Key Determinants of Service Quality in Retail Banking Evangelos Tsoukatos - Evmorfia Mastrojianni
1. build a retail-banking specific service quality scale, examine its item and factorial structure, asses its reliability and validity 2. compare the scale against metrics that are currently used in banking 3. deepen understanding of service quality determinants and provide input to the ever persisting debate over generic against setting/industry/time specific service quality metrics With respect to practice, the better service firms (in this case banks) are equipped to appraise service delivery performance the more effective they become in monitoring quality strategies and assessing their impact. 2
This line of business is neglected by bankers with respect to faceto-face service. For a long number of years now banks invest heavily on automated means of retail service delivery However, to the eyes of customers, retail and corporate alike, the unique selling proposition (Kotler, 1997) of a bank remains face-to-face banking. This is especially true in times of financial crises, when customers trust to banking institutions is put to danger. Yet, customer retention in banking depends on trust and confidence Customers need to be assured as they consider safety no more guaranteed by regulatory control. 3
Service quality is a measure of how well the service level delivered matches customer expectations. Delivering quality service means conforming to customer expectations on a consistent basis Service quality is the result of the comparison between delivered and expected service performance. Customers perceive the relative inferiority/superiority of services by comparing a firm s actual performance with their expectations, shaped by experience and/or memories. The result of this comparison is perceived service quality related but not equivalent to satisfaction of the overall excellence or superiority of a service. 4
In a state of market complexity it is imperative for banks to achieve customer longevity which can only be accomplished through service excellence. Despite service superiority s importance, the banking industry is short of a bank-specific widely recognised instrument for service quality assessment Quality studies in traditional face-to-face banking have mainly adopted the five dimensional SERVQUAL/SERVPERF BSQ: Bahia and Nantel (2000) proposed BSQ as an alternative to SERVQUAL although they recognised their study s main limitation, notably its implementation only in the French language. BSQ was subsequently used in numerous occasions including some settings in Greece SYSTRA-SQ: Aldlaigan and Buttle (2002) proposed that customers evaluate SQ at two levels: organisational and transactional and reported that the parsimony, reliability and validity of SYSTRA-SQ suggest that the measure is of high utility to the banking industry. 5
A five sections research instrument demographics, service expectations, service performance, overall satisfaction and word-of-mouth communication Service expectations and performance scores were measured in identical 7- point Likert scales as was overall satisfaction and intentions to recommend the bank to friends and relatives.. 6
The battery of service attributes was built through reviewing the literature and conducting two focus groups of banking customers and employees. The first stage, literature review, produced an extensive list of 70 items from SERVQUAL, customized SERVQUAL and bank-specific service quality measures such as BSQ. Focus groups discussions reduced the list down to 31 items that were incorporated in the research instrument after being several times translated back and forth from Greek to English for ensuring functional equivalence. 7
Data was collected in early spring 2008, from a convenience sample of n=91 retail-banking customers, in three downtown Athens branches of the two leading Greek banks. Respondents were adults, residing in Greece but not necessarily Greek nationals, consumers of both traditional and automated retail banking services who had at least one face-to-face banking transaction during the last month. Convenience sampling is very common in service-quality/customer-satisfaction research mainly due to random sampling requirements for population homogeneity, hardly met in practice, and high costs associated with locating chosen population items The major weakness of convenience sampling is that it does not provide any built-in means of eliminating or assessing sampling bias. There has been no evidence, however, that the aforementioned sample deviates in any respect from the general population regarding customers attitudes towards retail banking. 8
The research instrument was administered through personal interviews conducted on the spot inside branch premises. T o minimise bias, caused by poor service, prospective respondents were approached and interviewed prior to conducting their intended transactions. Despite the questionnaire s complexity the response rate reached 50.27%. The method of personal interviews is superior to self-administered questionnaires in perceptual or attitudinal surveys, while face-to-face administration maximizes response rates and field researchers availability to answer respondents questions 9
Performance-minus-expectations scores (Q i = P i E i ) of service attributes were appropriately analysed for: o a) initial scale purification, o b) unveiling the metric s underlying structure and o c) assessing its reliability and validity. Statistical treatment involved Cronbach s alpha Reliability Analysis, Exploratory Factor Analysis and Linear Regression Analysis. All analyses were executed by using the SPSS for Windows statistical package. 10
Table 1. Rotated Component Matrix(a) Assurance and Empathy 21.52% of Variance α=0.931 Overall α = 0.966 Effectiveness 20.82% of Variance α=0.942 Understanding customers individual needs.823 Understanding customers individual goals.788 Having customers interest at heart.712 Easy access to service personnel.705 Employees instilling confidence in customers.665 Courteousness.654.581 Understanding customers problems.645 Avoid technical jargon when talking to customers.574 Innovative products and services.741 Error free statements, bills etc..738 Prompt Service.670 Employees knowing exactly what they re doing.666 Employees Professionalism.659 Inform customers exactly when they will be served.655 Component Reliability 16.44% of Variance α=0.894 Confidence 13.99% of Variance α=0.896 Employees well trained in using technology.648 High quality of services.611 Clear terms in contracts.796 Full range of products and services.761 Competitive Pricing.707 Keeping time promises.668 Respond honestly to customers requirements.647 Doing the service right the first time.559 Customers confidence to documents, statements etc.869 Secure filling systems.836 Safe use of alternative service channels.674 Customers feeling safe in their transactions.593 Customers confidence to the service.572 Extraction Method: Principal Component Analysis. Rotation Method: Varimax with Kaiser Normalization. a Rotation converged in 9 iterations. 11
The alpha scores of 0.931, 0.942, 0.894 and 0.896 for the dimensions Assurance/Empathy, Effectiveness, Reliability and Confidence respectively indicate high internal consistency of the BANQUAL-R measure. Construct validity is secured on the grounds that the battery of service attributes comes from theoretically well supported literature sources including SERVQUAL and others that have been used in a multitude of settings around the world. The nomological/predictive validity of the scale is assessed by examining the association of service quality with other constructs to which SQ is theoretically related. In this case Customer Satisfaction (CS) and Word of Mouth communication (WOM) Both regression models produced statistically significant adjusted R-square scores of 0.529 and 0.362 respectively, indicating that SQ scores, produced by BANQUAL-R can indeed predict CS and WOM as theory suggests 12
The 27-item BANQUAL-R scale consists of 12 SERVQUAL, 7 BSQ, 2 common in SERVQUAl and BSQ and 6 setting-specific items. In this respect, the metric is a hybrid of the SERVQUAL and BSQ scales. T he factorial structure of BANQUAL-R consists of SERVQUAL s Empathy, and Assurance, BSQ s Effectiveness, Reliability which is common in SERVQUAL and BSQ and finally Confidence. The latter is mainly associated with service innovation leading to increased intangibility of previously tangible service components such as physical records, archives etc. Assurance/Empathy and Effectiveness are the primary and secondary factors respectively (explaining very similar variance percentages) while Reliability and Confidence are in the third and last positions, explaining 16.44% and 13.99% of variance. 13
In the majority of SERVQUAL applications Reliability is found to be the most important of dimensions, interchangeably followed by Responsiveness and Assurance while Empathy is usually more important only from Tangibles. Yet, in this study Assurance/Empathy is the most important service element closely followed by Effectiveness (also including certain Responsiveness items), while Reliability is in the third position. In the BSQ study Effectiveness was found to be the most important dimension, followed by Assurance, Access, Price, Tangibles, Service Portfolio and Reliability. It is clear that the dimensionality of BANQUAL-R indicates a different set of priorities than those mentioned by the SERVQUAL and BSQ studies. 14
The importance ranking of BANQUAL-R dimensions indicates that what retail customers need more is personal attention and human contact both instilling trust and confidence in customers. This is inconsistent with both the majority of SERVQUAL studies and the BSQ study in which Effectiveness was found to be the most important of dimensions. The first position of Assurance/Empathy in this study is consistent with previous research findings that in financial crises situations, such as the one at the core of which we currently are, customers trust to the service is put to danger and needs to be reinstated as safety is no more taken for granted. The positioning of Effectiveness in the second position, with similar importance to that of Assurance/Empathy, provides support to Bahia and Nantel s (2001) similar argument and can be interpreted as a banking-specific feature of service quality. Regarding Reliability, there is no solid evidence on why it is so low in importance in both the BSQ and the BANQUAL-R studies. A plausible explanation might be that banks are by-definition considered reliable and so, Reliability is taken for granted. Similarly, the last position of Confidence implies that customers have finally overcome their concerns regarding the use of ICT in banking transactions.