Diagnosing the impact of retail bank customers’ perceived justice on their service recovery satisfaction and post-purchase behaviours: an empirical study in financial centre of middle east

Abstract

The banking sector ranks among the top three sectors in terms of frequency of complaints. The purpose of this study is to assess the effects of perceived justice on recovery satisfaction and to examine the relationships between recovery satisfaction and customer relationship variables in the banking sector where there are lack of empirical studies. Empirical observations were made through questionnaires conducted with 178 retail bank customers in Dubai. The data are analysed through partial least squares (PLS) approach to path modelling to estimate the measurement and structural parameters. The results revealed that distributive justice (DJ), procedural justice (PJ) and interactional justice (IJ) had significant positive effects on service recovery satisfaction, while IJ has the strongest effect with respect to others. Also, service recovery satisfaction had a significant positive effect on trust. The structural model results also revealed that customers’ perceptions of trust had a significant positive effect on their WOM communication and repurchase intentions. The results of the study show that the service recovery satisfaction is a vital mediating variable between perceived justice of customers and customer relationship variables. Furthermore, the mediational role of trust between service recovery satisfaction and future intentions is extensive.

Keywords:

• Perceived justice
• service recovery satisfaction
• customer relationship variables
• retail bank customers
• Dubai

JEL classifications:

• M31
• M16

1. Introduction

In today’s highly competitive global environment service firms’ success depends on their capacity to consistently deliver satisfying consumption experiences (Karatepe, 2012; Patterson, Cowley, & Prasongsukarn, 2006). Achieving consistent and error-free services should be a goal in service businesses. However, mistakes, failures and complaints are inevitable due to the nature of service firms (Boshoff, 2005; Del Rio-Lanza, Vazquez-Casielles, & Diaz-Martin, 2009; Zeithaml, Bitner, & Gremler, 2009). Although service businesses do their best to meet customers’ expectations, they often fail to satisfy their customers, who tend to be more demanding and less loyal than ever before (Kim, Kim, & Kim, 2009).

Intense competition makes customers more demanding and unlikely to forgive, so service recovery is more crucial than ever before (Nadiri, 2011). A service failure is defined as ‘any service-related mishaps or problems that occur during a consumer’s experience with the firm’ (Maxham, 2001, p. 11). Service recovery has also been defined as the service provider’s action when something goes wrong (Grönroos, 1988). Service failures cause detrimental outcomes, like a decline in customer confidence, negative word-of-mouth (WOM) and permanent loss of customers (Yavas, Karatepe, Avci, & Tekinkus, 2003) or continued patronage of the same service provider, albeit with dissatisfaction (Kim et al., 2009). Thus, from a managerial point of view, achieving fair handling of customer complaints is a matter of profitable management (Chebat & Slusarczyk, 2005).

The banking sector ranks among the top three sectors in terms of frequency of customers’ complaints, so service failures are frequent occurrences in the delivery of financial services (Chen, Yu-Chih Liu, Shin Sheu, & Yang, 2012; Lewis & Spryrakopoulos, 2001; Shemwell & Yavas, 1999; Yavas & Yasin, 2001). Managers of financial service institutes should accept that service failure and recovery encounters are critical moments of truth in their quest to satisfy and retain customers and that customers are more emotionally involved in recovery services than routine services (Hultén, 2012; Smith & Bolton, 2002). Thus, in the banking sector, implementations of well-executed service recoveries are vital for increasing customer satisfaction, building customer relationships and preventing customer defections.

Although service recovery has strategic importance for the success of service providers, a growing number of researchers have identified service recovery as a rather neglected aspect of service marketing and one that deserves much greater research attention (Komunda & Osarenkhoe, 2012; Nikbin, Armesh, Heydari, & Jalalkamali, 2011).

The justice theory of Adams (1963) has been perceived as a useful tool in recognising customers’ satisfaction, or otherwise, after service failure. According to Adams, perceived justice is critical for studying a person’s reaction to a conflict situation (Vázquez-Casielles, Álvarez, & Martin, 2010). Researchers have utilised justice theory as the main framework for examining service recovery procedures (Kwon & Jang, 2012; Mccoll-Kennedy & Sparks, 2003; Nikbin, Ismail, Marimuthu, & Armesh, 2012). Thus, the theory has gained popularity in explaining how customers evaluate service providers’ reactions to service failure/recovery. Perceived justice is a multidimensional concept comprising three dimensions: DJ, PJ and IJ. Perceived justice as a driver of emotions is relatively new in the service recovery context and contributions are very limited (Del Rio-Lanza et al., 2009). There is much to learn about how a service provider’s recovery efforts affect subsequent customers’ recovery satisfaction. Also, there is still a need for solid empirical research regarding the impact of organisational responses to customers’ complaints (Davidow, 2000, 2003; Komunda & Osarenkhoe, 2012). According to various researchers (Kim et al., 2009; Maxham & Netemeyer, 2002), there is a need for more empirical research regarding the effects of complainants’ perceptions of justice on satisfaction and intent. Therefore, there is interest in continuing to explore the relative influences of the dimensions of perceived justice on recovery satisfaction (Del Rio-Lanza et al., 2009). Given the importance of relationship marketing in ongoing service industries, such analyses are needed to determine if satisfaction gains realised by offering justice in service recovery affect the post-purchase behaviour of customers. Throughout the last decade, relationship marketing has been one of the most important paradigms in marketing literature (Morgan & Hunt, 1994; Weun, Beatty, & Jones, 2004). Due to the nature of services, they are especially conducive to relationship marketing (Weun et al., 2004). In the literature there are findings that support successful recovery strategies which have positive impacts on customers’ satisfaction and contribute to building customer relationship (Hart, Heskett, & Sasser, 1990; Smith, Bolton, & Wagner, 1999). Also, research results prove that service recovery satisfaction has a positive effect on customers’ trust towards the supplier (Kim et al., 2009; Ok, Back, & Shanklin, 2005). Both WOM and revisit intentions as a post-purchase behaviour are accepted as critical factors that influence the image of a company (Reichheld & Sasser, 1990). Thus, successful recovery efforts not only improve customers’ levels of satisfaction but also contribute to achieve positive WOM behaviour and revisit intention of customers (Maxham, 2001; Maxham & Netemeyer, 2002; Susskind, 2002; Swanson & Kelley, 2001) and ultimately to reach company’s profitability targets as well (Hogan, Lemon, & Libai, 2003; Rust, Lemon, & Zeithaml, 2004). So, these issues justify the necessity of having a study that covers customer relationship variables like trust, WOM and revisit intention.

Although some studies have been performed on this issue in hospitality-industry settings (DeWitt, Nguyen, & Marshall, 2008; Karatepe, 2006; Yuksel, Kilinc, & Yuksel, 2006), there is a lack of studies that have attempted to adopt the relationship marketing approach to explain the relationship between justice perceptions with service recovery, recovery satisfaction and the three relationship outcomes of trust, WOM and repurchase intention (Holloway & Wang, 2015; Kim et al., 2009). Furthermore, there are almost no studies that analyse the relationship in financial services industry. Additionally, the global financial crises had some negative effects on the banking industry and markets the experienced the bankruptcy of some well-known financial institutions. This makes customers much more sensitive in their relations with banks and the relationship becomes fragile (Matute-Vallejo, Bravo, & Pina, 2011). Thus, well-executed service recoveries are vital for increasing customer satisfaction, building customer relationships and preventing customer defections in bank management, so this is one of the most important areas of this study.

Therefore, this study aims to assess the effects of perceived justice on recovery satisfaction and to examine the relationships between recovery satisfaction and customer relationship variables (trust, WOM and revisit intention) in the banking sector, where banks are among the most vulnerable to service failure (Chebat, Amor, & Davidow, 2010) and, more importantly, bank customers consider service recovery as the most important factor of global satisfaction. The study also concentrates on assessing the mediating role of trust between recovery satisfaction and complainants’ future intentions (WOM/repurchase intention).

In the next section, the conceptual model and hypotheses are presented. This is followed by discussions of the method and results of the empirical study. The study concludes with the implications of the results and avenues for future research.

Figure 1. The conceptual model. Source: Author.

2. Theoretical background and research hypotheses

Complaint satisfaction is a necessary prerequisite for customer retention (Stauss, 2002). A business with the ability to react to service failure effectively and implement some form of service recovery will be in a much better position to retain profitable customers (Michel & Meuter, 2008).

2.1 Perceived justice and satisfaction with service recovery

Justice perception has been portrayed as a three-dimensional phenomenon consisting of DJ, PJ and IJ (Greenberg, 1987). A synthesis of the literature reveals that service failure and recovery have presented considerable evidence of the suitability of the concept of justice as a basis for understanding the process of service recovery and its outcomes (Blodgett, Hill, & Tax, 1997; Goodwin & Ross, 1992; Knox & van Oest, 2014; Smith et al., 1999; Tax, Brown, & Chandrashekaran, 1998). Post-purchase satisfaction has been considered a focal point, linking previous purchase beliefs to post-purchase cognitive structures, communications and repurchase behaviour (Harris, Grewal, Mohr, & Bernhardt, 2006). According to the literature, a consumer’s satisfaction with complaint handling/service recovery results from an evaluation of the aspects involved in the final result (regarding DJ), the process that originated such a result (PJ) and the way the consumer was treated during the episode (IJ) (Blodgett et al., 1997; Goodwin & Ross, 1992; Mattila, 2006; Siu, Zhang, & Yau, 2013; Smith et al., 1999; Tax et al., 1998).

2.2 Distributive justice

DJ occurs when the customer perceives the company’s effort in solving the problem after service failure (Smith et al., 1999; Tax et al., 1998). DJ is more tangible and easier to quantify (Chang & Hsiao, 2008) and generally deals with outcomes offered to customers during service recovery, like monetary rewards, discounts, coupons and offering products/services free of charge during the service failure (Mattila, 2001; Sparks & McColl-Kennedy, 2001). Atonement is the most important organisational response to a customer complaint associated with DJ (Tax et al., 1998).

The justice, fairness, need, value and reward of outcomes are the measurement tools for DJ in service recovery (Blodgett et al., 1997; Chebat & Slusarczyk, 2005; Smith et al., 1999; Wirtz & Mattila, 2004). There are studies in the literature that find a positive relationship between DJ and satisfaction with service recovery (Fang, Luo, & Jiang, 2013; Maxham & Netemeyer, 2002; Smith & Bolton, 1998) and a positive association with complaint handling (Homburg & Fürst, 2005; Karatepe, 2006; Sharma & Dwivedi, 2014). Thus, the following hypothesis is proposed:

Hypothesis 1: The distributive justice perception of the bank’s customers has a positive effect on their service recovery satisfaction.

2.3 Procedural justice

PJ is ‘the perceived fairness of policies, procedures, and criteria used by decision-makers to arrive at the outcome of a dispute or negotiation’ (Blodgett et al., 1997, p. 189). In terms of effective service recovery, PJ refers to the customer’s perception of justice for the several phases of procedures and processes required to recover the failed service (Choi & Choi, 2014; Mattila, 2001). PJ should accomplish certain procedures such as ‘refund policies’ or ‘time to receive refund (timing)’ for the successful outcome of an exchange (Lind & Tyler, 1988; Thibaut & Walker, 1975). In addition, the timing concept is the main core of PJ within a customer complaint situation (Smith et al., 1999).

It is remarkable that a service provider can increase customer satisfaction with service recovery by improving their awareness about PJ (Vázquez-Casielles et al., 2010). There is obvious evidence that indicates that perceived PJ does significantly influence customer satisfaction via complaint handling (Homburg & Fürst, 2005; Karatepe, 2006; Wu & Huang, 2015). Thus, the following hypothesis is proposed:

Hypothesis 2: The procedural justice perception of the bank’s customers has a positive effect on their service recovery satisfaction.

2.4 Interactional justice

IJ concentrates mainly on the interaction between an employee and a customer through communication during a complaint episode (Santos & Fernandes, 2008). According to Sparks and McColl-Kennedy (2001, p. 52),

‘IJ in service recovery is related to the way customers involved in a failed service are handled, and it means the evaluation of the degree to which the customers have experienced justice in human interactions from the employees of service firms during the recovery process.’

Some researchers believe that IJ has the most significant influence on customer satisfaction during service recovery (Homburg & Fürst, 2005; Maxham & Netemeyer, 2002). The literature review suggests that there is a positive relationship between IJ and customer satisfaction during service recovery (Goodwin & Ross, 1992; Siu et al., 2013; Tax et al., 1998; Wu, 2013). Thus, the following hypothesis is proposed:

Hypothesis 3: The interactional justice perception of the bank’s customers has a positive effect on their service recovery satisfaction.

2.5 Recovery satisfaction and trust

The importance of trust in the field of marketing is critically important for both buyers and sellers. Trust is defined as ‘a willingness to rely on an exchange partner in whom one has confidence’ (Moorman, Deshpande, & Zaltman, 1993, p. 90). In service marketing, trust is an important element in continuing the relationship between the customer and the service provider, as making a decision to purchase an item occurs before experiencing the service (Berry & Parasuraman, 1991). Therefore, trust is highlighted as one of the crucial ingredients for developing strong and long-term relationships between consumers and suppliers (Garbarino & Johnson, 1999; Morgan & Hunt, 1994; Tax et al., 1998).

Service recovery is an activity performed by organisations in order to pay particular attention to customers’ complaints regarding a service failure (Spreng, Gilbert, & Robert, 1995). A good service recovery process is expected to eliminate customer anger, motivate customers and promote the retention of customers by satisfying them (Etzel & Silverman, 1981; Hart et al., 1990). Therefore, recovery satisfaction will have a positive influence on the trust of customers. Trust is built among customers by allowing them to feel that their voices are heard and that they are valued by organisations. Therefore, service recovery in an organisation can be changed to recovery satisfaction by stimulating the trust of customers.

The significant role of satisfaction in trust after service recovery and the influence of satisfaction as a key element on perceptions of customers regarding complaint handling in the future have been shown by many studies (Kau & Loh, 2006; Santos & Fernandes, 2008; Simon, 2013; Tax et al., 1998). Thus, the following hypothesis is proposed:

Hypothesis 4: The service recovery satisfaction of the bank’s customers has a positive effect on their trust.

2.6 Trust and behavioural intentions: WOM and revisit intention

Customers pursue providers whom they will be able to trust and to whom they can remain loyal to minimise the risk of their purchases (Rundle-Thiele & Bennett, 2001). Additionally, Morgan and Hunt (1994) proposed a model of relationship marketing where trust is conceptualised as the key variable to the development of long-term customer relationships. A provider’s successful service recovery contributes to build customer trust (Tax et al., 1998). The reason for this is simple; trust is a kind of guarantee that a company will continue to offer a consistent and competent performance. According to relationship marketing theory, customers are more likely to continue in a relationship with a service provider if they trust that provider (Shemwell, Cronin, & Bullard, 1994). Santos and Fernandes (2008) proposed that complainants’ trust in a provider will lead to repeat purchases and positive WOM communication about the provider. So, companies may use trust as an important marketing tool to facilitate customer loyalty.

Customer satisfaction/dissatisfaction is an antecedent affecting behavioural intentions (Anderson & Sullivan, 1993; Oliver, 1981). Also satisfaction is not a single driving force for customers that lead them to behave positively toward a service provider (Hoffman, Kelley, & Rotalsky, 1995). Thus, identifying mediating variables between customer satisfaction and behavioural intentions is of interest. This study suggests that customer’s trust mediates between service recovery satisfaction and behavioural intentions.

In the literature, revisit intention and WOM communication have been used as dimensions of loyalty (Lam, Shankar, & Murthy, 2004; Sirdeshmukh, Singh, & Sabol, 2002). Thus, this study tries to analyse the effect of customers’ trust on their revisit intentions and WOM communication.

A revisit intention takes place when a person is motivated to remain in the relationship, and, as explained so far, trust can be considered as a strong element influencing and empowering this relationship (Garbarino & Johnson, 1999; Oh, 2002; Yen, 2009). WOM is one of the aspects of post-purchase behaviour and occurs as people keep on sharing their assessments of their experiences and the findings of many studies have supported that trust is positively associated with WOM (Garbarino & Johnson, 1999; Kim, Han, & Lee, 2001; Kim & Park, 2013; Oh, 2002; Simon, 2013). Therefore, the following hypotheses are proposed:

Hypothesis 5: Trust has a positive effect on bank customers’ word-of-mouth.

Hypothesis 6: Trust has a positive effect on bank customers’ revisit intentions.

2.7 Recovery satisfaction and behavioural intentions

Customers’ behavioural intentions are consequences of their level of satisfaction and influences customers’ relations with companies in future. In other words, an increase in customer satisfaction should lead to an increase in customer loyalty (Heskett, Jones, Loveman, Sasser, & Schlesinger, 1994). Companies’ success and fairness in handling customers’ complaints not only contributes to recovery satisfaction but also to effective service recovery, and can turn an unfavourable service experience into a favourable one, consequently leading to repurchase intention and positive WOM intention (Spreng et al.,1995). Thus, the following hypotheses are proposed.

Hypothesis 7: Bank customers’ service recovery satisfaction has a positive effect on their word-of-mouth.

Hypothesis 8: Bank customers’ service recovery satisfaction has a positive effect on their revisit intentions.

3. Methodology

The United Arab Emirates (UAE) is an emerging country that has achieved continuous economic growth in the last 25 years (Dayan, Al-Tamimi, & Elhadji, 2008). The study was carried out in Dubai, one of the emirates of the UAE, which is one of the world’s fastest-growing economies, with a per capita income of US$49,000 (Central Intelligence Agency, 2013). Dubai has emerged as a global city and a business hub. Although Dubai’s economy was built on the oil industry, the emirate’s model of business drives its economy, with the effect that its main revenues are now from tourism, real estate and financial services, similar to those of Western countries. The strategic plan of the government is to make Dubai a ‘globally leading Arab city’ and a ‘global city’ in 2015. Furthermore, services such as financial, tourism, transport and trade services contribute 74% of Dubai’s GDP growth (Balakrishnan, 2008). As of the end of 2012, the number of banks in the UAE was 51 (23 UAE banks and 28 foreign banks). According to UAE Central Bank reports, both Islamic and conventional banks increased their number of branches during 2012 by around 10% with respect to the previous year, reaching a total of 948 (including head offices, banking service units, branches, etc.). Due to the rapid increase in competition between banks in the Dubai, service recovery has become a critically competitive issue. Intense competition and many alternatives might encourage customers in this market to change banks and this provides a favourable position for customers to ask for better service recovery. The international structure of its population and the competitive environment of the banking sector of Dubai make this geographical area an appropriate research field for this study in order to contribute to the literature.

3.1 Survey instrument

The survey instrument was designed to measure the constructs specified in the conceptual model. A four-item DJ scale was adapted from Smith et al. (1999). A three-item PJ scale was adapted from Rupp and Cropanzano (2002). IJ was measured using seven items, where five items were adopted from Severt (2002) and two from Smith et al. (1999). The four-item recovery satisfaction scale was adopted from Brown, Cowles, and Tuten (1996), as well as Maxham and Netemeyer (2002). The questions regarding trust consisted of four items adopted from Morgan and Hunt (1994) and Wong and Sohal (2002). The WOM scale was measured by a two-item scale adapted from Mattila (2001) and Wong and Sohal (2002). The two-item revisit intention scale was adapted from Mattila (2001) and Maxham and Netemeyer (2002). Respondents were asked to indicate their agreement/disagreement with these statements using a seven-point Likert-type scale (1 = ‘strongly disagree’; 7 = ‘strongly agree’). Finally, the survey also included questions regarding certain demographic characteristics of the respondents, like age, gender, nationality, education level, marital status, family income and occupation.

The survey instrument was applied in three languages; English, Persian and Arabic. The original instrument was prepared in English and then translated into Farsi Persian and Arabic in line with the back-translation method (McGorry, 2000). The instrument was finalised based upon feedback from a pilot sample of 20 respondents. Table 2 reflects the instrument that was used to collect data base on developed conceptual model.

3.2 Sample and data collection

The population of the study was retail customers of conventional banks in Dubai who did experience a service failure. The hypothesised relationships were tested by collecting data from respondents who had experienced a service failure and reported their complaints to their financial institutions during the previous three-month period. A survey method was used by collecting data through using a questionnaire that was developed to measure respondents’ perceptions on constructs that forms the conceptual model of the study. All the respondents were carefully screened to ensure their eligibility to be involved the study. While choosing the sample for the study the judgmental sampling procedure was used (Judd, Smith, & Kidder, 1991). Questionnaires were filled out by respondents in a self-administered manner and sent back to researchers. A total of 220 survey instruments were administered during April and May 2012 by trained researchers. The overall response rate was 92% (202 survey instruments), 178 of which were usable.

3.3 Data analysis

Surveyed data entered into SPSS 16.0 to carry out descriptive statistics, while Smart-PLS 2.0 M3 software was used to test path model. Smart-PLS (Ringle, Wende, & Will, 2005) employed since it allows us to have a simultaneous assessment of the structural component and measurement component to be carried out in a single model like other structural equation approaches (e.g. LISREL) (Halawi & McCarthy, 2008). The SmartPLS has certain advantages over others like; it is more users friendly and has more extensive online support (Temme, Kreis, & Hildebrandt, 2006). Unlike the covariance-based structural equations modelling (e.g. LISREL and AMOS), PLS does not require multivariate normal data (Jain, Malhotra, & Guan, 2012; Lee & Tsang, 2001) and considered to be appropriate in building causal modelling for future testing purposes (Teo et al., 2012). Additionally, PLS does not create identification problem with the use of formative indicators and let the operationalisation of formative scales (Chin, 2010). So, it gained acceptance among many researchers and started to be used extensively in management researches (Chin, 1998; Felix & Garcia-Vega, 2012). One of the main advantages of PLS that extends its usage among researchers is that it is believed to have the ability to estimate research models with a smaller sample sizes (Jain et al., 2012; Tenenhaus, Vinzi, Chatelin, & Lauro, 2005).

4. Results

4.1 Respondents’ demographic profile

The following table reflects the demographic breakdown of respondents.

Table 1. Demographic breakdown of respondents.

	Frequency	Percentage (%)
Gender
Female	69	39
Male	109	61
Age
20–34	117	66
35–44	42	24
45–60	17	9
61 and above	2	1
Maritual Status
Single	91	51
Married	80	45
Divorced or Widowed	7	4
Nationality
Philippine	44	25
India	31	17
Iran	24	13
UAE	18	10
Syria	9	5
Other	52	30
Education Level
Degree Lower than High School High School	39	22
Degree	102	57
Bachelor Degree	14	8
Graduate Degree	23	13
Monthly Family Income Level
Less than $2500	71	40
$2501–$5000	45	25
$5,001–$7500	22	12
$7501–$10,000	19	11
$10,001 and above	21	12
Occupation
Employee	92	52
Manager	33	18
Business Owner	20	11
Doctor/Lawyer/Consultant	16	9
Retired or Housewives	17	10

Source: Author.

4.2 Descriptive statistics

Table 2 shows the means and standard deviations of the composite measures of the model construct. Frequency analysis of the 30 items indicated no problems with regard to ‘floor’ or ‘ceiling’ effects in the measurements. The usable response number (n = 178) exceeded the recommended minimum of 30 required for estimation of this model using PLS (Chin, 1998).

Table 2. Convergent validity of constructs.

Variable											Factor Loading
Distributive Justice
ic =	0.93	Factor mean score =		4.66	α =	0.93	SD = 1.24		AVE =	0.65
The outcome I received was fair.											0.8275
I did not get what I was interested in. ®											0.7249
In resolving the problem, the bank gave me what I needed.											0.8409
The outcome I received was not right. ®											0.742
The bank’s employees were willing to provide what I needed when my requests were reasonable.											0.8413
When I brought valid requests, the bank’s employees would provide fair outcomes.											0.8508
When I consider my contribution to this bank as a customer, I am happy with the responses I get from the employees.											0.8137
Interactional Justice
ic =	0.93	Factor mean score =		4.91	α =	0.93	SD = 1.24		AVE =	0.66
The bank’s employees were courteous to me.											0.656
The bank’s personnel were honest with me.											0.8222
The bank’s personnel showed a real interest in trying to be fair.											0.8385
The bank’s employees showed concern.											0.8218
The bank’s employees tried to help me.											0.8552
The bank’s employees’ communication with me was appropriate.											0.8348
The bank’s employees put the proper effort into resolving my problem.											0.8212
Procedural Justice
ic =	0.97	Factor mean score =		4.57	α =	0.94	SD = 1.54		AVE =	0.60
My complaint was handled in a very timely manner.											0.9149
My complaint was not resolved as quickly as it should have been. ®											0.914
I often had to apply to the related bank’s employees for the resolution of my problem. ®											0.8761
Recovery Satisfaction
ic =	0.92	Factor mean score =		4.60	α =	0.89	SD = 1.44		AVE =	0.75
Overall, I am satisfied with the service I received.											0.8801
I am satisfied with the manner in which the service failure was resolved.											0.8841
This bank’s response to the service failure was better than expected.											0.8243
I now have a more positive attitude toward this bank.											0.8746
Trust
ic =	0.93	Factor mean score =		4.68	α =	0.89	SD = 1.40		AVE =	0.76
This bank’s employees can be trusted at all times.											0.861
This bank’s employees have a high level of integrity											0.8645
This bank’s employees make every effort to fulfil the promises made to its customers.											0.9012
Overall, this bank’s is reliable.											0.8612
Repurchase Intention
ic =	0.97	Factor mean score =		4.59	α =	0.94	SD = 1.75		AVE =	0.94
I consider this bank my first choice for work with.											0.9708
I consider this bank to work with when I need banking services.											0.9711
Word of Mouth
ic =	0.95	Factor mean score =		4.64	α =	0.92	SD = 1.62		AVE =	0.88
I would say positive things about this bank to other people.											0.9708
I would recommend this bank to someone who seeks your advice.											0.9711
I would encourage friends and relatives to use this bank.											0.9385

Notes: ‘ic’ is international consistency measure; ‘α’ is Cronbach’s alpha; ‘AVE’ is average variance extracted; ‘®’ means that these items were reversed scored. Source: Author.

4.3 Model estimation

The PLS that was used to test the model, is a structural equation modelling (SEM). SEM is a multivariate technique that combines multiple regression and factor analysis simultaneously to predict a series of interrelated dependence relationships. The measurement model was tested by estimation of the internal consistency and validity (convergent and discriminant) of the instrument items.

The composite reliability measure of internal consistency and the average variance extracted (AVE) were used to assess the composite reliability of each block of indicators measuring a given construct. The internal consistency measure was used in preference to Cronbach’s alpha because Cronbach’s alpha assumes parallel measures and represents a lower bound of composite reliability, whereas the internal consistency measure is unaffected by scale length and is a more general measure (Chin, 1998). All the reliability measures were above the recommended level of 0.70 (see Table 1), thus indicating adequate internal consistency (Nunnally, 1978). The AVE scores for both tables varied from 0.6 to 0.94: above the minimum threshold of 0.5 (Chin, 1998).

All the factors and their loadings are listed in Table 2. In this table, all of the factor loadings are satisfactory in terms of being above the cut-off value of 0.5 (Hair, Bush, & Ortinau, 2000; Stevens, 1992). Also, the values of the composite reliability and average extracted variance satisfy the threshold values of 0.7 and 0.5, respectively, demonstrating good internal consistency and suggesting good convergent validity and reliability (Fornell & Larcker, 1981).

Adequate discriminant validity was confirmed by comparing the average variance shared between a construct and its measures (AVE), which revealed values greater than the recommended value of 0.5 (Fornell & Larcker, 1981). In addition, adequate discriminant validity was confirmed when the square root of the AVE for each construct was shown to be larger than the correlation between the construct and any other construct in the model (Table 3).

Table 3. Discriminant validity of constructs.

	DJ	IJ	PJ	RI	RS	Trust	WOM
Distributive Justice	0.807
Interactional Justice	0.5.541	0.81
Procedural Justice	0.549	0.4.032	0.777
Repurchase Intention	0.3.861	0.5.618	0.2.981	0.971
Recovery Satisfaction	0.4.735	0.5.561	0.3.655	0.681	0.866
Trust	0.4.836	0.658	0.3.734	0.6.685	0.7.022	0.872
Word of Mouth	0.4.275	0.5.886	0.3.326	0.7.702	0.6.998	0.7.073	0.936

Note: Square root of AVE in the diagonal. Source: Author.

In summary, the measurement models’ results provided support for the reliability and validity (convergent and discriminant) of the measures used in the study.

Different to other SEM techniques, the PLS structural model is mainly evaluated by using Goodness-of-Fit (GOF) index (Tenenhaus et al., 2005). Tenenhaus et al. (2005) proposed a global fit measure where GOF (0 < GOF < 1) is defined as the geometric mean of the average communality and average R² (for endogenous constructs). For this model the GOF index was 0.7162 which was satisfying the cut-off values used globally (Wetzels et al., 2009). Thus, this result implies that the model had acceptable predictive relevance.

Before conducting further tests on the structural model, the test for multicollinearity was achieved. One common approach to detect multicollinearity is based on the variance inflation factors (VIF). All VIF values range from 1.29 to 2.09 that are far below the recommended threshold 10 that indicate there is no cause of concern about the multicollinearity.

4.4 Testing of hypotheses

As shown in Table 4, the Smart PLS software provided the R² statistic for each endogenous construct in the model and the standardised coefficients, thus indicating the percentage of each construct’s variance in the model, as well as the strengths of the relationships between constructs. In accordance with Chin (1998), ‘bootstrapping’ (300 resamples) was applied to produce standard errors and t-statistics to assess the statistical significance of the standardised coefficients.

Table 4. Structural model results.

	Proposed Effect	Path Coef.	T-value	Significance
Effect on Recovery Sat. (R-Square 0.70)
H1. Distributive Justice	+	0.347	6.29	0.00
H2. Procedural Justice	+	0.218	4.02	0.00
H3. Interactional Justice	+	0.380	5.81	0.00
Effect on Trust (R-Square 0.64)
H4. Recovery Satisfaction	+	0.801	34.28	0.00
Effect on WOM(R-Square 0.69)
H5. Trust	+	0.399	5.84	0.00
H7. Recovery Satisfaction	+	0.477	6.90	0.00
Effect on Repurchase Int. (R-Square 0.67)
H6. Trust	+	0.396	5.56	0.00
H8. Recovery Satisfaction	+	0.465	6.26	0.00
Indirect Effects
Distributive Justice → Trust		0.278	6.49	0.00
Procedural Justice → Trust		0.175	4.13	0.00
Interactional Justice → Trust		0.304	5.43	0.00
Distributive Justice → WOM		0.277	6.23	0.00
Procedural Justice → WOM		0.174	4.06	0.00
Interactional Justice → WOM		0.303	5.71	0.00
Recovery Satisfaction → WOM		0.796	32.45	0.00
Distributive Justice → Repurchase Int.		0.271	5.99	0.00
Procedural Justice → Repurchase Int.		0.170	4.06	0.00
Interactional Justice → Repurchase Int.		0.297	5.75	0.00
Recovery Satisfaction → Repurchase Int.		0.781	26.44	0.00

Notes: p-values < 0,01. Source: Author.

The results in Table 4 depict that the structural model explained 70% of the variance in the ‘service recovery satisfaction’ construct. As can be seen from the results, DJ, PJ and IJ have significant positive effects on service recovery satisfaction. Thus, H1, H2 and H3 are supported. It is apparent that the construct of ‘service recovery satisfaction’ had a significant positive effect on ‘trust’, so H4 is supported. The structural model results also reveal that customers’ perceptions of ‘trust’ had a significant positive effect on their WOM communication and intention to revisit. Therefore, support exists for both H5 and H6. The results also indicate that the construct ‘service recovery satisfaction’ had a significant positive effect on customers’ WOM communication and intention to revisit. Thus, both H7 and H8 are supported. Each and every one of the model’s eight hypotheses is supported (Figure 2).

Figure 2. Structural model. Source: Author.

Besides direct effects of the proposed model, the indirect effects were also analysed to get more in depth analysis of customers’ further intentions with service recovery.

According to results presented in Table 4, perceived DJ (0.278), PJ (0.175) and IJ (0.304) influenced trust positively through service recovery satisfaction. Once three justice perceptions were compared, PJ and DJ had relatively smaller effect on trust through service recovery satisfaction and IJ has ample effect. The mediational roles of service recovery satisfaction and trust between perceived justice of customers and future intentions are presented in Table 3 as well. According to the results, all of the perceived DJ, PJ, and IJ had important and positive effect on both WOM and repurchase intentions of customers through service recovery satisfaction and trust. The effect of the perceived IJ on both WOM communication and repurchase intentions of customers found to be the greatest with respect to others.

The results about the direct effect of customers’ service recovery satisfaction on their WOM communication was significant (0.477), as well as the indirect effect of it through trust was found to be very important (0.796). On the other hand, the direct effect of service recovery satisfaction on repurchase intention was noticeable (0.465) while its indirect effect through trust was found to be apparent (0.781). Thus, results point out that the mediational role of trust between service recovery satisfaction and future intentions are important.

5. Discussion

A structural model was proposed to explore the influence of retail bank customers’ justice perceptions on their service recovery satisfaction and in turn the effect of service recovery satisfaction on customers’ trust perceptions and its impact on customers’ future intentions/post-purchase behaviour in the banking sector of Dubai.

The findings of the study show that all of the distributive, procedural and IJs have a positive significant effect on customers’ service recovery satisfaction. The results regarding IJ show that it has a stronger effect on recovery satisfaction with respect to others that are consistent with some of the findings (Davidow, 2003; Karatepe, 2006) as well as inconsistent with others (Maxham & Netemeyer, 2002, 2003; Smith & Bolton, 1998). This indicates that the fair interpersonal treatment of customers has a vital effect on their recovery satisfaction. Thus, bank staff should be ready to express their apologies, must use empathy and attentiveness and be courteous towards their customers who had experience a service failure. These reactions contribute to achieve the proper IJ among complainants. Currently, apologies in service marketing are recognised as a tool for maintaining the relationships of customers with a firm after service failure. The other aspect of IJ is communication style, through which employees and managers (especially front line employees) are able to recover the service failure by having the experience and knowledge to communicate with customers. The importance of IJ in the satisfaction of customers after service failure has been observed in this study. This study shows that employees will be able to satisfy customers through IJ by satisfying their emotional feelings first. Similarly, a study conducted by Tax et al. (1998) supports the significant influence of IJ on the emotional satisfaction of customers. A study done by Goodwin and Ross (1992, p. 516) also supports the findings of this study by stating that ‘… offering an apology and giving customers the chance to express their feelings with real atonement will improve perceived justice and satisfaction’. In addition, a study by Dayan et al. (2008) is another important source of evidence supporting the IJ findings of this study, in which the importance of IJ in the satisfaction of customers with service recovery has been found.

The results of this study also show that the effect of DJ is stronger than that of PJ. This implies that in order to regain satisfaction from displeased customers both bank managers and staff would consider fair distributive treatment like refunds, special discounts, upgrading status/services and offering credit cards with no commission.

Although in this study PJ has a significant positive effect on recovery satisfaction, it has the lowest effect with respect to others. PJ significant positive effect on recovery satisfaction might contradict with a study done by Dayan et al. (2008), this relationship is consistent with those of Chebat and Slusarczyk (2005), which also found the importance of timing in PJ. Although the concept of timing is the major attribute of PJ in a customer complaints context, this study considers the perception of PJ in a broader perspective, with respect to Dayan et al. (2008) who just operationalised PJ in terms of timing. This might be the reason why the present study is closer to a study done in Canada (Chebat & Slusarczyk, 2005) than the study done by Dayan et al. (2008) in the UAE. Additionally, the differences between this study and Dayan et al.’s study might be due to the influences of the participants. This is because the current study is being tested on participants from Dubai, who are of a wider range of nationalities regarding the financial market, while the study of Dayan et al. involves participants from different parts of UAE that might not reflect the cosmopolitan structure of Dubai. In addition, empowerment could be an important factor, causing the results of the current study to demonstrate the significant influence of PJ on recovery satisfaction. Empowerment provides an opportunity for employees to solve service failure.

The results of the structural model indicate the influence of recovery satisfaction on trust. The importance of recovery satisfaction in trust has been explained widely. It is obvious now that recovery satisfaction has a significant influence on the trust of customers. This is because recovery satisfaction is a key element in stimulating the emotional feelings of customers. The results of this study, which indicate the significant influence of recovery satisfaction on trust, are because customers in Dubai are highly satisfied in terms of recovery satisfaction. Thus, recovery satisfaction cannot always be viewed as a cost to an organisation because trust creates a feeling among customers of them being valuable. Support for this finding comes from the study by Santos and Fernandes (2008), who observe the significant influence of post-complaint satisfaction on trust. Overall, trust can be considered as a key element for the success of organisations.

Another important point derived from this study is the influence of trust on post-purchase behaviour (repurchase intention and WOM). Thus, firms fair and successful complaint handling efforts influence the post-purchase behaviour of their customers and this demonstrate the importance of relationship marketing in banking sector.

The structural model results in Table 4 show that the path coefficient of WOM is 0.399 and there is a strong positive relationship between trust and WOM within a 99% confidence interval. This result is also similar to the findings of a study done by Hong and Rim (2010). On the other hand, the path coefficient of the repurchase intention which is the other aspect of post-purchase behaviour, is 0.396 and there is a significant relation between trust and repurchase intention within a 99% confidence interval. This is supported by the findings from a study by Hamid (2008), in which a relationship between trust and repurchase intention was found.

The significant influence of trust on post-purchase behaviour, in terms of it having a positive effect on WOM and repurchase intention, has also been found in this study. Yen (2009) shows the importance of loyalty in the repurchase intentions of customers, which is an important support to the present study. Therefore, trust can be considered as an important and essential element promoting the retention of customers. So, the reason behind finding significant effects for perceived justice components might be due to the importance of trust in financial service institutions in Dubai. The trust that is created among the customers is due to the efforts of the firm in encouraging and building trust among its customers. Therefore, the trust created by the firm encourages and motivates the repurchase intentions of the customers. In terms of trust, it is possible to come up with the conclusion that trust might be reciprocal, in order to enhance positive WOM and promote post-purchase behaviour.

This study also finds out a significant positive effect of service recovery satisfaction on future intentions of customers. This supports the idea that effective service recovery efforts of organisations after service failure not only correct service failure but also help to maintain strong relationships (Ha & Jang, 2009; Kim & Kandampully, 2011).

The results of this study also indicate that service recovery satisfaction is a vital mediating variable between perceived justice by customers and customer relationship variables. Additionally, this study’s results also demonstrate the substantial mediating role of trust between service recovery satisfaction and future intentions (WOM/repurchase intention) of customers. In other words, mangers/firms who are able to establish a strong trust between their institute and customers, benefit from a long-term relationship that will lead to loyalty, which is an important asset in today’s highly competitive market conditions.

6. Conclusion

This study aims to assess the effects of perceived justice on recovery satisfaction and to examine the relationships between recovery satisfaction and customer relationship variables (trust, WOM and revisit intention) in the banking sector. Another significant aspect of this study is that there are lack of studies that examine this relationship in the banking sector, where banks are among the most vulnerable to service failure (Chebat et al., 2010) and, more importantly, bank customers consider service recovery as the most important factor of global satisfaction (Hall, 1997). So, it is vital for service managers to develop appropriate strategies that will lead to effective service recovery that will be perceived fairly among their customers who experience service failure.

In today’s highly competitive and dynamic market conditions, establishing and maintaining good relations with customers is a must in order to be competitive (Ramachandran & Chidambaram, 2012) Thus, effective service recovery efforts of banks have considerable positive effects on customers’ positive WOM communication as opposite to angry customers (Makdessian, 2004). Through effective service recovery, managers not only decrease the number of dissatisfied customers, but increase the number of satisfied customers, which will lead to an increase in sales, loyalty and positive WOM communication. The results of this study also prove that complainants who perceives justice in a bank’s recovery efforts have higher service recovery satisfaction and trust towards that institute, and this will consequently lead to positive post-purchase behaviour.

So, the effectiveness of service recovery encourages customer loyalty. Therefore, service recovery increases ‘customers’ perceptive value’, ‘satisfactory feelings’ and ‘loyalty and credit’ (Bitner, Booms, & Tetreault, 1990; Brown et al., 1996; Cong, Yan, & Wang, 2007; Lewis & McCann, 2004). Deregulation in the service industry portrays the new perspective of the future. Competition in this industry is incrementally increasing. These circumstances may occur as a result of changes in both customers’ perceptions and customers’ behaviour. Unlike previously, the power is not in the hands of companies, but in the hands of the customer, who now determine the production of services in the market. Perhaps many years ago there were only a few firms fulfilling the needs and wants of customers, and customers were forced to buy from those firms due to a shortage of production services. However, today it is a different marketplace because of the complexity of service production. In today’s world, it is really difficult to convince people to consume regularly from a particular firm. Therefore, the responsibility of service managers has dramatically increased as a result of the competitive environment in service sectors.

If a service provider cannot satisfy a consumer after service failure, the customer will no longer purchase from the company. There are many service recovery strategies that managers can utilise to retain customers. According to Zeithaml, Bitner, and Gremler (2006), ‘doing it right the first time’, ‘learning from recovery experiences’ and ‘learning from lost customers’ are strategies that can be crucial in helping the management to avoid customers abandoning the company. It is necessary for the management to understand that services should be delivered to customers correctly first time. This strategy prevents the extra cost of service recovery, as the consumer will be satisfied during the initial point of service delivery. On the other hand, there are many uncontrollable situations in the business environment; therefore, managers should predict that service failure is unavoidable and that this is the time to implement subsequent strategies. Finding the reasons for customers leaving, doing research on them and using the experience from service recovery are very useful for managers in terms of improving their ability to manage their company when service recovery is necessary.

Like many other industries, the financial services industry also relies on an imaginary term in marketing: namely, customer satisfaction. The managers in financial services use this term to attract customers to use more and more of their services. Service recovery in financial services is an essential function of a service delivery strategy, as this sector deals with the investment and capital of people. With the emergence of advanced technology such as e-banking services, mobile banking and ATM services (and the decrease in face-to-face delivery systems), the number of service failures may increase and service recovery plays a vital role in financial organisations. The customers’ education and experiences allow them to interpret and evaluate the effectiveness of the service delivery. How the bank handles their complaints after service failure determines their perspective in terms of working with that bank in the future. These are the main points that management should be concerned with, as well as how to use them to be successful in the organisation.

In this study, both the validity and reliability of the data have been measured by using PLS. Further, all the relationships in the PLS analysis have been tested and the results have shown a significant influence for all hypotheses. However, there are certain limitations within the study. One limitation of this study is that a specific service industry chosen for empirical study. Therefore, it is not possible to say that the outcomes of this study can be generalised for all service sectors. For further research, a similar model might examine corporate customers. A future study might test similar models that include commitment as well, which reflects the effectiveness of service recovery. This study’s model might be tested on other service sectors to analyse whether customers’ perceptions of justice and their future intentions do vary according to service sectors. Finally, this study did not employ an in depth interview, a scenario-based experiment or a critical incident technique to analyse the relationship. One of these techniques might use in future to further examine the relationship in the conceptual model.

From a practical point of view, it is not easy to establish a relationship where competition is increasing on global bases and customers have wide variety of choices in terms of financial service institutions. Today, customers are the most valuable assets of all type of businesses. Thus, in the existing consumer economy, gaining new customers and keeping existing ones is the one of the most crucial efforts of any businesses. Like almost all service businesses, financial service institutions are also trying to establish good relationships with their customers. People tend to continue to work with same financial service institution if they receive satisfactory service level that match their expectations. The services that banks provide are related with the wealth of people, so customers are much more conscious about receiving better and more secure services. It is so vital for financial service institutions to achieve customer satisfaction that will help them to retain customers loyal when they experience service failure.

Although service failures are not inevitable, banks should try to do their best to eliminate or minimise the occurrence of service failures where there is an intense competition. Thus, it is important for bank managers to monitor the quality of services provided, train employees regularly to provide better services and set working standards that are some of the accepted measures that management might use to reduce failures. Like other service sectors, financial service providers should treat their customers as individuals and find individual solutions for their expectations. In order to treat customers as individuals, banks use a structure where customers and staff have direct contact. This might be a tool that bank managers could use to measure customers’ levels of satisfaction or how well their expectations were matched by the institute.

The results of this study also prove that perceived complaint justice has a strong effect on service recovery satisfaction and behavioural intentions. So, customers’ perceptions of how fair and how well they are treated in service recovery after they informed an institute about their complaint, will affect their satisfaction toward the bank as well as their intentions to continue to work with that bank and recommend it to others. In other words, it seems to be that it is compulsory for banks to handle and respond complaints in the right manner.

According to results, IJ had the strongest effect among other justice dimensions on service recovery satisfaction. This dimension is related to employees’ manner, and especially due to cultural reasons in some Middle East countries peoples’ intentions related to an institute rely heavily on their relationship with that institute and its staff. Thus, employees’ poor manner may be risky for a bank as customers may pass on negative WOM communications and will not want to work with that bank. It is important for bank managers to encourage their staff to act courteously while dealing with customers who are angry due to service failure.

Another important dimension that has a strong effect on service recovery satisfaction is DJ. This dimension is related with outcome of the recovery effort. So, bank management should focus on the outcome of complaint handling and try to understand how satisfied the complainant is from service recovery efforts. It is important to give a message to the complainant that management is sorry about the failure and they are ready to compensate it. Sometimes that compensation might be in monetary terms but sometimes an apology and explanation might be helpful to calm the complainant. Depending on how serious the failure is, management should take appropriate action that will lead recovery satisfaction.

In this research, although the PJ is the dimension that has the least strong effect on service recovery satisfaction, it is necessary to pay attention to service recovery procedures. Banks should train their staff to respond to customers’ complaints quickly. Thus, clarifying policies about how staff should deal with complaints and giving authority to staff about how to handle complaints immediately will positively influence customers’ perceptions of justice. Also banks should make clear to the complainant the policies and practices that the bank will follow to handle a failure and convince the complainant that it will be most appropriate for both parties.

Today’s consumer economy forces businesses, whatever the industry, to attract and have long-term relationships with customers and to be competitive. This is also true for banks and banks managers are encouraging their staff to meet their customers’ expectations in a highly customised and responsive manner. Such a behaviour enables any bank to achieve a higher level of customer satisfaction. So, satisfied customers buy more, are in contact more frequently with bank, undertake cross-selling and provide positive WOM communication. Thus, attracting, keeping and enhancing the customer relationship is crucial for the ongoing success of a bank. As discussed previously, service failures are inevitable. Identifying service failures, handling complaints in a fair manner and providing satisfactory service recovery is vital for a sustainable competitive position. In order to achieve these, banks should closely follow up how well they satisfy the expectations of their customers through their services and they must try to provide services that will better satisfy their customers with respect to their rivals. This study’s results are expected to contribute ideas to bank managers when setting up their strategies related to the handling of customers’ complaints and the importance and impact of various justice dimensions on customers’ service recovery satisfaction and their further behaviour. Another important contribution of this study is, since it was applied on a sample with a diverse nature, that the results allow bank managers to predict possible effects of their strategies on a global basis.

Disclosure statement

No potential conflict of interest was reported by the author.