Individual Employees' Service Failures and Customer Satisfaction with the Firm

Abstract

Purpose: Do close linkages among employees during service encounters always enhance customer satisfaction? Drawing on literature in social psychology, this research argues that under certain circumstances close linkages among employees undermine customer satisfaction. More specifically, this research explores a service failure context and shows that higher task interaction among employees during service encounters leads to higher perceived firm entitativity, resulting in an individual employee's service failure being detrimental to customer satisfaction with the firm. Data and research methodology: A series of experiments using scenarios across different service contexts was used in order to test hypotheses. Data was collected via Amazon Mechanical Turk and the models were tested using Hayes PROCESS. Results: The results show that interactions among employees during service result in 1) lower satisfaction with the firm and 2) lower revisit intentions in the future when there is a service failure by an individual employee. Following the main effects analysis, mediation analysis shows that the effect of employee interaction on customer satisfaction with the firm and revisit intention is mediated by perceived firm entitativity. Implications: By examining contexts where employee interaction may be detrimental to firms, this research provides novel insights on how to manage communications among service employees.

1. Introduction 

When employees receive training, they are often told that group unity is a vital component to positive outcomes in various ways. Inside many companies, leadership promotes employee motivation by enhancing productivity (Hanaysha & Tahir, 2016), job performance (Jiang, 2010), job satisfaction (Khuong & Tien, 2013), and self-efficacy (Griffin, Patterson, & West, 2001). Outside, it conveys an image of the company that highlights employees working together as a unified group, symbolizing a friendly working environment and well-trained employees.

However, is group unity always a good thing? Given that the previous literature focuses mostly on the positive side of group unity and that most organizations focus on strengthening it among employees, little is known about the possible undesirable consequences of group unity. This research explores circumstances where commonly presumed positive outcomes from group unity and find how it can result in at least partially negative outcomes. Specifically, our research examines whether the effect of group unity holds in such negative circumstances. As an indicator of a group unity, we examine employees’ task-related interactions during a service encounter and study how higher task interactions lead to lower customer satisfaction with the firm in the context of service failure by an individual employee through a series of scenario-based experiments. The effects of close linkages among employees have mostly been studied in the context of their work environment (e.g., Jiang, 2010), while the service literature mostly examines the effects of customer–employee interactions on customer satisfaction (e.g., Bendapudi & Leone, 2002; Bitner, Booms, & Tetreault, 1990; Khoa, Nguyen, & Nguyen, 2020; Kim, 2019) or employee performance (Kim, Hur, & Kim, 2019). Our research contributes to the service literature by exploring linkages among employee interactions, perceptions of group unity (i.e., entitativity), and customer satisfaction with the firm in the context of service failure. This research shows that task interaction among employees lead customers to perceive them as highly entitative (i.e., a unified group). It predicts that when employees in a firm are seen as highly entitative, an individual’s service failure will be more detrimental to customer satisfaction with the firm.

2. Theoretical Framework and Hypotheses Development

2.1. Entitativity

Most researchers in the fields of social and political psychology adopt Campbell’s definition of entitativity in their studies (e.g., Yzerbyt, Castano, Leyens, & Paladino, 2000; Lickel, Hamilton, & Sherman, 2001; Smith, Faro, & Burson, 2013). They see entitativity as “the degree of having the nature of an entity, of having real existence” (Campbell, 1958). In other words, entitativity refers to the extent to which a collection of people is perceived as a unified group (i.e., perceptions of groupness).

Campbell (1958) identifies members’ proximity, similarities, common fate, and pregnance as four elements that increase the perception of entitativity. Building on his work, social psychology researchers have demonstrated that physical and behavioral factors play a role in the perception of entitativity. Physical properties, such as members’ similarities, create perceptions of groupness (Dasgupta, Banaji, & Abelson, 1999). Lickel, Hamilton, Wieczorkowska, Lewis, Sherman, and Uhles (2000) show that factors such as interactions, common goals, common outcomes, similarities, and permeability are closely associated with perceptions of entitativity.

Social psychology scholars argue that processing information about individuals and groups works quite differently (e.g., Lickel et al., 2000). However, when a group of people is perceived to be highly entitative, perceivers tend to process information about that group in the same manner as they process information about individuals (McConnell, Sherman, & Hamilton, 1997; Yzerbyt, Rogier, & Fiske, 1998). Moreover, if a group is highly consistent among its members in terms of attitude, behavior, and systems, judgments about the group are stronger and more robust than those about less consistent groups (Palmatier, Scheer, & Steenkamp, 2007). For highly entitative groups, people also make more accurate judgements (Dasgupta et al., 1999), have a greater tendency to retrieve and use trait abstractions (Johnson & Queller, 2003), and more often use stereotypical judgements concerning their members (Spencer-Rodgers, Hamilton, & Sherman, 2007; Crawford, Sherman, & Hamilton, 2002). These findings suggest that judgments about individuals in a group are transferred to judgments about the group when the group is entitative.

Perceptions of entitativity can also have negative consequences. Roets and Van Hiel (2011) demonstrate that beliefs about the uniformity, informativeness, and inherent core of racial groups increase racial prejudices towards them. Moreover, perceptions of entitativity induce more negative responses towards negative groups. For example, Castano, Sacchi, and Gries (2003) argue that while entitativity can increase friendliness towards allied countries, it may also result in an exaggeration of the extent to which enemy countries are perceived as harmful. In a similar vein, entitativity perceptions amplify negative perceptions of misbehaving groups. People tend to think that highly entitative groups deserve greater retribution for perceived wrongdoings and are capable of more successful retaliation than groups with low entitativity (Newheiser & Dovidio, 2015; Stenstrom, Lickel, Denson, & Miller, 2008).

2.2. Implications of Entitativity in Marketing

While the research by marketing scholars is still at an early stage, scholars also find that employees of service firms are perceived as group members (e.g., Lickel et al., 2000). Recent work by marketing scholars has mostly focused on the positive impact of entiativity on firms. Nelson and Bowen (2000) and Tu, Yeh, Chuang, Chen, and Hu (2011) show that appearance and behavioral cues among members in a highly entitative team also signals professionalism. High entitativity reduces mind attribution to individuals (i.e., tendency to make judgements about individual members), making judgements about members as a whole (Morewedge, Chandler, Smith, Schwarz, & Schooler, 2013). Recently, Wang, Hoegg, and Dahl (2018) found that appearance (i.e., wearing the same outfit) and behavioral cues (i.e., explicit coordination) lead to enhanced perceptions of entitativity among salespeople in a team and that higher entitativity enhances customer satisfaction.

However, perceptions of entitativity lead to negative consequences in the case of negative events. Smith et al. (2013) identified a condition where entitativity has an opposite (i.e., negative) effect on charitable giving when the victim shows negative traits. Such findings show that in Myoung-Jin CHAE / Journal of Distribution Science 18-8 (2020) 35-45 37 service contexts entitativity is not always a good thing. To broaden our understanding of the impact of entitativity during negative events, the present study examines an individual employee’s service failure and how perceptions of entitativity shaped firm evaluations.

2.3. Interconnections among Employees in S ervice Encounters

Prior research has focused on the role of employee–customer interactions in service contexts and how such interactions influence customer satisfaction. Researchers have argued that interpersonal factors such as response to failures and prompted actions are closely related to customer satisfaction (Bitner, Booms, & Tetreault, 1990) and that customers’ expectations regarding relationship continuity enhance customer satisfaction (Hess, Ganesan, & Klein, 2003). Some researchers have argued that an employee–organization relationship provides an important buffer to service failures, mitigating the negative impact of these failures on customer satisfaction (Berry, 1995; Goodman, Fichman, & Snyder, 1995). Maxham III and Netemeyer (2003) suggested that employee perceptions of shared values and organizational justice are drivers of customer satisfaction. Although ample marketing literature has shown that customer-employee or employee-organization relationships matter, the influence of interconnections among employees on customer satisfaction is not wellresearched.

2.4. The Impact of Task Interaction among Employees on Customer Satisfaction

For the purposes of this study, task interaction refers to the communication among members in a service encounter (i.e., a customer’s total service experience during their visit). Previous literature has identified interaction among members as a driver of perceived entitativity (Menese, Ortega, Navarro, & de Quijano, 2008; Lickel et al., 2000). Building on this insight, this study suggests that customers’ perceptions of entitativity vary depending on the level of interaction among the employees observed during the service encounter. More specifically, as employees have more conversations with each other during work, customers are more likely to perceive them as a unified group, indicating in turn that they are more likely to see each employee as a member of the group (i.e., firm) rather than as an individual. However, when customers observe low interaction among employees, they are more likely to see the employee as an individual (i.e., apart from the group).

Imagine that a customer visits a hair salon to receive a haircut service. During her visit, she meets at least a few employees for different services — i.e, the assistant who washes her hair, the stylist who cuts it, and the receptionist who accepts her payment and schedules her next appointment. As the customer interacts with different employees, she may also observe varying levels of interactions among them. More interactions among employees lead customers to perceive that employees are working together as a unified group (i.e., entitativity). According to the previous literature, this article predicts that in cases of high entitativity the customer will consider each employee “a member of the salon” rather than “just an individual,” and will judge the hair salon (i.e., group of employees) similar to how they judge individual employees. In such cases, any action or event by an individual employee will be recalled as and considered “their” or “the salon’s” act, rather than “the employee’s” act. 

Then, the assistant made a mistake when washing the customer’s hair and her experience at the hair salon ended up being unpleasant. She would likely judge other employees in the same manner as the employee who made a mistake and evaluate other employees negatively based on the assistant’s mistake, as evaluations of members are transferred to each other. Assuming that she perceived the salon employees to be highly entitative, she would likely think that the unpleasant experience had been caused by “the hair salon” or “employees at the salon” rather than “the assistant” or “the employee” and attribute the service failure to the hair salon rather than the assistant, thinking that the hair salon is responsible for the failure. This research contends that such negative attribution to the group leads to lower satisfaction with the hair salon. On such bases, this article hypothesizes that amid a service failure by an individual employee, task interaction of the broader groups of employees leads to lower levels of customer satisfaction with the firm and such lower levels of satisfaction derive from perceptions of entitativity. Formally, 

H1A: Following a service failure by an individual employee, customer satisfaction with a firm is lower when task interaction among employees is higher. 

H1B: Perceived firm entitativity mediates the relationship between task interaction and customer satisfaction with the firm. 

2.5. The Impact of Task Interaction among Employees on Revisit Intention

This article also contends that task interaction among employees influences customers’ revisit intention in the future. Revisit intention has been studied in various contexts such as retail business and tourism services as they indicate a firm’s sales and profits. Moreover, some prior research has linked revisit intention to customer satisfaction in many occasions. When customers are satisfied, they are likely to repeat purchase, and when they are dissatisfied, they are likely to purchase alternative options (Oliver & Swan, 1989). In the restaurant industry, customer satisfaction is known to be a strong predictor of intention to repurchase (Oh, 2000). Although dimensions of satisfaction and revisit intention vary by contexts, the literature supports a strong positive relationship between customer satisfaction and revisit intention. Therefore, 

H2A: Following a service failure by an individual employee, revisit intention is lower when task interaction among employees is higher.

H2B: Perceived firm entitativity mediates the relationship between task interaction and revisit intention. 

Figure 1A and 1B presents a conceptual framework that reflects these hypotheses regarding the impact of task interaction among employees on 1) customer satisfaction 2) and revisit intention.

Figure 1A: Satisfaction with the Firm

Figure 1B: Revisit Intention 

3. Methodology

To test the hypotheses outlined previously, this article presents three studies (i.e., study 1A, 1B, and 2). Addressing hypotheses 1A and 2A (i.e., impact of employee interaction on customer satisfaction and revisit intention), Study 1A shows that higher task interaction leads to lower customer satisfaction and revisit intention. Study 1B provides partial support for the mediation effects in hypotheses 1B and 2B by manipulating the level of entitativity and showing that higher entitativity results in lower customer satisfaction and revisit intention. Building on the results from Study 1A and 1B, Study 2 demonstrates the role of entitativity as a mediator by measuring entitativity and outcome variables together. It also shows that higher task interaction results in perceived entitativity, which lowers customer satisfaction and revisit intention (H1B & H2B). 

The scenarios in three studies take place in two different settings — a hotel and a coffee shop. These two settings are appropriate for several reasons. First, many people visit them regularly for service transactions, implying a high degree of familiarity. Second, and by extension, these settings have often been used in the service literature (e.g., Bitner, Booms, & Mohr, 1994; Galbrich, 2010). Third, the kind of service failure settings presented here are widely familiar even if people haven’t experienced them directly in the past. Above and beyond these reasons, testing the hypotheses in two different settings renders the test results generalizable to different service circumstances. 

3.1. Overview of Study 1

Study 1A uses a hotel setting to test the effect of individual service employees’ task interaction on 1) customer satisfaction with the firm and 2) future intention to visit in a hotel scenario. Using a service failure scenario with different levels of employee interaction, it explores whether different interaction levels lead to different levels in our outcome variables. Study 1B involves a scenario with manipulated entitativity levels but consistent employee interaction levels. It measures the difference in customer satisfaction with the firm and future intention to visit to examine partial evidence that perceived employee entitativity serves as a mediator. 

3.2. Pilot Study

First, we created a pilot study to test whether our scenario was properly developed. Participants read descriptions of how hotel employees interacted (i.e., high or low) during a customer’s visit. In the high task-interaction condition, the hotel employees were described as having casual conversations (e.g., “After the attendant leaves your room, you settle down and say to yourself that the hotel’s employees seem to be busy with their own work while being friendly to each other.”) In the low task-interaction condition, the hotel employees were without conversations (e.g., “After the attendant leaves your room, you settle down and say to yourself that the hotel’s employees seem to be busy with their own work.”) The scenario used is a shorter version of the full scenario in the main study (Study 1A), which contains descriptions of hotel employees’ responsibilities and interactions without a service failure story. After reading the scenario, participants were asked to rate the level of task interaction, task interchangeability, and task Myoung-Jin CHAE / Journal of Distribution Science 18-8 (2020) 35-45 39 interdependence that they observed. Task interaction was measured using a single-item scale to see whether the manipulation was successful. Task interchangeability and task interdependence were likewise measured using single-item scales to rule out the possibility that other variables influence outcomes of the study, thereby potentially confounding the results of the study. All responses were captured on seven-point Likert-type scales. 

The pilot study showed that the two conditions differed in terms of the level of employees’ task interaction (Mhigh= 6.13, SDhigh= 1.03 ; Mlow= 4.14, SDlow= 1.66; F(1,111)=57.876, p= < .001). However, it did not show significant differences in task interchangeability (p= .954) and task interdependence (p=.139). The results suggest that the manipulation was successful and does not generate differences in other factors such as task interchangeability and task interdependence. 

3.3. Study 1A

3.3.1.Procedure

The purpose of this study is to examine the relationship between employee interactions and 1) customer satisfaction with the firm (H1A) and 2) future intention to visit (H2A). To test these hypotheses, the same scenario used in the pilot study was extended to include a service failure episode in the last paragraph. Participants were recruited through Amazon Mechanical Turk and required to read descriptions of hotel employees’ behavior during a customer’s visit, specifically focusing on each employee’s responsibility when serving customers. They were then asked to read a description of an unsatisfactory service experience, i.e., a receptionist’s mistake in assigning rooms and then having a bad attitude when dealing with her mistake afterwards. Like the pilot study, the scenario had two conditions that differed in terms of level of employee task interaction (i.e., high/low). 

3.3.2. Method

In exchange for monetary compensation, one-hundred and fifty-eight respondents from Amazon Mechanical Turk (Mage = 32.7, 41% female) participated in this between-subjects experiment. Participants were randomly assigned to one of the two conditions. After reading the scenario about the hotel employee’s service failure, they were asked to rate the performance of the receptionist who delivered the unsatisfactory service, the performance of the hotel, and how likely they would be to visit the hotel again in the future (See Table 1). To measure each of these criteria, a three-item scale from Maxham III and Netemeyer (2002) was adapted, with the three items for each measure collapsed into a single measure of satisfaction with the receptionist (α = .81), satisfaction with the hotel (α = .79), and revisit intention (α = .81). The scale is appropriate because the items were adapted from Maxham III and Netemeyer (2002), where they ask about diverse aspects of 40 Individual Employees’ Service Failures and Customer Satisfaction with the Firm services, including reversed-order items. To examine whether the scenario is realistic, participants were asked about the frequency of similar experiences in the past. Among participants, 66.4% responded that they experienced them more than occasionally (>4 on seven-point Likert scale, 1 = never, 7 = very frequently). All responses were captured on seven-point Likert-type scales. 

Table 1: Measure

3.3.3. Results and Discussions

The first study provides empirical evidence supporting the negative impact of employees’ task interaction on firms (in the case of a service failure by an individual employee). In support of H1A, the study found a significant difference in participants’ satisfaction with the hotel’s performance in the two conditions. A one-way ANOVA reveals that participants were less satisfied with the hotel’s performance when task interaction was high compared to when task interaction was low (M_high= 2.88, SD_high= 1.17; M_low= 3.25, SD_low= 1.43; F(1, 156)= 3.13, p= .079 ). Consistent with H2A, participants’ hotel revisit intentions were also different between the two conditions at a 90% significance level, showing that participants were less likely to revisit the hotel in the future in the high interaction condition than in the low (M_high= 2.65, SD_high= 1.67; M_low= 3.00, SD_low= 1.45; F(1, 156)= 2.79, p= .097). In addition, participants’ satisfaction with the receptionist’s performance differed between the two conditions (M_high= 2.52, SD_high= 1.18; M_low= 3.14, SD_low= 1.49; F(1, 156)= 8.28, p= .005), suggesting that participants evaluated the receptionist more negatively when interaction was high (vs. low), even when they experienced the same failure experience by the receptionist. It provides evidence that task interaction among employees has a negative impact on customers’ overall experience with the firm when negative service encounters occur (see Figure 2A & 2B). 

Figure 2A: Satisfaction with the Hotel by Employee Interaction Level

Figure 2B: Revisit Intentions in the Future by Employee Interaction Level 

3.4. Study 1B

3.4.1. Procedure

This study attempted to examine the mechanism behind the relationship between entitativity and satisfaction with the firm. By manipulating firm entitativity, we expected to find partial evidence for entitativity as a mediator.

First suggested by Baron and Kenny (1986), measuring mediators and dependent variables together in a technique that is widely used among social science researchers focused on mediation analysis. Spencer, Zenna, and Fong (2005) provide support for manipulating entitativity, arguing that manipulations of mediators may offer stronger evidence for causal relationships than measuring mediators and then testing mediation effects. This assertion is reasonable given the properties of experiments, which provide support for causal relationships; thus, the results in this study show evidence of the effect of entitativity as a mediator. 

For manipulation, the hotel scenario used in the control (i.e., low task interaction) condition was applied, leaving out the direct descriptions of employees’ task interactions. In the high-entitativity condition, a statement that “the employees in the hotel were trained to work together as a team” was added. In the low-entitativity condition, a statement that “the employees in this hotel were trained to individually rather than as a team” was used. Consistent with the previous study, multiple-item scales from the previous literature were adapted to the three measures in study 1A: satisfaction with the receptionist (α = .91), satisfaction with the hotel (α = .73), and revisit intention (α = .71). 

3.4.2. Method

Ninety-one respondents recruited from Amazon’s Mechanical Turk service participated in this between-subjects experiment for monetary compensation (Mage = 32.6, 40% female). Participants were randomly assigned to one of the two hotel conditions developed for this study. After reading the given scenario, they were asked to provide their opinions on several questions. 

3.4.3. Results and Discussion

As expected, this study found that satisfaction with the hotel was lower in the high-entitativity condition than in the low-entitativity condition (Mhigh= 2.23, SDhigh= .81; Mlow= 2.69, SDlow= 1.02; F(1, 89)= 5.66, p= .020). Consistent with the results obtained from the previous study, the effect of entitativity on revisit intention was also significant at a 90% level (Mhigh= 1.92, SDhigh= .85; Mlow= 2.25, SDlow= .97; F(1, 89)= 2.96, p= .089). Moreover, satisfaction with the receptionist was different between the high- and low-entitativity conditions (Mhigh= 1.86, SDhigh= 1.09; Mlow= 2.47, SDlow= 1.16; F(1, 89)= 6.80, p= .011). These results support the idea that higher entitativity results in lower satisfaction with the firm when customers receive poor service from an individual employee. Based on the results of study 1A (i.e., higher interaction among employees results in lower customer satisfaction and revisit intention) and study 1B (i.e., higher entitativity perceptions result in lower customer satisfaction and revisit intention), the prediction that entitativity mediates the impact of employees’ task interactions on customer satisfaction with the firm and revisit intention was partially supported (See Figure 3A & 3B).

Figure 3A: Satisfaction with the Hotel by Perceived Entitativity Level 

Figure 3B: Revisit Intentions in the Future by Perceived Entitativity Level 

3.5. Overview of Study 2

The purpose of study 2 was to 1) directly examine the role of perceived entitativity as a mediator using mediation analysis and 2) extend our findings in study 1 to a different service context (i.e., coffee shop) using a different type of service failure. Study 2 involved a scenario where the individual service failure is more salient, i.e., the failure is more easily attributed to one individual rather than the firm (i.e., coffee shop), to test whether the effects still held.

3.6. Pilot Study

To test whether the scenario was developed appropriately, participants read descriptions of coffee shop employees during a customer’s visit. Similar to Study 1A, the scenario, which was a shorter version of the full scenario used in study 2 without a service failure story, had two conditions that differed only in the level of employee task interaction (i.e., high/low). Like in the previous study, participants were asked to read the scenario and then rate the level of task interaction, task interchangeability, and task interdependence.

The two conditions differed in terms of employees’ task interactions (M_high= 6.2, SD_high= 1.24 ; M_low= 4.18, SD_low= 2.02; F(1,35)=12.69, p= .001). However, there was no significant differences in terms of task interchangeability (p= .945) and task interdependence (p=.102). The results suggest that the manipulation was successful and did not generate differences in other factors such as task interchangeability and task interdependence.

3.7. Study 2

3.7.1. Procedure

The purpose of this study was to examine 1) the relationship between employee interaction and consumers’ satisfaction with the firm and intention to revisit (H1A & H2A) and 2) the role of perceived entitativity as a mediator (H1B & H2B). Recruited through Amazon Mechanical Turk, participants read descriptions of employees during a customer’s visit to a coffee shop. Specifically, they read about each employee’s responsibility when serving customers using the same scenario used in the pilot study with a service failure story in the last paragraph. The scenario described an unsatisfactory service experience (a male barista forgetting a customer’s order and then handing that person cold and bad-tasting coffee after a long wait), which had two conditions that differed in terms of employee task interaction (i.e., high/low). In the high-task- interaction condition, the employees of the coffee shop were described as having conversations (e.g., “The staff members in this coffee shop seem to be busy with their own work while having a lot of casual conversations with each other.”) In the low-task-interaction condition, the employees of the coffee shop were described without conversations (e.g., “The staff members in this coffee shop seem to be busy with their own work while having a lot of casual conversations with each other.”)

3.7.2. Method

In exchange for monetary compensation, two-hundred and forty-four respondents on Mechanical Turk participated in this between-subjects experiment (Mage = 38.4, 58% female). Participants were randomly assigned to one of the two conditions described earlier. After reading the scenario, they were asked to rate 1) their satisfaction with the coffee shop, 2) revisit intention, and 3) perceived level of entitativity. As in Study 1A and 1B, multiple-item scales were adapted from the previous literature to the three outcome variables: satisfaction with the firm (α= .75), future intention to visit (α= .86), and satisfaction with the employee (α= .82). Perceived entitativity was measured using a four-item scale adapted from Crump et al. (2010), and the four items were again collapsed into a single measure of entitativity (α= 0.94; see Table 1). Lastly, as a proxy for reality of the scenario, participants rated the frequency of similar experiences in the past. In total, 57.8% responded that they had such experiences more than occasionally (>4 on seven-point Likert scale, 1 = never, 7 = very frequently).

3.7.3. Results and Discussions

The results provide evidence of the main effects in our hypotheses 1A and 2A. As expected, the results revealed that the two conditions differ in terms of satisfaction with the firm (Mhigh= 2.14, SDhigh= 1.23; Mlow= 2.44, SDlow= 1.38; F(1,242)= 3.17, p= .076). Participants’ revisit intentions were also lower in the high (vs. low) interaction condition (Mhigh= 2.23, SDhigh= 1.35; Mlow= 2.65, SDlow= 1.38; F(1,242)=5.84, p= .016). In addition, satisfaction with the employee was lower in the high (vs. low) interaction condition (Mhigh= 2.09, SDhigh= 1.33; Mlow= 2.40, SDlow= 1.37; F(1,242)=3.10, p= .080). These findings suggest that consumers’ experience with the coffee shop was more negatively evaluated in the high interaction condition (vs. low). In addition, the results show that perceptions of entitativity differ in the two conditions (Mhigh= 4.64, SDhigh= 1.41; Mlow= 3.36, SDlow= 1.52; F(1,242)=46.44, p< .000), which indicate that task interaction among employees is related to the perception of entitativity (see Figure 4A & 4B).

Figure 4A: Perceived Entitativity as a Mediator in the Relationship between Employee Interaction and Satisfaction with the Coffee Shop 

Figure 4B: Perceived Entitativity as a Mediator in the Relationship between Employee Interaction and Revisit Intentions 

3.7.4. Results and Discussions

This study tested the proposed mediating role of perceived entitativity in the relationship between employee interactions and 1) customer satisfaction with the firm and 2) future revisit intentions. Using the bootstrapping procedure (PROCESS model 4; Hayes 2017; Preacher & Hayes 2008) with 5,000 samples, the study found that, consistent with the prediction in H1B, the effect of employee interactions on customer satisfaction with the firm was mediated by perceived entitativity (point estimate = .34, 95% CI = [ .18, .53]). It also found that, in support of H2B, the effect of perceived entitativity as a mediator between employee interactions and revisit intentions was significant (point estimate = .44, 95% CI = [ .27, .65]).

4. General Discussions

This research examined the impact of employees’ task interaction on consumers’ 1) satisfaction with the firm 2) revisit intentions. Further, the research examined the role of perceived firm entitativity to understand the mechanisms of this relationship. The first study (i.e., Study 1A) tested the main effects in hypotheses to study if higher task interaction results in 1) lower satisfaction with the firm (H1A) and 2) lower revisit intentions (H2A). The study found evidence that in firms with greater task interaction among employees, customers are less satisfied with the firm and less willing to revisit it after they experience a service failure by an individual employee.

The second study (i.e., Study 1B) tested if the results in the first study come derived from the framework in which it is grounded — perceived level of firm entitativity. In the second study (study 1B), we manipulated the entitativity level in the scenario rather than the level of task interaction and found significant differences in 1) satisfaction with the firm and 2) revisit intentions between the two conditions (entitativity: high vs. low). These findings partially support the role of entitativity as a mediator.

Lastly, this research developed a new scenario in different service contexts and directly examined the role of perceived entitativity as a mediator. Through the coffee shop scenario, evidence of the main effect relationships between employee task interaction and 1) customer satisfaction with the firm and 2) revisit intentions was found. Moreover, the mediation analysis results showed that consumers perceive a higher level of entitativity in the high interaction condition and, as a result, are 1) less satisfied with the firm and 2) less willing to visit in the future. In addition, the coffee shop scenario indicates that our study findings from Study 1 are generalizable to other service contexts in which consumers experience different types of service failures. In addition, we created a scenario in which the service failure is attributed to an individual employee (i.e., the barista) rather than the firm to a greater extent and examined whether the effect still holds.

We argue that our research makes the following theoretical contributions. While previous research findings show positive effects of groupness of employees (i.e., entitativity) in the marketing literature, this research advances our knowledge of entitativity by finding the condition where entitativity plays a negative role. Based on the idea in the social psychology literature that for negative circumstances (e.g., Castano et al., 2003), entitativity plays a negative role in the group perception, we explore the context of service failures and find that higher perceived entitativity leads to lower customer satisfaction with the firm. Thus, this research provides a unique theoretical contribution to the service literature by advancing our understanding of the perception of groupness which could also lead to greater negative consequences, in contrast to what has been known. Moreover, our research finds a behavioral cue for entitativity by exploring the impact of employee interactions during the service encounter and provides novel insights on how specifically entitativity can play a negative role in service contexts.

5. Managerial Implications

It is well known among social scientist and practitioners that firms’ delivering positive images to their customers depends on an impression that employees in a firm work together as a cohesive group and were trained together under the same mission. However, this research suggests that when creating such an impression firms need to be more careful about its possible negative consequences. Specifically, firms need to consider the idea of strategically managing levels of interaction in service encounters if their employees often deliver unsatisfactory service. While group cohesiveness and good communication among employees certainly enhance productivity at work, firms might want to consider employee communication across different circumstances separately. If employee interactions are not necessary (i.e., casual non-work-related conversations), firms need to advise their workers to reduce them during work hours.