The Influence of Banking Service Customers ’ Satisfaction Level on the Perception of Switching Costs and on Behavioural Loyalty

This article presents the results of a research aiming to evaluate whether retail banking customers’ satisfaction level has some impact on the perception of switching costs and on customers’ loyalty. Loyalty was measured through behaviors traditionally associated with this construct: intention to repurchase, propensity to recommend service and price sensitivity. The results of the research conducted in May 2007 with 1000 retail banking clients in the cities of Rio de Janeiro and São Paulo showed that, the higher the customers’ satisfaction with banking services, the higher the propensity to adopt classical behavior related to loyalty such as the intention to remain as customers for the next 12 months, to recommend the bank to relatives and friends and to be willing to pay a higher price for the offered services than their bank’s competitors. On the other hand, no significant difference was observed on perceptions of bank switching costs as a result of customers’ satisfaction level. These results provide financial institution managers with a better understanding of the relation between satisfaction and loyalty behavior on the part of customers. Such understanding is paramount for banks’ good performance in the long run. Moreover, attention should be drawn to elements that may not be overlooked in service projects and their operations in order to achieve quality excellence perceived by customers. In this study such perception precedes satisfaction.


InTroduCTIon
In the past decade, the banking market in Brazil has gone through deep and fast changes such as Real Plan economic policy, a drop in the number of banks operating in the country, the arrival of various foreign institutions and the drop in the number of state-owned banks in the market as a result of privatizations at the beginning of year 2000.Such changes on the Brazilian scenario made costumers' loyalty achievement one of the leading concerns for banking institution managers due to the highly positive impact of the latter on the organizations' financial results as attested by several empirical studies developed with different types of services (REICHHELD and SASSER, 1990;HESKETT, JONES, LOVEMAN and SASSER 1994;HALLOWELL, 1996).Therefore, bank managers need to guide their strategies in order to boost their customers' satisfaction and loyalty through service quality improvement.
Despite the belief that satisfaction is a critical determinant to achieve customers' loyalty, empirical studies conducted under the same theme showed that the relation between satisfaction and loyalty is neither linear nor simple (JONES and SASSER, 1995;GASTAL and LUCE, 2005): a high level of satisfaction is not always enough to generate loyalty behavior.Hence, researchers suggest that other potentially influential factors be analyzed in this relation such The FLagShIP reSearCh JournaL oF InTernaTIonaL ConFerenCe oF The ProduCTIon and oPeraTIonS ManageMenT SoCIeTy as costs associated by costumers with service provision also known as switch-related costs.
According to the literature, switching costs refer to the perception that the costumer has about the magnitude of additional costs necessary to end the current relationship as well as to guarantee an alternative relationship.These costs can be related to additional efforts, lost time and/or financial expenditures in the service supplier switch.(JONES, MOTHERS-BAUGH and BEATTY, 2002;BURNHAM, FRELS and MAHAJAN, 2003;PATTERSON and SMITH, 2003;YANAMANDRAM and WHITE, 2006).
Historically, bank clients have shown low switching behavior.However, according to Maddil, Feeney, Riding and Haines Jr. (2002) understanding, this behavior is not related to positive feelings regarding their banks, but, instead, to the belief that there is not much difference among the banking products offered by diverse institutions.Moreover, these clients can perceive switching as costly for several factors: the need to change the account number to receive payment; the need to "learn" about products and services of the new bank; the loss of the rapport established with the former account manager and the loss of possible financial benefits resulting from the long term relationship with the institution.
The present competitive environment, however, has provoked changes in bank customers' behavior: the latter have been seeking other institutions and have had accounts in several banks, depending on the kind of products and offered services (LEVESQUE and MCDOUGALL, 1996); BLOEMER, RUYTER and WETZELS, 2002).Therefore, taking into consideration the mediator role of bank switching costs perceived by bank customers is fundamental for the perfect understanding of the relation between customers' satisfaction and loyalty in this sector.
Studies as the ones carried out by Jones et al. (2002), Patterson and Smith (2003) and Lopez, Round and Olivan (2006), show that the assessment of bank switching costs perception on the part of customers may vary according to their satisfaction level: bank switching costs become particularly important when satisfaction with the service proves to be lower, thus, in a way, minimizing customers' reactions to a lower desired performance.On the other hand, high satisfaction levels with the service may reduce the benefits perceived with supplier switches, thus increasing service purchase intention.Alternatively, dissatisfied costumers may perceive significant ben-efits in supplier switch and, therefore, they might be more prone to quit than satisfied customers (JONES, MOTHERSBAUGH and BEATTY, 2000).
To better assess these issues, this article is intent on presenting the results of an empirical study carried out with 1,000 retail bank clients in the cities of Rio de Janeiro and São Paulo.It aims to assess whether bank switch-related cost perception and loyalty, represented by behaviors associated with price sensitivity , intention to recommend the service and intention to repurchase, are actually influenced by customers' satisfaction level.A potential contribution of the study is related to the service project and to the necessary operations for its execution.If satisfaction really exerts influence upon the variables used such as loyalty representatives, it proves necessary to identify the relevant attributes with operations leading to customers' satisfaction so as to assure that resources employed in services are designed in accordance with the performance expected from those attributes.
After such a brief introduction, the theoretical background containing the research hypotheses will be presented.Next, the methodological procedures used in this study will be described.Finally, the results of the research will be presented and discussed and the article will be wound up with the main conclusions about the study.

2.1-Customers' satisfaction and its antecedents
In order to define customer's satisfaction in the service sector, it proves indispensable to speak about quality.The direction of causality between both has been the object of noteworthy discussion in the academic environment.While some advocate that quality precedes satisfaction (e.g., CRONIN and TAYLOR, 1992), others defend satisfaction as being one of quality antecedents (HESKETT et al., 1994;ANDERSON and MITTAL, 2000).In the present study, quality in service is understood as satisfaction antecedent.
Quality in service has been defined as a construct composed of a technical dimension and a functional or relational dimension: the former centers upon what is delivered by the company to the customer -result, whereas the latter refers to the way the service is performed -process (GRONRÖOS, 1984).Besides, empirical researches have indicated a strong relation between these two quality dimensions and customer's satisfaction.The acknowledgment that quality technical aspects are critical to customers' satisfaction has made service operations become more and more important so that customers' satisfaction and loyalty may be achieved (CRONIN and TAYLOR, 1992;LASSAR, MANOLIS and WINSOR, 2000;MADDERN, MAULL, SMART and BAKER, 2007;TONTINI and SILVEIRA, 2007).
In addition to technical aspects, functional quality has also proven important to achieve customers' satisfaction.The result of the study driven by Lassar et al., (2000) indicates that bank clients do not overlook how the service was performed (functional quality) even when asked specifically about their satisfaction towards service technical aspects.This suggests that the functional dimension also influences clients' satisfaction significantly.
Customer's satisfaction, in turn, is understood as an emotional reaction, related to specific transactions.Also, it is strongly associated with the corroboration of prior expectations (OLIVER, 1997) and there is sufficient evidence to suggest that customer's satisfaction may be seen as an attitude (LEVESQUE and MCDOUGALL, 1996).According to service literature, satisfaction is the result of a customer's perception of the value received in a transaction or relationship, compared with price and acquisition costs.Moreover, it is relative to the value expected from transactions with competitors (REICHHELD and SASSER, 1990;REICHHELD, 1996;HALLOWELL, 1996).
Owing to the nature of the financial service activity, in which costumers in general do not regard the financial product as a tangible, complete product, the service that accompanies the product is very important to determine general satisfaction towards the institution.Consequently, by modeling general satisfaction for financial service companies, it proves important to include both the attributes linked to products and the ones related to services (KRISHNAN, RAMASWAMY, MEYER and DAMIEN, 1999).

2.2-Loyalty in services
According to Reichheld (1993), one of the greatest researchers on the theme, loyalty in service results from customer's belief that the quantity of the value received from a supplier is higher than the one made available by other suppliers.
Importance on loyalty is acknowledged by Haksever, Yield, Russel and Murdick (2000), for they defend that in a service project and in the operations that will carry out that service, employees and clients' loyalty development mechanisms should be included.Citing "The Service Profit Chain Model" (Heskett et al., 1994), the authors recommend a series of precautions to obtain productive and loyal employees, a pre-requisite to generate higher value service for costumers whom the company intends to preserve throughout time.
In the service literature, loyalty is defined as a multidimensional construct, composed of three dimensions.The first one is a behavioral dimension, which is characterized by repurchase behaviors, switch and exclusiveness in relation to a service supplier.The second dimension is named attitudinal, and it has impact upon the intention to recommend the service to others.The third dimension has arisen more recently in the literature and is related to cognitive aspects as to be willing to pay more, to consider the supplier as the only option to be taken into account and to identify with the service supplier (ZEITHAML, BERRY and PARASURAMAN, 1996;HALLOWELL, 1996;BLOEMER et. al, 1999;BUTCHER, SPARKES and O´CALLAGHAN, 2001;RANAWEERA and NEELY, 2003).
Those three aspects are present in a thirteen-point scale proposed by Zeithaml et al., Berry and Parasuraman (1996), whose factor analysis generated five dimensions: (1) loyalty to the company; (2) propensity to change; (3) willingness to pay more; (4) external reaction to the problem; and (5) internal reaction to the problem.Conceptually, Bloemer et al. (1999) argue that a solution of five factors did not provide us with a standard of unmistakable and consistent dimensions.Then, in an empirical study, those authors replicated the set of behavioral intentions proposed by Zeithaml et al. (1996), analyzing different kinds of services.The results pointed out to the existence of four distinct loyalty dimensions in service, which were kept consistent among the four different kinds of services assessed: word-of-mouth promotion, intention to repurchase, price sensitivity and complaint behavior.
Subsequently, a study carried out by Jones and Taylor (2007) with 348 service company costumers identified the existence of two loyalty dimensions: behavioral (intention to repurchase, intention to change and intention to give exclusivity to the service supplier); and attitude/cognitive (preference over a supplier, advocate the company, willingness to pay more, identification with service supplier).Therefore, it is observed that loyalty operationalization in the service sector should consider behavioral, attitudinal and cognitive aspects for its measurement.In the present study, loyalty will be represented through three dimensions: Intention to repurchase (in the form of relationship continuity); price sensitivity; and intention to recommend the service to third parties (word-of-mouth promotion), bearing in mind that these dimensions are more frequently used as loyalty measures in the service literature.

2.3-Bank Switching costs
Researchers such as Reichheld and Sasser (1990), Gummesson (1993), Heskett et al. (1994), Storbacka, Strandvik and Grönroos (1994), Stuart and Tax (2004), among others, propose that satisfaction influences costumers' loyalty, which, in turn, affects the profit value of the company.At the beginning of the 90's, authors such as Reichheld and Sasser defended that high levels of costumer's satisfaction would generate high levels of retention, as described in a retail bank case that increased clients' retention rate in 5% besides having boosted profits up to 85% (RE-ICHHELD and SASSER, 1990;REICHHELD, 1993 and1996).
However, high levels of satisfaction sometimes are accompanied by a continuous growth in costumers' desertion.In the research carried out by Reichheld in 1993, results showed that between 65% and 85% of the clients that switched suppliers said to be satisfied with the old supplier.Thus, it is possible to affirm that satisfaction is not always sufficient to generate loyalty behaviors on the part of the clients and that retention may depend on additional factors such as the costs associated by clients referring to service supplier switch.
Switching costs can be related to additional efforts, lost time and/or financial expenditures in switching service suppliers (JONES et al., 2002;BURNHAM et al., 2003;PATTERSON and SMITH, 2003).As services are experimental by nature, and in general cannot be evaluated before the purchase, the costumer may perceive switching service suppliers a considerable risk.Besides those factors, clients' habits and inertia should be taken into account as aspects that may complicate the switch decision.
Switching costs can be grouped in three broad cat-egories: procedure costs -which include time and effort expenses associated with evaluation costs, training and setup; financial costs -which involve the loss of benefits and financial resources; and relational costs -which are related to the loss of personal relationship and with the mark that can generate emotional and psychological discomfort (BURN-HAM et al., 2003).As Sharma and Patterson (2000) highlight, when social ties, personal understanding and confidence are built between the costumer and the service supplier throughout a certain period of time, a barrier of psychological exit cost may come up, even when the performance of the main service is below what is deemed as satisfactory.
Distinguishing and understanding the several switch-related cost dimensions prove to be important because different strategies can be necessary to manage a loyalty program efficiently.However, to perfectly understand switching costs, the kind of service being analyzed ought to be taken into account, since service characteristics such as heterogeneity and difficulty in standardization can generate differences in consumers' perceptions regarding switching costs, depending on the service (JONES et al., 2002).
In a study carried out by Jones et al. (2002) with banking service and hairdressing salons, differences both in switching cost perception level and in the relationship strength between switching costs and intention to purchase were found.For example, perceptions of set-up costs and pre-switching search costs and evaluation costs were higher in hairdressing salons than in banks.Moreover, they were also more strongly associated with intention to repurchase in hairdressing salons than in banks, confirming that the relation between intention to repurchase and switching costs vary, to a certain extent, according to the kind of service.
In another publication, Patterson and Smith (2003) confirmed the hypothesis that switching costs vary among kinds of service, as well as in their impact on intention to repurchase.In sum, there seem to be important differences among the three kinds of service studied: travel agencies, medical service and hairdressing salons.
Studies carried out in Brazil attest the mediator role played by switching costs between satisfaction and costumers' loyalty.Gastal and Luce (2004) investigated satisfaction and switching costs as loyalty antecedents in the mobile phone sector.The results pointed out that both satisfaction and switching costs have positive impact on loyalty and that the magnitude impact of switching costs on loyalty is higher than what is deemed as satisfactory.
In general terms, regardless of the kind of service being analyzed, high levels of satisfaction with the basic service should reduce the perceived benefits from switching to another service supplier (ANDERSON and SULLIVAN, 1993).In a study carried out by Jones et al. (2000) as far as service is concerned, the authors concluded that only when the level of satisfaction is below a determined level, will costumers begin to consider, how much they will be affected by switching costs.Those results suggest that, depending on the level of customers' satisfaction, their perception of switching costs, or of the rank of difficulty in switching suppliers, may vary.High levels of satisfaction with the basic service should reduce the perception of benefits when supplier switching happens, thus increasing the intention to remain as a client.Alternatively, dissatisfied costumers should perceive significant benefits in switching suppliers.Therefore, they will be more prone to desert than satisfied clients.

-Satisfaction, Loyalty and Switching costs in banking services
Several studies conducted with bank customers confirm that the technical and relational dimensions of quality are important antecedents of customer satisfaction.From extensive literature review, Krishnan et al. (1999) identified four key factors that influence bank customers' satisfaction: (1) the quality of the types of products and services, (2) the quality of financial reports and account statements, (3) service quality offered in automated telephone systems, and (4) service quality offered through direct contact in the agencies.It is noteworthy to state that the first three factors are associated with quality technical aspects, whereas the latter is associated with the last relational dimension.
A research conducted by Hallowell (1996) sought to investigate the relation among Satisfaction, Loyalty and Profitability in a retail bank.The author surveyed 12,000 bank customers in 59 divisions, business units composed of multiple agencies.The results indicated the existence of a relation between customer satisfaction and retention (explained variance of 37%) and between the latter and institution profitability (explained variance of 40%).
In another study, Gremler and Brown (1999) conducted a survey with 3,400 bank customers and 921 customers in a clinic aiming to investigate whether or not loyal customers are those who actually recommended the company.The results did not vary significantly between the two types of services investigated.Besides, they indicated that loyal customers most highly recommend the service and that the number of recommendations increases with: the length of time of the relationship with the company, the amount of money invested in the service and the number of services used.Furthermore, we observed that the number of recommendations varies significantly according to costumers' age: the older they are, the larger the number of recommendations will be.Madill and colleagues (2002) argue that strong relationships between banks and customers have advantages on both sides: for financial institutions, the advantages include the ability to maximize profits by reducing risks and improving information flow; for customers, the advantages consist of more access to credit, more favorable rates on loans, less stress and greater convenience.Therefore, the development of satisfactory relationships with customers is expected to result in greater satisfaction on the latter's part, thus leading customers to make recommendations to others and decreasing their likelihood of switching to another financial institution's service.
The results of the research developed more recently in France by N'Goala (2007) corroborate this view.The author tried to investigate the factors that can lead to resistance from clients in switching banks.About 2,000 clients were surveyed and the results indicated that affective commitment, combined with strong relationships, is the main background in detecting resistance in switching banks.Consequently, N'Goala suggests that banks strive to develop positive feelings in customers towards the institution.Also in Brazil, studies indicate that developing relationships with customers is strategic for financial institutions.In some field research conducted by Damke and Pereira (2004) with 100 customers of a credit cooperative, customers' knowledge , relationship marketing and market orientation are noticeably strategic resources that impact positively on customers' loyalty in relation to the company.The results also showed that relationship affects the quality of customer services.Similarly, the results of the research conducted by Perin, Sampaio, Brei and Porto (2004) with internet banking users highlighted the importance of confidence in the company-client relationship as far as perception of value by customers and their loyalty reach towards the institution were concerned.
As far as the role of switching costs in the relation between satisfaction and bank customers' loyalty , Matos, Henry and Rosa (2007) sought to investigate the different roles of switching costs (antecedent, mediator and moderator) in the relation between satisfaction and loyalty attitudinal and behavioral components .The authors conducted a survey with 22,000 customers of a major Brazilian bank.The main results showed that switch-related cost is an important loyalty antecedent, which plays a mediatior role between satisfaction and loyalty, especially in attitudinal loyalty.Its moderator role also proved significant, with greater emphasis on satisfaction influence on behavioral loyalty.
Historically, it is possible to state that bank customers have shown low switch performance, due more to the belief that there is not much difference among banking products than to positive feelings in relation to banks (MADILL et al., 2002).However, in the current competitive environment, this behavior has been changing: bank customers have sought other institutions and kept accounts in various banks, depending on the type of products and services offered (LEVESQUE and MCDOUGALL, 1996).

reSearCh hyPoTheSIS
In the literature review, the constructs known as satisfaction, loyalty and switching costs were characterized.Moreover, a report of the results of empirical studies investigating the relation among these constructs was made.Zeithaml et al.´s (1996) and Bloemer et al.´s researches allowed for the enunciation of the first hypothesis of this study: H1: The more satisfied the customer is, the great-• er will be his susceptibility to pay a higher price for services.et al.'s (1999) andMADILL et al.'s (2002) contributions helped to outline the second hypothesis:

Bloemer
H2: The more satisfied the customer is, the more • likely he will be to recommend the service to third parties (word-of-mouth advertising).MADILL et al. (2002) and N'Goala (2007) published studies that made possible to come up with the third hypothesis of this study: H3: The more satisfied the customer is, the less • likely he will be to switch banks.
Finally, studies by Anderson and Sullivan (1993) and Jones et al. (2000) served as the basis for the enunciation of the fourth hypothesis: • H4: The more satisfied the customer, the greater his perception of switching costs due to supplier switch.

objective of the research and operationalization of the constructs
This study aims to assess whether or not the perception of fairness and switch-related cost, represented by associated behaviors -price sensitivity, intention to recommend the service and intention to repurchase (to remain as a costumer) -are influenced by customer satisfaction level.
In other words, the study aims, firstly, to assess the influence of satisfaction on loyalty, investigating how different levels of customer satisfaction influence their intention to remain as clients, to accept paying more for the service or to do positive mouthto-mouth advertising about the bank.Meanwhile, the study aims to test how relevant it is to represent loyalty in the banking service environment through the three dimensions, which are most frequently found in services in the literature, and which are, herein, referred to as loyalty measures: Intention to repurchase, price sensitivity and intention to recommend the service to third parties.
To measure the independent variable of the study, that is, customer's satisfaction level, it was asked that the respondents identify their main retail bank, and that they assess their overall satisfaction in the following areas: banking costumer service, products and services.A 10 point scale was offered in order to assess the overall satisfaction level, in which the measure was operationalized through the following question: "Considering all aspects, costumer service, products and services that you use in your bank, what is your overall satisfaction degree from 1 to 10, in which 10 means Very Satisfied?" To measure the dependent variables: perception of switch-related cost, price sensitivity, and intention to recommend (mouth-to-mouth), a five-point Likert scale (from 1 -totally disagree, to 5 -totally agree) was adopted for the interviewees to indicate their level of agreement with the sentences below: "If I wanted to switch my current bank for another bank, it would be very laborious."(Switching cost Perception) "If I could save 10% using the services of another bank, I would switch my current bank for another."(Price Sensitivity -in this case, a price variation within a 10% order was pre-defined for evaluation purposes in order to make it tangible for the magnitude of the economy on the price of the service) "When relatives and friends ask about my bank today, I say good things about it" (Intention to Recommend) The fourth dependent variable "The intention to buy", translated as the intention to continue being customers of the bank, was measured by the following question, also using a five-point Likert scale: Over the next 12 months, you would say: Of course I will not switch banks; 1.
it is unlikely that I will switch; 2.
I may switch, I may not switch; 3.
It is likely that I will switch; 4.
Of course I will switch banks.5.

Sample and data collection
The hypotheses proposed were tested from a research survey conducted with 1,000 retail bank customers, 500 of whom were from Rio de Janeiro and 500 of whom were from São Paulo.All of them belonged to A and B socio-economic classes, recognized through demographic filters according to the Brazilian Economic Classification Criterion, which was applied early in the questionnaire.The interviewees were randomly selected, based on the landline telephone directory database from the two cities.
The interviews were conducted by telephone in May 2007, using structured questionnaire, applied by means of CATI system (Computer-Assisted Telephone Interview), that is, computer-assisted.If the costumer used more than one bank, it was asked that he consider the most used or the main bank.Only customers who had more than three months of experience with the bank were interviewed.
Most respondents are female (61%), 30% from Class A and 70 % from Class B, according to the Brazilian Criterion.As far as the length of the relationship was concerned, 70% of the respondents have had an account at their banks for over five years.The distribution of respondents by age groups shows that almost half of the respondents (48.2%) are between 35 and 54 years old, 30% are between 16 and 34 years old and the other 22% are over 54 years old.

Statistical procedures
For the application of statistical tests associated with the variables switching costs perception, price sensitivity and willingness to recommend, customers were clustered into two groups: YES (4 and 5) and NO (1 and 2).Answer number 3 (Neither agree nor disagree) was not considered.Similarly, to apply statistical tests to the variable intention to buy, customers were clustered into two classes: YES, I want to switch (4 and 5) and NO, I do not want to switch (1 and 2).Answer number 3 (I may switch, I may not switch) was not considered.
The collected data were processed using SPSS 12.0 software, applying the following procedures: (a) descriptive statistics (average and standard deviation) of the marks given to satisfaction, (b) Mann-Whitney test to compare satisfaction levels between two groups created for each hypothesis, as described above.This test is used to check whether two independent samples were taken from populations with equal averages.This is an interesting alternative to the parametric test for average equality, for the Mann-Whitney test does not require considerations on population distribution.The only requirement of this test is that the observations be measured in numerical or ordinal scale (SIEGEL AND CASTEL-LANE, 1988).

verification of the hypotheses
Using a grading scale from 1 to 10, where 1 is very dissatisfied and 10 is very satisfied, the interviewees were asked to give grades to their overall satisfaction degree towards the bank.The overall satisfac-tion grade average given by clients was 8.0.Mann-Whitney tests were used to compare satisfaction levels between the two groups for each hypothesis.
Frequency distribution of the responses of each measure, per group of respondents (YES group and NO group), the respective overall satisfaction grade average, and the significance level of the test are presented in Table 1 below The two groups taken into account to test hypothesis H1 (price sensitivity) were clients who would switch banks and those who would not if they could save 10% on the price of services.The test pointed out a significant difference among satisfaction levels in the groups, with the average of the group that would not switch the bank (8.5) higher than those who would (7.66).Thus, hypothesis H1, which says the more satisfied the customer, the greater his pre-disposition to pay a higher price for their services, is supported (U = 50043 p = 0000).The results also allow us to observe that the average satisfaction of 7.66 was not enough to ensure that customers were willing to pay a price premium; namely, the client should be very satisfied to be willing to pay a higher price for bank services.
Hypothesis H2 test (mouth-to-mouth advertising) took into account the group of customers who intended to recommend the bank to others as well as those who did not have such intention.The result of the test indicated the existence of significant difference in the satisfaction level between the groups, the latter being higher in the first group described, thus confirming hypothesis H2, which states that the more satisfied the customer is ,the more likely he will be to recommend the service to others (U = 12927 p = 0000).
To test hypothesis H3, the intention to repurchase, the groups were divided between those who want and those who not want to switch banks over the next 12 months.The result of the test showed significant difference between satisfaction levels of the two groups considered.The group not willing to switch banks showed significantly higher satisfaction level than the other.Therefore, this result supports hypothesis H3, which states that the more satisfied the customer is, the less likely he will switch banks (U = 24634 p = 0000).
Finally, to test hypothesis H4, related to switch-related cost perception, the two groups taken into account were customers who found it laborious to switch banks and those who did not find it hard to do so.The satisfaction level was not significantly different between the groups, indicating that customer satisfaction level does not affect the client's switching cost perception.Thus, such finding does not support hypothesis H4 (U = 85852.5,p = 0858).

discussion
The results provide some interesting insights related to the complex relation among satisfaction, switching cost perception and costumer's loyalty.The latter was evaluated in terms of intentions and future behavior.
In this study, an attempt was made to assess the influence of satisfaction on loyalty in the banking sector.It was operationalized based upon three dimensions often found in the literature -Intention to repurchase, price sensitivity and intention to recommend the service to others.Additionally, it was intended to assess whether customers' satisfaction level influences their perception of switching costs, resulting in high impact on clients' degree of difficulty in switching banks.
By starting the discussion through the only hypothesis that was not verified (H4), it became conspicuous that the interviewees did not present different perceptions on switching costs as far as costumers' satisfaction level is concerned.Quite the contrary, both the customers who expressed difficulty in switching banks and those who did not, interestingly forming two groups of almost equal size, possessed the same level of satisfaction towards the bank services.This result contradicts Jones et al.´s findings (2000), which suggest that, depending on customers' satisfaction level, their perception of switching costs may vary, concluding that only when satisfaction is low, do switching costs play an important role in the intention to repurchase.The findings also contradict Anderson and Sullivan's study (1993), in which high satisfaction levels with basic service should reduce the perceived benefits in switching to another service provider.
Although the literature reports that, traditionally, banking service market has been characterized by relative inertia on the part of clients (Madill et al., 2002), this study seems to point towards a direction proposed by Levesque and McDougall (1996) for whom the current competitive environment altered this behavior: Through the results of this study, it was found out that, regardless of their satisfaction level with the bank, customers are considering switching to other providers with less difficulty and lower perception of switching costs than what would be expected.
The feeling of a happy relationship with a service provider can motivate costumers to continue with the same provider and recommend it to other cus-tomers.Thus, customer satisfaction is expected to have a positive effect on loyalty dimensions.The results confirmed such an effect (H2), in that different levels of satisfaction generated impacts on intention and behavioural loyalty.It was found that the higher the customers' satisfaction level is, the more inclined they will be to conduct positive and friendly intention towards their bank.Confirming what has already been validated by several authors (REICH-HELD and SASSER, 1990, REICHHELD, 1993, Hallowell, 1996;JONES et al., 2000), the study indicated that, the higher the customers' satisfaction level, the greater their propensity to repurchase.In other words, they will be more prone to continue to do business with their current service provider: their main bank.
Similarly, it was possible to verify an association (H1 and H3) between the levels of satisfaction and the two other evaluated dimensions related to loyalty; that is, price sensitivity and willingness to recommend the bank to others.The result of the research showed that, the higher bank customers' satisfaction level is, the greater their acceptance will be to pay a higher price for their services, and the greater their intention to say good things about the bank to others will be.These results are aligned with Zeithaml et al.'s (1996), Gremler and Brown's (1999) and Bloemer et al.'s (1999) findings, except for the fact the complaint behavior construct was not tested in this study.
The three tested variables -intention to repurchase, price sensitivity and propensity to recommend the service -proved to be affected by customers' satisfaction level, reinforcing them both as dimensions linked to loyalty and as substitute measures for that construct.
An interesting aspect revealed by the research involves the relationship between satisfaction level and loyalty dimension called "price sensitivity." Heskett et al. (1994) suggest that loyalty may increase rapidly after customers' satisfaction exceeds a certain limit; namely, there are increasing returns in scale in the relation between satisfaction and customers' loyalty.Being consistent with the "limit" argument, the survey found that very satisfied customers are more inclined to be loyal to the bank than those who are merely satisfied, especially as far as price sensitivity dimension is concerned.The predisposition to pay a higher price for banking services was perceived at a higher satisfaction level (8.5), whereas a slightly lower satisfaction level (7.66) was not enough to ensure that customers were willing to pay a price premium for the service.In other words, only being satisfied does not count.Instead, it is paramount to be very satisfied in order to be willing to pay a higher price for bank services.

Managerial Implications
In practical terms, the results confirm the importance of satisfaction in the process of retention and customers' loyalty, contributing to the growing evidence on the importance of customers' satisfaction.
Consistent with other previous research, satisfaction ought to remain being the primary strategic focus of financial service providers owing to its strong impact on retention and on some behaviors associated with customers' loyalty.Costumers' satisfaction may generate a variety of positive effects such as intention to repurchase and willingness to recommend the service, which were also verified in this study.
In addition, the study also showed that customers with high satisfaction levels are more likely to pay a higher price for services.From the managerial point of view, these findings emphasize the important role that customers' satisfaction plays in price rise implementation.
Another important practical implication is the possibility to gain new customers through positive advertising done by satisfied customers.Word-ofmouth recommendation is the most focused and most targeted marketing form.Most people relate to other people who have different affinities.Therefore, if someone recommends a service to a friend who has the same tastes and needs, this information will probably be somehow useful or interesting.If the bank has a profitable customer, it is likely that he will recommend the bank products or services to someone with a similar profile, needs and desires.That is why word-of-mouth advertising proves so important because it can increase the bank's costumers with a given interest profile at a minimum effort, providing that the company focuses upon the strategy to keep their target customers with a high satisfaction level.
By keeping loyal customers, banks may make use of three benefits.Firstly, the cost of acquiring customers should fall because it will not be necessary to replace deserters and customers because new customers will emerge as a result of word-of-mouth advertising done by loyal customers.Secondly, a loyal and long term customer is less price sensitive.Therefore, profit margins may be higher.Finally, a loyal customer will better respond to the suggestion to buy more products and services from the bank, expanding its business scope, and thus generating more profits for the company.

ConCLuSIon
This study presented a research conducted with retail bank customers aiming to assess the existence of the relation between customer satisfaction and switch-related cost perception.Also, it aimed to assess the relation between customer satisfaction and loyalty, the latter being measured by future behaviors associated with this construct.
As a general rule, customer satisfaction and loyalty are closely related to satisfaction, serving as an antecedent to loyalty, thus preventing supplier switch and consolidating customers' retention.
The results were able to show that the more the customer is satisfied with the bank services, the more loyal he will be, adopting traditional behaviors associated with loyalty, which are: to remain as a client of the same bank (intention to repurchase), to recommend the bank to relatives and friends (word-ofmouth) and to be willing to pay a price higher than competitors' for the services offered (price sensitivity).
This reinforces the assumption that, in the banking context as well as in services in general, it is important to satisfy customers to grant their loyalty.Consequently, it is highly recommended that companies develop good relationships and a client-oriented climate to identify customers' genuine needs, thus developing products and services to meet such needs.Accordingly, it is noteworthy to highlight the role of business services to create greater value for the customer.In order to ensure interpersonal relationship quality and interactions between customers and services provided by banks, it has proven essential to invest in front line staff selection and training as well as in equipment that will provide quick answers tailor-made for the client, who, in turn, will need to interact with it.
Despite the existence of literature studies showing that the level of customers' satisfaction may influence their perception of switching costs, the results of the research did not confirm this trend.There was no significant difference in perceptions of bank switching costs as a result of customers' satisfaction level.Both customers who expressed difficulty in switching banks, and those who did not, showed a similar satisfaction level towards the bank.
The research findings also contradict the traditional market trend in banking services related to a certain amount of inertia on the part of customers, who, even being dissatisfied with services, would remain as customers due to switching costs.The research revealed that, apparently, regardless of their satisfaction level, customers do not find it very difficult or costly to switch banks.
One should bear in mind that bank customers currently have each time more a wider range of choices, and therefore, it has become easier to switch suppliers.Hence, there is a growing need to continue to develop a strategy to keep customers satisfied.Satisfied customers tend to recommend their bank more often and pay a premium price for bank services.Besides, they are more likely not to switch banks, at least in the short term.Because of that, satisfied customers tend to gather their transactions in just one bank and respond positively to offers of new products and services, thereby enlarging their relationship and business scope with the company.This confirms Reichheld and Sasser' thesis (1990) that high satisfaction levels lead to high loyalty levels, which, in turn, lead to profit increase by the means aforementioned.
Therefore, creating high satisfaction levels is absolutely necessary for customer base maintenance and for the financial institution's consequent permanence in the market.Costumers' growing loyalty is the most important factor to ensure a good bank performance in the long run.To do so, service quality and customer satisfaction have proven to be key aspects to ensuring it.
In spite of a limited number of constructs that can be used as substitutes for customer loyalty, it should be enhanced that the variables in this study are not exhaustive.It is recommended that future researchers study other possible variables associated with customer loyalty in the banking service environment and that they develop a model that will enable to show the existing relationships among these variables.
It is also noteworthy to state that the results of this study cannot be generalized to other services, including financial, or banking services in other cultures.However, it would be interesting to replicate it in other contexts.Finally, it is noteworthy to men-tion that aiming to facilitate the implementation of the questionnaire and to increase interviewees' willingness to answer it, a small number of levels in the scale were used.These scales could be expanded to give greater accuracy and reliability to behavioral aspects and future intentions related to loyalty, as seen in the study by Zeithaml, Berry and Parasuraman (1996), aforementioned.

note:
1-Brazil Economic Classification Criterion is a classification used to estimate urban people and families' purchasing power, defining walks of life that meet mainly think -tanks' segmentation needs (as far as purchasing power), which are its largest users.For further information, one may turn to: http:// www.abep.org/codigosguias/ABEP_CCEB.pdf.

Table 1 . Frequency of responses per group of respondents, overall satisfaction grade average and signifi- cance level of the test
: