The Emerald Research Register for this journal is available at http://www.emeraldinsight.

com/researchregister

The current issue and full text archive of this journal is available at http://www.emeraldinsight.com/0956-4233.htm

IJSIM 14,1

36

The effect of service guarantees on service recovery

Department of Business and Economics, Service Research Center, Karlstad University, Karlstad, Sweden
Keywords Service operations, Service quality, Customer care, Guarantees Abstract Service guarantees have been attributed the benefit of improving the overall service of a service provider. However, little research has been carried out within the area. This article focuses on one aspect of the service guarantee, the effects that service guarantees may have on service recovery. Critical incident data were collected using the critical incident interview technique with customers of RadissonSAS, a worldwide hotel chain using a service guarantee. One contribution of this article is that the interviews convey that the implicit guarantee may serve as a risk reducer, which contradicts and adds to previous research. Previous research states that only the explicit guarantee has these benefits. In this case, the guarantee does not reduce risk in the purchase or consumption stage, but after the consumption when the service has failed, as the customer finds out about the guarantee in the recovery situation. Another contribution of this article is that service guarantees are found to influence the outcome of service recovery as they affect how employees behave to recover the customer.

Sara Bjorlin Liden and Per Skalen

International Journal of Service Industry Management Vol. 14 No. 1, 2003 pp. 36-58 # MCB UP Limited 0956-4233 DOI 10.1108/09564230310465985

Introduction Service guarantees have been claimed to be a key to success in terms of improving the process of service recovery, employee performance, providing data on service failures, developing measures for customer satisfaction and setting performance standards (Hart, 1988; Maher, 1992; Ettorre, 1994). These statements are based on experiences from a few companies, such as Nordstrom, Domino's Pizza, or Lands End Hotels, which successfully have implemented service guarantees on their offerings (Hart, 1998). However, neither methodological approaches nor theoretical standpoints are presented to authenticate the findings (Hart, 1988; 1995; 1998; Heskett et al., 1990; Maher, 1992; Ettorre, 1994). Nevertheless, such anecdotes have become the archetype for guarantee design and development. Despite the mentioned shortcomings and academic rigidity in much research on service guarantees, a few recent articles have presented in depth research on the service guarantee as a quality signal (Tucci and Talaga, 1997; Wirtz et al., 2000) and guarantee design matters (Donath, 1997; McDougall et al., 1998; Hill et al., 2000). This research conveys insight into how the guarantee may affect customers prior to purchase, in reducing the risks of an uncertain service. However, post-purchase experiences of a service guarantee, i.e. in the recovery
The authors would like to thank Professor Bo Edvardsson, Karlstad University, Professor Tore Strandvik, Swedish School of Economics and Business Administration in Helsinki and the two anonymous IJSIM reviewers for their useful comments.

context, have not yet been examined. Nevertheless, service guarantees are Effect of service presumed to have a positive effect on service recovery (Tax and Brown, 2000), guarantees on especially as they may communicate to customers that employees take service recovery responsibility for their failures, and may set performance standards that systemize how employees deal with such failures. In this article, we view the service guarantee as a tool to systematize and formalize the recovery process. 37 However, little empirical research confirms or contradicts whether the guarantee affects these issues. Therefore, the aim of this article is to study the possible effect that service guarantees may have on service recovery. More specifically, we address an issue that has been called for in recent articles (Bolton and Drew, 1995; Ostrom and Iacobucci, 1998): what role the guarantee has on customers' subsequent behavior in relation to the service provider. In order to do so, we focus on the guarantee in the situation when it is part of the service recovery efforts of aiding a negative critical incident. However, the guarantee should not be separated from its service context and be assumed to be the sole contributor to recovery. We argue that the service guarantee can be viewed as one tool to aid a critical incident in the service recovery context. However, several additional matters may affect customers' future intentions, such as the length of relationship (Liljander and Strandvik, 1995), the criticality if the critical incident (Edvardsson and Strandvik, 2000) and the behavior of front-line employees (Parasuraman et al., 1985; 1988; 1994; Zeithaml et al., 1990; Edvardsson, 1996). In this paper, the service recovery context refers to the process that begins when the company becomes aware that dissatisfaction has occurred, to the situation when the problem has been solved, and/or the customer has been reimbursed to achieve satisfaction. Because of the complexity of such a process, we argue that it is important not to study service guarantees as an isolated matter. Empirically, this article is based on negative critical incident interviews with customers of RadissonSAS, which is a worldwide hotel chain serving mainly the business segment. The entire organization has used a ``100 percent guest satisfaction guarantee'' for several years, with a focus on improving the service and the guarantee to better serve customers. The outline of the paper is as follows: first, we make a theoretical overview concerning service guarantees and service recovery. Thereafter, our interpretative methodological standpoint and qualitative methods are explained. Then, the empirical findings are presented, followed by an analysis and interpretation of how a service guarantee can aid or obstruct in recovering customers that have experienced service failures. We conclude the article by discussing our main findings and directions for future research. Theoretical framework The nature of services and the need for service recovery The intangible nature of services makes the management of expectations very important as service companies seek to influence customers' perception of the service (Parasuraman et al., 1985; 1988; 1994). In fact:

IJSIM 14,1

When prospective customers can't experience the product [or service] in advance, they are asked to buy what are essentially promises, so called promises of satisfaction. Even tangible, testable, palpable, smellable products are, before they are bought, largely just promises (Levitt, 1981, p. 96).

38

These promises are delivered to customers by front-line employees in what is referred to as ``the moment of truth'' (Edvardsson, 1996; Gronroos, 2000). The moment of truth is often characterized as a critical incident (Edvardsson and Strandvik, 2000) where research has found that behavior of front personnel often influence whether the outcome is positive or negative. A few traits significantly seem to influence customer perceived quality: . Empathy. Concerns the caring and individualized attention that the front personnel provide for its customers (Parasuraman et al., 1985; 1988; 1994). . Adaptability. The ability of the front personnel to meet the needs of different customers (Edvardsson, 1996). . Responsiveness. Regards the readiness to help customers and to provide prompt service. The dimension emphasizes attentiveness in dealing with the customer requests, complaints and demands and seems to be of significant importance in a service recovery context (Parasuraman et al., 1985; 1988; 1994). . Trust and reliability. The ability of the front personnel to behave in a way that customers rely on them (Zeithaml et al., 1990; Morgan and Hunt, 1994; Rousseau et al., 1998). Customers' expectations of the service are subjective and based on needs and desires that they expect to fulfill using the service (Gronroos, 1992). The difficulty of actually delivering according to these expectations creates a need for a support system for service failures, if there is a long-term approach to the customer relationship. Service recovery is consistent with recognizing that service failures have occurred but also doing something to correct it (Sasser et al., 1990; Johnston, 1995), and its primary purpose is to handle the moments of truth (Edvardsson, 1996) as efficiently as possible. Brown et al. (1996, p. 32) propose a simple yet clarifying definition of service recovery which is in accordance with the view of service recovery applied in the present article; service recovery, they say, is consistent with ``. . . fixing or compensating for service failure . . .'' In order to preserve or enhance the perceived service quality after a failure, companies need a service recovery strategy. Empirical research on service recovery is, however, still scarce. The most prominent conclusion is that the actions taken by the front-line employees will have a significant effect on the outcome of service recovery (Zemke, 1995; Brown et al., 1996; Tax and Brown, 1998; Webster and Sundaram, 1998; Boshoff and Leong, 1998; Boshoff and Allen, 2000). Brown et al. (1996) argue that front-line employees should be provided with resources and authority to handle the problem if the customer

relationship after a service failure is to be improved or unchanged in terms of Effect of service customer perceived service quality and satisfaction. In addition, Tax and guarantees on Brown (1998) emphasize the importance of fair and just treatment of customers service recovery when a service failure has occurred. Accordingly, Boshoff and colleagues (Boshoff and Leong, 1998; Boshoff and Allen, 2000) have used quantitative methods to show that empowerment of front-line employees has a positive 39 impact on the outcome of service recovery. Previous research on service recovery also concludes that the customers' reaction is highly dependent on the graveness of service failure (Bitner et al., 1990; Hoffman et al., 1995; Barlow and Moller, 1996; Webster and Sundaram, 1998; Blomqvist et al., 1999). Webster and Sundaram (1998) have shown that the perceived importance of successful service delivery in a given service encounter will have a great impact on the service recovery outcome. In other words, a high-critical situation will be harder to recover compared to a lowcritical situation. In a relationship context it can be argued that customers, who intend to interact with the company on more than one occasion, have a longterm interest in the quality of the service, and wish to prevent the incident occurring again. Thus, the customers' assessment of the relation whether it is worth keeping or not influences the outcome of the recovery process, which indicates that neither the critical incident, nor the recovery situation is to be viewed as isolated from prior experience, or in other terms, the pre-history of the relationship (Boshoff and Leong, 1998; Tax and Brown, 1998). To conclude, service recovery literature also states that the kind/type of measures taken by front-line employees will affect the service recovery outcome. One suggestion is that a positive outcome will occur if the front-line employees ``fix the problem'' (Brown et al., 1996; Boshoff and Leong, 1998). Hoffman et al. (1995) have, however, shown that customers of restaurants prefer some kind of economic compensation (free food, discount or a coupon) compared to a replacement or correction, which is in accordance with Webster and Sundaram (1998) who demonstrated that customers prefer economic compensation in low-critical situations. In high-critical situations, however, customer prefers that the service provider fix the problem (Webster and Sundaram, 1998). Research is concordant that any effort taken is preferable to no effort at all. In short, the most important conclusion of the scarce but growing body of empirical service recovery research is that the behavior of front-line employees will have a great effect on the outcome of service recovery. Service guarantee in a service recovery context To guarantee a service simply means to present measures for service quality and to offer compensation[1] in cases when the promised quality is not achieved. A service guarantee can thus serve as a means for service recovery (Tax and Brown, 1998), i.e. ``fixing or compensating for service failure . . .'' (Brown et al., 1996, p. 32). To the design, the guarantee is either explicit; put in writing and actively advocated, or implicit; used only when customers mention that they are not

IJSIM 14,1

40

satisfied. The explicit guarantee is based on knowledge it is marketed and communicates messages displaying that customer satisfaction is a priority in the organization (Hart, 1998). It also communicates the service quality level that customers can expect from the provider. The implicit guarantee is based on experiences the company must teach its customers that it is devoted to customer satisfaction. Ideally, the implicit guarantee functions as an internal quality statement, that improves the internal quality and efficiency, which eventually is believed to raise customer satisfaction. Moreover, customers' expectations may also be influenced by another service guarantee design issue: the terms for when the guarantee may be invoked. In this respect, the unconditional guarantee is considered superior, as it does not stipulate when the customer may, or may not, invoke the guarantee (Maher, 1992; Ettorre, 1994; Hart, 1998). The specific guarantee conditions state when the guarantee is applicable to the service. Although the comprehensive aim of a service guarantee is to achieve higher rates of customer satisfaction and to offer compensation, previous research identifies several additional and more specific purposes of a service guarantee. First, they may be a mean to communicate a certain quality level to customers, as customers believe that low quality providers cannot afford to honor a high quality guarantee (Boulding and Kirmani, 1993; Tucci and Talaga, 1997). Second, service guarantees may reduce the customer's uncertainties when purchasing an unfamiliar service or they may minimize the negative consequences of the dysfunctional service. Here, it clarifies what customers rightfully can expect of a service, and it presents customers with an incentive to complain (Barlow and Moller, 1996; McDougall et al., 1998). That is, a service guarantee may have a risk-reducing function (Berry, 1995). Third, guarantees may serve as an impetus to improve the internal organizational processes by guiding the service recovery actions taken by the front-line employees (Maher, 1992; Hart, 1998). When set expectations fall short the guarantee may thus serve as a recovery method that prevents service failures having fatal impacts on the customer relationship (Zemke, 1995; Heskett et al., 1997; Tax and Brown, 1998; Hart, 1998). In the next section we outline in what ways service guarantees may serve as a mean of service recovery from the standpoint of our relationship approach. Service guarantees as a means of service recovery We conceptualize service guarantees to function as a means for recovery in two ways: first, it seems plausible that both the explicit and the implicit guarantee will affect service recovery because previous research assumes that guarantees will guide the behavior of front-line personnel when the customers complain (Maher, 1992; Hart, 1998). Such empowerment of front-line employees will have a positive effect on service recovery (Boshoff and Leong, 1998; Boshoff and Allen, 2000), as companies should train their people to act on service failures with the help of the service guarantee. Also the training, or possibly the design

of the guarantee, may enhance the ability of front-line employees to act on a Effect of service service failure. Clearly, the guarantee statement itself serves as a guideline for guarantees on employees as it states what is to be delivered to the customer. service recovery According to Tax and Brown (1998), customers should perceive that they receive fair treatment if the outcome of service recovery is to be satisfactory. A distinction is made between fair outcomes, fair processes and fair interactions:
Outcome fairness concerns the results customers receive from complaints. Procedural fairness refers to the policies, rules and timeliness of the complaint process. Interactional fairness focuses on the interpersonal treatment received during the complaint process (Tax and Brown, 1998, p. 79).

41

Fair outcomes are synonymous with having the right compensation; depending on the inconvenience caused, the right refund, the right correction, etc. Fair processes are processes that are clear to the customers and that the actual problem is dealt with quickly. Fair interactions are described as:
Demonstrating politeness, concern, and honesty; providing an explanation for the failure; and making a genuine effort to resolve the problem (Tax and Brown, 1998, p. 81) .

It can be argued that the three types of fairness will promote the quality factors outlined in the beginning of the theory section responsiveness, empathy, adaptability trust and reliability and that lack of fairness will not. Possibly, the utilization of service guarantees in companies will positively affect the three types of fairness proposed by Tax and Brown (1998). The second manner in which service guarantees may serve as a means of service recovery is that they reduce the risk of both the service firm and the customer (Berry, 1995). The risk of the service firm will be reduced by the mitigating effect that explicit service guarantees may have on switching from relationships. Singh (1990) has found that high levels of perceived probability of successful complaint treatment are associated with lower levels of exit, i.e. switching behavior. In accordance with previous research, the explicit service guarantee will communicate to customers that the service provider takes complaining seriously and therefore enhance the customers perceived probability of successful complaint. Besides the service guarantee's effect on switching, it is probable that the explicit guarantee will lower the customer perceived risk (Barlow and Moller, 1996; McDougall et al., 1998). According to prospect theory (Kahneman and Tversky, 1979) the degree of risk that customers connect with a purchase will affect their decision. Customers make their decisions on the basis of previous experiences in similar, but also different, conditions. Rust et al. (1999, p. 85) question a number of assumptions from research on perceived service quality with the help of prospect theory, displaying that:
Subjects did not necessarily choose that brand with the greatest expected performance. Rather, they balanced the brand's expected performance against its variability in performance.

IJSIM 14,1

42

By promising a good service, the explicit guarantee may thus lower the customer perceived quality variance, which, according to prospect theory, is synonymous with lowering the customer perceived risk and increasing the likeliness of customers to choose that service provider. Studying the effect of service guarantees on service recovery with a relationship approach puts a focus on perceived behavior of front-line employees in the complaint situation, and on the risk reducing effects of service guarantees. Compared to the general definition of service guarantees outlined above, guarantees are in the present article defined as a means for service recovery in terms of lowering risk and to guide service recovery actions taken by the front-line employees in order to preserve relationships. Method The aim of this article is to study the possible effects of service guarantees on service recovery. Focus is on the risk-reducing effect of service guarantees and how guarantees influence employee service recovery behavior. As argued in the introduction, research in this area is lacking, which is why we choose to conduct exploratory research. To study new phenomena with an exploratory interest often calls for a qualitative methodology (Alvesson and Skoldberg, 2000). Qualitative methodology and data gathering technique Within qualitative methodology a division is often made between functionalist and interpretative approaches of analyzing and interpreting data (Burrell and Morgan, 1979). One important line of demarcation between these approaches is that a functionalist approach regards the results as objective truths, but the interpretative approach holds them to be inter-subjective truths (Alvesson and Skoldberg, 2000). Considering the inapt research on service guarantees, and our explanatory interest, we argue that it is difficult to express objectives about the nature of service guarantees and its effect on service recovery. Instead, we have conducted in-depth studies of a specific empirical situation in order to understand the customers' subjective perception of the phenomenon of service guarantees. Thus, in this article we draw on the interpretative approach. As some type of negative incident proceeds the invoking of the guarantee, the critical incident technique (CIT) was a natural choice to gather data and structure the interviews (Edvardsson, 1988; 1992; 1996; Roos and Strandvik, 1996; Edvardsson and Strandvik, 2000). However, the presentation of customers' behavior subsequent to the incident and the invoking of the guarantee are inspired by the narrative approach (Van Mannen, 1988; Czarniawska, 1998; 1999). Besides the correspondence with our interpretative approach, we favor a presentation that allows the opinions of several customers to stand out. Therefore, two stories ``the comedy'' and ``the tragedy'' about the effect of service guarantees on service recovery are presented, both inspired by literary theory and fiction which is consistent with the narrative method

(Czarniawska, 1998; 1999). Czarniawska argues that narratives can be used in Effect of service several manners: guarantees on
Narrative enters organization studies in at least four forms: organizational research that is written in a storylike fashion (``tales from the field,'' to paraphrase Van Mannen, 1988); organizational research that collects organizational stories (tales of the field); organizational research that conceptualizes organizational life as story making and organizational theory as story reading (interpretative approaches); and a disciplinary reflection that takes the form of literary critique (Czarniawska, 1998, pp. 13-14).

service recovery

43

In this article we conceptualize our empirical findings as research presented in a story like fashion. In this specific approach, the researcher converts stories collected from the field (the CIT interviews) to one story, or several, which convincingly reason according to the thesis presented in the analysis. A critical incident is defined as:
A specific, unusual incident or situation which deviates from how things normally are, or how the customer expects it to be (Edvardsson, 1996, p. 196).

Although critical incidents may be of both positive and negative character, this paper focuses on the negative critical incidents as these incidents are necessary for the service guarantee to be invoked. The research design is in agreement with the design proposed by Edvardsson (1988; 1992; 1996), Roos and Strandvik (1996) and Roos (1998; 1999). Practical method and sampling technique Critical incident interviews were conducted with 19 customers from northern and southern America, Europe, Asia and Australia. Written documentation of both complaint and compensation was available in RadissonSAS' database. Several selection criteria determined the choice of interviewees the customer had to: . have stayed at RadissonSAS within the last year; . have invoked the 100 percent guest satisfaction guarantee; . have been refunded or compensated for their inconveniences; and . have a good image of the complaint situation that was similar to that presented in the database. The customers were called either at work or at home and were briefly retold of the negative critical incident when they had stayed at a Swedish or Norwegian RadissonSAS hotel, and what inconvenience caused them to invoke the 100 percent service guarantee. This recollection of the incident was necessary to ensure that the perceptions of the complaint process referred to the specific incident that resulted in a guarantee reimbursement. Most interviewees were executives traveling on behalf of their companies and as the respondents had varying origins, the language used was English, or Swedish when possible. The open questions allowed the interviewees to describe the course of events in their own terms, only led by areas such as; the nature of their relationship to the

IJSIM 14,1

44

hotel prior to the incident; what they felt was the reason for their complaint; how the entire process was handled; and what they considered the result of the inconvenience (Edvardsson and Strandvik, 2000) after the guarantee allowed them a reimbursement. The CIT interviews were then transcribed and interpreted by the researchers. After reading through the 19 transcribed critical incidents it became clear that the customers' relationship with RadissonSAS differed in some respects. The frequency of relationships, i.e. how often the customers stay at RadissonSAS, was one differing variable and another was the character of the relationship, i.e. the customers' attitude toward RadissonSAS, before and after the critical incident. The customer's answers were placed in a table (Table I) that displayed how the relationship had changed, to positive/neutral or negative, after the critical incident. Guided by these variables and results, both authors separately interpreted how each respondent fit the model. The transcribed interviews were summed and two narratives, ``the comedy'' and ``the tragedy'', were constructed. Thereafter, the narratives were interpreted and conclusions in relation to previous research were made. Empirical case In the empirical section of the article we introduce the 100 percent guest satisfaction guarantee at RadissonSAS. Thereafter two ideal-type empirical stories are created about the two clusters of respondents that are identified in Table I, presented as ``the comedy'' and ``the tragedy''. The 100 percent guest satisfaction guarantee at RadissonSAS The 100 percent guest satisfaction guarantee is marketed to customers on ``flyers'' that are placed all over the hotels; in the rooms, in the restaurants, in the lobbies and so on. The flyers tell the customers in bold print that ``Our goal at RadissonSAS is 100 percent guest satisfaction'' and in small print that ``If you aren't satisfied with something, please let us know and we'll make it right or you won't pay''. A brochure from a three-day training program for personnel that was held prior to the introduction of the guarantee tells us what it stipulates, who can utilize it and when to use it. RadissonSAS's primary aim with the guarantee is to improve customer retention, satisfaction and loyalty. In terms of employee performance, the guarantee stipulates that personnel should try to correct mistakes but if the customer is still not satisfied a refund (some kind of economic reimbursement) should be offered. All employees are authorized to utilize the guarantee at any time, but they also have a responsibility to interpret when it is appropriate to apply the guarantee. Some guidelines are presented in the brochure; the guarantee should be used when a customer has a serious problem (if the hotel is responsible for the problem) and if the customer cannot be satisfied through other measures. First, personnel are advised to listen to the customer and apologize for the caused inconvenience. Second, if the customer still appears dissatisfied, personnel should try to find a solution to the problem

Character of relationship and relationship change after critical incident 7 3 3 1 1 6 1 6 6 6 6 6 8 6 6 6 3 1 6 6 1 6 6 1 5 6 6 6 6 7 6 1 1 6 6 1 2 1

Frequent

Frequency of relationship Occasional Seldom

Total 10 3 4 3 8 1 2 5 0 0 0 3 18

Positive

Relationship change Positive Unchanged Negative

Neutral

Relationship change Positive Unchanged Negative

Negative

Relationship change Positive Unchanged Negative

Total

Note: One interview was deleted because the respondent could not remember the incident reported

Effect of service guarantees on service recovery

Table I. Frequency and character of relationship before and after a critical incident at RadissonSAS

45

IJSIM 14,1

46

and if the solution(s) presented does not satisfy the customer, personnel are advised to present the final step of the guarantee a refund. The 100 percent guest satisfaction guarantee is from RadissonSAS's standpoint explicit and unconditional in its character; it aims at satisfying customers, which should be achieved partly through empowered personnel. Making up for unsatisfactory services by a second try or a refund seems to be the essence of the guarantee. Frequency and character of customer relationship before and after a critical incident In order to understand what characterizes the interviewed customers' relationships with RadissonSAS before and after the critical incidents, their responses were placed in a table. Building on recent findings from Johnston and Fern (1999), who argue that customers have different expectations for different levels of failure, and Tax and Brown (1998) who state the importance of the role of prior experience, we chose to present the customers after their previous level of interaction with RadissonSAS. The table reveals that frequent and occasional guests of RadissonSAS appear more content after the critical incident compared to customers staying more seldom at RadissonSAS. There may be several explanations to this approach. Perhaps frequent and occasional guests have developed strong and positive bonds with personnel at RadissonSAS and know to tell employees when problems appear (Liljander and Strandvik, 1995). Another interpretation may be that personnel are apt to help familiar customers. Or, customers who return may do so because previous experiences at RadissonSAS were favorable, and those customers are therefore lenient towards a mishap under fair circumstances (Roos, 1999; Tax and Brown, 1998). To learn more about the reasons underlying our findings, two empirical stories are presented and interpreted. The first story is about the frequent and occasional guests and is named ``the comedy'', as this story presents a happy end, i.e. customers perceive the relationship with RadissonSAS as indifferent or more positive, despite the incident. The second story concerns customers who seldom or never previously have visited RadissonSAS. It is named ``the tragedy'', because it ends with the customer still being dissatisfied. Both stories present the criticality of the incident, behavior of front-line employees, and impact of the service guarantee. The former two areas have proved important in the recovery process (Bitner et al., 1990; Hoffman et al., 1995; Zemke, 1995; Webster and Sundaram, 1998) and the latter area has been proposed to systemize and clarify responsibilities in the recovery process (Tax and Brown, 1998). To interpret the empirical case from these standpoints is thus in line with previous research and our theoretical discussion. The comedy The comedy is an ideal type story on how interaction between customers and RadissonSAS develops in relation to the critical incident. In the group of 11

respondents, four report a more positive relation to RadissonSAS after the Effect of service incident, five mention a relationship indifferent from the incident, and two state guarantees on that the relationship developed negatively as a result of the incident. service recovery The criticality of the incident. The negativity of critical incidents reported varied in proportions. Objectively judged, most of them do not seem very negative, i.e. RadissonSAS forgot to dry-clean a set of clothes, the air 47 conditioning did not work, or the cleaning of a room was not satisfactory. These occurrences did not negatively affect customers' relationship to RadissonSAS. Thus, if the proportions of the critical incident are low it can be argued that a critical incident has a positive or indifferent effect on the relationship. Edvardsson and Strandvik (2000), describing how the criticality of the critical incident directly influences how customers perceive the situation, supports this line of reasoning. However, incidents that may seem superficial at a first glance may have great importance to a specific customer. For example, one incident involved a relaxation lounge that was closed, when a customer's sole reason for visiting the hotel was the availability of such a lounge:
I wasn't at all satisfied with the compensation I got. I went there to relax, not to save SEK2,000.

This quotation displays the difficulty in labeling an incident as of great importance or as inferior, as the perception of a situation varies from respondent to respondent. It further displays the difficulty of turning dissatisfaction into satisfaction without improving the conditions that caused the incident. In the case of the customer complaint above, the availability of the relaxation lounge is not measurable in monetary compensation. The behavior of the front-line employees. The front-line employees' actions in connection to the incident, as well as the criticality of it, have, according to previous research, a large effect on the success of recovery. It seems important that the front-line employee displays concern for the inconvenience that customers have experienced. One customer whose room had not been properly cleaned pointed out that:
I was unhappy about the room, but I was happy with the way it was dealt with afterwards. I wasn't made to feel like it was my fault. I wasn't looking for compensation, I was looking for satisfaction. To let someone know that I was unhappy, and I felt that I got that.

Neither this citation, nor the previous example concerning the relaxation lounge, shows any relation between the scope of compensation and satisfaction. In fact, one customer who complained about a noisy fan that made it impossible to make phone calls in the room, said:
When I complained the following morning, I said I would move because I was not happy. Then I was told that if I would stay another night they wouldnt charge me for the room, and so they put me in another room . . . they were very apologetic and I remember being treated very well.

These statements identify the importance that service personnel have empathy with customers and interpret what kind of reimbursement or measures should

IJSIM 14,1

48

be undertaken, and if an honest apology would please more than a refund. In addition, the compensation should probably be adjusted to suit the customer concerned and the criticality of the incident. The criticality of the critical incident seems important also when it comes to possibilities of turning a negative critical incident into an improved relation. When the incident is of little criticality, it seems quite easy for the front personnel to act in a way that satisfies the customer, resulting in an unaffected or more positive relationship. The difficulty is how to turn a grave failure into a satisfying experience. We have identified that in incidents with high criticality, customers state that the front personnel did not handle the incident appropriately. This seems to result in a negative customer relationship. However, most of the critical incidents seem to create positive or neutral outcomes in terms of customer attitudes to the relationship, something that we argue is the result of the frequent interaction that characterizes the relationship. Such interaction seems to make the relationship less vulnerable to service failures. The impact of the service guarantee. A feature that distinguishes RadissonSAS from other service providers is that it uses a service guarantee. RadissonSAS guarantees 100 percent customer satisfaction i.e. if customers are not satisfied with any aspect of the service, RadissonSAS will correct the mistake or reimburse the customer if the solution was not satisfactory. As was mentioned in the theory section, literature concerning service guarantees states that a primary benefit of a service guarantee is its great impact on marketing. Interestingly, none of the frequent or occasional customers knew of the guarantee until they were complaining and were enlightened about it by frontline employees. Customers seem to have two basic different attitudes towards the service guarantee: satisfied customers (positive or indifferent customers) consider a service guarantee to be a positive strategy for service companies. They believe that the guarantee contributed to the successful service recovery process:
. . . you know that the company co-operates if any problem should arise. And, if personnel know that the customers should be 100 percent satisfied, it puts pressure on them and they must be service minded . . .

On the other hand, those who state that the relationship was affected in a negative way by the incident also have a negative attitude towards the service guarantee. These customers state that they have been inadequately compensated and consider the guarantee to be a rhetorical device. To sum up, although the comedy often is a happy story the reasons are difficult to understand the criticality of the negative critical incident, the front-line employees, the interaction between service provider and customer all affect the customers, in this case satisfactory perception of the incident. In addition, these satisfied customers state that the service guarantee was important for their sense of recovery. All these factors together seem to have

positive effects on the relationship when degree of interaction between parties Effect of service is high. guarantees on The tragedy The tragedy is, like the comedy, an ideal type story about how interaction between the customers and RadissonSAS develops in relation to the critical incident. As is implied by the name, ``the tragedy'' often has negative outcomes, i.e. customers are dissatisfied with the refund or other measures of compensation that RadissonSAS provides them with after a critical incident has occurred. In total seven respondents experience what we name ``the tragedy'', which has two distinguishing features: first, they are dissatisfied with the service provided (one is neutral) and second, they seldom interact with RadissonSAS (Table I). The criticality of the incident. The negativity of critical incidents varied in proportion also in ``the tragedy''. A logical hypothesis would be that the criticality of the critical incident is larger for the customers constituting ``the tragedy'' compared to those constituting ``the comedy''. Yet when examining the critical incidents, this seems only partly to be the case. Some critical incidents reported seem to be greater than those reported in ``the comedy''. For example, a couple who went to RadissonSAS to start out their honeymoon straight after the wedding ceremony did not receive their midnight snack of cheese and wine, nor did they get their breakfast delivered to their room as ordered in advance. However, other critical incidents are comparable to those reported in ``the comedy''. For example, one customer complained about receiving cold food, which seems comparable to a suit that had not been drycleaned properly (which was reported from one customer in ``the comedy''). Nevertheless, the former had a negative outcome but the latter a positive outcome, where the customer's attitude towards RadissonSAS was improved. This fact indicates that the customer's judgment of the criticality of the incident is rather subjective, but also that something other than the incident itself influences whether the outcome is positive or negative. We suggest that the previous relationship history may be one explanation. Most customers in ``the tragedy'' revealed dissatisfaction concerning the compensation, which seems to influence how they perceive the relationship. Respondents of this group mostly received refunds, for example a discount on the room or in the restaurant. For example, the telephone bill was reduced after one customer complained that the pre-ordered breakfast was late, and once it arrived, it was cold:
I didn't really consider the discount as compensation. I could have gotten an apology instead. That would probably have been sufficient.

service recovery

49

Again, the importance of finding a compensation that satisfies the customer and relates to the incident seems a very important but difficult task. The behavior of the front-line employees. Although economic reimbursement seem to dominate how RadissonSAS makes up for failures, the concern and

IJSIM 14,1

empathy of employees seem to be very important also for customers experiencing ``the tragedy''. Several respondents' mention that a primary reason for their dissatisfaction with the service provided by RadissonSAS, i.e. the critical incident itself, was inapt treatment by employees. The CEO of a large company was dissatisfied after interacting with front-line employees:
They probably thought they only lost one customer, when they in fact potentially lost 150 customers, because that's the amount of people I could have put in that hotel in a night. They were dealing with somebody that could have put potentially several hundred thousand pounds worth of business their way, but because of very, very poor service they now have nothing.

50

Other customers felt that they had to argue with the employee before their inconvenience was taken seriously. To conclude, improper handling of complaints by the front-line employees seems to have a large effect on the status of the relationship after a critical incident has occurred. The impact of the service guarantee. None of the customers in ``the tragedy'' knew of the guarantee before complaining. Although this was true also in ``the comedy'', the difference lies in that there the customers were told about the 100 percent guest guarantee at the moment when they complained. Interestingly, the customers who seldom visit RadissonSAS do not recall being told of the service guarantee. This fact could be the result of the inconvenience of arguing with front-line employees before their situation was taken seriously, something that most customers stated as an unsatisfactory experience. Overall, the customers of this group display skepticism or mistrust towards a service guarantee: at least in comparison to the customers in ``the comedy''. For example, one customer thinks that RadissonSAS should concentrate its effort on hiring the right people instead of using service guarantees as guidelines for their behavior. Another customer believes that the guarantee gives the company a delusive impression RadissonSAS guarantees 100 percent satisfaction but cannot keep the promise. Analysis and interpretation The aim of this article is to study the possible effects of service guarantees on service recovery. In the presentation of our empirical findings certain patterns have evolved that may explain whether the outcome of the recovery process is to be considered positive or negative. Furthermore, based on previous research, we have argued that it is plausible that the service guarantee can be a means for service recovery in a relationship context in at least two ways; to serve as a risk-reducer and to influence the behavior of the front-line employees in the complaint situation. The following section presents our reasoning on these two issues. The service guarantee as a risk-reducer In the theory section we argued that explicit service guarantees could function as risk-reducers for both the service provider and the customer. For the service provider, the guarantee would have a mitigating effect on customers switching

from relationships as it would enhance the customer perceived probability of Effect of service successful complaint, which according to Singh (1990) is associated with lower guarantees on risk of switching from relationships. In addition, in promising a good service service recovery the explicit guarantee lowers the customer perceived quality variance, which is synonymous with lowering the customer perceived risk and increasing the likelihood that customers choose a service provider who guarantees its services 51 (Kahneman and Tversky, 1979). Service guarantee research (McDougall et al., 1998; Wirtz et al., 2000) led us to believe that only the explicit guarantee will have these two risk-reducing effects. However, the RadissonSAS case reveals that the customers were unaware of the guarantee; it was implicit, even though RadissonSAS thought it was explicit. So, if the guarantee does not seem to fill any of the theoretically argued benefits, why do we give credence to it affecting the outcome of the recovery process from a risk perspective? Because according to the interviewed customers, especially to those represented in ``the comedy'', the service guarantee seems to be of importance when they complain. These customers explicitly mention their appreciation of the guarantee. How come customers speak positively of and say that they benefit from the guarantee although they were not aware of the guarantee? One customer told us that when he called to complain, the employee told him that there was a service guarantee for failures such as the one reported by him. He was positively surprised by this fact, and that he did not have to pay for the service. And more importantly, he said that if there had not been a guarantee, which contributed to the satisfaction he felt after the complaint, he would probably not have stayed at the hotel again. One plausible interpretation is that when the customer is about to complain, the customer perceives the situation as unpleasant, and perhaps also sees a risk that his or her complaint will not be considered important. In this very moment the employee informs the customer of the guarantee, and the customer is relieved to hear that the company has a pre-planned procedure for failures, that in fact others have complained as well, and the company does consider the complaint important. This interpretation would be supported by the fact that most customers in ``the tragedy'' report that they have not been told of the guarantee, which in part may explain their perception of not being taken seriously. Also, the frustration of having to argue with personnel could have influenced the customers not to pay attention to, or care for, the information about the guarantee. Furthermore, research has shown that customers feel very vulnerable in the complaint situation (Zemke, 1995; Barlow and Moller, 1996) and service recovery literature emphasizes the importance of easing this process for the customers (Tax and Brown, 1998). Thus, the explanation of customer satisfaction with the seemingly unfamiliar guarantee may lie in the fact that the front-line employee presents the guarantee of satisfaction at the time when customers need it the most as they speak up against the company. Hart (1998) argues that customers who are unaware of the implicit guarantee's presence, as well as first-time customers, must experience a failure before learning about the guarantees' existence. We interpret the customers'

IJSIM 14,1

52

appreciation of a guarantee, which they become familiar with only in the actual complaint situation, as a sign that the implicit guarantee may serve more purposes for customers than has been suggested by previous research. It seems that implicit guarantees have a mitigating effect on the customer's interest in switching from a relationship, because it enhance the customer perceived probability of successful complaint (Singh, 1990) when they need it the most, that is, when they actually complain. How the service guarantee affects employee behavior The employees at RadissonSAS are instructed to present an apology to complaining customers, and thereafter try to correct the mistake by replacing, exchanging or reproducing the service that has failed to please the customer, with the help of the guarantee. According to Tax and Brown (1998, p. 80) the apology is important, as ``fair procedures begin with the firm assuming responsibility for the failure''. If the measures fail to satisfy the customer an economic reimbursement, ranging from partial to full reimbursement, would apply. The essence of the guarantee at RadissonSAS is to compensate customers economically if they cannot be pleased otherwise. In the theory section we argued that it seemed plausible that both the explicit and the implicit guarantee will enhance service recovery. The rational for this assumption was that previous research (Maher, 1992; Hart, 1998) argues that guarantees will guide the behavior of front-line personnel when the customers complain. It can thus be assumed that customers perceive that they get a fair treatment which according to Tax and Brown (1998) is important if the outcome of service recovery is going to be satisfactory if the service provider uses service guarantees. In order to analyze if the service guarantee at RadissonSAS had a positive influence on the service recovery actions carried out by the front-line employees we use the distinction between the three kinds of fairness proposed by Tax and Brown (1998) fair processes, outcomes and interaction which promote responsive, adaptable, trustworthy, emphatic and reliable behavior by front-line employees. Interestingly, in several situations where front-line employees act in accordance with the last step of the guarantee (economic reimbursement), their effort does not satisfy the customer, i.e. process fairness was low. In fact, the interviews revealed that several customers did not ask for a refund. Rather they seem to be satisfied when they perceived the behavior of personnel to be empathic and responsive what Tax and Brown (1998) name interaction fairness. These findings imply that the guarantee at RadissonSAS put too much emphasis on compensation and too little on fair behavior. One plausible interpretation is that the customers in ``the tragedy'' were dissatisfied because front-line employees relied on the refund to solve the problem and satisfy the customer. If the service guarantee is inadequately designed or not fully understood by front-line employees, the use of the guarantee lowers customer satisfaction after a critical incident as personnel seem to forget how they should treat a dissatisfied customer. It also implies that a refund cannot be considered

compensation, which only is achieved by added value such as empathy or Effect of service compensation beyond what the customer has already invested. According to guarantees on our interpretation, the service guarantee at RadissonSAS does not promote service recovery process fairness. Another matter is that the compensation itself seems to influence customers' perception of the recovery process. The standpoint is that RadissonSAS offers 53 an unconditional guarantee, which is considered superior to other guarantees (Hart, 1988; Maher, 1992; Ettorre, 1994). However, also this type of guarantee requires employees to be responsive to what best will satisfy each individual customer. Still, how employees go about selecting an appropriate compensation for dissatisfied customers is not guided by the guarantee. However, the interviews suggest that customers react negatively to compensation if it does not stand in relation to the service failure, which indicates that the outcome fairness was low as well. For example, when RadissonSAS had forgotten to repair the air-conditioning in a room, or delayed the breakfast for a customer, these customers were offered a reduction on the phone bill to compensate for their inconvenience. These customers did not appreciate the efforts made by the front-line employees. Nor did customers appreciate being offered a reduction, but not full refund, for service failures. One possible interpretation, which is supported by previous research on service recovery (Brown et al., 1996; Boshoff and Leong, 1998), is that the customers remained dissatisfied as they were only compensated for the problem; no solution was offered to correct the problem that caused the dissatisfaction. Upon the question of whether the personnel were dedicated to his problem, one customer who was compensated with two nights for free, felt that:
Well . . . they did not fix the problem so they weren't very committed!

We also found that in several instances, the interaction fairness was unsatisfactory. When the customer perceived the behavior of the front-line employee as unfair there seems to be a lack of adaptability. Some customers were satisfied when receiving a price reduction, but others would have settled with an apology. In the latter cases the employees especially seem to have lacked empathy and responsiveness. The empirical stories reveal that an outstanding feature to understand how customers perceive the service recovery process is the presence or lack of empathy displayed by the front-personnel. We also interpret that fair interactions relate to how employees respond in the situation when customers complain. Certain customers revealed that they had to argue with the employees before their inconvenience was taken seriously and said that this had considerable negative effect on the service recovery process. The last example indicates that lack of responsiveness, and presumably that presence of responsiveness, have a great effect on the service recovery process. To conclude, the service guarantees seem to affect the behavior of front-line employees. From a customer perspective it is, however, a risk that the

IJSIM 14,1

54

guarantee affects the behavior in an unfair manner. From our interpretation it can be argued that the 100 percent guest guarantee at RadissonSAS makes the front-line employee more apt to give customer economic reimbursement instead of individualized care and attention. A service guarantee might, though, direct attention from empathic, responsive and adaptable behavior, all important for effective service recovery, to instrumental behavior in the form of economic reimbursement. As Maher (1992) and Hart (1998) argue, a service guarantee clarifies for employees what is expected of them, but it seems that the service guarantee may stress the economic compensation rather than the empathetic behavior, and therefore it could complicate, or even prevent an effective service recovery process. Conclusions and future research The focus in the article has been on the role of service guarantees in the service recovery process. More precisely the focus has been on the risk-reducing effects of service guarantees and the impact they have on service recovery. One conclusion is that service guarantees seem to have a risk-reducing attribute, which has not been recognized prior to the research presented here. Previous research has found that explicit guarantees have a risk-reducing effect while implicit guarantees do not reduce risk (Hart, 1998; McDougall et al., 1998; Wirtz et al., 2000). Our research indicates that none of the interviewed customers were aware of the guarantee although the company believes that its presence is known. Still, the guarantee serves as a risk-reduction tool. In the moment of complaining the uncertainty is great for customers, and the guarantee is a signal that the company takes them and their complaint seriously. Therefore, we have argued that the implicit guarantee has a mitigating effect on customers switching from relationships. Another conclusion is that service guarantees can impact the behavior of front-line employees in the service recovery process. The 100 percent service guarantee at RadissonSAS emphasizes that the customer is offered a refund when service failures occur. When employees act in accordance with the guarantee, customers with a ``seldom'' relationship to RadissonSAS perceive such behavior as unfair the guarantee seems to negatively influence the service recovery process, and thereby the relationship. Customers with a more frequent and positive relationship to RadissonSAS are not that vulnerable to this kind of instrumental behavior, meaning that the focus is on economic compensation, as their relationship is characterized by trust. The customers that already have seen proof that the company is devoted to service quality seem less annoyed by the fact that they are compensated economically, than customers that are new to the service and focus on how this one specific situation is treated. Therefore, one suggestion is that it seems important to consider what kind of relationship the company has to a dissatisfied customer when initiating the guarantee and the recovery process. Our research indicates that the service guarantee may actually lower customer satisfaction after an

incident, if the employee relies too heavily on the possibility of compensating Effect of service the customer with economic means. guarantees on Therefore, a service guarantee should stimulate ``fair recovery behavior'' service recovery from the front-line employee. An important part of the service recovery process is not economic reimbursement, but empathy and responsiveness of employees. If service guarantees are designed to stimulate adaptive, empathic and 55 responsive behavior, all-important quality factors (Parasuraman et al., 1985; 1988; 1994; Edvardsson, 1996), it may contribute to improved service recovery processes. Previous research on service guarantees has argued that explicit guarantees encourage customers to complain, express their dissatisfaction, and that guarantees clarify customer expectations. Our analysis did not focus on these topics because the 100 percent guest satisfaction guarantee at RadissonSAS should be considered implicit, as the customers we have interviewed are unaware of it prior to the complaint situation. In future research, however, it would be interesting to study a guarantee that is explicit both to the company and its customers. Answers to the following questions would further widen the knowledge of how service guarantees affect customers: . How do customers perceive the recovery situation when they have expectations on, and knowledge of, the guarantee prior to the incident? . Does the explicit guarantee encourage customers to complain and express their dissatisfaction? . Does the guarantee clarify customer expectations? It would also be interesting to deepen the knowledge of service guarantees' (both implicit and explicit) risk-reducing effects with the help of switching (Forbes et al., 1986; Singh, 1990) and choice literature (Rust et al., 1999).
Note 1. In this article a distinction is made between compensation and refund. A refund refers solely to economic reimbursement of a customer's expenses. With compensation, we refer to additional efforts, beyond a refund or a replacement, such as empathic behavior of front personnel or services that express condolences to the dissatisfied customer. References Alvesson, M. and Skoldberg, K. (2000), Reflexive Methodology, Sage, London. Barlow, J. and Moller, C. (1996), A Complaint is a Gift Using Customer Feedback as a Strategic Tool, Berrett Koehler Publishers, San Francisco, CA. Berry, L. (1995), On Great Services: A Framework for Action, Free Press, New York, NY. Bitner, M.J., Boom, B.H. and Tetrault, M.S. (1990), ``The service encounter: diagnosing favorable and unfavorable incidents'', Journal of Marketing, Vol. 54 No. 1, pp. 71-84. Blomqvist, R., Dahl, J., Haeger, T. and Storbacka, K. (1999), Det kundnara foretaget, Liber AB, Malmo. Bolton, R.N. and Drew, J.H. (1995), ``Factors influencing customer's assessments of service quality and their invocation of a service warranty'', in Swartz, T.A. et al. (Eds), Advances in

IJSIM 14,1

56

Service Marketing and Management: Research and Practice, Vol. 4, JAI Press, Greenwich, CT, pp. 1-23. Boshoff, C. and Allen, J. (2000), ``The influence of selected antecedents on frontline staff's perceptions of service recovery performance'', International Journal of Service Industry Management, Vol. 11 No. 1, pp. 63-90. Boshoff, C. and Leong, J. (1998) ``Empowerment, attribution and apologizing as dimensions of service recovery: an experimental study'', International Journal of Service Industry Management, Vol. 9 No. 1, pp. 24-47. Brown, S.W., Cowles, D.L. and Tuten, T.L. (1996), ``Service recovery: its value and limitations as a retail strategy'', International Journal of Service Industry Management, Vol. 7 No. 5, pp. 32-46. Boulding, W. and Kirmani, A. (1993), ``A consumer-side experimental examination of signaling theory: do consumers perceive warranties as signals of quality?'', Journal of Consumer Research, Vol. 20, June, pp. 111-23. Burrell, G. and Morgan, G. (1979), Sociological Paradisms and Organizationa Analysis: Elements of the Sociology of Corporate Life, Ashgate, Brookfield, VT. Czarniawska, B. (1998), A Narrative Approach to Organization Studies, Qualitative Research Methods Series, No. 43, Sage, London. Czarniawska, B. (1999), Writing Management: Organizational Theory as a Literary Genre, Oxford University Press, Oxford. Donath, N. (1997), ``How Israeli hotel chain ISROTEL developed its service guarantee'', Managing Service Quality, Vol. 7 No. 2, pp. 87-91. Edvardsson, B. (1988), ``Service quality in customer relationships: a study of critical incidents in mechanical engineering companies'', The Service Industries Journal, Vol. 8 No. 4, pp. 48-51. Edvardsson, B. (1992), ``Service breakdowns: a study of critical incidents in an airline'', International Journal of Service Industry Management, Vol. 3 No. 4, pp. 17-29. Edvardsson, B. (1996), Kvalitet och tjansteutveckling, studentlitteratur, Lund. Edvardsson, B. and Strandvik, T. (2000) ``Is a critical incident critical for a customer relationship'', Managing Service Quality, Vol. 10 No. 2, pp. 82-91. Ettorre, B. (1994), ``Phenomenal promises that mean business'', Management Review, Vol. 83 No. 3, pp. 18-23. Forbes, J., Tse, D. and Taylor, S. (1986), ``Toward a model of consumer post-choice response behavior'', Advances in Consumer Research, Vol. 13, pp. 658-61. Gronroos, C. (1992), Service Management: ledning, strategi, marknadsforing i servicekonkurrens, ISL Forlag, Goteborg. Gronroos, C. (2000), Service Management and Marketing: A Customer Relationship Management Approach, 2nd ed., Wiley, Chichester. Hart, C. (1988), ``The power of unconditional service guarantees'', Harvard Business Review, Vol. 66, July/August, pp. 54-62. Hart, C. (1995), ``The power of internal guarantees'', Harvard Business Review, Vol. 73 No. 1, pp. 64-73. Hart, C. (1998), Extraordinary Guarantees Achieving Breakthrough Gains in Quality & Customer Satisfaction, Spire Group, Brookline, MA. Heskett, J., Sasser, E. and Hart, C. (1990), Service Breakthroughs: Changing the Rules of the Game, The Free Press, New York, NY. Heskett, J., Sasser, E. and Schlesinger, L. (1997), The Service Profit Chain: How Leading Companies Link Profit and Growth to Loyalty, Satisfaction, and Value, The Free Press, New York, NY.

Hill, A., Hays, J. and Naveh, E. (2000), ``A model for optimal delivery time guarantees'', Journal of Service Research, Vol. 2 No. 3 , pp. 254-64. Hoffman, D.K., Kelley, S.W. and Rotalsky, H.M. (1995), ``Tracking service failure and employee recovery effort'', Journal of Services Marketing, Vol. 9 No. 2, pp. 49-61. Johnston, R. (1995), ``Service failure and recovery: impact, attributes and process'', Advances in Services Marketing and Management: Research and Practice, Vol. 4, pp. 211-28. Johnston, R. and Fern, A. (1999), ``Service recovery strategies for single and double deviation scenarios'', The Service Industries Journal, Vol. 19 No. 2, pp. 69-82. Kahneman, D. and Tversky, T. (1979), ``Prospect theory: an analysis of decision under risk'', Econometria, Vol. 47 No. 2, pp. 263-92. Levitt, T. (1981), ``Marketing intangible products and product intangibles'', Harvard Business Review, Vol. 59, May/June, pp. 94-102. Liljander, V. and Strandvik, T. (1995), ``The nature of customer relationships in services'', Advances in Services Marketing and Management, Vol. 4, pp. 141-67. Maher, D. (1992), ``Service guarantees'', Manage, Vol. 43 No. 4, pp. 22-4. McDougall, G., Levesque, T. and VanderPlaat, P. (1998), ``Designing the service guarantee: unconditional or specific?'', The Journal of Services Marketing, Vol. 12 No. 4, pp. 278-93. Morgan, R.M. and Hunt, S.D. (1994), ``The commitment-trust theory of relationship marketing'', Journal of Marketing, Vol. 58, July, pp. 20-38. Ostrom, A. and Iacobucci, D. (1998), ``The effect of guarantees on consumers' evaluation of services'', The Journal of Services Marketing, Vol. 12 No. 5, pp. 362-78. Parasuraman, A., Zeithaml, V.A. and Berry, L.L. (1985), ``A conceptual model of service quality and its implications for future research'', Journal of Marketing, Vol. 49 No. 3, pp. 41-50. Parasuraman, A., Zeithaml, V.A. and Berry, L.L. (1988) ``SERVQUAL: a multiple-item scale for measuring consumer perceptions of service quality'', Journal of Retailing, Vol. 64 No. 1, pp. 12-40. Parasuraman, A., Zeithaml, V.A. and Berry, L.L. (1994), ``Reassessment of expectations as a comparison standard in measuring service quality: implications for further research'', Journal of Marketing, Vol. 58 No. 1, pp. 111-24. Roos, I. (1998), Customer Switching Behavior in Retailing, Research report No. 41, Swedish School of Economics and Business Administration, Helsinki. Roos, I. (1999), Switching Paths in Customer Relationships, publication of Swedish School of Economics and Business Administration, Helsinki. Roos, I. and Strandvik, T. (1996), Diagnosing the Termination of Customer Relationships, Working paper No. 335, Swedish School of Economics and Business Administration, Helsinki. Rousseau, D.M., Sitkin, S.B., Burt, R.S. and Camerer, C. (1998), ``Not so different after all: a cross discipline view of trust'', Academy of Mangement Review, Vol. 23 No. 3, pp. 393-404. Rust, R., Inman, J., Jia, J. and Zahorik, A. (1999), ``What you don't know about customer perceived quality: the role of customer expectation distributions'', Marketing Science, Vol. 18 No. 1, pp. 77-92. Sasser, E., Hart, C. and Heskett, J. (1990), The Service Management Course: Cases and Readings, The Free Press, New York, NY. Singh, J. (1990), ``Voice, exit, and negative word-of-mouth behaviors: an investigation across three service categories'', Journal of the Academy of Marketing Science, Vol. 18 No. 1, pp. 1-15. Tax, S. and Brown, S. (1998), ``Recovering and learning from service failure'', Sloan Management Review, Vol. 40 No. 1, pp. 75-88.

Effect of service guarantees on service recovery 57

IJSIM 14,1

58

Tax, S. and Brown, S. (2000), ``Service recovery research insights and practices'' in Swartz, T. and Iacobucci, D. (Eds), Handbook of Services Marketing and Management, Sage Publications, Thousand Oaks, CA. Tucci, L.A. and Talaga, J. (1997), ``Service guarantees and consumers' evaluation of services'', The Journal of Services Marketing, Vol. 11 No. 1, pp. 10-18. Van Mannen, J. (1988), Tales of the Field: On Writing Etnography, The University of Chicago Press, Chicago, IL. Wirtz, J., Kum, D. and Sheang Lee, K. (2000), ``Should a firm with a reputation for outstanding service quality offer a service guarantee?'', Journal of Services Marketing, Vol. 14 No. 6, pp. 502-12. Webster, C. and Sundaram, D.S. (1998), ``Service consumption criticality in failure recovery'', Journal of Business Research, Vol. 41, pp. 153-9. Zeithaml, V., Parasuraman, A. and Berry, L. (1990), Delivering Quality Service Balancing Customer Perceptions and Expectations, The Free Press, New York, NY. Zemke, R. (1995), Service Recovery; Fixing Broken Customers, Productivity Press, Portland, OR.