Keywords

1 Introduction

Services are typically provided by front-line employees, meaning the ability of employees to perform these services correctly is important for service providers to maintain high service quality [1]. The capability required in the workplace is generally called competency. Boyatzis defined competencies as the personal characteristics that lead to successful performance in a job role [2].

Research on the competencies of front-line employees has hitherto focused on two specific competencies. The first is technical competency [3,4,5], which refers to having the knowledge to accomplish the job. The second is emotional competency [4,5,6,7], which means the capability to act in a manner that displays an understanding of people’s feelings. These competencies represent the knowledge and behavioral tendencies of front-line employees.

Value in customer service is generated when an appropriate interaction is carried out in various contexts involving customers and external environments, such as customers’ emotions or priority of the service in the workplace. An appropriate interaction is based on cognition, meaning the employee properly understands the context of the service and considers his/her behavior. Value co-creation is achieved by using different resources and integrating them according to context [8]. Therefore, the conventional way of viewing competencies, in which knowledge and service behavior are treated without context, cannot explain the reason behind resource integration. Consequently, it is necessary to understand competency from the aspect of cognition by enabling appropriate resource integration.

In this study, we define “cognitive competency” as the personal characteristics of cognitive processes that enable appropriate resource integration. Understanding cognitive competency can help clarify employees’ capabilities required to take appropriate actions to perform a certain job. Therefore, the purpose of this study is to identify features that represent cognitive competencies utilized by front-line employees in the hospitality industry.

We created a questionnaire survey to quantitatively identify the characteristics of cognitive competency. Specifically, the survey collects the cognition of front-line employees during service. First, we proposed a cognitive process model to be used during service that can serve as a framework for creating questionnaire survey questions. The qualitative grounded theory approach (GTA) was used to create the model. The proposed method is shown in Fig. 1 as follows. First, in study 1, retrospective interviews were conducted to create a model that represents the cognitive process of front-line employees during service. This interview was conducted referring to footage of their customer service during an experimental environment. We then analyzed the interview data based on GTA. Next, study 2 identified the features representing cognitive competencies. A descriptive questionnaire survey to obtain cognition during service was created in accordance with the cognitive framework in study 1 and was answered by 155 front-line employees. The survey data were assigned semantic codes and the number of codes was counted, enabling the use of the quantitative principal component analysis.

Fig. 1.
figure 1

Overview of the proposed methodology.

Full size image

This analysis contributes to hospitality research by extending the understanding of existing competency concepts from the viewpoint of the cognitive competency that causes service behaviors. Furthermore, by focusing on cognition, we have derived a new concept of service behavior⁠—serving not to serve⁠—that cannot be expressed using behavioral characteristics.

2 Literature Review

2.1 Service Interaction

In recent years, the idea of value co-creation has become widespread, where experience gained as a result of the interactions between service providers and customers is a source of value [9]. Companies that provide services aim to maintain and improve market competitiveness by providing customers with better experiences [10]. Since service behaviors have a significant impact on customer sentiment and decision-making [11, 12], value is co-created when employees and customers interact and influence each other [13,14,15]. Additionally, differences in employees’ capabilities can create heterogeneity of the provided service and its quality [16]. For this reason, knowledge that contributes to maintaining a sufficient level of competencies for front-line employees is necessary to enhance customer value experience.

Several studies have tried to determine how to enable front-line employees to provide better service. For example, in recent years, there has been a movement towards more effective service provision using technology [17,18,19]. Researchers have indicated that it is necessary to understand the behaviors and characteristics of employees. For example, Victorino et al. [20] pointed out the importance of studying the service provision process by understanding customers and employees from psychological and emotional perspectives. In addition, Subramony et al. [12] indicated that research should explore the temporal and dynamic aspects of emotional labor. Such research is crucial to explain how employees learn and adapt to emotional display regulations. The dynamic aspect means that employees control their emotion sequentially during customer service. This is due to the real-time interactions between employees and customers in addition to employees’ personalities.

Therefore, the competency of front-line employees is important to ensure a highly satisfactory customer experience. In light of this, it is increasingly important to understand cognitive competencies to deal with the dynamic aspects of service interactions.

2.2 Competency of Front-Line Staff

The competencies necessary for front-line hospitality industry employees can be divided into technical competencies [4, 5] and emotional competencies [3,4,5,6,7]. Technical competency refers to the knowledge required in each industry, while emotional competency refers to the capability to act with an understanding of emotions, including others’ and their own [3].

Competency depends highly on individuals, thus research on emotional competency is becoming increasingly important. Researchers have been studying emotional competency in the context of customer satisfaction [3, 6] and service recovery [5, 7]. Employees with high emotional competency have the ability to increase customer satisfaction and loyalty [3, 6]. Previous studies have shown that emotional competencies are important in improving customer satisfaction.

The previously mentioned studies examined competencies regarding knowledge and behavioral characteristics based on understanding human emotion. However, since customer service is provided under various contexts, such as customers’ situation and external environment, the extant understanding of competency cannot fully describe the competencies that front-line employees utilize to create value in their working place. Therefore, this study focuses on the cognitive competencies that front-line employees use to understand context, which includes customers and the external environments.

3 Study 1

3.1 Data Collection and Sample

In order to create a model that represents the cognition of front-line employees during service, retrospective interviews were conducted with front-line employees in a mock-up experiment in study 1. The interview data were analyzed based on GTA. We chose flight attendants because they have to provide customer service to passengers by considering passenger safety, hence cognitive competence is crucial in their service.

To record the cognition of flight attendants, we conducted recording experiments in a cabin mock-up that imitates the actual cabin environment. The experiment participants included three junior employees who had experience of less than three years and senior employees with more than 10 years of experience. The recorded customer service time was approximately 25 min and the customer service actions included serving drinks, cup collection and those provided during passenger boarding phase. The entire in-flight environment during the service and interactions between the flight attendants and passengers were recorded to capture the customer service process.

In this experiment, the persona and scenario were prepared for each passenger role in order to make the experimental environment realistic. These settings enabled the flight attendants to carry out in-flight service as usual.

After recording the cabin mock-up service, the retrospective interviews were carried out. The flight attendants were asked to recall what they were thinking when performing the service, while referring to the footage recorded during the experiment. A semi-structured interview was used, which is an interview method where the interviewer decides beforehand what to broadly ask, but he/she is often guided by the answers of the respondents.

3.2 Analysis

GTA was used to analyze the interviews. GTA is a qualitative research method developed in the field of sociology, which reveals the abstract theory concerning the process of interaction and phenomena between characters, based on the researchers ground the analysis on the data [21, 22].

Data analysis was completed in three steps. First, semantic codes are attached to the interview data based on a deep understanding of these data. Second, these semantic codes are grouped into similar semantic categories. Third, researchers identify the relationships between categories by which the theory was generated.

We consider that in-flight services, including safety services and interactions between flight attendants and passengers, depend on the cognitive competencies of the flight attendants. Therefore, our aim was to model the theory embedded in the cognition during the in-flight service by using GTA to analyze interview data.

3.3 Results

The created categories and codes are detailed in Table 1, along with their hierarchical relationship. Based on Table 1, we created flight attendants’ cognition model during service in Fig. 2. We used the business process model notation (BPMN) to illustrate the flight attendants’ cognitive process generated in the GTA. In particular, the process was formalized using BPMN for the large category “Cognition of decision making for customer service behavior” shown in the bottom of Table 1.

Table 1. Created categories, codes, and sub-code.
Full size table
Fig. 2.
figure 2

Cognition model during service.

Full size image

The cognitive process model shown in Fig. 2 can be divided into two parts, the upper and lower halves. The lower half of Fig. 2 corresponds to “Cognition of customer service environment” in Table 1. The hierarchical relationship of the large categories, categories, codes, and sub-codes are described in the diagrams of the lower half of Fig. 2. The upper half of Fig. 2, corresponds to “Cognition of decision making of customer service behavior.” During decision making, codes and sub-codes belonging to the large category “Cognition regarding customer service environment” represented as dotted arrows in Fig. 2.

Here, we explain flight attendants’ decision-making process in detail. After identifying customers’ behaviors, flight attendants make decisions about their service behaviors. This decision making is represented by the two ramifications in Fig. 2, namely “should observe a passenger more or not” and “whether they should interact with a passenger or not.” If they decide not to observe the passenger, they proceed to the next ramification, otherwise the decision is taken once more at the subsequent evaluation. If they choose to interact with a passenger, they will decide to provide “service behavior,” and otherwise they “Do not serve.” As mentioned above, in each ramification, “Cognition about customer service environment” in the lower half of Fig. 2 is referred to, but at the same time, “Examine service behavior,” positioning at the lowest in the upper half of Fig. 2, is also referred to. Though the “Examine service behavior” belongs to the large category “Cognition regarding decision making of customer service behavior,” it is freely referred to in actual decision making.”

Therefore, employees made decisions about whether or not they should interact with a customer. Interestingly, the employee has a choice of behavior even when service behaviors are not performed from the customer’s viewpoint but only in employees’ cognition. We call this phenomenon “serving not to serve.”

4 Study 2

4.1 Questionnaire Survey

Development of Questionnaire Survey.

In order to identify the cognition of flight attendants during service, a descriptive questionnaire survey was developed based on the structure of the cognitive model during service. Respondents first watched videos about passengers’ behaviors. Two different videos showing passengers on board were viewed. The scenarios in the two videos were as follows:

  • Scenario 1: a passenger who is sweating because he has boarded in a hurry.

  • Scenario 2: a frustrated lady who is on board with her children.

Second, respondents choose whether they should serve the passenger or not. Finally, respondents explained in writing why they chose that option. As such, respondents’ cognition leading to a decision could be identified.

The decision “not to serve the passenger” was rarely chosen in the three pilot surveys. Therefore, we divided the questionnaire into two parts: (1) cognition when respondents decide to serve after watching the video and (2) the cognition about “serve not to serve” by asking about their past experience when the respondent decided not to serve.

Collect Questionnaire Survey Data.

The questionnaire was given to flight attendants, and it took respondents 20–30 min to complete. The number of valid responses was 155. In addition, respondents noted their years of experience at the beginning of the questionnaire. The information on the years of experience is required to identify cognitive competency. Table 2 shows the number of respondents by year. The cognitive data derived from scenarios 1 and 2 are integrated and analyzed.

Table 2. Number of respondents by years of experience.
Full size table

Extract Variables of Cognition.

Structured semantic codes enable qualitative data to be analyzed quantitatively [23]. In the analysis of structured semantic codes, code frequency is used as a variable [23]. Code frequency means how many times a specific code appears in one answer and makes it possible to understand which semantic concepts are frequent or rare. In this study, each code frequency is used as a variable to understand how frequently each respondent considers a semantic concept. These variables quantitively represent the characteristics of cognition extracted from questionnaire survey.

4.2 Analysis

Extraction of Composite Variables Representing Cognitive Competencies.

When capturing the cognitive competencies of 155 flight attendants, it is difficult to extract features of cognition by using each variable of code frequency because the number of variables is too large. Therefore, we reduced the dimensions of the variables using principal component analysis to understand which frequency of semantic codes can effectively explain respondents’ heterogeneity. The newly created composite variables could explain cognitive competency.

The questionnaire survey included two parts. The first part referred to cognition leading to decision making and the second part to cognition leading to “serving not to serve.” We analyzed the two parts separately.

Verification of Variables Representing Cognitive Competencies.

The feature of cognitive competency cannot be judged only by the score of the composite variable obtained by principal component analysis. Therefore, k-means clustering was performed using composite variables whose cumulative contribution rates were above 0.70. The tendency of the years of experience in each cluster was used as a criterion to judge cognitive competency. The number of clusters was determined by referring to gap statistics.

4.3 Results

Part 1 of Questionnaire.

Part 1 of the questionnaire extracts cognition referred to when respondents decided to serve. In this section, we show the results of the analysis from the viewpoint of the number of semantic codes, namely code frequency.

Results of Principal Component Analysis.

Table 3 summarizes the information up to the fourth principal component with a cumulative contribution rate exceeding 0.70. The right-hand column shows the interpretation of each composite variable.

Table 3. Result of principal component analysis for part 1.
Full size table

For example, FR1_PC1, which is the first principal component, is the variable that can best represent the data with a contribution rate of 0.37. For example, when the value of FR1_PC1 is large, it there is a large amount of cognition regarding safety and the physical condition of passengers.

The composite variables can effectively represent the cognitive competencies of flight attendants. By cluster analysis with information on the years of experience, researchers examined if the composite variables are effective as a feature of cognitive competency.

Results of Cluster Analysis.

Clustering was performed by the k-means method using the four composite variables in Table 1. From the calculation of the gap statistic, the appropriate number of clusters is k = 2 and the data are divided appropriately. Table 4 shows the position of centroid of each cluster. The data are divided into two by composite variable FR1_PC1, which indicates the amount of consideration regarding the safety and the physical condition of passengers. The properties of the cluster can be interpreted as follows:

Table 4. Value of centroid (part 1).
Full size table

  • Cluster 1 has a sufficient but relatively less consideration of safety and the physical condition of passengers.

  • Cluster 2 has a high consideration level of safety and the physical condition of passengers.

Table 5 shows the number of respondents allocated to each cluster by years of experience. Cluster 1 consists of 80 respondents and cluster 2 of 75 respondents. From Table 5, the relationship between cluster allocation and years of experience can be interpreted as follows:

Table 5. Number of respondents allocated to each cluster by years of experience (part 1).
Full size table

  • Most respondents with up to 9 years of experience belong to cluster 1.

  • Most respondents with experience of 10 years and above belong to cluster 2.

That is, the clusters created by the composite variable have a strong reliance on the years of experience.

Further, senior employees with more than 10 years of experience consider safety and the physical condition of passengers more than juniors do. Therefore, the cognitive competency represented by composite variable FR1_PC1 refers to estimating passenger safety and physical condition, that is, risk perception. As a result, cognitive competency in part 1 is the amount of cognition about the safety and physical condition of passengers.

Figure 3 shows the respondents for the first and second composite variables and the clusters to which they belong are shown. Again, the clusters are separated by risk perception, namely composite variable FR1_PC1.

Fig. 3.
figure 3

Plot of respondents for first and second composite variable (part 1).

Full size image

Part 2 of Questionnaire.

In the second part of the questionnaire, we identified the cognition when respondents decided not to serve passengers, namely “serving not to serve.” This second part of the questionnaire was developed because it was not possible to extract the cognition about the decision not to serve customers from the first part.

Results of Principal Component Analysis.

Table 6 summarizes the information up to the fourth principal component with a cumulative contribution rate exceeding 0.70. The right-hand column shows the interpretation of each composite variable.

Table 6. The result of principal component analysis for part 1.
Full size table

For example, FR2_PC1, which is the first principal component, is a variable that can represent the data collectively with a contribution rate of 0.50. For example, when the value of FR2_PC1 is large, a high consideration of the passenger’s emotions and interpretation of their behaviors exists when making the decision not to serve passengers.

The proposed composite variables are effective to represent the cognitive competencies of flight attendants. By cluster analysis using information on the years of experience, researchers examined if the composite variables are effective as cognitive competency features.

Result of Cluster Analysis.

Clustering was performed by the k-means method using the four composite variables shown in Table 6. From the calculation of the gap statistic, k = 4 is the most suitable partition for the data. Table 7 shows the position of the centroid of each cluster. The data are divided into four clusters by composite variable FR2_PC1, which represents the amount of consideration regarding passenger’s emotions and behaviors.

Table 7. Value of centroid (part 2).
Full size table

The properties of the cluster can be interpreted as follows:

  • When cluster 1 decides not to serve, it has the least amount of consideration regarding passenger’s emotions and behaviors.

  • Cluster 2 has the second lowest amount of consideration regarding passenger’s emotions and behaviors when deciding not to serve.

  • When cluster 3 decides not to serve, it has the second largest amount of consideration regarding passenger’s emotions and behaviors.

  • Cluster 4 has the largest amount of consideration regarding passenger’s emotions and behaviors when deciding not to serve.

Table 8 shows the number of respondents allocated to each cluster by years of experience. From Table 8, the tendency of cluster regarding years of experience can be interpreted as follows:

Table 8. Number of respondents allocated to each cluster by years of experience (part 2).
Full size table

  • Most respondents with up to 9 years of experience belong to clusters 1 and 2.

  • Most respondents with 10 or more years of experience belong to clusters 3 and 4.

In this way, the cluster membership by the composite variables has a strong relationship with the years of experience. A summary of the cluster’s properties is as follows:

  • Cluster 1 consists of flight attendants with up to 9 years of experience that have the least amount of consideration regarding passenger’s emotions and behaviors.

  • Cluster 2 mostly consists of flight attendants with up to 9 years of experience and some with more than 10 years of experience. They have the second least amount of consideration regarding passenger’s emotions and behaviors.

  • Cluster 3 mostly consists of flight attendants with more than 10 years of experience and some with up to 9 years of experience. They have a relatively large amount of consideration regarding passenger’s emotions and behaviors.

  • Cluster 4 mostly consists of flight attendants with more than 10 years of experience and some with up to 9 years of experience. They have the largest amount of consideration regarding passenger’s emotions and behaviors.

From the above, seniors understand more than juniors do about the emotions and behavior of passengers when they decide not to serve. Therefore, the cognitive competence represented by composite variable FR2_PC1 is thoughtfulness for passengers when deciding not to serve.

Figure 4 shows the respondents for the first and second composite variables and the clusters to which they belong. Clusters are separated by combined variable FR2_PC1 which represents the thoughtfulness for passengers when deciding not to serve.

Fig. 4.
figure 4

Plot of respondents for first and second composite variables (part 2).

Full size image

5 Discussion

The theoretical and practical contributions of this research are as follows.

5.1 Extending the Concept of Competency in Hospitality Research

This study extended the concept of competency by clarifying the meaning of cognitive competency. The competencies of hospitality industry employees in previous studies have been divided into technical [4, 5] and emotional competencies [3,4,5]. These two competencies together represent the degree of achieving goals.

Unlike this previous understanding of competency, the cognitive competency proposed in this study uses the features of the process to achieve a result, namely cognition, thus making it possible to describe the competencies that enable front-line employees to adapt to dynamically changing environments in the actual working place.

We created a cognitive model that comprehensively describes the cognitive process of front-line employees by conducting retrospective interviews using videos. By constructing the model, it was possible to perform quantitative analysis on employees’ cognition. As a result, the cognitive features could be expressed quantitatively by the frequency of the semantic codes. In the proposed model, we clarified the structure of cognition during service, where the environment is taken into account in service behavior decision-making. Furthermore, we also identified the components constituting cognition by detailed categories, codes, and sub-codes.

Based on the above, we constructed a cognitive model during service that covers the cognitive components and processes during in-flight service, by which the cognitive competencies in the hospitality industry can be identified.

5.2 A New Competency of Hospitality from the Cognitive Perspective

In this study, we clarified the concept of “serving not to serve,” meaning that front-line employees do not serve based on their considerations about customers. This concept is novel to service research, as it has not focused on cognition but on actual behavior. In previous research on service operations, the focus was on the behavior of employees and customers, highlighting the importance of adopting an interdisciplinary approach, such as utilizing knowledge from psychology and emotions [20].

This study clarifies the cognitive competencies of front-line employees by identifying the cognitive processes during service and analyzing their features. In this way, the concept of “serving not to serve” was demonstrated by the data. As previously mentioned, this concept cannot be identified in studies that analyze employee behavior, such as technical and emotional competencies. Further, there exists a significant difference in the perceptions of employees and customers during service. Even when customers think a service is not being provided, employees intendedly decide not to serve customers based on contextual awareness. This research contributes to the understanding of the concept of hospitality by clarifying the concept of “serving not to serve” using cognition.

5.3 Indicating Differences in Cognitive Competencies

From the quantitative analysis, it there are different cognitive competencies required to serve and not to serve. Subramony et al. [12] indicated that the dynamic aspect of service should be reviewed to understand how employees in industries requiring emotional labor are adapting their feelings and thoughts and how they are developing these skills. The dynamic aspect means that employees change their emotional control method sequentially during customer service. This is due to real-time interaction between employees and customers and employee personality.

This study clarified the disparity of cognitive competency based on their years of experience. During service, the cognitive competency of “risk perception” is important, and is front-line employees decide not to serve, the competency of “thoughtfulness for passengers” is important. These cognitive competencies have strong correlations with the years of experience. Therefore, the more experienced the employees are, the higher their competencies. Promoting employee training with a focus on these cognitive competencies can improve the quality of front-line employees’ hospitality to create competitive advantage.

Although our study only used two scenarios in the questionnaire, the results will likely not change in another scenario. The results from watching the videos indicated risk perception that relate less to the actual scenarios. Therefore, our findings are not affected by the considered scenarios.

5.4 Improving Employees’ Hospitality

As a practical contribution, this study pointed out the importance of training for increasing the cognitive competency in the hospitality industry. The cognitive competency is a core competency for the hospitality service quality by determining employees’ performance according to the change of context.

Applying technologies for employee education will be of benefit to managers to improve employees’ cognitive competency. For example, recording what employees see is needed to understand their cognition. Therefore, the recording devices such as smart glasses must be useful for understanding the cognitive competence. Furthermore, e-learning using smart devices can be used for employee training. The cognitive competency is a difficult factor for learners (employees) to monitor by themselves. Therefore, service managers should introduce e-learning systems indicating the real-time competency level of learners.