Customers' Responses to Customer Orientation of Service Employees in Full‐Service Restaurants: A Relational Benefits Perspective

Abstract

The interpersonal interaction between service employees and customers influences service quality, customer satisfaction, and customer loyalty. The importance of maintaining an enduring relationship with customers is generally accepted in the marketing literature and field. In an effort to incorporate these two convictions, this study was designed to investigate how the customer orientation of service employees (COSE) affects customers' perceptions of relational benefits and ultimately contributes to repurchase intention in the full‐service restaurant context. Findings indicate that relational benefits, enhanced by COSE, positively influence customers' favorable inequity perception, customer satisfaction, and repurchase intention.

Keywords:

• customer orientation of service employees (COSE)
• relational benefits
• customer satisfaction
• favorable inequity
• repurchase intention
• full‐service restaurant

INTRODUCTION

The fundamental tenet of marketing is that all of a firm's activities should be directed toward ensuring customer satisfaction and creating mutually beneficial, long‐term relationships with the market (Kotler, 1980). In other words, firms should and will benefit by providing customers with satisfactory services and products that contribute to long‐term relationships. One way to facilitate this tenet is customer orientation1—specifically, putting customers' interests first (Saxe and Weitz, 1982). In this sense, a customer‐oriented business culture is preferable for firms that seek success in the market (e.g., Houston, 1986; Parasuraman, 1987).

In the service context, individual service employees are the last people to implement the firm's marketing efforts and the first people to directly interact with the customer (Brown, Mowen, Donavan, and Licata, 2002). In other words, service employees actualize the firm's customer orientation at the front line and at the same time acquire feedback from customers and provide it to the management as market intelligence. As such, the service firm's ability to be customer‐oriented primarily depends on individual service employees' customer orientation.

Given the experiential and interactive nature of service and the key role played by service employees in service provision, the customer orientation of service employees (COSE) is critical to service businesses, including restaurants. However, no study yet has focused on the effects of COSE on customers' responses to restaurants. In the hospitality context, Susskind and colleagues (2000, 2007) adopted the construct of the customer (guest) orientation of service employees. However, they tested customer orientation in the organizational behavior context. In addition, they adopted Kelley's (1992) definition of ‘customer orientation’—employees' general commitment toward customer satisfaction—which is distinct from COSE.

The objective of this study was to assess how COSE influences customers' repurchase intention by examining customers' perceptions of relational benefits, favorable inequity, and satisfaction. The basic notion of this study is that COSE leads customers to stay in long‐term relationships with restaurants because they benefit by doing so.

REVIEW OF LITERATURE AND HYPOTHESES

Customer Orientation of Service Employees (COSE)

Customer orientation at the employee level is the extent to which an employee seeks to increase long‐term customer satisfaction (Saxe and Weitz, 1982). In service businesses, the service employee and the service are often regarded as synonymous from the customer's perspective (Bowen and Schneider, 1985) and the interaction with the service employee is the most important influence on customers' service quality assessment (Brown and Swartz, 1989) and on overall service satisfaction (Crosby and Stephens, 1987; Hennig‐Thurau and Thurau, 2003). Thus, COSE is critical to customers' evaluations of the service and the service firm.

Hennig‐Thurau and Thurau (2003) defined COSE as “the behavior of service employees when serving the needs and wishes of existing and prospect customers” (p.27) and suggested three dimensions of COSE. Building on the conceptualization of Hennig‐Thurau and Thurau (2003), Hennig‐Thurau (2004) outlined four dimensions of COSE and emphasized that all four dimensions are required to a certain degree to enable employees to behave in a customer‐oriented manner. These dimensions are technical skills, social skills, motivation, and employees' self‐perceived decision‐making authority.

According to his conceptualization, the service employee's technical skills refer to the knowledge and skills that the employee should have in order to fulfill the customer's needs during the personal service interaction (cf. Argyle, 1967). In the food‐service context, such technical skills may include the employee's knowledge of menus, parts of recipes, ingredients, and such, and skills in communicating with the kitchen, serving customers, managing multiple tables, and such. This dimension of COSE is fundamental if the service employee is to meet the customer's needs.

Hennig‐Thurau (2004) defined the concept of social skills as the service employee's ability to take the customer's perspective during service interactions visually, cognitively, and emotionally. Using these skills, the employee can empathize with the customer's perceptions, thoughts, and feelings. Customers are not always able or willing to express what they perceive, think, or feel, especially during the short duration of service interactions. An understanding of the customers' perspective is essential to a proper, timely understanding of customer needs. In this sense, employees require social skills to perform in a customer‐oriented manner.

Motivation is what actually drives behavior (Hennig‐Thurau and Thurau, 2003). The service employee's motivation leads the employee to exert technical and social skills in performing in a customer‐oriented fashion (Hennig‐Thurau, 2004). Hennig‐Thurau (2004) argued that the three elements of the employee's motivation are: “(a) a positive valence of customer‐oriented behavior and the consequences associated with such behavior on the part of the employee; (b) the employee's self‐perception of being able to behave in a customer‐oriented way; (c) and his/her expectations of reaching the desired outcome through engaging in such behavior (e.g., happy customers, rewards from the employer)” (p.463). An employee must manifest all three elements are required to a certain extent in order to engage in customer‐oriented behaviors (Hennig‐Thurau and Thurau, 2003).

Lastly, to some degree, the service employee should feel authorized to make decisions about issues that concern customers' interests and needs—further evidence of an ability to behave in a customer‐oriented manner (Hennig‐Thurau, 2004). The lack of the employee's self‐perceived decision‐making authority may demotivate the employee and prevent him/her from fully utilizing skills to fulfill customers' needs. Self‐perceived authority differs from empowerment in that empowerment is a type of formal or objective authority given by the supervisor and/or organization whereas self‐perceived authority is a subjective concept (Hennig‐Thurau, 2004; Hennig‐Thurau and Thurau, 2003).

Relational Benefits for Customers as Outcomes of COSE

The concept of relational benefits is based on the belief that the customer remains in a long‐term relationship only when the relationship benefits him/her (cf. Gwinner, Gremler, and Bitner, 1998). Long‐term relationships with customers are advantageous to the firm (e.g., Berry, 1995; Heskett et al., 1994). To be successful over the long term, the firm should be able to promote customer benefits to keep customers in long‐term relationships. The concept of gaining by benefiting customers is closely tied to the concept of customer orientation—specifically, putting the customer's interests first (Saxe and Weitz, 1982). COSE focuses on how customer orientation is executed at the service employee level (cf. Hennig‐Thurau and Thurau, 2003).

Relational benefits (from the customer perspective) are defined as “those benefits customers receive from long‐term relationships above and beyond the core service performance” (Gwinner et al., 1998: 102). Gwinner et al. (1998) developed and empirically supported a typology of three relational benefits: confidence, social, and special treatment. Confidence benefits include reduced anxiety and comfort resulting from faith in the trustworthiness of the service employee and his/her performance; social benefits include pleasure and comfort stemming from familiarity with and recognition by the service employee; and special treatment benefits take the form of price breaks, additional services, and faster service. All of these relational benefits would be enhanced when the service employee performs reliably (technical skills), takes adequate interest in the customer's perspective (social skills), has the will to fulfill the customer's needs (motivation), and perceives enough authority to take action in the customer's best interests (self‐perceived authority).

Hypothesis 1a: COSE has a positive effect on confidence benefits.

Hypothesis 1b: COSE has a positive effect on social benefits.

Hypothesis 1c: COSE has a positive effect on special treatment benefits.

Social Benefits as an Antecedent of Confidence Benefits and Special Treatment Benefits

As implied earlier, the customer perceives social benefits from an interpersonal relationship with the service provider. The intangible nature of service poses more uncertainty relating to quality and thus more risk to the customer. Hence, the customer‐employee interaction is particularly salient in customers' evaluation of the service (Bitner, Booms, and Tetreault, 1990) and the social bonding between the customer and employee can serve to increase the customer's dependence on the service provider and build the customer's trust (Bendapudi and Berry, 1997). Therefore, if the customer has developed a close relationship with the service employee, he/she would have more confidence in the employee's performance.

Hypothesis 2a: Social benefits have a positive effect on confidence benefits.

Special treatment benefits are associated with special consideration to customers (Patterson and Smith, 2001). When the customer has developed a close relationship with the service employee (indicating more social benefits), his/her preferences are likely to be better understood by the employee (Gwinner et al., 1998) and thus he/she is likely to receive more customized services while others receive standardized services (Gutek, 1995). Further, if the customer is not personally recognized by the service employee, the customer is likely to receive less special treatment because in these cases, the customer is a mere role occupant to the employee (Gutek et al., 2000), so that the employee has no personal feelings for the customer.

Hypothesis 2b: Social benefits have a positive effect on special treatment benefits.

Favorable Inequity as an Outcome of Relational Benefits

Oliver and Swan (1989b) argued that the concept of fairness includes what is fair and what is hedonically desirable to the customer, thus encompassing both fairness and preference. “Preference” refers to maximizing one's own outcome relative to that of the other (i.e., positive or favorable inequity) (Oliver and Swan, 1989a). In a commercial exchange, when individuals perceive that they can maximize their outcomes by behaving inequitably, they are likely to do so (Walster, Walster, and Berscheid, 1978). Thus, customers actively seek exchanges or exchange relationships that are advantageous to them. Relational benefits are likely to satisfy customers' favorable inequity needs because such benefits are highly exclusive to long‐term relational customers (Gwinner et al., 1998).

Hypothesis 3a: Confidence benefits have a positive effect on favorable inequity.

Hypothesis 4a: Social benefits have a positive effect on favorable inequity.

Hypothesis 5a: Special treatment benefits have a positive effect on favorable inequity.

Customer Satisfaction as an Outcome of Relational Benefits and Favorable Inequity

Cultivating relationships with customers has long been thought to increase customer satisfaction (e.g., Berry and Parasuraman, 1991; Czepiel, 1990). Repeatedly, relational benefits are extra benefits above and beyond the core service that customers receive as a result of having developed long‐term relationships with the firm. Thus, long‐term relationships add extra values to customers' service consumption and thereby contribute to customer satisfaction. Specifically, when customers have strong confidence in their service expectations, they tend to assimilate satisfaction levels to match expectation levels to reduce dissonance (Hennig‐Thurau, Gwinner, and Gremler, 2002; Yi and La, 2003). Thus, Hypothesis 3b follows as below.

Hypothesis 3b: Confidence benefits have a positive effect on customer satisfaction.

The customer's interpersonal relationships with service employees are a key element of the overall service offering (Czepiel, 1990; Berry, 1995). These relationships serve to fill the basic human needs of recognition and feeling important (Jackson, 1993) and in turn contribute to customer satisfaction (Guenzi and Pelloni, 2004). Interestingly, Gutek, Bhappu, Liao‐Troth, and Cherry (1999) found that customers who have a close personal relationship with a specific employee show higher levels of satisfaction than those who have a relationship only with the firm. Thus, Hypothesis 4b follows as below.

Hypothesis 4b: Social benefits have a positive effect on customer satisfaction.

In that special treatment benefits offer monetary and non‐monetary advantages to customers, with regard to customer satisfaction, the effect of special treatment benefits is closely related to that of favorable inequity. In other words, special treatment benefits increase outcome or decrease input in favor of the customer, and thus generate favorable inequity to the customer. More specifically, additional service, faster service, and higher priority increase customer outcomes while price breaks decrease customer inputs. On top of the actual advantageous outcome/input ratio, special treatments would give the customer a feeling of importance (Jackson, 1993). Thus, Hypotheses 5b and 6 follow as below.

Hypothesis 5b: Special treatment benefits have a positive effect on customer satisfaction.

Hypothesis 6: Favorable inequity has a positive effect on customer satisfaction.

Lastly, numerous studies have provided evidence of the significant influence of customer satisfaction on customer loyalty to the firm—it is intuitively apparent by itself. Particularly in the restaurant context, Oh (2002) and Ok, Back, and Shanklin (2005) reported significant direct effects of customer satisfaction on customers' repurchase intention.

Hypothesis 7: Customer satisfaction has a positive effect on repurchase intention.

CONCEPTUAL MODEL

In Figure 1, the hypothesized relationships are identified. The conceptual model indicates that restaurants in which service employees are perceived to be customer‐oriented receive positive responses from customers. First, COSE is positively related to customers' perceptions of relational benefits. That is, the more customers perceive COSE, the more they perceive social benefits, confidence benefits, and special treatment benefits in their relationship with the restaurant. Second, increased perception of relational benefits positively affects customers' perception of favorable inequity and satisfaction, which in turn enhances customers' repurchase intentions. In sum, COSE leads customers to revisit the restaurant by enhancing customers' perceptions of relational benefits, favorable inequity, and satisfaction.

Figure 1 Conceptual Model. Note: TS: technical skills, SS: social skills, MT: motivation, DMA: decision‐making authority, COSE: customer orientation of service employee, CB: confidence benefits, SB: social benefits, STB: special treatment benefits, FI: favorable inequity, CS: customer satisfaction, RI: repurchase intention. * Shaded circles represent first‐order factors.

METHODOLOGY

Data Collection

A self‐report questionnaire was distributed by email to 2,956 staff and faculty members of a major public university in a midwestern state of the United States. Participants were provided with a brief explanation of the term ‘full‐service restaurant,’ which was the context of this study, and then asked to name the full‐service restaurant at which they dine most frequently. Then, they were asked to answer the remaining questions based on their experiences with the restaurant they named. To identify the length and intensity of their patronage of the restaurant, they were asked how long they had been a customer of the restaurant and how often they currently dine there.

Of the 2,956 distributed questionnaires, a total of 545 were completed and returned (a response rate of 18.4%). After eliminating those with missing information, 465 questionnaires remained for analysis (a usable response rate of 15.7%). Further, because the concept of relational benefits is not applicable to short‐term and/or non‐relational customers, analysis was confined to responses from customers who reported a first dining experience in the restaurant at least 12 months previously and currently dine at the restaurant at least once every 3 months. This left 401 completed questionnaires.

Measures

Significant discrepant results have been reported relating to the degree to which customers assessed service personnel as being customer‐oriented compared to self‐assessments (Daniel and Darby, 1997; Dunlap, Dotson, and Chambers, 1988; Michaels and Day, 1985). More specifically, service personnel tend to perceive themselves as being more customer‐oriented than their customers perceive them to be (Bove and Johnson, 2000). Accordingly, measuring COSE as perceived by customers is more adequate than measuring it as perceived by employees themselves or their supervisors because the target audience for the firm's marketing efforts are its customers.

To measure the constructs in our conceptual framework, validated scales from the literature were adapted to the food‐service setting, as follows:

	COSE was measured with 12 items previously employed by Hennig‐Thurau (2004);
	Relational benefits were measured with 12 items borrowed from Gremler and Gwinner (2000);
	Favorable inequity was measured with two items adapted from Oliver and Swan (1989a);
	Customer satisfaction was measured with three items drawn from Hennig‐Thurau et al. (2002) and Oliver (1980); and
	Repurchase intention was measured with two items drawn from Hellier, Geursen, Carr, and Rickard (2003).

All items were assessed on 7‐point Likert‐type scales ranging from 1 (‘strongly disagree’) to 7 (‘strongly agree’). The measurement items are shown in Table 1. The initial questionnaire was modified after a pre‐test (using a sample of 10 graduate students and faculty members in a hospitality department) and a pilot test (using a sample of 70 undergraduate students in two hospitality classes). Based on the pilot test, reliability of measurements was estimated using Cronbach's alpha. All of the alpha values exceeded the conventional cut‐off point of .70 (Nunnally, 1978).

Table 1. Profile of the Sample

		Frequency	%			Frequency	%
Gender	Male	173	43.5	Income	Under $25,000	26	6.7
	Female	225	56.5		$25,000–$39,999	59	15.2
Age	19–24	20	5.1		$40,000–$54,999	81	20.9
	25–34	66	16.7		$55,000–$69,999	59	15.2
	35–44	92	23.3		$70,000–$84,999	62	16.0
	45–54	117	29.6		$85,000–$99,999	34	8.8
	55–64	88	22.3		Over $100,000	66	17.1
	65 or older	12	3.0	Education	High school degree	38	9.5
					Associate degree	17	4.3
					Bachelor degree	97	24.3
					Master degree	103	25.8
					Doctoral degree	144	36.1

Profile of the Sample

The sample (n  =  401) in the analysis was 56.5% female (n  =  225). The mean age of respondents was 45.2 years (range 19 to 76 years). The largest age category was 45 to 54 years (29.6%), followed by 35 to 44 years (23.3%) and 55 to 64 years (22.3%). In terms of income, the respondents were fairly evenly distributed, with the largest group (20.9%) reporting an income between $40,000 and $54,999 and the smallest group (6.7%) reporting an income of $25,000 or less. The respondents were highly educated overall. The largest group had doctoral degrees (36.1%), followed by Masters (25.8%) and Bachelors (24.3%).

ANALYSIS AND RESULTS

Measurement Model

The measurement model provided a fairly good fit to the data in the full‐service restaurant setting (NFI  =  .938; TLI  =  .966; CFI  =  .970; RMSEA  =  .046; χ ²(408)  =  756.9, p < .001) (Hu and Bentler, 1999). Table 1 shows the specific items and the first‐order factors of COSE (technical skills, social skills, motivation, decision‐making authority) employed in this study, together with their standardized factor loadings.

The factor loadings were equal to or greater than .710 and all were significant at p < .001 with the t‐values, not shown, ranging from 13.9 to 32.6. The factor loadings for the first‐order factors of COSE were .885 or above except for ‘decision‐making authority’ (.605). The t‐values of the first‐order factors ranged from 11.6 to 21.3.

Table 2 shows the descriptive statistics and associated measures for the constructs. The convergent and discriminant validity of the scales were tested via a confirmatory factor analysis, as recommended by Anderson and Gerbing (1988). As noted above, all indicators loaded on the proposed constructs significantly at p < .001. Average variance extracted (AVE) was greater than the .50 cutoff for all constructs (Bagozzi and Yi, 1988). On the basis of the factor loadings and AVE estimates, convergent validity among the measurement‐scale items was thus established (Fornell and Larcker, 1981). Strong discriminant validity was demonstrated when the squared correlation (R ²) between a pair of constructs was found to be less than the AVE for each construct (Fornell and Larcker, 1981), except for that between ‘confidence benefits’ and ‘COSE’. Because the R ² between ‘confidence benefits’ and ‘COSE’ was slightly higher (.694) than the AVE of ‘confidence benefits’ (.672), discriminant validity between them was established by combining them into one construct and then performing a χ ² difference test on the values obtained from the combined and uncombined models (Bagozzi and Yi, 1988). The resulting χ ² difference was 257.9 (df  =  6), which is significant at p < .001. Thus, discriminant validity between ‘confidence benefits’ and ‘COSE’ was also confirmed. Then, adequate internal consistency of the scales was confirmed by computing composite reliabilities, which were all well above the recommended value of .70 (Hair, Anderson, Tatham, and Black, 1998). Additionally, because some correlations between each construct of ‘COSE,’ ‘confidence benefits,’ ‘customer satisfaction,’ and ‘repurchase intention’ were as high as .794 or above, we checked the possibility of multicollinearity. As a result, all the items for the four constructs showed tolerance levels of .203 or higher, which are above the recommended cutoff of .100 (Tabachnick and Fidell, 2007). Thus, no multicollinearity problem was found.

Table 2. Items and Loadings

Construct and Scale Item	Standardized Loading^a
COSE
Technical skills: The employees of the restaurant…	.885
…have a high level of job knowledge.	.854
…are experts in their job.	.871
…highly competent.	.920
Social skills: The employees of the restaurant…	.968
…have extensive social skills.	.745
…are able to consider their customers' perspective.	.834
…know hot to treat a customer well.	.909
Motivation: The employees of the restaurant…	.992
…show strong commitment to their job.	.869
…do their best to fulfill their customers' needs.	.893
…are always highly motivated.	.826
Decision‐making authority: The employees of the restaurant…	.605
…are allowed to decide autonomously in customer matters.	.845
…have appropriate room for maneuver in solving customer problems.	.932
…do not need to ask their superior for permission in the case of customer requests.	.850
Confidence benefits
Based on all my experiences with the restaurant…
…I believe there is less risk that something will go wrong	.771
…I feel I can trust the employee(s).	.862
…I have more confidence the service will be performed correctly.	.919
…I have less anxiety when I decide to dine out at this restaurant.	.710
Social benefits
Based on all my experiences with the restaurant…
…I am recognized by (a) certain employee(s).	.819
…I am familiar with the employee(s) who provide(s) the service.	.852
…I have developed a friendship with the employee(s).	.965
…the employee(s) know(s) my name.	.893
Special treatment benefits
Based on all my experiences with the restaurant…
…I receive better prices or special deals that most customers don't.	.902
…they provide services to me that they don't provide to most other customers.	.940
…I receive faster service than most other customers.	.917
…they pay extra attention to my needs.	.748
Favorable inequity
Thinking of all my experiences with the restaurant…
…overall, I have been treated more than fairly by this restaurant	.846
…overall, the food and service I received from this restaurant have been more than my expense, time, and effort.	.828
Customer satisfaction
Considering all my experiences with the restaurant…
…my choice to dine out at this restaurant has been a wise one.	.915
…I am delighted with this restaurant's service.	.881
…overall, I am satisfied with this restaurant.	.910
Repurchase intention
All things considered, as long as the present service continues…
…I would keep dining out at this restaurant in the future.	.949
…I would dine out at this restaurant at least at current frequency in the future.	.873
Note: ^a All factor loadings are significant (p < .001).1. Bold figures represent first‐order factor loadings.

Structural Model

The proposed model provided a fairly good fit to the data in the foodservice setting (NFI  =  .934; TLI  =  .964; CFI  =  .967; RMSEA  =  .048; χ ²  =  796.0, df  =  416, p < .001) (Hu and Bentler, 1999). Further, the modification indices indicated that no other paths are necessary for a better model fit. Figure 2 shows the path coefficients and t‐values for significant paths (at p < .05) in the model. In addition, mediation tests revealed that three of the four insignificant paths in the model were actually meditated by other variable(s). The results of the mediation tests were discussed in a later separate section.

Figure 2 Test Results of the Proposed Model. Note: COSE: customer orientation of service employee, CB: confidence benefits, SB: social benefits, STB: special treatment benefits, FI: favorable inequity, CS: customer satisfaction, RI: repurchase intention. *p < .05, otherwise p < .001 1. Numbers in parentheses are the t‐values. 2. Numbers outside of parentheses are the standardized path coefficients. 3. Dotted arrows indicate nonsignificant paths (p > .05).

Hypotheses Testing

According to the test results in Figure 2, COSE had positive and significant effects on confidence benefits and social benefits (H1a and H1b supported) but not on special treatment benefits (H1c not supported). As proposed, as well, social benefits showed positive and significant effects on confidence benefits and special treatment benefits (H2a and H2b supported). In the meantime, social benefits' effects on favorable inequity and customer satisfaction were insignificant (H4a and H4b not supported). Confidence benefits positively and significantly affected both favorable inequity and customer satisfaction as hypothesized (H3a and H3b supported) whereas special treatment benefits positively and significantly affected only favorable inequity, but not customer satisfaction (H4a supported, H4b not supported). The effect of favorable inequity on customer satisfaction was positive and significant (H6 supported). Lastly, customer satisfaction showed a positive and significant affect on repurchase intention (H7 supported). Overall, nine out of 13 hypothesized paths were directly supported by the data.

Mediation Testing

When the path from social benefits to special treatment benefits was constrained to zero, the effect of COSE on special treatment benefits (H1c) was significant at p < .001 (β  =  .195, t  =  3.72). Thus, social benefits fully meditated the path between COSE and special treatment benefits in the model. When both paths from confidence benefits and favorable inequity to customer satisfaction were controlled, the path from social benefits to customer satisfaction (H4b) was significant at p < .001 (β  =  .341, t  =  4.59). However, when either of the paths was released, the effect of social benefits on customer satisfaction became insignificant. Thus, both confidence benefits and favorable inequity worked as full mediators in the path from social benefits to customer satisfaction. When the paths from confidence benefits and special treatment benefits to favorable inequity were controlled, social benefits showed a significant effect on favorable inequity (H4a) at p < .001 (β  =  .431, t  =  7.19). However, when confidence benefits alone was added as a mediator in the model, the effect of social benefits on favorable inequity significantly decreased (β  =  .256, t  =  4.94, p < .01) at p < .001 based on the χ ² difference (Δχ ²(1)  =  74.8). Similarly, when only special treatment benefits was added to the model, the effect significantly decreased (β  =  .221, t  =  2.91, p < .001) at p < .01 based on the χ ² difference (Δχ ²(1)  =  15.1). Therefore, confidence benefits and special treatment benefits individually partially mediated and together fully mediated the effect of social benefits on favorable inequity. Meantime, even when the effect of favorable inequity on customer satisfaction was constrained to zero, the path from special treatment benefits to customer satisfaction (H5b) was still insignificant. Thus, H5b was simply not supported by the data. In addition, when the effect of customer satisfaction to repurchase intention was set to zero, favorable inequity showed a significant effect on repurchase intention (although not hypothesized) (β  =  .836, t  =  10.87, p < .001). Therefore, customer satisfaction worked as a full mediator between favorable inequity and repurchase intention as the conceptual model indicates.

Standardized Total Effects of Predictor Variables

To include only significant effects in calculating total effects, the insignificant paths were removed from the model. Scrutinizing the total effects of each predictor variable revealed that COSE affects all of the endogenous variables directly and/or indirectly as shown in Table 3. Among the effects of COSE, those on confidence benefits, customer satisfaction, and repurchase intention were remarkable as .855 (.830 + .025), .710, and .601, respectively. From these results, the clear effects of COSE on the critical service outcome variables were confirmed. These findings are consistent with the contention that the service employee has a decisive impact on the customer's service experience due to the service employee's central role in service provision (e.g., Gremler and Gwinner, 2000; Reynolds and Beatty, 1999). Among the three types of relational benefits, confidence benefits showed the largest effects on favorable inequity, customer satisfaction, and repurchase intention as .491, .821 (.752 + .069), and .694. Such large effects seem to be related to the fact that confidence benefits are inherently tied to the quality of the core service (Gwinner et al., 1998)—the food quality in this study context. Another notable effect is that of social benefits on special treatment benefits (.681). This result leads us to suppose that the customer's receipt of special treatment benefits mainly depends on his/her interpersonal relationship with the service employee. Unlike the initial expectation, favorable inequity showed limited effects on customer satisfaction and repurchase intention as .140 and .118, respectively. Lastly, customer satisfaction exerted a prominent effect on repurchase intention (.846). To some extent, a dominant effect of customer satisfaction also seems to be associated with the determinant influence of food quality on customers' restaurant choices.

Table 3. Descriptive Statistics and Associated Measures

	No. of items	Mean (Std dev.)	AVE	COSE	CB	SB	STB	FI	CS	RI
COSE	12	5.27 (.90)	.768	.928^a	.833^b	.297	.182	.531	.794	.600
CB	4	5.48 (1.00)	.672	.694^c	.890	.330	.219	.543	.812	.664
SB	4	3.47 (1.75)	.781	.088	.109	.934	.680	.410	.298	.215
STB	4	2.35 (1.43)	.774	.033	.048	.462	.932	.431	.189	.118
FI	2	4.85 (1.47)	.701	.282	.295	.168	.186	.824	.326	.458
CS	3	5.92 (.97)	.814	.630	.659	.089	.036	.571	.929	.851
RI	2	6.14 (.95)	.831	.360	.441	.046	.014	.210	.724	.908
Goodness‐of‐fit statistics:
χ ^2 (408)  =  756.9, p < .001
χ ^2/df  =  1.86
NFI  =  .938; TLI  =  .966; CFI  =  .970
RMSEA  =  .046
Note: COSE: customer orientation of service employee, CB: confidence benefits, SB: social benefits, STB: special treatment benefits, FI: favorable inequity, CS: customer satisfaction, RI: repurchase intention, AVE: average variance extracted estimate, NFI: normed fit index, TLI: Tucker‐Lewis index, CFI: comparative fit index, RMSEA: root mean square error of approximation.^a Composite reliabilities are along the diagonal, ^b Correlations are above the diagonal, and ^c Squared correlations are below the diagonal.

Table 4. Standardized Effects of Predictor Variables on Outcome Variables

Effects	SB		CB		STB		FI		CS		RI
	Direct	Indirect	Direct	Indirect	Direct	Indirect	Direct	Indirect	Direct	Indirect	Direct	Indirect
COSE	.297	–	.830	.025	–	.202	–	.485	–	.710	–	.601
SB	–	–	.085	–	.681	–	–	.260	–	.100	–	.085
CB	–	–	–	–	–	–	.491	–	.752	.069	–	.694
STB	–	–	–	–	–	–	.322	–	–	.045	–	.038
FI	–	–	–	–	–	–	–	–	.140	–	–	.118
CS	–	–	–	–	–	–	–	–	–	–	.846	–
Note: COSE: customer orientation of service employee, CB: confidence benefits, SB: social benefits, STB: special treatment benefits, FI: favorable inequity, CS: customer satisfaction, RI: repurchase intention.

Amounts of Variance Explained

According to the squared multiple correlation (SMC; i.e., R² ), COSE explained only 8.9% of the total variance in social benefits. Its true effect was found on confidence benefits. Together with social benefits, COSE explained 73.7% of the total variance in confidence benefits and 46.3% of total variance in special treatment benefits. However, considering those R² s and the standardized effects, discussed in the previous section, we could conclude that COSE explained a major portion of the total variance in confidence benefits whereas social benefits did so for special treatment benefits. These results imply that COSE is a critical factor in explaining the customer's relational benefits by itself as well as together with social benefits. Next, COSE and relational benefits together explained 41.3% of the total variance in favorable inequity and in turn favorable inequity together with COSE and relational benefits accounted for 70.6% of the total variance in customer satisfaction. Lastly, 71.5% of the total variance in repurchase intention was explained by all the predictor variables. Overall, the R² s demonstrated that relational benefits influenced by COSE explain fairly large amounts of the total variances of the key service outcome variables of favorable inequity, customer satisfaction, and repurchase intention in the full‐service restaurant setting.

DISCUSSION AND IMPLICATIONS

Discussion

The findings clearly support the main contention of this study—the customer orientation of individual service employees enhances to customers' perceptions of relational benefits in their relationship with the restaurant and ultimately contributes to customers' long‐term relationship orientation toward the restaurant. Thus, the findings provide direct explanations for how customer orientation at the individual employee level contributes to the firm's long‐term success. Related with it, the significance of the findings should be noted as provided below.

First, as posited, COSE directly and/or indirectly enhanced the customer's perceptions of all three types of relational benefits. In a sense, these results explain why three out of five service quality dimensions in Parasuraman, Zeithaml, and Berry's (1988) SERVQUAL measure directly or indirectly deal with employees' behaviors (i.e., responsiveness, assurance, and empathy). In this regard, it might be argued that service employees' customer‐oriented behaviors promote customers' perceptions of relational benefits ‘by improving service quality.’ Thus, this study provides evidence for Hennig‐Thurau and Thurau's (2003) argument that COSE is a crucial variable that enables a service firm to provide high‐quality service.

Second, the weak effect of social benefits on confidence benefits might be due to customers' strong focus on food quality in their restaurant choices. This result is in accordance with Gwinner et al.'s (1998) acknowledgment that confidence benefits are inherently tied to core service performance (i.e., food quality in this study context). Thus, this result might lead one to conclude that the customer's confidence in the restaurant is more centered on food quality than on pleasure and comfort stemming from service interactions. On the other hand, social benefits had a strong effect on special treatment benefits, implying that when customers have a close personal relationship with a service employee, they are treated as ‘true’ relational customers. Thus, if a service employee is highly customer‐oriented and thus able and willing to develop close relationships with customers, the employee is more likely to make the customers feel special and thus retain them in relationships with the restaurant.

Third, confidence benefits and special treatment benefits enhanced customers' favorable inequity perception. These results indicate that confidence benefits and special treatment benefits improve customers' outcome‐input ratio by increasing the outcomes or decreasing the inputs. However, social benefits affected favorable inequity only indirectly through confidence benefits and special treatment benefits, implying that social benefits are secondary to the other types of benefits in relation to customers' favorable inequity perception. Customers seem to feel that close customer‐employee relationships are mutually beneficial. That is, social benefits are as beneficial to the employee as they are to the customer.

Fourth, favorable inequity mediated the effects of relational benefits on customer satisfaction. Only confidence benefits showed a strong, direct effect on customer satisfaction. This result is consistent with Gwinner et al.'s (1998) findings, which showed that confidence benefits are the most important relational benefits from customers' perspective. A long‐term customer is likely to develop confidence in both the food quality and the employee's service quality. Thus, such a strong effect of confidence benefits on customer satisfaction seems to be influenced by both the food quality and the employee's service quality. On the other hand, the effects of social benefits and special treatment benefits on favorable inequity and customer satisfaction were relatively weak in general. The reason seems to be the dominant effects of confidence benefits on favorable inequity and customer satisfaction.

Lastly, customer satisfaction showed a very strong effect on repurchase intention. This remarkably strong effect seems due to the fact that the samples in this study were composed of relational customers. That is, customer satisfaction enhanced by relational benefits can result in the retention of relational customers. Interestingly, Patterson and Smith (2001) argued that the provision of relational benefits can serve as powerful barriers to switching.

Implications

In the organizational behavior context, research has revealed that service employees' customer orientation positively influences employees' job satisfaction, organizational commitment, altruism (Donavan, Brown, and Mowen, 2004; in the bank context), and self‐ratings of overall performance (Brown et al., 2002; in the food‐service context). As such, the customer orientation of service employees has noteworthy implications for the employees themselves as well as for their customers as revealed in this study. That is, service employees' customer orientation promotes organizational functioning as well as relationship marketing. Thus, by training, motivating, and giving sufficient authority to service employees, restaurants are more likely to satisfy both employees and customers and thereby retain them longer. Surely, lower employee turnover and more customer retention significantly contribute to a restaurant's bottom line.

Kelley (1992) emphasized the importance of ‘organizational climate for service’ as a means of promoting the customer orientation of service employees. That is, the organization's customer‐oriented focus on service delivery and quality encourages service employees' customer orientation. Thus, the management should be customer‐oriented if they want their employees to be customer‐oriented. In the short term, cost reduction and fast table turnover can surely improve the bottom line. However, in the long term, retaining customers by providing beneficial relationships to them is a more effective cost‐saving and profit‐boosting strategy.

In terms of employees' personality traits, Brown et al. (2002) revealed that emotional stability, agreeability, and need for activity influence employees' customer orientation. In that their study was conducted in the food‐service setting, their findings have direct implications for food‐service businesses. Specifically, emotionally stable employees are more likely to be motivated to improve their technical and social skills. Agreeable employees are more likely to understand customers' perspectives and attempt to solve customers' problems (cf. Hogan, Hogan, and Busch, 1984). A high need for activity would lead employees to complete more tasks and do more things (cf. Buss, 1991). Thus, the management would be better advised to devise a selection procedure that ensures that their employees possess such personality traits.

Lastly, the management should check which dimensions of COSE are needed by their current employees. The remedy should begin with their shortcomings. If technical skills are most needed, a more thorough test may be employed to precisely identify which technical skills are required by each employee. Among others, role‐playing would be a useful means to improve employees' social skills by giving them insight into customers' perspectives. Numerous ways to promote employees' motivation and self‐perceived authority (or empowerment) may be found in the organizational behavior literature.

LIMITATIONS AND FUTURE RESEARCH

In light of the implications of this study, three limitations should be acknowledged. First, this study tested the conceptual model only in the full‐service restaurant context. In that COSE and long‐term customer relationships are critical to hospitality organizations, future research might extend the findings of this study to other hospitality organizations, especially those in which customer life‐time value is greater than others such as in hotels, airlines, and casinos. Second, the measures used in this study were adopted from studies of other service categories or general service. Thus, future research might tailor some measures to the food‐service business. For instance, some forms of special treatment benefits might be less or more relevant to the food‐service context than others. Third, the conceptual model of this study was tested in just one geographical area, and our samples tended to be more highly educated than average restaurant customers. These limit the generalizability of the findings of this study. Accordingly, future research may further verify or extend the current conceptual model with more diverse customers in various geographical areas.

Another suggestion for future research is to measure customers' responses to ‘service employees’ in relation to COSE levels and the effects of such responses on employees' job attitude and satisfaction. Supposedly, COSE and customers' favorable responses to service employees will create synergies in service encounters and relationships. Further, it also would be interesting and meaningful to investigate how the core service quality (e.g., food quality in the food‐service context) interacts with enhanced perceptions of relational benefits in terms of customer satisfaction and loyalty intention. Interestingly, Gwinner et al. (1998) suggested that if customers are receiving important relational benefits, they may remain in a relationship even when they perceive the core service attributes to be less than superior. Patterson and Smith (2001) concurred, arguing that the provision of relational benefits can serve as a powerful barrier to switching.

Notes

1. “The terms market oriented, market driven, and customer focused tend to be considered synonymous” (Brady and Cronin, 2001: 241).