FACULTY OF EDUCATION AND BUSINESS STUDIES
Department of Business and Economic Studies
A search for the level of Service Recovery required for the Service Recovery Paradox to occur in the Hotel Industry
Where Service Recovery Meets its Paradox
Beatrice Nylander & Andreas Edström
2021
Student thesis, Bachelor degree, 15 HE credits Business Administration
Study Programme in Business Administration Supervisor: Patrik Sörqvist
Examiner: Aihie Osarenkhoe
I Acknowledgements
We would like to thank our supervisor Patrik Sörqvist. We are very grateful for the valuable feedback you have provided along the way. Your creative input definitely made this examination paper more solid than it would have been without it.
We also want to thank all 190 participants who each and every one contributed to this paper.
Thank you for sharing your valuable thoughts and your precious time to make this possible.
II
Sammanfattning
Titel: Where Service Recovery Meets its Paradox - A search for the level of Service Recovery required for the Service Recovery Paradox to occur in the Hotel Industry
Nivå: Examensarbete på Grundnivå (kandidatexamen) i ämnet företagsekonomi Författare: Beatrice Nylander & Andreas Edström
Handledare: Patrik Sörqvist Datum: 2021 – juni
Syfte: Syftet med studien var att, inom hotellindustrin, undersöka vid vilken nivå av service recovery som får service recovery paradox att uppstå.
Metod: Två enkätundersökningar har genomförts där 190 respondenter introducerades för ett scenario-baserat servicemisslyckande i hotellindustrin. Enkät 1 samlade in data om vilken mängd monetär kompensation kunderna ansåg sig behöva efter servicemisslyckandet för att i efterhand känna sig lite mer nöjda än vid serviceupplevelse utan misslyckande (punkten för service recovery paradox). I enkät 2 mättes sedan kundnöjdheten vid tre olika nivåer av kompensation efter servicemisslyckande De tre nivåerna var medelvärde från enkät 1 samt -1 SD och +1 SD. Detta med syftet att, i ett oberoende sampel, testa om punkten för service recovery paradox i enkät 1 fick paradoxen att uppstå i enkät 2.
Resultat & slutsats: Vid ett bestämt servicemisslyckande i hotellindustrin kunde nivån för när service recovery paradox uppstår hittas, genom monetär kompensation. Enkät 1 (N = 40) gav nivån SEK 1204, vilken sedan bekräftades som tillräcklig nivå för att uppnå service recovery paradox i enkät 2 (N = 150).
Examensarbetets bidrag: Servicemisslyckanden är oftast inte möjliga att helt undvika, särskilt inte inom hotellindustrin. Denna studie bidrar med förståelse för vilken nivå av monetär kompensation som kan komma att krävas för att uppnå service recovery paradox, detta vid ett specifikt servicemisslyckande inom hotellindustrin. Resultaten från studien bidrar till vägledning inom service recovery och ger hotel managers kunskap om vilken monetär kompensationsnivå de bör använda sig av för att återskapa kundnöjdheten, samt även undvika överkompensation, vid servicemisslyckanden.
Förslag till fortsatt forskning: Framtida forskning bör utgå ifrån att serviceindustrier och servicemisslyckanden alla är unika. Servicemisslyckanden behöver i forskningen beskrivas i detalj för att resultat ska kunna appliceras av managers. Förankringseffekten skulle även kunna undersökas för att, om möjligt, nå SRP men med en lägre kompensation. SRP-forskare som tar sig an just hotellindustrin kan undersöka SRP-nivån hos ytterligare kundsegment som reser; ensamma, i jobbet, med vänner eller med barn. Även om SRP kan uppstå vid recovery är det inte säkert att kunden har för avsikt att återkomma. En större utmaning för framtida forskning blir därför att mäta hur kundnöjdhet efter servicemisslyckande och recovery står sig på lång sikt. Ytterligare ett förslag är att utföra studien igen, men med riktiga servicemisslyckanden och äkta kompensation.
Nyckelord: service failure, service recovery, service recovery paradox, service recovery hotel industry, customer satisfaction
III
Abstract
Title: Where Service Recovery Meets its Paradox - A search for the level of Service Recovery required for the Service Recovery Paradox to occur in the Hotel Industry
Level: Student thesis, final assignment for Bachelor Degree in Business Administration Author: Beatrice Nylander & Andreas Edström
Supervisor: Patrik Sörqvist Date: 2021 – June
Aim: The purpose of this study was to, within the hotel industry, investigate at what level of service recovery the service recovery paradox will come into existence.
Method: Two surveys have been conducted where 190 respondents were introduced to a scenario- based service failure in the hotel industry. Survey 1 collected data about the amount of monetary compensation that customers felt they needed post service failure, in order to subsequently feel a little more satisfied than in the case of an error-free service experience (the point for service recovery paradox). In survey 2, customer satisfaction was then measured at three different levels of compensation after service failure. The three levels were the mean from survey 1, mean -1 SD, and mean +1 SD. This with the aim to, in an independent sample, test whether the point for service recovery paradox in survey 1 causes the paradox to arise in survey 2 or not.
Result & Conclusions: In the event of a specific service failure in the hotel industry, the level of service recovery paradox, through monetary compensation, was found. Survey 1 (N = 40) determined the compensation level to be SEK 1204. The same compensation level was confirmed as a sufficient level to achieve the service recovery paradox in survey 2 (N = 150).
Contribution of the thesis: Service failures are often not completely avoidable, especially not in the hotel industry. This study helps to understand the level of monetary compensation that may be required to achieve the service recovery paradox, in the event of a specific service failure in the hotel industry. The results from this study contribute to guidance in service recovery and give hotel managers knowledge of what monetary compensation level they should apply to restore customer satisfaction and also avoid overcompensation, in case of service failures.
Suggestions for future research: Future researchers should assume that service industries and service failures are all unique. Failure scenarios need to be described in detail in order for results to be applicable by managers. Future research may also examine the anchoring effect in order to avoid overcompensation if possible. SRP researchers who examine the hotel industry in particular may investigate the SRP level of additional customer segments that travel; alone, at work, with friends or with children. Although SRP might occur because of service recovery, repurchase intentions might decrease. A major challenge for future researchers will therefore be to measure how customer satisfaction after service failure and recovery stands long term. Another suggestion is to conduct the study again, but with real service failures and with real compensation.
Keywords: service failure, service recovery, service recovery paradox, service recovery hotel industry, customer satisfaction
IV Table of Contents
List of Abbreviations and Definitions ... VII
1. Introduction ... 1
1.1. Problem Description ... 2
1.2. Purpose ... 4
1.3. Limitations ... 4
1.4. Disposition ... 4
2. A Review of the Literature ... 6
2.1. Service Failure ... 6
2.1.2. Turn the Dissatisfied Customer Loyal ... 6
2.2. Service Recovery ... 7
2.2.1. Key Stages of Recovery ... 8
2.2.2. Risk of Additional Failure ... 9
2.3. Failure and Recovery in the Hotel Industry ... 9
2.4. Service Recovery Paradox (SRP) ...11
2.4.1. Support for the Service Recovery Paradox ...11
2.4.2. Empirical Evidence for the Absence of the SRP ... 13
2.4.3. Questioning the Paradox ... 14
2.4.4. Elements Affecting the SRP ... 15
2.5. Framework of Hypotheses ... 17
3. Methodology ... 18
3.1. Quantitative Research & Experimental Method ... 18
3.2 Research Design – The Scenario-based Approach ... 18
3.2.1. Unavailable Room as Scenario ... 19
3.2.2. Design of Failure Scenario Explained ... 20
3.3. Data Collection Method ... 20
3.4. Operationalization ... 21
3.5 Quality ... 22
3.5.1. Reliability ... 22
3.5.2. Validity ... 23
3.5.3. Replication ... 23
4. Methods – Survey 1 ... 24
4.1. Survey 1 – Pre-testing ... 24
4.2. Survey 1 – Participants ... 25
5. Results – Survey 1 ... 26
5.1. Results – Survey 1 ... 26
5.2. Hypotheses for Study 2 ... 26
V
6. Methods – Survey 2... 27
6.1. Survey 2 – Variables and Design ... 27
6.2. Survey 2 – Pre-testing ... 28
6.3. Survey 2 – Participants ... 28
6.4. Data Analysis Method ... 29
6.5. Method Critiques ... 29
7. Results – Survey 2... 30
7.1. Pre-recovery ... 30
7.2. Post-recovery and Confirmation of Hypotheses ... 31
7.3. ANOVA ... 31
7.4. One-sample T-tests ... 31
7.5. Independent Samples T-tests ... 32
8. Discussion and Conclusions... 33
8.1. Recovery ... 33
8.2. A Unique Contribution ... 33
9. Conclusions... 36
9.1. Theoretical Contributions ... 36
9.2. Practical Implications... 36
9.3. Limitations and Future Research ... 37
References ... 39
Appendix 1 – Survey for Study 1 ... 45
Appendix 2 – Survey 1 for Study 2 – Group SEK 633 ... 49
Appendix 3 – Survey 2 for Study 2 – Group SEK 1204 ... 54
Appendix 4 – Survey 3 for Study 2 – Group SEK 1774 ... 59
VI List of Figures
Figure 1. The SRP Intersection Point ... 2
Figure 2. A possible process of customer satisfaction ... 27
Figure 3. Results from survey 2 ... 30
List of Tables Table 1. Top 10 most frequent types of Failure ... 10
Table 2. Top 10 most severe types of Failure ... 10
Table 3. Top 10 most frequent Recovery actions ... 10
Table 4. SRP could / is likely to occur when ... 15
Table 5. Elements affecting customer satisfaction ... 16
Table 6. One-sample t-tests results from survey 2 ... 31
VII List of Abbreviations and Definitions
Service Failure: A Service Failure is a service performance that fails to meet the customer’s expectations of the service (e.g. a delayed or canceled flight). When a service failure occurs customers usually expect compensation for the inconvenience.
Service Recovery: A Service Recovery refers to the action or actions (compensations) taken by a firm in response to a service failure (e.g. a voucher or future discount).
Service Recovery Paradox (SRP): The SRP occurs in the scenario where customer satisfaction ends up higher after a service failure followed by a successful service recovery, compared to if there were no service failure (error-free service).
Error-free service: An error-free service is a service performance delivered as planned. The service performance is matching the expectations the customer had when the service was paid for.
SRP-point: The monetary recovery amount needed for the SRP to occur.
1 1. Introduction
In this chapter, a brief background on SRP research is presented in order to shed light on the research gap that the study has investigated. The SRP phenomenon is explained in depth as it is necessary to understand the model for further reading. In addition, the purpose of the study, its limitations and the disposition are also presented.
The growth in service markets over the last decades is undeniable and services’ contribution to the gross domestic product (GDP) of the European Union was 72.9% in 2019 (European Commission, 2020). In the United States, the number reached above 80% (Norvell, Kumar & Dass 2018) and considering the growth in the service sector the attention to the sector’s issues and challenges are steadily increasing worldwide (Wilson, Zeithaml, Bitner & Gremler 2016).
Providing services to customers sometimes entails failing. Even if service firms offer high levels of excellence and have a clear customer orientation, it is impossible to eliminate all service failures (Fayos-Gardó, Moliner-Velázquez, Ruiz-Molina & Šerić, 2017). Firms must understand the importance of service recovery since service failure is a major determinant of customer loyalty and switching behavior (Norvell et al., 2018). Even small increases in customer retention can result in mayor impact on profitability. However, if a service recovery is not satisfying enough, the customer encounters two dissatisfying experiences within one service performance, e.g. a cancelled flight (service failure) followed by a compensation considered too low by the customer (failed service recovery). This double deviation issue will magnify the customer’s overall dissatisfaction (Norvell et al., 2018).
When a service goes wrong, the firm needs to take service recovery actions in order to try to restore customer satisfaction. This is when the aim of the service shifts from error-free service performance to getting the Service Recovery Paradox into existence (see Figure 1). The Service Recovery Paradox (SRP) happens when a customer ends up more satisfied after a service failure followed by a successful service recovery, in comparison to if the initial service would have been error-free (Wilson et al., 2016). Wilson et al. (2016) illustrate the SRP by the scenario where a hotel guest arrives to a hotel with no rooms available in the booked category. The guest gets upgraded to a suite and the customer reports extreme satisfaction over the recovery, a satisfaction higher than if the booked room would have been available.
The SRP is a debated phenomenon within service marketing and according to Wilson et al. (2016), some advocates the idea as an alternative, valid strategy (failing on purpose followed by excellent
2 service recovery) while others claim the disadvantages are too great and the risks with such an approach are too high.
Figure 1
1.1. Problem Description
When a service failure occurs, the need for service recovery actions is urgent and the existence of the SRP should be desirable. It is difficult for managers to apply SRP research results since conclusions are often diverse. According to Norvell et al. (2018) SRP research conclusions show mixed findings on how common the SRP is and several studies have had problems finding evidence for the SRP’s very existence.
When studying dining restaurant customers, Ok, Back and Chanklin (2007) found evidence suggesting that the SRP can indeed become manifest in some conditions. However, its occurrence depended on the recovery to be “exceptional”. Magnini, Ford and Markowski's (2007) contributions showed similar results when students were introduced to different failure scenarios related to the
Figure 1. The figure illustrates customer satisfaction during failure and recovery (red line) and during error-free service (dotted blue line). The SRP occurs where the red line exceeds the dotted blue line. This is where customer satisfaction caused by recovery exceeds the customer satisfaction at error-free service (modified by authors, from Kneip (2018)).
3 restaurant market. However, occurrence of the SRP depended on several contextual influences such as the history of the customer’s relationship and the severity of the service failure (Magnini et al., 2007).
When examining service failure in the airline industry, McCollough, Berry and Yadav (2000) found no evidence of the SRP. Customer satisfaction seemed to be lower after the service recovery efforts were made. Lin, Wang and Chang (2011) studied the online retailing market and found themselves among researchers who have found no support for the SRP.
Where support for the SRP has been found, researchers have claimed that the paradox is likely to occur when effective recovery communication is exercised (Mount, 2012) or when the customer has had no prior failure with the firm, the cause of the failure was viewed as unstable by the customer, and the customer perceived that the company had little control over the cause of the failure (Magnini et al., 2007). SRP is likely to occur only at the very highest levels of customer recovery ratings (Smith, Bolton & Wagner, 1999) and when the customer does not consider the failure as severe (Magnini et al., 2007). Other elements that affect the occurrence of the SRP are; the customers’ potential guilt feelings in the failure process (Cheng, Chang, Chuang & Liao, 2015), whether the examination of customer satisfaction is made over a longer time period or not (Norvell et al., 2018), level of recovery effort (Ok et al., 2007), the degree of consumer’s generosity (Krishna, Dangayach & Sharma, 2014) and consumer sociodemographic characteristics and shopping experience (Fayos-Gardó et al., 2017).
It seems that researchers have mainly focused on determining the existence (or absence) of the SRP at given compensation points, rather than examining when the SRP occurs or at what magnitude of service recovery the SRP comes into existence. The SRP is a rare event (Michel & Meuter, 2008) but it does not have to be. A problem by investigating if the SRP is real or not (at fixed compensation levels) is that results will always depend on numerous elements as the ones mentioned, therefore generalizing conclusions may become challenging.
Another problematic pattern within past research is that presenting participants to failure- and recovery scenarios will only generate conclusions on a case-by-case basis. The SRP will or will not occur and in the cases where no support is found, there is no way to know how close the provided compensation (recovery) is to where the paradox kicks in. Thus, managers cannot know what compensation is required to accomplish high customer satisfaction post failure, without overcompensating. SRP research needs to investigate adjustments of the efforts of recovery to define where the SRP appears (Allen, Brady, Robinson & Voorhees, 2015).
This study investigated the SRP from a new perspective regarding scenario presentation. The magnitude of compensation required for the SRP to occur in a specific hotel service failure scenario
4 was identified in this study. This was done by letting the customers decide the compensation the hotel had to take in the scenario, in order for their post failure satisfaction to exceed error-free service satisfaction (intersection point in Figure 1).
1.2. Purpose
The purpose with Study 1 (survey) was to identify the monetary amount required as compensation, in a specific failure scenario, for a customer to state that they would end up more satisfied than if the service experience would have been error-free. Study 1 aimed to provide a suggested compensation amount (in SEK) for the SRP to occur.
With an independent sample, study 2 (experimental design) aimed to test the level for SRP occurrence, provided by study 1. Another purpose with study 2 was to test the effect of compensation, something that was not done in study 1. In addition, study 2 investigated what the dose-response curve looks like for different compensation levels.
The studies’ joint aim was to identify the recovery needed for the SRP to occur for a specific failure scenario in the hotel industry.
1.3. Limitations
Differences between customers’ various ethnical and cultural backgrounds may affect results in service recovery evaluation (Ok et al., 2007) and Eastern cultures typically value interdependence, relationships and connectedness in comparison to Western customers (Van Vaerenbergh, Hazée &
Costers, 2018). This study only focused on customers in Western culture, specifically Swedish hotel guests.
Researchers see a need for testing the SRP in additional service settings (Magnini et al., 2007) since service industries and sectors clearly differ in e.g., customer experience management, customers as strategic assets, and the accompanying complaint management that should be undertaken (Morgeson III, Hult, Mithas, Keiningham & Fornell, 2020). This study investigated only one service industry, the hotel industry. Service providers need to look at the key types of service failures, e.g.
process failures vs. outcome failures (Nguyen & McColl-Kennedy, 2003). This study investigated an outcome failure scenario.
1.4. Disposition
It is important to understand the disposition of the study in order to follow the content chronologically.
By presenting two studies in one, the disposition of this study differs from the traditional chronology in bachelor thesis works consisting of only one study.
5 The chapters in this study are organized as followed. First, chapter 2 presents a literature review of service failure, service recovery, failure and recovery in the Hotel industry and the service recovery paradox (SRP). Chapter 2 ends with a framework of three hypotheses for study 2. In chapter 3 the methodology is described. Because this work contains two studies and study two was dependent on results from study 1, Chapter 4 presents only the methods in study 1 and is followed by a presentation of the results from study 1 in chapter 5. The methods in study 2 are then presented in chapter 6, followed by the results from study 2 in chapter 7. Chapter 8 finalizes the study by discussing the results relative to the literature, the unique contribution from this work, conclusions, and limitations along with suggestions for future research.
6 2. A Review of the Literature
Chapter 2 presents current research on service failure and service recovery. Failure and recovery in the Hotel Industry are introduced followed by mixed findings in SRP research.
2.1. Service Failure
Service failures are service-related mishaps or problems that occur during a consumer’s experience with a service provider (Michel & Meuter, 2008) and include (but are not limited to) unavailability of the service (no service personnel with the appropriate skills are available), exceptionally slow service, mistakes in the service (e.g., bank statement errors) (Spreng, Harrell & Mackoy, 1995).
Service failures in the service provider industry are impossible to avoid and cannot be neglected.
They are inevitable due to their nature (Ok et al., 2007) and are present even in the best run organizations, which means that service failures occur both at 5-star hotels and at the fast-food chain next door (Mount, 2012). A service failure happens when the service firm fails to provide the service for a customer. A service failure could lead to dissatisfaction, decline of customer confidence, negative word-of-mouth, loss of revenue and increase in cost and also affect the employee morale and employee performance (Lewis & McCann, 2004).
It is important to evaluate and try to recover every failure that might occur since service failures are critical incidents which definitely elicit a negative response from consumers (Hazarika, Dhaliwal,
& Rachna, 2019). If left unchecked, failures may lead to situations where dissatisfied customers may switch to another provider. Apart from a loss in revenue, loss in goodwill may occur as disgruntled customers tend to spread unhealthy, damaging word of mouth.
2.1.2. Turn the Dissatisfied Customer Loyal
Customer satisfaction is conditional upon a customer’s entire history with the firm. A failure in the early stage of the customer’s relationship with the firm will weigh more heavily on customer dissatisfaction because the customer has fewer successful service experiences to counterbalance the failure (Magnini et al., 2007). Therefore, it is crucial to gain loyalty from the customer through the first experience or handle the service failure in the best way possible. If the service failure were to happen a second time the customer is more likely to attribute the cause of that problem to the firm than when the customer experienced the first failure (Magnini et al., 2007).
7 2.2. Service Recovery
“A good recovery can turn angry, frustrated customers into loyal ones.
It can, in fact, create more goodwill than if things had gone smoothly in the first place.”
– Hart, Heskett and Sasser (1990, p. 148)
Service failures are inevitable in all service contexts (Wilson et al., 2016), which means that firms must strategize for the SRP to occur when these inevitable service failures take place. Managers need concrete guidance in creating these strategies when obvious SRP determinants as severity of failure (Krishna et al., 2014) and consumer sociodemographic characteristics (Fayos-Gardó et al., 2017) provides no solid approachable guidance.
When the service provider fails to meet a customer’s expectations and a failure occurs, this leads to a crucial breaking point between the customer and the provider. That is where the service recovery kicks in, defined as the action of a service provider in response to service failure (Krishna, Dangayach
& Jain, 2011). It is a systematic business process that must be designed properly and implemented in an organization to recognize customers with issues in order to address those issues to increase customer satisfaction and endorse customer retention.
If the service provider is able to recover a failure, this could, instead of a lasting failure, become a strong advantage in the loyalty process between the customer and the service provider. The service failure gives opportunity to the provider to make a strong relation with the customer by interacting (listening), understanding and taking the appropriate action (Krishna et al., 2014). Sometimes, however, there is no need for action after a service failure. Sometimes, the only thing the customers seek is an apology by the service provider (Nguyen & McColl-Kennedy, 2003). By apologizing directly to the customer and offer other promotions as a consolation for the service failure, the failure is no longer a threat, instead it could become an opportunity to promote (Gohary, Hamzelu &
Pourazizi, 2016).
A good service recovery enhances customers' perceptions of the firm's competence, the services already purchased, and the value of the organization's other offerings (Swanson & Kelley, 2001).
Thus, converting dissatisfied consumers to satisfied (i.e., recovered) customers is an important objective for most companies.
The service recovery is a multifaceted process which involves apologizing, explaining, making offers of compensation, and being courteous in the process. After the failure, accepting responsibility and explanation in combination with some form of compensation, such as cash back, a coupon, or a discount, is viewed by the customer in a favorable manner (Krishna et al., 2011). Customers who
8 have the opportunity to voice their views, with regards to the service failure, are more likely to express higher levels of satisfaction and successful service recovery consists of apologies, general acknowledgment, explanation, and some form of compensation (Krishna et al., 2011).
Even though, service recovery is recognized as an opportunity. It is an opportunity to turn a negative encounter into a positive encounter (Ok et al., 2007), which means that the service provider should do enough research in the service recovery to be able to use different techniques to apologize to their customer. Research has evaluated that external service recovery techniques can be defined as a blend of a) response, b) information, c) action, and d) compensation (Krishna et al., 2011). By using these techniques successfully, the service provider may be able to achieve competitive advantages, not only against other service providers but also by establishing positive word-of-mouth between existing and future customers.
2.2.1. Key Stages of Recovery
Bell and Zemke (1987) propose five dimensions for recovery i.e., apology, empathy, urgent reinstatement, symbolic atonement, and follow-up, while Bitner, Booms and Tetreault (1990) offer the process of recovery in four steps: problem acknowledgement, explanation of the reason, apology where appropriate, and compensation such as a free ticket, discount coupons etc. Correction and exceptional treatment are two additional steps in in a successful service recovery (Lewis & McCann, 2004).
Since service failures are inevitable and likely to occur more than once, Johnston and Michel (2008) argue for the importance of process improvement and that it involves four key stages. (1) Data collection: the collection of information about both complaints and operational recovery procedures so that data about problem areas are made available to managers. (2) Data analysis: the analysis of the data to identify problems and establish priorities. (3) Costing: assessing the cost and other implications of rectifying or improving the processes. (4) Improvement: improving, where appropriate, operational and organizational processes and assessing the impact after implementation.
By collecting data from complaints, the firm can help their employees facing service failures and handle them correctly.
Personalizing the apology based on the individual differences between customers are necessary (Cheng et al., 2015). When a waiter uses impolite words to respond to a customer’s demand, he or she might feel unhappy or even angry. In this failure scenario the customer possibly wants an apology (psychological recovery) from the waiter, rather than receiving a coupon. Similar to psychological loss, if the customer encounters an economic loss (e.g., an overcooked beefsteak), instead of an
9 apology, he or she might want a new beefsteak more than an apology. Thus, while one customer needs compensation, another customer may need an apology (Cheng et al., 2015).
A service failure is more likely to reach a successful resolution if the problem is solved promptly (Miller, Craighead & Karwan, 2000). The ideal is to identify and to solve the problem before the customer becomes aware of it. This also shows the importance of having a database of previous failures or following a “handbook” to proceed to solve the failure accordingly.
2.2.2. Risk of Additional Failure
If the service provider fails to provide the correct recovery, inappropriate service recovery will lead to a negative evaluation of the transaction and result in magnification of negative evaluation (Ok et al., 2007). Double deviation happens when customers experience two negative disconfirmations:
failure and unsuccessful recovery. This double deviation could magnify the customer’s overall dissatisfaction (Norvell et al. 2018).
Restoring customers’ negative evaluation of service performance, caused by not being able to meet customers’ expectations, is not an option but a necessity in building continual relationships with existing customers (Ok et al., 2007).
2.3. Failure and Recovery in the Hotel Industry
The hotel industry involves a high degree of interaction between employees and consumers and provides many opportunities for service failures to occur (Lewis & McCann, 2004). The hotel industry faces different challenges compared to other service providers. Hotels are characterized by continuous (24/7) operation and highly fluctuating demand, relative to constant rates of supply, which make service failure more likely to happen than in other industries. Thus, the hotel industry is more exposed to service failures (Lewis & McCann, 2004).
Customers in the hotel industry might experience service failure when the hotel does not meet expected standards such as clean and functional rooms, quality food and beverages, helpful and friendly employees, high level of security and efficient check in and check out service. The study by Lewis and McCann (2004) identified the most common hotel service failures (see table 1) followed by the failures considered most severe by hotel guests (see table 2). Finally, the study identified the most common service recovery actions taken by the hotels in the study (see table 3).
10 Hotel management should develop specific monetary compensation guidelines while training both full- and part-time employees to quickly and properly react to various service failure situations (Kim, Kim & Kim, 2009). A comprehensive training program can facilitate in selecting the appropriate level of compensation (e.g. room discount versus upgrading a room). Hotel managers need to train front
Table 2
The table lists how guests ranked the severity of the service failures (Lewis & McCann, 2014).
Note: Mean ratings; 1 = ”Very serious” to 5 = ”Not at all serious”.
. Top 10 most severe types of Failure 1 Room not clean (1,45)
2 Reservation missing (1,62)
3 Staff unhelpful and unfriendly (1,75) 4 Room locks appeared flimsy (1,82) 5 Staff inefficient (1,82)
6 Food and beverage not of high quality (1,84) 7 Staff would not put themselves out to help (1,85) 8 Restaurant staff unhelpful and unfriendly (1,89) 9 Bill incorrect (1,91)
10 Receptionist unfriendly and unhelpful (2,04)
Table 1
The table lists the most common service failures found by Lewis and McCann (2004) when investigating 149 service failures in the hotel industry. Percentage show how many of the 149 hotel guests experienced each type of failure.
Top 10 most frequent types of Failure 1 Slow restaurant service (61,7%)
2 Slow check-in/out (53,7%) 3 Staff inefficient (47,7%)
4 Receptionist unfriendly and unhelpful (47%) 5 Food and beverage not of high quality (47%) 6 Room not ready (46,3%)
7 Kept waiting for a table at breakfast (45%)
8 Items in room not working, e.g. TV, phone (43,6%) 9 Variety of food limited (43,6%)
10 Staff would not put themselves out to help (43,6%)
Table 3
The table lists the most common recovery actions taken by the hotels (Lewis & McCann, 2014). Parentheses shows number of times out of the 149 failures.
Top 10 most frequent Recovery actions 1 Apologized (68)
2 Corrected the problem (40) 3 Explanation provided (24) 4 Immediate action (23) 5 Did nothing (21)
6 Hotel took responsibility for the problem (20)
7 Followed up to see if I was satisfied with response to problem (10) 8 Redirected the complaint (9)
9 Compensation provided (9) 10 Exceptional treatment (4)
11 desk staff to offer proper arrays of compensations with similar values and let their guests select the best compensation (Kim et al., 2009).
2.4. Service Recovery Paradox (SRP)
The SRP-phenomenon was discussed as early as 1981 by Etzel and Silverman. The argument presented was that a customer who experiences an effective complaint-handling might end up becoming the best customer of the firm (Gohary et al., 2016). The term Service Recovery Paradox was, in 1992, invented by McCollough and Bharadwaj who implied that after facing a service failure and high service recovery, customer satisfaction might end up even higher than it was prior to the failure.
Wilson et al. (2016) describe the phenomenon as when a service failure occurs, customer satisfaction falls, and the firm takes service recovery actions. If the customer goes on to experience a high level of excellent service recovery, the customer satisfaction and repurchase intention might end up even higher than if there were no service failure to begin with. A third way of describing the paradox (Oh, 2003) is: {S | SR | F} > {S | NF}; that is, customer satisfaction (S), given a successful recovery (SR) and service failure (F), should be greater (>) than customer satisfaction, given no failure (NF) and thus no recovery attempts.
The SRP has been discussed as an alternative strategy to increase customer satisfaction (Wilson et al., 2016), to purposely provide service failure in order to take excellent service recovery actions and thereby increase customer satisfaction. This concept got the attention of both service marketing managers and researchers (Gohary et al., 2016) and for years researchers have been trying to standardize this phenomenon and identify the reasons why and how successful service recovery may cancel the impact of service failure (Krishna et al., 2014).
Research on the Service Recovery Paradox (SRP) is both scarce and contradictory (Fayos-Gardó, Moliner-Velázquez & Ruiz-Molina, 2015). While several researchers have failed to find support for the SRP (Berry & Yadav, 2000; Boshoff, 1997; Fayos-Gardó et al., 2017; Krishna et al., 2014; Lin et al., 2011; Maxham, 2001; McCollough, 2008; Michel & Meuter, 2008; and Norvell et al., 2018), others have found empirical evidence for the phenomenon (Fayos-Gardó et al., 2015; Garg, 2013;
Gohary et al., 2016; Magnini et al., 2007; Ok et al., 2007; and Smith et al., 1999).
2.4.1. Support for the Service Recovery Paradox
Dining restaurant customers in the American Midwest were asked about their initial overall satisfaction with the restaurant they named by choice (Ok et al., 2007). The customers were faced by
12 a failure scenario followed by one out of eight recovery scenarios. After the service recovery customers were asked about their post recovery overall satisfaction in order to test the SRP.
Conclusions were that if the customers are highly satisfied with the recovery efforts, their post- recovery overall satisfaction may end up higher than their initial overall satisfaction. The recovery must, however, be exceptional rather than good for the SRP to occur.
A similar method was conducted by Magnini et al. (2007) asking undergraduate students about their response to different service failure recovery situations. Students were instructed to read the scenario thoroughly and assume that the scene had just happened. Satisfaction, purchase intent, propensity to spread positive word of mouth were then rated by the students. Magnini et al. (2007) found the SRP to be a valid theory. However, the paradoxical post failure satisfaction increase seems to be dependent on a number of contextual influences. For the SRP to possibly occur the following four elements are required (Magnini et al., 2007); the failure should not be considered severe by the customer, the customer has had no prior failure with the company, the reason for failure was viewed as unstable by the customer and the customer perceived that the company had little control over the cause of the failure.
Customer satisfaction must increase for the SRP to exist and customer satisfaction, pre- and post- service failure recovery, is affected by the customer’s shopping experience and to consumer socio- demographic characteristics (Fayos-Gardó et al., 2015).
In the travel industry the SRP occurs when recovery incorporates cooperation and the value created goes beyond goods and responsibility (Gohary et al., 2016).
Customers in the banking industry led Garg (2013) to evidence for the SRP. Satisfaction and recommendation intentions were compared between two categories of customers; those who experienced error-free service and those who experienced failure and recovery. Selection resulted in 926 customers who had experienced one service failure with the bank within the last year. Service failure should be seen as an opportunity to appease customers by giving the customer a much better recovery experience than is expected. However, the risk by using the SRP as a strategy to increase customer satisfaction is high and failing to provide a recovery successful enough may result in the customer leaving the firm (Garg, 2013).
The SRP is likely to occur when excellent recoveries are compared with mediocre, error-free service transactions (Michel & Meuter, 2008). Feelings of guilt influence the SRP since the SRP- effect is much greater when customers with guilt feelings encounter service failure and recovery and customers without guilt may have little or no influence on the SRP (Cheng et al., 2015). The severity of the failure influences the fulfilment of the SRP in the sense that the less serious the problem is, the more likely the SRP is to occur. In contrast, it is more likely that a serious incident has a negative
13 impact on customer satisfaction after service recovery (Fayos-Gardó et al., 2017). The magnitude of failure is a critical factor for the SRP occurrence, along with the degree of consumers’ generosity (Krishna et al., 2014). Smith et al. (1999) found support for the SRP in the hotel industry. They too argue for the importance of failures causing low harm, and that support was only found at the very highest levels of customer recovery ratings.
2.4.2. Empirical Evidence for the Absence of the SRP
Numerous studies on several different service markets have examined the existence and reasons for the SRP to occur, resulting in diverse findings and several found no evidence for the phenomenon.
The examination of 8800 dining restaurant customers’ short-term attitudinal and long-term behavioral responses to service failures and recovery efforts (Norvell et al., 2018) generated somewhat contradictory results. The frequency of visitation of the customers who were successfully recovered was similar to those who never experienced failure, in the medium term. In the long run the frequency of visitation changed substantially, the customers who experienced error-free service returned to the brand significantly more than all other groups in the study. Conclusions suggest that even though successful recovery temporarily compensates for service failures, positive effects fade to negative long term. Thus, even if the SRP would occur in the short run, it is unlikely to hold over time (Norvell et al., 2018).
In a twin study investigating service failure in the airline industry (McCollough et al., 2000), SRP- coiner McCollough found no evidence for the existence of the paradox. Customer satisfaction was lower after service failure followed by recovery in comparison to error-free service, even when recovery performance was high. McCollough et al. (2000) argue that companies are better in the eyes of consumers when delivering error-free service instead of offering superior recovery. In another airline setting (Boshoff, 1997) consisting of 27 failure scenarios, where the supervisor immediately offered a refund of expenses and an additional free airline ticket, significantly higher post-recovery satisfaction than disconfirmation was observed in only one of the scenarios.
Maxham (2001) utilized both a field study (with recent complainants of an internet service) and experimental design (with students and a haircut experience) when examining the recovery paradox.
Results found no support for the SRP in either study. A similar conclusion was made in a study where nine service scenarios (eight failure-recovery and one error-free) were presented to online retail customers the error-free scenario resulted in both the highest customer satisfaction and repurchase intention (Lin et al., 2011).
When investigating the SRP within retail Fayos-Gardó et al. (2017) found no support for the SRP.
The high level of customer satisfaction generated by service recovery was not sufficient to get the
14 SRP into existence. Levels of satisfaction after the service recovery were significantly lower than prior the service failure (Fayos-Gardó et al., 2017).
Michel and Meuter (2008) aimed to find if the SRP is true or overrated by collecting data from 11929 customers in the banking industry. Only 0.53% (63) of the customers experienced a service failure and evaluated the recovery effort as much better than expected. Even though the data indicated that the SRP is true (and very rare), the findings of the study show that error-free, “very satisfying”
initial service performances are the best way to accomplish customer satisfaction (Michel & Meuter, 2008).
150 employees within the health care, banking and personal care service sector were targeted in order to find and validate the key factors required for the SRP to occur (Krishna et al., 2014). After questionnaire analysis Krishna et al. (2014) could not conclude the existence of the SRP. However, six exigencies that influence relative importance within the SRP were found; creating understanding, clearance of failure memory, reciprocating, building risk-taking confidence, fair communications that build relations and depending on human nature to forgive and forget. Krishna et al. (2014) advise that loss in health is to be recovered with health, respect with respect, money with money and so on, while money loss, comfort loss, time loss and loss of control are easier to compensate. It is difficult for firms to compensate for failures causing loss in health, respect, character and emotion. Offering monetary compensation for these losses may even worsen the situation.
The nature of service seems to play a role in the absence of the SRP since there are no evidence for the SRP when excellent recoveries are compared with excellent, error-free service transactions (Michel & Meuter, 2008).
2.4.3. Questioning the Paradox
By questioning the empirical evidence of the SRP, an attempt to reach an overall view was made in de Matos, Henrique, Alberto and Rossi’s (2007) meta-analysis, arguing empirical research testing the SRP has produced mixed results with only some studies supporting this paradox. Even though results showed that satisfaction increases after a high service recovery there were no evidence for repurchase intentions. De Matos et al. (2007) suggest that firms should not purposely provide failure followed by recovery as a strategy to reach higher customer satisfaction. The SRP is a real customer- satisfaction-phenomenon because customers are open to leave positive reviews after experiencing high recovery efforts. However, customer loyalty seems to be damaged after a service failure regardless of the recovery effort.
Since the effort for recovery has to be superior while the failure of service must be very modest for the SRP to arise, the phenomenon might be limited (McCollough, 2008). Analysis of 36000
15 customers’ complaint- and repurchase behavior (Soares, Zhang, Proença & Kandampully, 2017) showed support for the SRP, but only as a generational phenomenon. Customers born in the mid- eighties (Generation Y) seem to complain more about service failures than other groups, but at the same time Generation Y tend to repurchase after a successful service recovery. Soares et al. (2017) concluded Generation Y as the only group of customers who showed evidence for the SRP.
2.4.4. A Collected View of Elements Affecting the SRP
The literature review clearly shows that findings and conclusions from investigations concerning the SRP are diverse. Even though support for the SRP is found in several studies, the reasons or elements required for the paradox to occur are said to be many. Table 4 illustrates elements found to be directly connected to SRP occurrence while table 5 illustrates findings affecting customer satisfaction, which in turn may affect SRP occurrence.
Table 4
The table lists elements having a positive impact on the occurrence of the paradox.
Researchers SRP could / is likely to occur when
Gohary et al. (2016) Recovery incorporates cooperation and the value created goes beyond goods and responsibility.
Magnini et al. (2007) Failure is not considered by the customer to be severe, the customer has had no prior failure with the firm, the cause of the failure was viewed as unstable by the customer, and the customer perceived that the company had little control over the cause of the failure.
Michel & Meuter (2008) Excellent recoveries are compared with mediocre, error-free service transactions (but not when excellent recoveries are compared with excellent, error-free service transactions).
Ok et al. (2007) Customers’ recovery ratings are at the very highest.
Customers are highly satisfied with service recovery.
Recovery efforts are exceptional rather than just good.
Smith et al. (1999) Only at the very highest levels of customer recovery ratings.
16 Table 5
The table lists elements affecting customer satisfaction.
Researchers Elements affecting customer satisfaction Balaji & Sarkar (2013)
Betts, Wood & Tadisina (2011) Krishna et al. (2014)
McQuilken (2010) Smith et al. (1999)
Perceived severity of failure.
Bradley & Sparks (2012) Explanation type, explanation quality, failure magnitude and compensation.
Cambra-Fierro, Berbel-Pineda, Ruiz-Benítez &
Vázquez- Carrasco (2013) Customer gender.
Cheng et al. (2015) Feelings of guilt.
Choi & Choi (2012) Customer affection and perception of justice.
Fayos-Gardó et al., (2017) Consumer sociodemographic characteristics and shopping experience.
Krishna et al. (2014) Six exigencies that influence relative importance within the SRP;
creating understanding, clearance of failure memory, reciprocating, building risk-taking confidence, fair communications that build relations and depending on human nature to forgive and forget.
The degree of SRP depends on magnitude of service failure and the degree of consumer’s generosity.
Lii, Chien, Pant & Lee (2013) Type of service. Psychological distance to service and recovery (online or offline).
Mattila (2010) Gender and letting the customer choose his/her preferred recovery option.
Maxham III & Netemeyer (2002) Procedural and interactional justice.
Mount (2012) Effective recovery communications, the consumer is led to believe the service provider is fair (e.g. admits its mistakes, makes restitution, etc.).
Nikbin, Ismail & Marimuthu (2013) Failure attributions such as stability and controllability.
Frequency of failures. If the firm is responsible for the failure.
Customer loyalty.
Norvell et al. (2018) Time. Even if the SRP would occur in the short run, it is unlikely to hold over time.
Roschk, Müller & Gelbrich (2013) Customer age.
Seawright, Detienne, Bernhisel & Hoopes Larson
(2008) Degree of failure. Both psychological and tangible factors are important contributors to service recovery satisfaction.
Van Vaerenbergh, Vermeir & Larivière (2013) Customers’ opportunity to observe the service recovery.
17 2.5. Framework of Hypotheses
Since the aim with study 1 was to find the SRP-point according to the respondents, no hypotheses were formed for that study.
For study 2, three hypotheses were proposed. However, results from study 1 are required to fully formulate these hypotheses. Therefore, only the hypotheses framework is presented in this chapter.
H1: The mean monetary compensation found in study 1 will be sufficient to reach SRP.
H2: The mean monetary compensation found in study 1, - 1 SD, will not be sufficient to reach SRP.
H3: The mean monetary compensation found in study 1, + 1 SD, will be sufficient to reach SRP.
The complete description of the three hypotheses in study 2 are presented after the presentation of the results from study 1, in chapter 5.
18 3. Methodology
Chapter 3 gives an overview of the methodology, presenting the overall design and the quality of the work. Methods used for each study will be presented in chapters 4 and 6.
3.1. Quantitative Research & Experimental Method
The SRP has been studied and investigated thoroughly during the last couple of decades. Several SRP researchers such as Hazarika et al. (2019), Johnston and Michel (2008), Magini et al. (2007), and Ok et al. (2007) state that the paradox is in need for further research, and the majority of SRP studies have conducted quantitative research.
According to Bryman and Bell (2011), using a quantitative research method will hopefully help in the process of generalizing study results. Also, by using a quantitative research method it is easier to gather a large sample size through surveys. The method is very structured and therefore able to find a pattern in a theory or concept (Bryman & Bell, 2011), in this case the service recovery paradox (SRP). A quantitative research method is therefore regarded beneficial for a study like this.
Experimental studies often possess the ability to create strong confidence in the credibility of causal conclusions. These studies tend to show a high internal validity (Bryman & Bell, 2013). Since the experimental method is the only one enabling investigation of cause-and-effect relationships, the method used in study 2 was experimental. Manipulation of monetary compensation (independent variable) was made in order to establish cause-and-effect relationship to customer satisfaction (dependent variable).
3.2. Research Design – The Scenario-based Approach
SRP-researchers often apply a role-playing method using service failure-scenarios followed by proposed service recovery actions to find the SRP or measure customer satisfaction. This has been done through surveys or interviews. Interviews have also been conducted to investigate real customer experiences, e.g. customers have been asked to remember a service failure incident they have experienced and then asked about their satisfaction prior the experience compared to post-recovery.
The method of analyzing memories from real experiences has been questioned since memory bias might affect study results. Customers tend to remember extremes that are not representative of the experiences of the general population (Ok et al., 2007) and rationalization is likely to be present in surveys that rely on recall (McCollough et al., 2000).
There are several advantages in the SRP literature connected to applying a scenario-based method.
By using the role-playing scenarios as method, the internal validity of the study enhances due to the increase in control over manipulated variables and decrease in external, unpredicted elements (Ok et
19 al., 2007). The method is time-effective by enabling summarizing events that might otherwise unfold over days or weeks and it allows resource-intensive manipulations to be more easily operationalized (McCollough et al., 2000). There is also an ethical advantage to the method since parties involved in actual service failures are left out of the investigation (McCollough et al., 2000). To inflict fabricated service failures upon customers is undesirable to both managers and customers, this is avoided by a scenario approach (Cheng et al., 2015).
This research used a scenario-based approach to collect empirical material. The research design used a similar method as several previous studies have done within service recovery research (Lewis
& McCann, 2004; Magini et al., 2007; and Ok et al., 2007).
The advantage of vignette questions compared to attitude questions is that the decision is anchored in a concrete situation, thus reduces the risk of an unreflect answer (Bryman & Bell, 2011). Therefore, this research was conducted by presenting a service failure scenario by using vignette questionnaire measuring customer satisfaction as the dependent variable. The measuring scale ranged from -10 to +10 where the value 0 was set as a neutral rating (error-free service).
It is important that respondents are able to understand instructions easily in order to complete the questionnaire (Adams & Cox, 2008). As the sample comprised individuals with Swedish as native tongue, this survey was conducted in Swedish instead of English. It is important with a clear and easy structure of the questionnaire and respondents’ short attention span means that long-lasting questionnaires may generate less accurate results since people may rush to finish them (Adams &
Cox, 2008). Finishing either survey in this study took no longer than 5 min per respondent.
3.2.1. Unavailable Room as Scenario
26 types of service failures in the hotel industry were identified by Lewis and McCann (2004).
Customers were asked to rate the seriousness of the failures from 1 (very serious) to 5 (not at all serious). The failure “Room not clean” was considered most severe with a rating of 1,45. It is difficult to measure and grade the cleaning of a room as a recovery action, which also means that it is difficult to determine how the cleaning must be performed for the SRP to occur. Therefore, the most severe failure was not investigated in this study. The overall mean for” Reservation missing” was ranked second most serious with a rating of 1,62. A missing reservation is not opinionable, either the reservation is missing or not. This assisted in determining the scenario as a stated failure and
“Reservation missing” was therefore the failure investigated in this study.
The Ritz-Carlton Hotel Company has created an internal service quality indicator (SQI) were the 12 most serious failures are rated. Among the most frequent failures Ritz-Carlton rates” Missing guest preferences” and” Guest-room changes” among the top serious failure scenarios (Wilson et al., 2016).
20 With the aim to provide useful and relevant data for hotel managers this study investigated a specific service failure scenario considered as one of the most severe in the hotel industry.
3.2.2. Design of Failure Scenario Explained
Below is the failure scenario presented as in the surveys (translated from Swedish), followed by an argument for each part:
“It is time for you and your partner (1) to go away on a long-awaited mini-holiday consisting of a one (2) overnight stay. You have booked a double room and paid SEK 1500 (3) for the reservation. When you arrive to the hotel lobby to check in, the booked double room is occupied. There are no other available double rooms and no larger rooms either (4). You are therefore offered a smaller single room and the staff arranges for an extra separate single bed to be carried into the room.”
1: Due to the nature of the failure scenario, a single traveling person would not be as affected by staying in a single room in comparison to a couple. Failure is believed to be more severe when traveling with a partner than traveling alone.
2: Both a two night-stay and a seven night-stay was discussed as failure scenario. The one night-stay was chosen because the authors assume it to be the most frequent type of booking. Hence, results from a one night-stay-scenario seem to be more relevant for hotel managers.
3: The authors investigated the average price for a double room-stay in the larger cities of Sweden which resulted in the scenario booking having a price of SEK 1500.
4: When a type of room (e.g. double room) is booked but not available, the customer might be upgraded (Kim et al., 2009) to a more expensive or appealing room. This study measured gradable recovery compensation which would not have been possible with a single action recovery event, such as an upgrade of the room. Thus, no rooms more expensive or appealing than the double room was available in this failure scenario.
3.3. Data Collection Method
The primary data collection was made through two surveys using Google Forms. Survey 1 collected information regarding the respondents’ attitude towards a service failure and how it affects their customer satisfaction regarding their stay at a hotel. The respondents were also asked what monetary post-failure compensation they required from the hotel, for their customer satisfaction to just surpass
21 0 (satisfaction at error-free service). All questions in survey 1 was executed through a standardized questionnaire where all respondents received the same failure scenario and questions.
The questions in survey 2 were similar to the ones in survey 1. Same failure scenario was presented but instead of asking respondents about required monetary compensation, the respondents in survey 2 was offered a monetary compensation and asked to rate their personal satisfaction after receiving it.
A survey can be divided into different parts to make it easier for the respondents to understand it (Adams & Cox, 2008). To make the data from both survey 1 and survey 2 as accurate as possible, an introduction part was implemented to thoroughly illustrate and explain the customer satisfaction scale used.
According to Bryman and Bell (2011), surveys are cheap and an easy way to distribute information and to reach the respondents. The surveys were sent out through different social platforms and personal messages, and the respondents was able to access the survey through their smart phone or computer. The decision to send the surveys through different platforms was based on making it easier for the respondents and encourage them to participate in the survey easily. Bryman and Bell (2011) argue that surveys often suit the respondents' needs better, as they can be participated in when the respondents have time.
3.4. Operationalization
The design of the concepts that the researcher is interested in is called operationalization (Bryman &
Bell, 2013). Customer satisfaction is the concept measured in all SRP research. The SRP’s presence or absence is based on a customer’s experienced post-recovery-satisfaction and nothing else. Both study 1 and study 2 measured the concept of customer satisfaction. Study 1 measured customer satisfaction post-failure and asked each respondent to set the amount of compensation (SEK) required for the customer satisfaction to be restored and just surpass the customer satisfaction present with error-free service.
Study 2 not only measured post-failure customer satisfaction, it also measured post-recovery customer satisfaction. In study 2 the respondents in each of the three groups were given a set amount of monetary compensation (SEK 633, SEK 1204 or SEK 1774). Customer satisfaction was then measured after the amount of compensation was known by the respondent. Customer satisfaction (post-failure and post-recovery) was the only variable measured in this study, while the SRP was measured as a possible outcome of recovery efforts.
22 3.5. Quality
Three criteria are commonly used in determining the quality of research results; reliability, validity and replication (Bryman & Bell 2013). These criteria are discussed below.
3.5.1. Reliability
For a study to accomplish a high reliability the results of the study should ideally be the same as the results in a second study, conducted the same way as the first one (Bryman & Bell, 2013).
Study 1 may have been sufficient to show that the monetary compensation required for the SRP to occur in the given failure scenario is SEK 1204. Despite this, a second study (study 2) was conducted over three groups to reach a higher grade of reliability to the results of study 1.
In order to avoid misunderstandings and misinterpretations among the respondents (potentially causing unreliable data), a test-survey was made to collect flaws and errors in the survey design. The test-survey was answered by four respondents and no additional changes to the survey design had to be made.
To further strengthen the reliability of the data collection, the number of respondents was carefully considered since a larger amount of data would strengthen the reliability. A total of 190 completed survey submissions were collected (40+50+50+50). Respondents’ gender and age distributions are known. Data about what kind of hotel (1-5 stars) respondents usually stay at were also collected and all data was collected from Swedish hotel guests.
The nature of the SRP could, over time, counteract reliability in SRP research. The SRP occurrence depends on increased customer satisfaction, which in most cases is attainable by surpassing a customer’s expectations of recovery. SRP occurrence happens when value created by service recovery goes beyond goods and responsibility (Gohary et al., 2016). However, it is difficult to attain expected and excellent service recovery every time (Krishna et al., 2014) and few things can” go beyond” forever.
If hotels would apply the results from this study in order to reach SRP at the specific failure scenario investigated, over time, a new SRP point would be established since the old point (from this study) would be the recovery expected by hotel guests. It is fair to assume that the anchoring effect could make recovery expectations rise long term. Increased recovery expectations would not generate the same results as prior SRP studies, such as this one. Because of higher customer expectations, recovery efforts have to increase for the SRP to occur. Over time customer expectations will increase again and recovery levels have to increase for the hotels to reach the SRP.
Increase in recovery efforts can only go on for as long as economically costs are justifiable, hence there is a limit where recovery efforts cannot longer surpass the recovery expectations, hence the SRP
23 may be even more rare in future research and theoretically it could even cease to exist. If service firms would apply SRP research results and the mechanism mentioned above would take place, it is almost as if there could be a paradox to the Service Recovery Paradox. High reliability in SRP research studies, such as this one, could therefore theoretically be difficult to accomplish.
3.5.2. Validity
Levels of satisfaction are rarely properly compared within SRP research. For instance, Ok et al. (2007, pp. 675-676) encountered problems when analyzing collected data:” First, service recovery paradox was observable (though not all of them) only when customers considered the recovery effort highly satisfactory (not just satisfactory). Therefore, analysis should be separated into more categories than just unsatisfactory recovery and satisfactory recovery based on customers’ evaluations.”.
Analysis not only have to be separated into more categories. This study suggests that the interpretation of ratings (such as” satisfactory” vs” highly satisfactory”) might be too subjective to provide generalized conclusions. Therefore, study 1 applied tangible service recovery, measurable amounts of monetary compensation to set a fixed recovery point (1204 SEK). This fixed recovery level point provided study 2 with a neutral point of customer satisfaction (0) on a scale from -10 to +10.
3.5.3. Replication
If a study becomes impossible to replicate, its validity becomes questionable (Bryman & Bell, 2013).
To make the study completely replicable, the names of the respondents must have been collected as well as individual records of income and other information possibly considered private. Due to a combination of respect for respondents’ privacy and the belief that less personal information collected would result in a higher number of responses, this study did not collect information that might help identify individuals among the respondents. Without that information it is impossible to retrieve the same collection of data that has been used in this study. However, other researchers may accomplish a similar data collection by looking in to the details described in the methods conducted.
What helps replication of this study is the collected attributes of the respondents (hotel guests), such as; age, gender, nationality and standard of hotel they usually stay at. This information may provide guidance for replicating researchers in the selection process.
