1 Evaluation of customer satisfaction with services of a Micro-finance Institution : Empirical Evidence from WAGES’ customers in Togo Eddy BALEMBA Kanyurhi1 1 PhD Fellow and Senior Lecturer, Department of Finance and Management , Catholic University of Bukavu (DR Congo), P.0.Box 02 Cyangu, Rwanda , Tel : + 243 997 79 6868, + 243 85 3131 577, Email : [email protected] . He detains a Complementary master degree from European Microfinance Programme since 2009. The author is grateful towards professors Aad Van Tilburg and Déogratias Buganda for helpful comments they provided him to improve this paper’s quality. 2 Abstract. Nowadays, Micro-finance industry is affected by competition leading to customers’ switching across MFIS. Thus, MFIs are concerned about customer satisfaction and have to pay attention to understand their customers’ preferences to survive in a competitive environment. Using data from 353 Women and Association for Gain both Economic and Social (WAGES)’s customers in Togo, this study aimed to : (1) determine the main dimensions of financial services of Microfinance institution (2) determine the customer satisfaction level and (3) assess to what extent customer satisfaction is influenced by customers’ characteristics. Resorting to Factorial analysis, Customer Satisfaction Index and ANOVA, we found that responsiveness remains the most important dimension in micro-finance sector. Results reveal that customer’s branch, customer’s revenue and number of services accessed by customers strongly influence customer’s satisfaction. Results also indicate that the current customer satisfaction is equals to 71.2%. Key words: Micro-finance customers’ satisfaction dimensions, customer’s satisfaction Index, customer’ characteristics influence. 3 1. Introduction Microfinance industry is now affected by strong competition : “ commercial banks have begun to target MFIs’ traditional customers , new MFIs have continued to be created in microfinance industry, the microfinance clientele is becoming more sophisticated concerning the quality of service they require or expect”( Daubert 2002) . These factors may negatively affect the MFIs. In fact, the microfinance industry is losing customers because of both the aggressive competition and MFIs’ weakness to satisfy their clients (Urguizo 2006). This simple description shows why MFIs are concerned about customer satisfaction and retention. It justifies also why they must “pay attention to understand their customers’ preferences and priorities” (IFAD 2007) to survive in a competitive environment. The microfinance industry is quite slowly in becoming more “market oriented” and it seems that customer satisfaction is one of the important tools to run a business and to achieve the mission statement (on sustainability and outreach) in this sector. Customer satisfaction is an evaluative process, it is defined as “… a judgment that a product of service feature, or the product or service itself, provided (or is providing) a pleasurable level of consumption related fulfillment, including levels of under or over fulfillment” (Oliver 1997, 13) cited by ( Swaid 2007; Hom 2002). Customer satisfaction is “captured as positive feeling (satisfaction), indifference or negative feelings (dissatisfaction)” (Bhattacherjee, 2001) cited by (Swaid 2007). It is a short-term attitude that can readily change given a constellation of circumstances. Therefore, satisfaction is not a static idea and it changes as soon as a client finds a better deal that meets his expectations. In this perspective, firms must focus on customer satisfaction, studying and determining as soon as possible the customer satisfaction level, to adjust the product to customer needs. Indeed, customer satisfaction has great significance for the future of an institution and it is seen as a basis for securing market position and achieving other objectives of the institution (Koraus , 2002). This study focuses on customer satisfaction with respect to the services of Micro-finance institutions in Togo. Data have been extracted from one MFI (the Women Association and Gain for both Economic and Social: WAGES) for two reasons: (1) WAGES is the second biggest MFI in Togo; launched in 1994, it grew rapidly serving nowadays more than 64.710 4 clients divided in eight branches located in the Togolese’ biggest towns. Thus, studying WAGES’ case will provide an exhaustive insight on MFI’customer satisfaction issue in Togo. (2) WAGES is one of the “market oriented” MFIs with a well established marketing department aiming to develop new products on clients’ needs and competitors’ strategic actions basis. Thus, it was easy to access to WAGES’ database and installations what was impossible for other institutions mainly characterized by opaque structures. Togolese’MFI sector experienced big competition since 2007; active MFIs increased in number reaching 167 institutions (Togolese’s MFI network Report, 2007). The multiplication of financial institutions (MFIs and commercial banks), a weak diversification of products and the lack of entrance barriers (MFI sector) are among elements explaining hard competition in the sector. While many MFIs have been concerned by competition, WAGES has been highly affected losing almost 5 % of active clients. WAGES’ marketing department positively reacted by conducting a customer satisfaction survey in June 2008 aiming to indentify clients’ needs and complaints for defining management implications for operational and strategic actions. Using WAGES’ survey outcome, this study aims to determine MFIs’ customer satisfaction level in Togo. The objective is to determine the current customer satisfaction level, to understand the main dimensions of service from the customer point of view and, at the end, to assess if the customer satisfaction is influenced by some characteristics of customers like: number of services obtained, the location of customer, the age, the sex (gender), the school level and number of years spent as customer in the MFI. Customer satisfaction is a subject with a lot of interest in both the marketing and finance literature. The great emphasis on customer satisfaction has given birth to multiple studies and innovative methodologies to assess and to understand customer behavior. Parasuraman et al.1985, 1988) are among the most well-known researchers who assessed customer satisfaction using service quality (ServQuality) model. Today, ServQuality model is used, with few adaptations, in multiple sectors to assess both service perceptions and expectations “across a range of different service characteristics” (European Commission, 2008). The original model is adapted in function of the researchers’ needs and “moved” from the non financial sector to the financial sector. Therefore, it is feasible to assess customer satisfaction also in the financial sector. Thus, “the BANKSERV model was developed in Australia to measure service quality in retail banking as perceived by customers. It was adapted from SERVQUAL to specifically suit the Australian banking industry”(Pont and McQuilken 2002). 5 The CARTER model “was conceptualized as a proposed framework for measuring quality of services in Islamic banks (Othman and Owen 2001), etc. Innovative, structured and formal models to assess customer satisfaction in Microfinance are still rare. However, some studies tried to formalize the main items of satisfaction that can be measured in this particular field. Murray (2001) investigated MFI customer’s value through a comparative analysis from three continents. His results show that “MFI customers will continually push for higher loan amounts, faster turnaround times, lower loan requirements and lower prices”. This study takes inspiration from previous studies and tries to adapt the ServQual model to the micro-finance field. It uses a sample of 353 respondents from WAGES by a stratified and reasoning sample choice. Data were obtained by a survey in three stages: the first stage is a focus group. The objective assigned to it is to gather the most important dimensions of service from the customer’s point of view and these dimensions will be introduced in the questionnaire to be able to use a quantitative methodology. The second stage was a presurvey. In this case, the questionnaire has been proportionally submitted to 30 customers in function of number of customers by branch (Giannelloni and Vernette 2001). The third stage is the proper survey. A questionnaire, inspired by previous studies and customer answers, has been used to gather expectations and perceived service by customers. Data processing has been done by resorting to factorial analysis, Anova test and customer satisfaction Index. This study is divided in five sections. The first section presents the literature review. We define customer satisfaction, discuss the importance of customer satisfaction in relation of retention and loyalty, present a short review of measurement tools of customer satisfaction and provide some results related to previous studies. The second section focuses on methodology and hypothesis. Some techniques related to information gathering and processing are presented (focus group, pre-survey and sampling, survey, factorial analysis, Anova test, customer satisfaction Index). The third section presents data. Data relate to customers’ characteristics, different items constituting the customer satisfaction and WAGES’ financial products. The fourth section discusses results providing main dimensions of customer satisfaction, the current customer satisfaction level and assesses if customer satisfaction is related to their characteristics. The fifth section raises limitations, implications and future research from this study. Following section presents literature review. 6 2. Literature review We review customer satisfaction theories and provide some empirical results related to customer satisfaction in Micro-finance and Finance sectors (2). 2.1. Theoretical review. We define customer satisfaction, discuss the importance of customer satisfaction in relationship with retention and loyalty and present a short review of measurement tools of customer satisfaction. 2.1.1. Definition of customer satisfaction. According to existing definitions and approaches, customer satisfaction can be analyzed as a general/overall judgment that a customer makes after consuming a product or a service. Customer satisfaction is perceived as “psychological state (feeling) appearing after buying and consuming a product or service …” (Lendrevie and Lindon 1997) cited by (Merouane 2008/2009). Thus, customer satisfaction reflects “a pleasure resulting to product’s consumption, including under or over fulfilment level” (Oliver 1997, 13) cited by (Hom, 2002). According to Olivier’s argument , customer satisfaction does not mean only positive feeling, it could also lead to a negative or neutral feeling withdrew from consuming a product or a service. Briefly, “customer satisfaction is captured as positive feeling (satisfaction), indifference (neutral), or negative feelings (dissatisfaction)” (Bhattacherjee 2001) cited by (Swaid and Wigand 2007, Hom 2002). When approaching customer satisfaction as a feeling, it is important to note that it is mostly influenced by the customer’s experience with the firm and product. In this perspective, customer satisfaction is conceived as “the emotional state that occurs as a result of a customer’s interactions with the firm over time” (Anderson et al. 1994) Cited by Verhoef (2003, 32). In fact, customers are usually comparing the product received from the firm to their own expectations over time. If the product fulfils and performs the customer’s expectations, customer seems satisfied. Briefly, customer’s satisfaction is analyzed as a confirmation or not of customer expectations (Conchon et al. 2006). 7 Clearly, it seems that customer satisfaction is composed by “two components: client expectations and the perceived quality. Thus, a proper measure of satisfaction would include a separate assessment of both client expectations and the quality of provided service” (Office of the Comptroller General Evaluation and Audit Branch 1991). Parassuraman et al .1985, 1988, 1991) approached customer satisfaction in the same way by demonstrating that customer satisfaction is a function of “the difference scores or gaps between expectations and perceptions (P – E)”. According to them, customer satisfaction is only achieved “if actual perceived quality surpasses the consumer’s expectations”. Even if the Parasuraman et al. 1985 customer definition seems to be more dominant, it is now more criticized because of practical problems related to the gap “performance minus expectations” (Teas 1994, 132). Thus, an alternative measurement of customer satisfaction has been proposed estimating that customer satisfaction would be only obtained by focusing on actual perceived satisfaction (Corin and Taylor 1992; Teas 1994). In this perspective and contrary to Parasuraman et al’s approach, “Customer satisfaction is defined as customer’s overall evaluation of the performance of an offering to date (Jonhnson and Fornell 1991) cited by (Gustafsson et al.2005, 210). 2.1.2. Importance of customer satisfaction in firms: retention and loyalty. Loyalty and retention are often analyzed as direct consequences of customer satisfaction. The two terms express “the attachment a customer feels for a company’s people, products, and services. A loyal customer is someone who makes regular purchases, purchases across product and service lines, refers others, demonstrates an immunity to the pull of the competition” Griffin (1995) cited by ( Churchill 2002) . Financial and marketing studies have supposed that satisfied customers constitute an important asset of firm. Even if there is not much empirical evidence, it seems that customer satisfaction will enhance both customer loyalty and retention through repeated purchases, less price sensitivity and costs reduction. In fact, when customers are satisfied , they become more loyal and will increase their level of purchasing from the firm over time ( Anderson and Sullivan 1993, Reichheld 1996) cited by (Verhoef 2003, 33) directly, they will also recommend other customers to consume the firm’s products and services . Thus, “the positive word of mouth that satisfied customers generate influences other consumers’ future purchases” (Anderson 1996) cited by ( Gruca 8 and Lopo 2005, 116). Satisfied customers are also expected to be “less likely to defect to competing products as a result of lower prices” (Fornell et al. 1996). For this reason, “greater customer satisfaction may enable firm to charge higher prices or at least to better resist downward pressure on prices” (Anderson 1996 , Narayandas 1998) . Briefly , “ a satisfied customer reacts less sensitively to price changes and is prepared to pay a higher price for a service that corresponds to their requirements and conceived ideas” ( Korauš 2002, Anderson et al .2004 , 172). Customer satisfaction will also exert a positive impact on firm’s costs through retention. Indeed, by satisfying customers, firms and MFIs will lower their actual costs avoiding gaining new clients and making extra marketing expenses. This argument is more important in microfinance sector where clients and the MFI are supposed to act for a long-term relationship. In fact, “ with client retention, institutional costs decrease as the institution needs to do less marketing, less new client orientation, and fewer new client background checks, staff productivity increases because loan officers work with established clients whom they know well, clients income increases as loan sizes generally increase with experienced clients” ( Waterfield 2006; Korauš 2002). Brief, by satisfying their customers, “firms and MFIs will generate benefit for themselves beyond the present transaction and the current moment. These benefits will arise from the positive shaping of the satisfied customer’s future behaviour” (Anderson 1996) cited by (Gruca and Lopo 2005, 116). It means that if firms and MFIs in particular, “are able to meet customers’ needs successfully, then, household stability will increase. Customer household stability will in turn contribute to organization’s financial sustainability” (Asian Institute of Management 2005). 1.1.3. Evolution of customer satisfaction measurement There are many techniques and methods for measuring customer satisfaction. We will not review all existing methods. We will limit our attention to representative methods like ServQual, ServPerf and some adapted methods resulting to ServQual model. a) Service quality model: The ServQual model is considered as the pioneer model in customer satisfaction measurement. Developed by Parasuraman et al.1985, the model has been recognized as the most representative tool in approaching 9 customer satisfaction issues. The central idea is that service quality is “ a function of the difference scores or gaps between expectations and perceptions (P – E)”. “SERVQUAL contains 22 pairs Likert scale statements structured around five service quality dimensions in order to measure service quality (Cronin and Taylor 1992) : Reliability, Responsiveness, Assurance , Empathy , Tangibles )” (Bloemer , Ruyter et al. 1999). In this perspective, customer satisfaction is analyzed as multidimensional concept resulting from a comparative approach between customer’s expectations and perceived quality delivered by the firm (cf. Parasuraman et al. 1985). Thus, “a positive gap score implies that expectation have been met or exceeded and a negative score implies that expectations are not being met” (Parasuraman at al. 1988) cited by (Safakli, nd, Barnes 2005, Parasuraman, et al.1985). Now, ServQual model is analyzed and modified by some authors seeking to adapt it or to correct some mistakes it may be perceived to contain. Rethinking ServQuality has given birth to multiple others models among them ServPerf which we analyze below. b) The SERVPERF model: This model had been developed by Cronin and Taylor 1992 from ServQual model basis. The fundamental criticism launched to ServQual model by Cronin, Taylor and other authors like Teas concerns the gap scores (PE). In fact, those authors estimated that “there are serious problems in conceptualizing service quality as a difference score ….” (Cronin and Taylor 1992, 26). Thus , ServPerf model suggests “ that customer satisfaction with service is based only on “performance” rather than a gap between performance and expectations, with the performance-only scale termed SERVPERF” (Cronin & Taylor 1992, 1994) cited by ( Lowndes 2001). However, SERVPERF model is composed of the same 22 perception items included in SERVQUAL. “It excludes any consideration of expectation, which makes SERVPERF a more efficient measure in comparison to SERVQUAL (Lee, Lee and Yoo 2000; Buttle, 1996)” cited by (ANZMAC Conference Proceedings 2002). Empirical studies have confirmed a relative superiority of ServPerf to ServQual models. Thus, PZB 1994 argued that “ServPerf has greater construct validity and that ServPerf measures also exhibit convergent and discriminant validity” (Cronin and Taylor 1992). c) Adapted models from ServQual model: Today, ServQual model is adapted from marketing to finance sector. The principal ServQual model has been changed in 10 function of authors’ research and interest. Many modifications are have been done by diminishing or adding some items or dimensions to the original model. Thus, for example , the PAKSERV1 model is using “ SERVQUAL dimensions of tangibility, reliability and assurance but replaced the responsiveness and empathy dimensions with three new dimensions: Sincerity( consumer’s evaluation of the genuineness of the service personnel), Formality( consumer’s evaluation of social distance, form of address and ritual) and Personalization( consumer’s evaluation of individualize and individualized attention)” ( Saunders nd) . The BANKSERV2 model adopts a ‘perception-expectation’ approach to the measurement of service quality. “The model contains 17 items regrouped in four main factors: staff conduct, credibility, communication, access to Teller Services” (Pont and McQuilken 2002) for assessing customer satisfaction in bank sector in Australia . The CARTER model was been developed to adapt ServQual to Muslim culture enabling it to measure customer satisfaction in Islamic banks. “The model makes assumption that the cultural and religious influences were significantly rated and placed in the front by Islamic banks customers. CARTER model includes 6 dimensions with 34 items. It includes in addition to the compliance with Islamic law and principles all SERVQUAL five dimensions” (Othman et al. 2001). 2.1. Some empirical studies. IFAD (2007) studied customer satisfaction in rural micro-finance institutions in Uganda, Kenya and Tanzania. Combining qualitative (14 focus group of 71 clients) and quantitative approaches (209 interviews), this study assessed the determinants of customer satisfaction for rural customers accessing both credit and savings facilities. Results revealed that “ customers prefer unlimited access to their savings while on credit facilities, customers want to have access to loan amounts they actually apply for at a ‘reasonable’ price and on flexible repayment term conditions” . The study suggested also that surveyed customers were all satisfied exhibiting a Customer Satisfaction Index of 81%. The study concluded that “financial services should be delivered by courteous staffs that preferably are not being ‘changed /swapped”. 11 Murray (2001) concentrated his study on customer satisfaction levels using data from four MFIs affiliated to Women’s World Banking in three countries: Colombia (America), Bangladesh (Asia) and Uganda (Africa) with a total sample of 3,000 clients. Using Likert’s scale, the author took into account expectations and perceptions items plotting results on a two-axis grid. Results proved that customers are more satisfied by accessing higher loan amounts, faster turnaround times, lower loan requirements and lower prices. However, it seemed that customers preferring to develop a long-term relationship with the MFI want to be given preferential treatment while all customers are demanding increasing levels of customer service. Othman and Owen (2001) conducted a study about customer satisfaction in Islamic Banks by using the service quality model. Their study used a survey of 360 customers selected by Systematic Random Sampling. Using CARTER model scale, their results suggested that customer satisfaction in Islamic banks “should be measured through the proposed 34 items instead of reducing it into the original number of SERVQUAL’s five dimensions and their 22 items”. Their results indicated that in Islamic banks, managers and practitioners should be aware of cultural or religious dimension. Alhemound (2007) investigated customer satisfaction in the banking sector in Kuwait. His study used a sample of 605 randomly selected retail customers. Using descriptive statistics, Correlation and ANOVA tests, his results showed that, in general, customers in Kuwait are satisfied with services provided by retail banks. In this regard, customer satisfaction is mainly driven by: “availability of ATM in several locations, safety of funds, easy to use ATM and the quality of services provided….” Table n°1: Summary of results from empirical studies N° 1 Authors Sampling and Data Main results Ifad ( 2007) 280 customers from Uganda, Tanzania 81 % of customers are satisfied by services. and Kenya using qualitative (focus Unlimited access to savings, access to loan groups) and quantitative interviewers, amounts at a ‘reasonable’ price and on recurring to Customer satisfaction Index flexible repayment term conditions are (CSI) . significant dimensions of customer satisfaction. 12 2 Murray (2001) A total of 3000 customers were been Customers are more satisfied by : higher selected from three countries: loan amount, faster turnaround times , lower Colombia, Bangladesh and Uganda. loan requirements and lower prices The author used graphic analysis for results processing. 3 Othman and A sample of 360 customers selected by Customer satisfaction in Islamic banks is Owen (2001) the Systematic Random Sampling (SRS) defined towards proposed 34 items among system Was used while CARTER model cultural / religious dimensions are more was been used for data processing important. through factorial analysis. 4 Alhemound (2007) A sample of 605 retail customers Generally, customers are satisfied. Customer randomly selected. Statistics descriptive satisfaction is more influenced by the (Mean, Standard Deviation) and customer’s locations (Kuwaiti and non statistics tests used Correlation, ANOVA) processing. (Person’s Kuwaiti). for data Source: Built by ourselves from empirical studies material. 3. Methodology and hypotheses definition. We successively analyze techniques used to gather and process data. 3.1. Data collection techniques: focus groups, sample and survey. We firstly conducted 11 focus groups with 118 participants from eight branches. The objective was to gather representative attributes from the customers’ point of view. Using three guide questions and two criteria for meetings’ synthesis, we obtained 15 attributes as being the most important for clients (Netteret and Nigel Hill, 2005). We secondly conducted a pre-survey on 30 clients for sampling definition and questionnaire testing purpose. Clients have been proportionally extracted from the 64.710 active WAGES clients. This (M= 3.88; Standard deviation = 0.49) allowed us to determine the sample’ size: (0.49) n= [(1.96)2x 2 ] /(0.05)2 = 369 customers . Through a pre-survey, we also modified the questionnaire by canceling ambiguous questions and changing the syntax of badly formulated questions for ensuring best understanding by clients (Niraj et al., 2003). We resorted to a proportionally 13 stratified and reasoning sampling basis. The 369 clients have been extracted from 6 branches focusing on both time spent as WAGES’ clients (would be regular MFI’s member one year ago) and financial product (benefiting at least one of three financial products: loans, savings and Roscas). Data have been collected by a team composed of 24 loans officers’ during 21 days. From the 369 distributed surveys, 353 have been returned and were well completed, representing a 96 % response rate. Data have been collected through a questionnaire adapted from previous studies (Parasuraman et al. (1988), European Union (2007); Wetzels (1997), existing questionnaire at WAGES and focus interviews. The questionnaire contained three principal sections. The first section was designed to gather information about both customers’ expectations and perceptions. This section includes 32 questions. Customers’ expectations and perceptions have been gathered using a five point Likert scale ranging from “strongly agree =5” to “strongly disagree = 1”. For this section, “ the statements were administered to the respondents with the following instructions: We would like you to put yourself in place of an MFI customer and then respond to all of the following statements by checking the category which best reflects your opinion” ( Yavas 2006). Second section has been concentrated on identification of customer satisfaction level for WAGES’ specific products like savings, loans, Roscas, transfer, formation, etc. We used a performance Likert scale ranging from “Strongly satisfied=5” to “strongly dissatisfied =1” for measuring 18 elements which have been identified with the programme and marketing manager as the most important for the institution. The third section was designed for customers’ identification. Extracted variables were related to customers’ age, sex, location in terms of branch, number of years spent as customers, financial services received from Wages, business status, enrolment status and customers’ revenues. 3. 2. Data processing techniques. Three main techniques have been used for processing data. We firstly resorted to factorial analysis to “satisfy the need of identifying structure through data summarization and data reduction” allowing us to define customer satisfaction dimensions. Using this technique, we were able to “condense (summarize) the information contained in a number of original variables (items) into a smaller set of new, composite dimensions or factors with a minimum 14 loss of information” (Hair, Anderson, Tatham, & Black, 1998 cited by Sarreal ,2008). We used Parasuraman et al. (1988) scale measurement as reference . “The original scale had 22 item scales built on five dimensions. However, even if the “ 22 items are inclusive enough to cover the general service quality issues in various service businesses including bank service, they may not be specific enough to understand the context of specific quality concerns and their priorities in the minds of bank customers. Therefore, the blind use of the 22-item SERVQUAL instrument may limit the accuracy in understanding service quality terms of customers in a specific service business” (Lee, 2006) like MFIs. We modified the original model to adapt it to MFI sector and African context. Verhoef (2003) had proceeded in the same way by “adding four new items on Signh’s (1990) scale to fit the context of financial services”. Adopting Verhoef’s (2003) approach, referring to focus groups results ( aiming to gather clients points of view) and the existing questionnaire, we developed an embedded scale with 32 item-scales and six principal dimensions: Tangibles, comfort and appearance of personal and material (6), Reliability or capacity to accomplish task (4), Responsiveness, dynamism and willingness for helping customers (6), Insurance and confidence (5), Empathy and attention to clients (5); Price, costs and conditions (6). Although the five first dimensions are similar to the Parasuraman scale in terms of structure, the items’ content is deeply different form the original scale trying to incorporate and meet sector (MIFI) and context (Africa) matter. Thus, the tangibles dimension give insight about particular items related to loan officers’ using motorcycle and badge as tangibles items facilitating loans officers’ mobility and identification on field. The sixth dimension is a new one added to the original scale highlighting how do microfinance clients are strongly concerned by services pricing. We captured both customers’ expectations and performance but only performance items have been submitted to factor analysis (Corin and Taylor, 1992; Teas, 1994, HarrisonWalker, 2000) using structural coefficients (.30) , communalities (equal or above .50) , eigen values ( equal or above 1) , explained % of variance ( at least 60% ) , Cronbach’s Alpha ( ≥ .70) for obtaining optimal solution (Malhotra et al 2007; ; Carricano and Poujol, 2008; Ahmad and Sungip, 2008). H 1: Responsiveness, Empathy, Price, conditions and costs are expected to exhibit high level of extracted variance proving they are the most representative dimensions of customer satisfaction for WAGES’ clients. 15 We secondly resorted to customer satisfaction Index for determining WAGES’current customer satisfaction level expressed “as a single number in percentage that tells the supplier where he stands today…” (Bhave, 2002) using both customers’ expectations and customers’ performance (Kumar and Mahaptra, 2006). We obtained WAGES’ actual customers satisfaction “using an importance weighting based on an average of 1” inspired from both Bhave (2002), Netteret and Nigel Hill 2005 following three stages: First, “we calculated the average of all the weightings given by the customer. Second, we divided the individual weightings by this average to find the weighting on the basis of average of 1. Customer's higher priorities are weighted more than 1 and lower priorities less than 1. The averages of the Customers Importance Scores are calculated and each individual score is expressed as a factor of that average” (Kumar and Mahaptra 2006). Third, we expressed the average in percentage terms and obtained the actual customer satisfaction level. H 2: WAGES’ current customer satisfaction level is high , it also is a function of financial services (saving and credit). We thirty resorted to ANOVA test to assess whether one or a combination of customers’ qualitative characteristics (location, level of education, number of financial services accessed, customer’s revenue, etc) would have an effect on customer satisfaction measured as quantitative variable from significant factors and items after scale purification. We processed by preliminary test for ensuring ANOVA usage (Levene test ˃ .05 and Duncan test for identifying the means that were significantly different”) and chose significant variables by examining their attached probabilities (p-value equal or less than .05 used as a thumb to judge the relevance of statistics tests at five percent level of significant). H3: WAGES’ customers’ satisfaction is positively influenced by customers’ characteristics such customers’ branch, customer’ revenue, the number of financial services accessed by clients. 16 4. Data presentation. We discuss customers’ characteristics (1) and we provide insights related to customers’ expectations and perceptions (2). 4.1. Social, economical and demographical characteristics of the Sample. The majority of surveyed clients are women representing 64% of the sample while men constitute only 36 %. Surveyed customers are adults with an average age of 37 years. Surveyed clients achieved at least primary school and secondary for 70 % of cases since 5% have non graduated skills and trainings in various fields including sawed machine, joiner’s workshop, etc. 71 % of clients are active in a small business (food, clothes selling) whereas 25% own a small and informal firm producing mainly services like hair –cutting , restaurant , sewed units , etc . The importance of commercial and services activities can be explained by a rapid rotation cycle enabling clients to invest and earn money in short period for affording MFI’s weekly or monthly payment. 82 % of surveyed clients contract an individual loan whereas 18 % are member of a group lending. In average, surveyed customers are MFI’s clients since 3 years and four months. 60 % of clients earn monthly income above 43 Euro, the SMIG defined by the Togolese’s government. Only 10 % of Surveyed clients are poor in the sense of the above criterion. This situation is mainly due to the fact that the MFI is now mostly focused on individual clients. Is there a mission drift? 4.2. Customers’ expectations and perceptions evaluations. Data from customers’ expectations prove that the average expectation score is 4.55 for all customers. So, three principal dimensions appear to be of value for customers. These are: (1) Tangibles, (2) Price, costs and conditions and (3) Insurance and confidence. All three dimensions are given high average scores of 4.57 and 4.66, which are above 4.55 the total average score for the all sample. Customers weakly rated the three remained dimensions comparing to total average notes. Thus, they attributed an average score of 4.52 to Reliability, 4.50 to Responsiveness and 4.47 average score to Empathy dimensions which are all below 17 the total average for all dimensions (4.55). Briefly, surveyed customers are likely to value tangibles, Insurance and price dimensions; average importance to reliability and responsiveness dimensions whereas they exhibit less importance for empathy dimension. The insurance and confidence dimension seems to be the less important one on client’s point of view. Data related to customers’ perception exhibit an average perception score of 3.74. Three dimensions are highly appreciated by the sample: Insurance and confidence (1), Tangibles (2) and Responsiveness (3). In fact, the average perception attached to those dimensions exceeds 3.74, the total average perception rate for the total sample (4; 3.88 and 3.87 ˃ 3.74). Customers weakly rated perception of three dimensions: Reliability (1), Empathy (2) and Price, costs and conditions (3) attributing them average perception scores below than the total perception average score for all dimensions (3.66; 3.49 and 3.52 ˂ 3. 74). Connecting customers’ expectations and perceptions, we noted that for all dimensions perceptions average scores are below expectations average notes. There is then a gap between customers’ perceptions and expectations indicating that current services are not meeting customers’ expectations ( 1 .05 for price, costs and conditions dimension and 0.63 for responsiveness dimension) . It is clear that we met rational and more demanding customers, who are able to state clearly what they expect from an MFI. 5. Results and discussion We present the main dimensions of customer’s satisfaction (1); we highlight results related to the current customer satisfaction level (2) and assess customers’ characteristics influence on customer’s satisfaction (3). 5.1. Determination of customer’s principal items and dimensions Results from preliminary tests are strongly significant (KMO=.906, Bartlett = approx chisquare=2,262, df =190, sign =.000) allowing us to apply factor analysis. Using an iterative approach, we obtained a factor structure with 20 items (Bressolles, 2006; Parasuraman et al, 1998, P 24). Table below summarizes the main results about factor structure. 18 Table 2: Factor structure with principal dimensions and items after rotation N° Significant items Dimensions Principal factors/Composantes 1 1 .587 willingness Item 15 .749 .579 Item 16 .814 .728 Item 17 .707 .607 Item 18 .734 .694 Item 19. .680 .608 Item 22 .677 .573 for helping and Item 2 .676 .541 appearance of personal and Item 3 .575 .509 material Item 4 .764 .683 Item 5 .665 .630 α=.883 Item 8 .592 .581 Conditions Item 27 .795 .675 Item 28 .615 .596 .612 .567 Tangibles, comfort Item 29 Costs and Price Item 30 .669 .580 α=.890 Item 31 .614 .632 .762 .699 Item 32 5 5 .653 α=.886 4 4 Item 12 α=.882 3 3 Responsiveness, dynamism and customers 2 2 Communalities Empathy and attention to clients α=.889 Cronbach’s Alpha=.890 Item 7 .664 .700 Item 25 .755 .648 Eigen Values 7.016 1.827 1.380 1.161 1.032 Explained 35.072 9.135 6.902 5.804 5.158 Percentage variance Source: Computations inspired from Bressolles (2006, p29). Reliability and validity analysis showed that overall 20 remained items belong to the same concept of customer satisfaction applied to MFI. In fact, the Cronbach’s Alpha is high equal to .890 ( > .70) indicating a very good reliability of scale measurement , attesting that our scale is measuring what it was supposed to do, according to Parasuraman et al. ( 1988 , p 28) who state: “ the scale items capture key facets of the unobersevable construct being 19 measured. The Cronbach’s Alpha after item deletion is also high (>.80) proving that it is not possible to remove one item for improving the final solution. The 20 remaining items are regrouped in five dimensions showing that customer satisfaction with service in an MFI is a multidimensional construct that allows explaining of 62. 1 % of variance . This result does not confirm our first hypothesis which predicted a factor structure with six principal factors corresponding to the basic scale model. This is due to the fact that a number of factors from the original model have disappeared (Reliability) and corresponding items have been integrated in other factors (tangibles and Empathy). Responsiveness and insurance dimensions have been combined to form one dimension (Responsiveness) whereas price, costs and conditions singular dimension engendered two separate dimensions (Parasuraman et al, 1998, P 23). The first factor is named “responsiveness, dynamism and willingness for helping customers” combining responsiveness and insurance aspects. This factor is the most important and is explaining 35.1 % of the total variability. This result confirms the importance of human relationship in MFI’context where people strongly interact. In fact, as consumers “do not clearly differentiate the interaction aspects of reliability, responsiveness and assurance ….” (Harrison –Walker, 2000, P45), they will judge and appreciate the MFI service quality through the quality of the interaction they have with the MFI’s employees. The more kind, polite and competent will be the employees, the more customers will appreciate the service they will receive reinforcing insurance and confidence they have in the MFI. Briefly, the human interaction between customers and MFI’s front off officers will impact more customers’ satisfaction and enhance loyalty and retention. In fact, experience has proved that “banks customers’ attitudes towards the human provision of services and subsequent level of satisfaction will impact on bank switching” (Othman and Owen, 2001). The second dimension refers to the appearance of MFI’s employees, buildings and equipments. This factor is explaining 9.1 % of the total variance. The appearance of this dimension is a surprising result since Micro-finance institutions provide proximity services. So, we could not expect customers to attach so much importance to such items as external appearance of individuals, buildings and equipments. However, focus group interviews had raised tangible aspects as important dimension on customers’ point of view explaining that 20 customers are not receiving services on time because loans officers’ logistic is very poor. In fact, customers desire to be served by employees who have their owner’s motorcycle, who can clearly be distinguished by a badge or a pullover marked by MFI’s signs. Motorcycles enable loan officers to attend clients’ house quickly for collecting savings and weekly installments. It also enhances loan officers’ mobility on field, increase their productivity and indirectly reduce transactions costs (time and money saved) leading to great satisfaction level. By well identifying the MFI’s employees, the Badge is an authenticity’s indicator improving confidence and insurance mainly for new customers who do not know all employers of MFI. Tangibles aspects reflect also the MFI’s ability to respect time table given to client providing them seats when they are waiting for service. Those two aspects are more important for clients who are working daily to perform their business (time save) and who come from far way the MFI. The third factor is banking factor reflecting conditions to be fitted for accessing MFI’s financial services. It is appealed “Loan conditions” explaining 6.9 % of the total variance. The appearance of this dimension highlights the importance that customers attach to the loan term, loan amount and the grace period. These results indicate that MFI’ customers will be more satisfied with a credit with a long term maturity, with grace period and with loan amount increasing for each loan cycle. Hence, the strategic dynamic incentives will enable clients to undertake projects with great impact while allowing MFI to increase its revenues and efficiency. In fact, when customers access to great loan amount and to other advantages such as grace period, they are likely to be loyal and unlikely to switch to another institution (Murray, 2001) The fourth factor can be named “costs of financial services” from the MFI explaining 5.8 % of the total variance. It is related to the importance that customers attach to guarantee, original fees and transaction costs due to the limitation of amount customers can access (loan) or can contribute in Rosca. This factor reveals that customers will be more satisfied if some costs or conditions are released. For example, customers with growing business will be penalized when they intend to develop a large business because they cannot withdraw an amount around 763 Euro without having a physical guarantee. The same statement would be done for poor people who are acting in Rosca. As contribution amount is fixed and invariable, they cannot contribute an amount below the standard. Consequently, as they are solicited by daily needs, 21 they will spend money for food putting them in contribution delay. This result suggests a reexamination of costs procedure for both customers and MFI effectiveness. The fifth factor is an empathy factor showing attention on clients in some particular situations explaining 5.2 % of the total variance. Although its percentage of variance is low, it remains significant for clients. Since focus on social impact of Micro-finance becomes more important, MFI must develop loans and savings collection methods that valorize clients’ human rights. In this perspective, aggressive loans and savings collecting methods might not produce expected effects in the long run as customers may switch to avoid being injured by loans officers. Briefly, “why do punish customers when delay is due to illness? More flexibility is required”. 5.2. WAGES’ customer current satisfaction level. WAGES’ customers’ actual satisfaction has been obtained by the Customer Satisfaction Index using both customers’ expectations and perceptions on the 20 remaining items. Table below give us more explanation. Tableau n° 3 : Computing Customer Satisfaction Index N° Attributes Importance Perceived weighting weighting*score (weighting )(a) quality (average (Score)(b) (c)=(a)/Average of 1) (d)= (c)*(b) 1 Responsiveness 4.58 3.92 1.05 4.11 2 Tangibles 4.53 3.77 1.03 3.88 3 Empathy 4.48 3.15 1.02 3.21 4 Conditions 4.25 3.55 0.97 3.44 5 Costs 3.96 3.53 0.90 3.17 Average= 4.36 average CSI=3.56 =3.58 Source: computations Inspired from Bhave (2002), Netteret and Nigel Hill (2005). We obtained a Customer satisfaction Index of 71. 2 % (3.56/5= 0.712*100). This result is in line with our second hypothesis that predicted that WAGES’ customers will be highly satisfied. Results also indicate that customer satisfaction is a function of financial services. In fact, customers are more satisfied by savings products than credit products. Thus, for all 22 kinds of savings, the average satisfaction scores are above 3 .56 (71 .2 %), the global satisfaction rate for all services (3.9 ˃ 3.56). Results also prove that clients are more satisfied by the cash deposit saving (F= 6.27, p=.000) than other kind of savings (4.06 ˃ to 3.76, 3.87 and 3.91 for term deposit, mandatory savings and Roscas savings). Results show that for each category of savings, customers are more satisfied by the requested documents for account opening than the minimum savings balance required to access savings products. We observe a contrary behavior for credit products. In fact, for all kinds of loans, the average satisfaction score is below the total average score, 3.56 ˃ to 3.12, 3.14 and 3.17 for direct credit, special credit and Rosca credit. However, customers appreciate Rosca credit more than other kind of loans (3.17 ˃ 3.12 and 3.14 for direct credit and special credit). Three conditions are almost critical for all kind of loans and less appreciated by customers: Risk on loan, origination fees and guarantee (direct credit especially). In fact, the average satisfaction score attached to those elements are below the total average satisfaction scores attributed to each kind of loans (2.94, 2.91 and 3.11 ˂ 3.12 for direct credit; 2.91 and 2.97 ˂ 3.14 for special credit; 2.88 and 2.90 ˂ 3.17 for Rosca credit). It seems that those conditions are badly appreciated by clients and need to be adjusted for the future. Interest rate is apparently better appreciated by clients (average satisfaction score attached to it is high than the total average score attached to each kind of loans). This result confirms the theoretical hypothesis that “poor people and entrepreneurs in developing countries will afford a high interest rate” (Armandariz and Murdoch, 2005). 5.3. The influence of customers characteristics on customer’s satisfaction Primary tests attested that data fitted optimal conditions in which Anova test can be applied and provide sound results. The Levene’s test provided a significant result (0.47˃ 0.05, F= 1.032, df1=298, df2= 47) allowing to accept the hypothesis of variance’s homogenous intragroup. This result enabled us to apply ANOVA test on data basing the decision of the significance of variables on Fischer Test. Table below presents the important results from Anova test. 23 Table 4 : Results from Anova test Source Type III Sum df Mean Square F Sig. 35 ,591 2,134 .000 of Squares a Corrected Model 20,671 Intercept 44,291 1 44,291 160,020 .000 School level ,975 4 ,244 ,881 .476 Number of services 6,280 6 1,047 3,781 .001 Type of credit ,556 2 ,278 1,004 .368 Branch of customer 5,013 5 1,003 3,623 .003 Sex ,656 2 ,328 1,185 .307 Business ,785 3 ,262 ,945 .419 Revenue 4,448 5 ,890 3,214 .008 Age ,376 5 ,075 ,272 .928 Time spent as customer ,982 3 ,327 1,182 .317 Error 85,804 310 ,277 Total 4835,447 346 Corrected Total 106,474 345 a. R Squared = ,194 (Adjusted R Squared = ,103) Source : Computed using SPSS Results show that only three customers’ characteristics significantly influence the customer satisfaction level: the number of total services to which customers can access (F=3.78; p=.000) (1), the customers’ Branch (F=3.62; p=.003) (2) and the customer’ estimated revenue (F=3.21; p=.008) (3). Those three characteristics are explaining 19 % of variance. Results reveal that the higher the number of products to which clients have accessed, the less is the satisfaction level. Thus, when customers access to three services, the average satisfaction score is 3.43, it decreases from 3.43 to 3.20 when customers access to all products offered by the MFI. However, when customers access to only one product or at least to two products, the average satisfaction score goes up attending values of 4.12; 3.71 and 3. 60 when customer accesses only to savings, loan and Rosca credit only. Only combination of both credit and savings gives a high average satisfaction scores (3.68) than other possible combinations. 24 Results from Duncan test prove that customers from Lomé are less satisfied than clients from other branches exhibiting an average satisfaction rate below average satisfaction scores from other branches (3.61 ˂ 3.94 for Tsvevie, 3. 93 for Agoe, 3.88 for Sokode, 3.81 Agodrafo and 3.72 for Adidogome ). These results suggest that there is a variance in quality of services through branches conducting to different levels of satisfaction rate. Is there a problem of general management or a missing of marketing plan in branches? Results from Duncan test suggest also that average satisfaction scores are different from classes of revenues. Thus, very poor customers are less satisfied by services (mean = 3.51) whereas middle and high income customers are high satisfied by services (Mean = 3.77 for clients with a revenue range from 22 Euro to 60 euro and 3.63 for rich clients with revenue above 60 Euro). Do those results suggest segmentation marketing for products and services to customers in terms of revenues and branches? 6. Conclusions, implications, limitations and future research perspectives. This study pursued three objectives: identification of WAGES’ customer satisfaction principal items and dimensions (1), determination of the current level of WAGES’ customer satisfaction (2), test if the customer satisfaction is influenced by customer’s characteristics (3) . To achieve these objectives, factor analysis, Customer Satisfaction Index, and Anova test have been applied. Through an iterative process, we applied factorial analysis and we obtained a refined scale with 20 items represented by five principal dimensions: a) responsiveness, b) tangibles, c) conditions, d) costs and (e) empathy explaining 62% of the total variance of customer satisfaction. Among the retained dimensions, 3 belong to the standard scale whereas 2 are related to the conditions and costs of MFI services. The results confirm that responsiveness remain the most important dimension in micro-finance sector. Results from the Customer Satisfaction Index revealed that WAGES’ customer satisfaction level is high attending 71.2% and varying in function of specific financial services. In general, customers are more satisfied with saving services than with loans products. Results from Anova test revealed that customer’s branch, customer’s revenue and number of services accessed by customers influence customer satisfaction. 25 So, customer satisfaction is mainly driven by responsiveness items indicating that performing MFI will account more on their employees providing those desired skills enabling them to deliver tailored services with more kindness to customers. That behavior will influence customer satisfaction and enhance customer’s retention and loyalty. As customer satisfaction will vary with time, managers would periodically assess the current satisfaction level and defining the most modifications to be brought to services to allow it to fulfill the customer’s needs. The variance of satisfaction level depending on the number of services, customers’ revenue and branches have practical and marketing implications. It means that MFI will focus efforts to perform current services avoiding brutal diversification of products (consolidation strategy) whereas the difference in customer satisfaction in relating to revenue and branches suggests a segmentation strategy for lending methods and marketing implementation. It suggests also a level of monitoring on the way customer care is handled among the different groups. Although this study provided sound results, it has some limitations which are sources for future research. In fact, customers surveyed in this study are coming from one MFI. Thus, results do not allow any comparison. So, the first improvement of results might be built on a survey of two, three or more MFIs, test the variability of results and validity of the scale measurement enabling to move towards a scale measurement for Micro-finance sector in general. The second way is to conduct cluster analysis on branches and revenues to determine marketing strategies to be implemented for satisfying clients in different branches and with different revenue. 26 Bibliography 1. Ahmad, A., and Sungip , Z., (2008) , “An Assessment on Service Quality in Malaysia Insurance Industry”, 13 , Communications of the IBIMA, Volume 1, 2008. 2. Anderson, E. , Fornell, C., et al (2004) , “ customer satisfaction and shareholder value”, Journal of Marketing, 68 , 172-185. 3. Armendariz de Aghion, B., Morduch, J. (2005). The economics of microfinance, Boston: MIT Press. 4. Asian Institute of Management (2005), “listening to customers: key to being effective microfinance organizations”, Copyright 2005. 5. Barnes,B.,(2005 ), “Analysing Service Quality: The Case of Post-Graduate Chinese Students”, Leeds University Business School , 2. 6. Bhave, A., (2002), “Customer Satisfaction Measurement”, Quality and productivity Journal: February 2002 Issue, http: //WWW.Symphonytech.com. 7. Bloemer, J., Ko de Ruyter, et al (1999), “ Linking perceived service quality and service loyalty: a multi-dimensional perspective” , European Journal of Marketing, 33, 1082. 8. Bressolles, G ., « qualité de service électronique : [email protected] proposition d’une échelle de mesure appliquée aux sites marchands et des effets modérateurs » , Recherche et applications en Marketing , 21 , 21-45. 9. Carricano et Poujol (2008), Analyse des données avec SPSS, Collection Synthex , Pearson Education France. 27 10. Cassel , C ., (2006), “ Measuring customer satisfaction , a methodological guidance”, State-of the Art Project 11. Chandra Shil, N., and Rashed Osman, A., (2007), “Quality Function Deployment for Customer Satisfaction in Banking Services”, Daffodil International University Journal of Business and Economics, 2, 106-118. 12. Chong “Joanna” S.K. Lee(2006) , “customers’ definition of bank service quality : an exploratory analysis of top-of-definition”, Proceedings of Fordham University conference of the marketing of financial services , November 3rd and 4th , 2006, New York . 13. Conchon ,F., Eleouet et al, .(2006), « La Satisfaction face à l’Information en MarketingCommunication: approche exploratoire dans le cadre des patients et des médicaments de prescription » , Proposition de Communication au 5ème Congrès des Tendances du Marketing en Europe, Venise 20-21, janvier 2006. 14. Cronin , J., and Talyor , S., (1994), “SERVPERF versus SERVQUAL: Reconciling Performance –based and Perceptions –Minus-Expectations Measurement of Service Quality” , Journal of Marketing , 58, 125-131. 15. Daubert (2002) , le Marketing au service des IMF , 6 , BIM. 16. Eighth Annual Conference PSM Users’ Group (2004), “Measuring Customer Satisfaction and Perceptions”, Keystone, Colorado, 26-30 July 2004. 17. European Union (2008), European Primer on Customer satisfaction management, European public administration network . 18. Forell, C., Mithas , S., et al (2005) , “Customer satisfaction and stock prices : high returns , low risk”, Journal of Marketing, 70 , 3-14. 28 19. Giannelloni et Vernette (2001), Etudes de marché, Paris : Vuibert . 20. Godquin,M.,(nd), « Crises des institutions de microfinance et attitudes de leurs clients » , TEAM-CRNS Université Paris I Panthéon –Sorbonne . 21. Gruca, T., and Rego, L., (2005) ,” Customer satisfaction, cash flow and shareholder value”, Journal of Marketing, 69 , 115-130. 22. Gustafsson , A., Johnson , M., et al ( 2005), “ the effects of customer satisfaction , relationships commitment dimensions, and triggers on customers retention”, Journal of Marketing , 69, 210-218. 23. Harrison-Walker, J., (2000), “ service quality in the Hair Salon Industry” , Journal of Business Disciplines , 1527-151X/Vol .I , 38-52. 24. Hosver, B., and Onut,S, (nd), “ Customer Relationship Management in Banking Sector and A Model Design for Banking” , Performance Enhancement, Yildiz Technical and Masla University, Turkey. 25. IFAD(2007) , “Survey report on the relationship between customer satisfaction in rural finance and technological innovation usage by rural finance institutions (RFIs)”, A fourcountry study of selected RFIs in East and Southern Africa . 26. Korauš (2002), “Quality of services in the financial sector and types of client behavior” , BIATEC, Volume X. 29 27. Kumar, D., and. Mahaptra , D.M.,(2006) , “measuring client satisfaction : Developing and Implementing Good Client Satisfaction Measurement and Monitoring Practices”, Office of the Comptroller General Evaluation and Audit Branch, April 2006. 28. Lowndes, M., (2001), “Do distinct ServQual dimensions emerge from mystery shopping data-a test of convergent validity” , Canadian Journal of Program Evaluation, 16, 41-53. 29. Malhotra , N . , Décaudin , J . et al (2007) . Etudes Marketing avec SPSS, Paris : Pearson Eduction France . 30. Merouane , A., (2008/2009 ), « La fidélisation et la satisfaction des clients », Institut d’Administration des entreprises –IAE, Université de France-Comte, Année universitaire :2008/2009. 31. Morgan, A., Anderson et al (2005) , “Understanding firms : customer satisfaction information use”, Journal of Marketing, 69 , 115-130. 32. Murray (2001), “what do MFI customers Value? A comparative analysis from three continents”, the IVth Interamerican Conference on Microenterprise , Dominican Republic, November 14th to 16th, 2001. 33. Netteret , O ., and Hill, N .,(2005) , « La Mesure de la Satisfaction Client (MSC) : Comment aider votre entreprise à faire parfaitement ce qui compte le plus pour les clients ». 34. Niraj, R., Foster, G., et al,(2003) , Implications of Satisfaction Program”. “Understanding Customer Level Profitability 30 35. Noor, N., and Muhamad, A., (2005), Individual Factors that Predict CustomerOrientation Behavior of Malaysian Life Insurance Agents, Journal Pengurusan , 24, 125-149. 36. Oksana, M., and Plepys, A.( 2003) , “Customer satisfaction: review of literature and application to the product-service systems”, Final report to the Society for Non- Traditional Technology, the international institute for industrial environmental economics , Lund University Sweden . 37. Othman, A., and Owen,L.(2001) , “The multi dimensionality of Carter model to measure customer service quality (SQ) in Islamic banking industry : a study in Kuwait finance House” , International Journal of Islamic Financial Services , 3 . 38. Parasuraman, Zeithaml, & Berry (1985), “A Conceptual Model of Service Quality and Its Implications for Future Research”, Journal of Marketing , 49 , 41-50. 39. Parasuraman, A., Zeithaml, et al (1988), “Servqual : A multiple –Item scale for measuring consumer perceptions of service quality” , Journal of retailing , 64, 12-40 . 40. Pont, M., and McQuilken, L. , (2005), “Testing the Fit of the BANKSERV Model to BANKPERF Data”, ANZMAC 2002 conference proceedings , 2, 861-867. 41. Safakli,O.,(nd), “ Measuring service quality of commercial banks towards Smes in Northern Cyprus” , Near East University, Department of Banking and Finance, Nicosia – Northern Cyprus, Mersin 10 Turkey . 42. Sarreal, E, ( 2008) , Customer Satisfaction and Service Quality in High-Contact Service Firm, Business & Economics Review, Volume 17 Number 1 January 2008, Business Management Department, De La Salle University-Manila. 31 43. Saunders, S.,(nd), “Confirmatory Factor Analysis of the PAKSERV Service Quality Construct in a Non- Western Cultural Context”, Monash University. 44. Swaid, S., and Wigand, R., (2007), “ Key Dimensions of E-commerce Service Quality and Its Relationships to Satisfaction and Loyalty”, 20th Bled Conference Mergence: Merging and Emerging Technologies, Processes, and Institutions, June 4 - 6, 2007; Bled, Slovenia. 45. Teas , K., (1994), “Expectations as a comparison standard in measuring service quality : An assessment of a reassessment” , Journal of Marketing , 58, 132-139. 46. Tuli , K., Kotli, A , et al (2007), “Rethinking customer solutions : from product bundless to relational processes”, Journal of Marketing, 71 , 1-17. 47. Verhoef, P.,(2003), “ Understanding the effect of customer relationship Management efforts on customer retention and customer share development”, Journal of Marketing , 67 , 30-45. 48. Véronique, (nd) , « La Gestion de la Relation Client dans la banque », Cermatie de Tours, Université de Tours . 49. Waterfield , C., (2008) , “The Challenges of Measuring Client Retention”, The SEEP Network. 50. Wetzels , M.,(1998), “ Service quality in Customer –Employee relationships : an Empirical study in the After –Sales Services context” , Maastricht : DATAWYSE. 51. Wright,G., Cracknell, D., et al (2003), “Strategic Marketing for Micro- Finance Institutions”, MicroSave, market-led solutions for financial services . 32 52. Yavas , U.,(2006) , “ How similar are frontline bank employees’ perceptions of service quality to their customers? A study of female customers and employees in Turkey”, Received (in revised form): 16th August, 2006.