Restaurant Crowding and Perceptions of Service Quality: The Role of Consumption Goals and Attributions

Abstract

This study examines the moderating effect of consumptions goals on the perceived crowding-service quality relationship in the context of restaurant experiences. It also investigates the effect of attributions for crowding on customer evaluations of service quality. The study found that, when the consumption goal is primarily utilitarian in nature, a non-crowded restaurant environment results in higher service quality evaluations. Conversely, for hedonic consumption goals, higher service quality evaluations are associated with a crowded environment. Findings also indicate that customer attributions for crowding have a direct effect on service quality ratings. Managerial implications and future research directions are discussed.

KEYWORDS:

• crowding
• consumption goals
• service quality
• attributions
• restaurants

INTRODUCTION

Delivering quality service is considered an essential strategy for success and survival in a competitive environment such as that of the restaurant industry (Dawkins & Reichheld, 1990; Parasuraman, Zeithaml, & Berry, 1985; Reichheld & Sasser, 1990; Zeithaml, Parasuraman, & Berry, 1990). A positive relationship between service quality and a number of key outcome variables has been indicated in the literature including value perceptions (e.g., Gale, 1994; Parasuraman & Grewal, 2000; Wakefield & Barnes, 1996), satisfaction (e.g., Anderson & Fornell, 1994; Andreassen, 1998; Dabholkar, Shepherd, & Thorpe, 2000) and behavioral intentions (e.g., Athanassopoulos, 2000; Boulding, Kalra, Staelin, & Zeithaml, 1993; Zeithaml, Berry, & Parasuraman, 1996). Given the critical role of service quality in driving these outcomes, it is important that restaurant operators engineer the service encounter to enhance customer perceptions of service quality. The service encounter can be defined as the encompassing all aspects of the service firm with which the consumer interacts—including its personnel, its physical facilities, and other tangible elements—during a given period of time (Shostack, 1985). This article focuses on the customer's interaction with the physical environment and in that context examines the effect of crowding on customers' global evaluations of service quality.

Crowding can be defined either from a physical (number of persons in a given space) or from a psychological perspective (perceived crowding). This study focuses on the latter dimension of crowding. Prior research on crowding has focused primarily on the retail environment. Research has shown that the level of in-store crowding perceived by shoppers can affect their patronage decisions as well as satisfaction with the overall shopping activity (e.g., Eroglu & Machleit, 1990; Eroglu, Machleit, & Feldman Barr, 2005; Machleit, Eroglu, & Mantel, 2000; Machleit, Kellaris, & Eroglu, 1994). Potential buyers may even deviate from their planned shopping experience (e.g., spend less money than planned or even leave the store without making a purchase) when the store is crowded (Harrell, Hutt, & Anderson, 1980). A negative relationship between perceived crowding and pleasure has also been demonstrated in other service settings (e.g., banks and bars; Hui & Bateson, 1990, 1991).

We expand this research by investigating the relationship between crowding and customer perceptions of service quality and we examine this relationship in the context of extended service encounters in a restaurant setting. Price, Arnould, and Tierney (1995) differentiate between brief and extended service encounters, with brief service encounters typically lasting less than 10 minutes (e.g., customer interaction with a bank teller), while extended service encounters extend over longer periods of time, often lasting more than 30 minutes (e.g., customer interaction with a waiter or river guide). First, we examine the moderating effect of consumption goals on the crowding-service quality relationship. Consumer goals can be broadly classified as either utilitarian (i.e., functional) or hedonic (i.e., experiential) in nature. Here we seek to answer the question: does the nature of the goal matter when customers are making a service quality judgment in a crowded restaurant environment? Furthermore, we are interested in the role of attributions in determining the crowding-service quality relationship. Specifically, if customers attribute crowding to the restaurant firm, will they hold this against the firm when making a service quality judgment?

The structure of this article is as follows. First, we review the literature relevant to our research hypotheses. We then present a description of our research methodology and empirical results. We conclude with a discussion and managerial implications.

CONCEPTUAL BACKGROUND

In this study we rely on Brady and Cronin's (2001) conceptualization of service quality as a multi-level, multi-dimensional construct. At the uppermost level are customers' overall perceptions of service quality. At the next level are the three dimensions of service quality: interaction quality, physical environment quality, and outcome quality. The lowest level comprises nine subdimensions: attitude, behavior and expertise of service personnel (interaction quality), ambient conditions, design and social factors (physical environment quality), waiting time, tangibles, and valence (outcome quality). This multilevel conceptualization allows for analysis at several levels of abstraction. If the focus is on identifying specific areas that are in need of attention, it is appropriate to use measures that capture each individual dimension of service quality. If, on the other hand, the focus is on prediction, measures for overall service quality evaluations can be applied (Dabholkar et al., 2000). Given that the purpose of this study is to examine the crowding-service quality relationship, we rely on customers' overall evaluations of service quality.

Prior studies in crowding have clearly demonstrated a negative impact of crowding on customers' emotional responses and satisfaction (e.g., Eroglu & Machleit, 1990; Eroglu, Machleit, & Feldman Barr, 2005; Harrell et al., 1980; Hui & Bateson, 1991; Machleit, et al., 1994; Machleit, et al., 2000). We extend this literature to the service quality domain, proposing that the negativity effect of crowding on service quality perceptions might be limited to situations where overcrowding hinders consumers' goal attainment. Prior research shows that consumers enter consumption situations with specific goals in mind (Bitner, 1992) and these goals can be roughly classified into two types: hedonic and utilitarian (Dhar & Wertenbroch, 2000). Hedonic goals focus on the consumption experience (Holbrook & Hirschman, 1982), thus reflecting the need for pleasure, fun, and excitement (e.g., a relaxing dinner). Utilitarian goals, on the other hand, are mainly instrumental or functional in nature (e.g., a quick meal at a restaurant to satisfy hunger). Previous studies further demonstrate that crowded environments hinder the goal attainment of task-oriented shoppers (Erologu & Harrell, 1986; Erologu & Machleit, 1990; Erologu,, Machleit, & Feldman Barr, 2005). Hence we expect that service quality perceptions will be negatively affected when the consumption goal is utilitarian. In other words, when the customer is in for a quick bite over lunch, and the restaurant is crowded, service quality inferences might be negatively impacted. Conversely, we expect that when the consumer goal is a hedonic experience, a crowded environment will enhance service quality perceptions. With leisure services, such as restaurants, consumers expect a certain level of excitement or emotional arousal (Wakefield & Blodgett, 1994). We propose that this is particularly true when the primary goal for a dining experience is hedonic in nature. The excitement and ambience (provided by other consumers in the servicescape) is a desired attribute. Therefore, we hypothesize:

H1: Consumer goals moderate the crowding-service quality relationship for restaurant experiences. When the goal is hedonic, service quality ratings will be higher when the restaurant is crowded. Conversely, when the goal is utilitarian, service quality ratings will be higher when the restaurant is not crowded.

Attributions have an important impact on consumer reactions to service failures (e.g., Folkes, 1988; Mattila & Patterson, 2004; Weiner, 2000), yet the concept of globality has been largely ignored (Hess, Ganesan, & Klein, 2007). In this study we are interested in how consumer perceptions of crowding influence their perceptions of the entire organization (generalized to the restaurant firm as opposed to the front-line employee or manager). Global attributions have been studied in psychology (Bradbury & Fincham, 1990) and in consumer psychology (Folkes & Patrick, 2003; Hess, et al., 2007; Matta & Folkes, 2005). Folkes and Patrick (2003) found that consumers easily generalize positive employee behavior to the entire organization. We extend this research by examining the role of global attributions in determining the crowding-service quality relationship. When customers perceive that a restaurant is crowded, they may reason that it is not the restaurant firm's fault. For example, sports events such as football games might bring an unexpected number of visitors to town, thus making it difficult for restaurants to handle the customer flow. On the other hand, customers may believe that the restaurant firm should have anticipated the crowding and done something about it. For example, customers may assess that restaurant operators should anticipate a high level of demand on a Saturday night or on a special occasion such as Mother's Day. When customers evaluate that the restaurant firm should have anticipated, and managed, crowding, we propose that customers may hold this against the restaurant firm when making an overall service quality judgment. Therefore, we hypothesize:

H2: The attributions that customers make about the restaurant firm's role in crowding have a direct effect on service quality perceptions.

METHODOLOGY

Study Design and Subjects

A 2 (Crowding: crowded or not crowded) × 2 (Goal: utilitarian or hedonic) × 2 (Service Level: bad or good) factorial, between-subjects design, was used to test the hypotheses. Service level was manipulated at two levels in the experiment, good and bad service, to control for the potential effect of encounter valence on the crowding-service quality relationship. Subjects were randomly assigned to a written scenario describing an experience in a restaurant. They were then shown a photograph depicting the interior of the restaurant. The subject pool was composed of undergraduate students enrolled in a large state university in the northeastern United States. Students were considered to be suitable subjects for the study as they are real-life consumers of restaurant services. Of the respondents (n = 198), 50.5% (n = 100) were male and the age range was 18 to 25 years of age.

Procedure

Subjects were given a written scenario describing an experience in a casual restaurant environment. In the utilitarian goal condition, subjects were asked to imagine that it is a Friday evening and they have decided, while en route by car to Philadelphia from Boston, to stop at a restaurant for a quick dinner because they are hungry. In the hedonic goal condition, subjects are told that they meet a friend for dinner on a Friday night. After an intense work week they are looking forward relaxing and socializing with their friend. In the good (bad) service level condition, the server is attentive (inattentive) and prompt (slow). At the end of the scenario subjects were shown a photograph depicting the interior of the restaurant while they were there. In the no-crowding condition the photograph showed the restaurant with a low customer volume (some vacant tables) and in the crowded condition all tables were occupied.

Measures

Service quality

We used a two-item, 7-point Likert scale adapted from Brady and Cronin (2001) and Dabholkar, Shepherd, and Thorpe (2000) to measure global perceptions of service quality (r = 0.79, p < 0.01): “The overall quality of this restaurant is” (poor/outstanding), “This restaurant offers excellent service” (very strongly agree/very strongly disagree).

Attribution

We used a 3-item, 7-point Likert scale (very strongly agree/very strongly disagree) adapted from Hui and Toffoli (2002) and Russell (1982) to measure attribution for crowding (Cronbach's α = 0.80): “It is the restaurant's fault that the place was crowded”, “The restaurant should have anticipated the crowds and done something about”, “Little could have been done by the restaurant to stop the crowding”.

Manipulation checks

We used a 5-item, 7-point Likert scale (very strongly agree/very strongly disagree) adapted from Machleit, Meyer, and Eroglu (2005) as a manipulation check for crowding (Cronbach's α = 0.80): “The restaurant seemed very crowded to me”, “The restaurant was a little too busy”, “There were a lot of customers in the restaurant”, “The restaurant seemed very spacious” and “I felt cramped in this dining space”.

We used two items to ensure that our manipulation for goals was effective: “When I entered the restaurant, my main goal was to have a quick dinner” (utilitarian goal) and “When I entered the restaurant, my main goal was to have a relaxing dining experience” (hedonic goal).

We used a 1-item, 7-point Likert scale (poor/outstanding) as a manipulation check for service level: “The quality of the service provided by this restaurant is”.

Tolerance of crowding

Since consumer tolerance of crowding can influence their perceptions of retail environments (Machleit, et al., 2005), a tolerance of crowding measure was included as a covariate in the ANOVA analysis. A 3-item, 7-point Likert scale (very strongly agree/very strongly disagree) adapted from Machleit, Meyer, and Eroglu (2005) was used to measure tolerance of crowding (Cronbach's α = 0.65): “I avoid crowded places whenever possible”, “A crowded restaurant doesn't really bother me”, and “If I see a restaurant that is crowded, I won't even go inside”.

The questionnaire was pre-tested and revised before its final administration. Following the pre-test, a number of changes were made to the layout of the questionnaire to facilitate ease of completion.

DATA ANALYSIS

Manipulation Checks

As expected, subjects in the crowded condition exhibited significantly higher ratings on the crowding scale than the subjects in the not crowded condition (mean ratings of 5.6 and 2.6 for crowding and no crowding respectively, t = 19.61, p < 0.001). Mean ratings for the goal item “When I entered the restaurant, my main goal was to have a quick meal” were significantly higher for the utilitarian goal condition than the hedonic goal condition (mean ratings of 4.7 and 3.2 for utilitarian and hedonic goals respectively, t = 6.85, p < 0.001). Mean ratings for the goal item “When I entered the restaurant, my main goal was to have a relaxing dining experience” were significantly higher for the hedonic goal condition than the utilitarian goal condition (mean ratings of 5.4 and 4.4 for hedonic and utilitarian respectively, t = −4.95, p < 0.001). Finally, subjects in the good service level condition exhibited significantly higher ratings on the service level measure than the subjects in the bad service level condition (mean ratings of 5.9 and 2.5 for good and bad service levels respectively, t = 20.23, p < 0.001). Taken together, these results indicate that our manipulations were effective.

Results

We used ANCOVA with perceived service quality as the dependent variable and tolerance for crowding as a covariate to test our hypotheses. The cell means for perceived service quality by crowding, goal and service level conditions are shown in Table 1.

TABLE 1 Means for Perceived Service Quality

Service level	Crowded condition	Goal condition	Perceived service quality
Good	Crowded	Utilitarian	5.46
		Hedonic	5.33
	Not crowded	Utilitarian	5.69
		Hedonic	5.21
Bad	Crowded	Utilitarian	3.02
		Hedonic	3.07
	Not crowded	Utilitarian	3.60
		Hedonic	2.68

We used the crowding, goal and service level conditions as the independent variables to test H1. The results are provided in Table 2. As expected, the two-way interaction between goal and crowding was significant (F=5.01, p < 0.05; Figure 1). Specifically, when the goal was utilitarian in nature, significantly higher service quality ratings were attained in the not crowded condition (Not Crowded: M = 4.77, Crowded: M = 4.10, t = −2.09, p < 0.05). Conversely, when the goal was hedonic in nature, significantly higher service quality ratings were attained in the crowded condition (Not Crowded: M = 3.76, Crowded: M = 4.45, t = 2.24, p < 0.05).These results support H1.

TABLE 2 ANCOVA Results on Perceived Service Quality

Source of variation	df	Sum of squares	Mean square	F	Sig.
Hypothesis 1:
Tolerance of crowding	1	0.12	0.12	0.11	ns
Service level (SL)	1	257.35	257.35	251.48	p < 0.001
Goal	1	6.38	6.38	6.24	p < 0.05
Crowding	1	0.25	0.25	0.24	ns
SL*goal	1	0.21	0.21	0.21	ns
SL*crowding	1	0.04	0.04	0.04	ns
Goal*crowding	1	5.13	5.13	5.01	p < 0.05
SL*crowding*goal	1	1.13	1.13	1.10	ns
Error	187	191.37	1.02
Total	196	4068.50
Hypothesis 2:
Tolerance of crowding	1	0.48	0.48	0.49	ns
Attribution for crowding	1	7.06	7.06	7.20	p < 0.01
Service level (SL)	1	143.16	143.16	146.00	p < 0.001
Goal	1	0.07	0.07	0.07	ns
SL*goal	1	0.08	0.08	0.09	ns
Error	102	100.01	0.98
Total	108	2253.25

FIGURE 1 Interaction effect of goal and crowding on perceived service quality.

To test H2, an ANCOVA was performed on the crowded subsample (N = 109), with attribution for crowding added as an independent variable. The results are provided in Table 2. As suggested by the plot of the goal-crowding interaction in Figure 1, the two-way interaction between goal and crowding was insignificant (F = .09, p > 0.1). The effect of crowding on perceptions of service quality was relatively consistent across goal type. As expected, the main effect for attribution for crowding was significant (F = 7.20, p < 0.01). Attribution for crowding has a direct significant effect on perceptions of service quality. This supports H2.

DISCUSSION

The relationship between perceived crowding and consumption goals has been examined in retail environments, with prior research indicating that crowded environments hinder the goal attainment of task-oriented shoppers (Eroglu & Harrell, 1986; Eroglu & Machleit, 1990, Eroglu, et al., 2005). A negative relationship between crowding and hedonic shopping values (Babin, Darden, & Griffin, 1994; Eroglu, et al., 2005) has also been demonstrated. This study extends prior research by examining the moderating effect of consumption goals in the crowding-service quality relationship in the context of extended service encounters in a restaurant setting. Findings indicate that, when the primary goal for a dining experience was utilitarian in nature, significantly lower service quality ratings were associated with a crowded restaurant than a non-crowded restaurant, regardless of service level or customer tolerance of crowding. This is consistent with the notion that crowded environments hinder the goal attainment of task-oriented consumers (Eroglu & Harrell, 1986; Eroglu & Machleit, 1990, Eroglu, et al., 2005) and this negatively affects service quality inferences, while a non-crowded environment facilitates the goal attainment for utilitarian experiences. On the other hand, when the primary goal for the dining experience was hedonic in nature, significantly higher service quality ratings were associated with a crowded restaurant environment than a non-crowded environment, regardless of service level or customer tolerance of crowding. This supports the idea that when the primary goal for a dining experience is hedonic in nature, the servicescape is an important component of the experience. Specifically, the excitement and ambience provided by the presence of other consumers contribute positively to consumers' quality inferences.

In this study we were also concerned with the effect of customers' attributions for crowding on service quality evaluations. Our findings indicate that customers' attributions regarding the firm's role in crowding have a direct effect on service quality perceptions. Prior research has demonstrated that attributions have an important impact on consumer reactions to service failures (e.g., Folkes, 1988; Mattila & Patterson, 2004; Weiner, 2000). This study contributes to our understanding of the role of attributions in customers' reactions to the servicescape. If consumers perceive that the restaurant had control over crowding and failed to do anything about it, then service quality ratings are likely to suffer.

MANAGERIAL IMPLICATIONS

In an increasingly competitive operating environment, restaurant operators must strive to engineer the service encounter to enhance customer perceptions of service quality. This study revealed that a crowded environment can enhance perceptions of service quality when the primary goal for the dining experience is hedonic. However, when the goal is utilitarian, a non-crowded environment will provoke a more positive evaluation of service quality than a crowded restaurant. So, how can a restaurant operator balance the desire for a more crowded environment to contribute to the ambience sought by customers during hedonic dining experiences, but also, create the impression of a less crowded environment to appeal to those diners with more utilitarian dining goals? The key may be to focus on the layout and design of the physical space within the restaurant. Specifically, management could consider dividing the service area into small, discrete sections such that each section appears comfortably busy as opposed to crowded. For example, restaurant operators could fill small areas with diners in careful sequence so that each section maintains a good ambience without being crowded. This design should appeal to those seeking a hedonic experience. Equally, for dining experiences driven by utilitarian goals, customers are more likely to perceive that their goals can be attained when the actual volume of customers in the restaurant is obscured. Demand controls such as reservations or special price promotions can also be used by restaurant operators to shift demand to off-peak demand periods. Implemented effectively, these approaches to demand management will not only eliminate over-crowding during peak demand periods, they should also lead to increase revenue and capacity utilization during off-peak periods.

Study findings also indicated that customer attributions for crowding have a direct effect on their perceptions of service quality. Specifically, quality evaluations are affected when customers attribute crowding to the restaurant firm. So, how can management diminish these attributions and give customers the impression that they are making an effort to manage the increase in customers? Increasing staffing levels may achieve this, although operators need to assess the profit impact of this and establish optimal staffing levels. Depending on the underlying reason for crowding, restaurant operators might be justified explaining the cause of crowding to their patrons and, in doing so, make sure that they understand that the cause was beyond their control. However, explanations need to be sincere and not sound like excuses for poor performance.

LIMITATIONS AND EXTENSIONS

This study was limited to one type of restaurant (casual restaurants). Future research across other types of restaurants and extended encounters in other service environments is needed to establish the generalizability of our findings. While tolerance for crowding was controlled for in this study, demographic variables such as age group may also influence customers' reactions to crowding. This should be addresses in future research. An interesting research area to consider is that of an optimal crowding level. Is there an optimal crowding level associated with different types of restaurants (e.g., luxury vs. casual restaurant) and restaurant images (a “hip” restaurant vs. the local diner)? If consumers hold different expectations regarding crowding levels across different types of service environments, how can management manipulate this to improve customer perceptions of service quality?