficient to generate customer engagement, although this is reinforced by customer satisfaction. Satisfaction is vital for achieving connected customers. However, our results suggest that the actions of employees could suffice to generate customer engagement, even though satisfaction reinforces the effect. This is a groundbreaking result and holds important implications for business practice.
Furthermore, we have examined whether the fact that a complaint has been made moderates the role of frontline employees in the level of customer engagement. Specifically, when a complaint has been made the impact of these employees is greater than in other types of situations such as those (i) characterized by a certain level of initial satisfaction or (ii) where no complaint was registered. This makes sense as during a complaint and service recovery process the number of employee–customer interactions is greater and the actions taken by the employees can even substitute the proposed solution and the real service quality. Again, this finding has important practical implications, in so much as the employees become spokespeople of the company.
This study also highlights the importance of employee–customer interactions in order to satisfy customers and generate their engagement. Aspects such as training or motivation determine employee performance and are crucial to understanding the success of a company, more so in the case of the service industry. Employees who are better prepared and more motivated can generate increased profits for the company thanks to the positive effects on non-transactional behaviour of customers.
From a practical point of view, companies should understand that the performance of their employees is important both as a way to generate satisfaction as well as a means of acquiring customers who are willing to build ties with the company. In this regard, companies should consider developing a closer rapport with their customers and employees, and think carefully about the selection criteria, training, motivation and salary conditions they use to manage the workforce.
This study represents a first attempt to explicitly examine the role of frontline employees and the level of customer engagement. Despite the encouraging results obtained, we should recognize a series of limitations. Firstly, the employee–customer interaction is not the only communication channel which companies use and this could affect the overall assessment of how the firm manages its relationships with customers. The new technologies may have much to say, as authors such as Wigley and Lewis (2012), Waters, Burnett, Lamm, and Lucas (2009) or Yang and Kang (2009), point out. Secondly, owing to the fact that this is a pioneer study and the size of the simple, we should be cautious with the robustness of the results obtained. Furthermore, the lack of consensus on how to measure the engagement construct makes our task of establishing a generally approved proposal difficult. The scale used in this study may be perhaps too basic but it has enabled us to empirically test a phenomenon that has not been tested to date. Future investigations could take up the challenge of proposing a more comprehensive scale. This would require a larger study, although our data show a definite path of inquiry. For example, it could be interesting to compare results with different sectors. Furthermore, the data are cross-sectional and are based on the opinions of the interviewees. We should note the possible presence of the common method bias. This bias refers to the proportion of variance of the variables related to the measurement method (Podsakoff, Mackenzie, & Lee, 2003). In order to counter this possible bias Podsakoff et al. (2003) recommend using procedural and/or statistical strategies. Both strategies have been simultaneously used in several researches in marketing (e.g., López-Sánchez, González, & Santos, 2010). In terms of procedure, which affects the basis of the study and tries to eliminate – or, at least minimize – the impact of this bias, we designed the study such that (i) we guaranteed the anonymity of the respondents, (ii) clarified that they were no right or wrong answers, (iii) used previously validated scales and (iv) through the use of various pre-tests with different reference groups deleted possible ambiguities from the scales and ensured that the items were simple, specific and concise. In terms of the strategies adopted for statistical analysis, we used Harman's single factor test. In the factorial analysis no single factor was identified that explained the variance of all the items, suggesting that it is unlikely that bias exists on the basis of using a single method. The principal factor explains 39.10% of the variance. Thus, given that there is no single factor that explains more than 50% of the variance, the data obtained can be accepted as valid (Podsakoff & Organ, 1986).
With regard to possible future lines of investigation we believe it would be interesting to study the influence of other channels of communication to assess which ones are most effective, and thus, increase the efficiency of resource allocation. It would also be interesting to replicate the study in different industries to determine the extent to which the nature of the product/services and the competitive structure of the sector can moderate the proposed relationships.
FundingThe authors express their gratitude for the financial support received from the Spanish Government CICYT (ECO 2011-23027), from the Regional Government and FEDER's funding (Generés S09) and the national grant FPU10/4448.
Conflict of interestThe authors declare no conflict of interest.
