You are on page 1of 11

Investigating drivers of bank loyalty: the complex

relationship between image, service quality and


satisfaction
Josée Bloemer
Associate Professor of Marketing, Department of Applied Economics, Limburg
University Centre, Diepenbeek, Belgium
Ko de Ruyter
Associate Professor of Marketing, Faculty of Economics and Business
Administration, Maastricht University, Maastricht, The Netherlands
Pascal Peeters
Research Assistant, Faculty of Economics and Business Administration,
Maastricht University, Maastricht, The Netherlands

This article investigates how this article we address this issue. We propose
image, perceived service Introduction a model that describes the relationship
quality and satisfaction deter- During the past decade, the financial services between service quality, satisfaction with a
mine loyalty in a retail bank sector has undergone drastic changes, result- bank and loyalty, taking into account the
setting at the global construct ing in a market place which is characterised effect of the image that a bank has in the
level, as well as the level of by intense competition, little growth in pri- market.
construct dimensions. At the mary demand and increased deregulation. In Our article unfolds as follows. First of all,
global level the results of a the new market place, the occurrence of com- we offer a brief outline of the construct of
large-scale empirical study mitted and often inherited relationships loyalty. Next, we will differentiate conceptu-
reveal that image is indirectly between a customer and his or her bank is ally between service quality and satisfaction
related to bank loyalty via becoming increasingly scarce (Levesque and as determinants of customer loyalty. Subse-
perceived quality. In turn, McDougall, 1996). Several strategies have quently, we will introduce the construct of
service quality is both directly been attempted to retain customers. In order image and focus on the relationship between
and indirectly related to bank to increase customer loyalty, many banks image, satisfaction, service quality and loy-
loyalty via satisfaction. The have introduced innovative products and alty by formulating a set of formal hypothe-
latter has a direct effect on services (Meidan, 1996). However, as such ses. After dealing with the conceptual issues,
bank loyalty. At the level of innovations are frequently followed by simi- we will discuss the results of an empirical
the dimensions underlying lar charges, it has been argued that a more study that was undertaken to test our
aforementioned constructs, it viable approach for banks is to focus on less research hypotheses for the setting of retail
becomes clear that reliability tangible and less easy-to-imitate determi- banking. In conclusion, we will address the
(a quality dimension) and nants of customer loyalty such as customer theoretical as well as the managerial implica-
position in the market (an evaluative judgements like service quality tions of our findings on the relationship
image dimension) are rela- and satisfaction (Worcester, 1997; Yavas and between image, service quality, satisfaction
tively important drivers of Shemwell, 1996). Surprisingly, however, while and loyalty.
retail bank loyalty. there has been a large number of studies that
focused on service quality and satisfaction
issues (Lewis, 1993), research on the relation- Service loyalty
ship between satisfaction, service quality and Research into customer loyalty has focused
loyalty in retail banking has remained lim- primarily on product-related or brand loyalty,
ited. Yet, in the present environment of whereas loyalty to service organisations has
increased competition with rapid market remained underexposed (Gremler and
entry of new service concepts and formats, Brown, 1996). Frequently, a high positive
the challenge of increasing loyalty also pre- correlation between the constructs of satis-
sents a challenge of a more in-depth under- faction and quality and product loyalty is
standing of the complex relationship between reported. Likewise, with regards to service
aforementioned types of customer evaluative loyalty, perceived service quality as well as
judgements and loyalty. satisfaction have been identified as key
In addition, there is some evidence that antecedents in banking as well as in other
loyalty may also be determined by image service industries (Dick and Basu, 1994;
(Mazursky and Jacoby, 1986; Murphy, 1996; Lewis, 1993). However, there are a number of
International Journal of
Osman, 1993). Again, it has remained unclear reasons why findings in the field of product
Bank Marketing whether there is a direct relationship loyalty cannot be generalised to service loy-
16/7 [1998] 276–286 between image and loyalty or whether this alty and more research into specific service
© MCB University Press relationship is mediated by, for instance, sectors is needed (Gremler and Brown, 1996;
[ISSN 0265-2323] Keaveney, 1995). Service loyalty is more
satisfaction and perceived service quality. In
[ 276 ]
Josée Bloemer, Ko de Ruyter dependent on the development of interper- the concept of commitment plays a central
and Pascal Peeters sonal relationships as opposed to loyalty with role (Morgan and Hunt, 1994; Shemwell et al.,
Investigating drivers of bank tangible products (Berry, 1983), for person-to- 1994). Commitment in service provider-cus-
loyalty: the complex person interactions form an essential ele- tomer relationships has been defined as “an
relationship between image,
service quality and ment in the marketing of services (Crosby et implicit or explicit pledge of relational conti-
satisfaction al., 1990; Czepiel 1990; Czepiel and Gilmore, nuity between exchange partners” (Dwyer et
International Journal of 1987; Surprenant and Solomon, 1987). Fur- al., 1987, p. 19). Likewise, Moorman et al.
Bank Marketing thermore, the influence of perceived risk is (1992) define commitment as an enduring
16/7 [1998] 276–286 greater in the case of services, as customer desire to maintain a valued relationship.
loyalty may act as a barrier to customer Parties identify commitment among
switching behaviour (Guiltinan, 1989; Klem- exchange partners as key to achieving valu-
perer, 1987; Zeithaml, 1981). Indeed, it has able outcomes for themselves, and they
been demonstrated that loyalty is more preva- endeavour to develop and maintain this pre-
lent among service customers than among cious attribute in their relationship (Morgan
customers of tangible products (Snyder, 1986). and Hunt, 1994). We propose that bank com-
In the services context, intangible attributes mitment is a necessary condition for bank
such as reliability and confidence may play a loyalty to occur. In case of absence of bank
major role in building or maintaining loyalty commitment, a patron to a bank is merely
(Dick and Basu, 1994). spuriously loyal, i.e. repeat visiting behav-
As most research originated from the field iour is directed by inertia (Dick and Basu,
of packaged consumer goods (Jacoby and 1994).
Chestnut, 1978), a strong emphasis has been We define bank commitment as: the pledg-
on behavioural measures. In a services con- ing or binding of an individual to his/her
text, loyalty is frequently defined as observed bank choice (Kiesler, 1968; Lastovicka and
behaviour (Liljander and Strandvik, 1995). Gardner, 1977). As a result of explicit and
Meidan (1996, p. 31) argues that the degree of extensive decision-making, as well as evalua-
loyalty in banking can be gauged by “track- tive processes, a consumer becomes commit-
ing customer’s accounts over a defined time ted to the bank and therefore, by definition,
period and noting the degree of continuity in becomes bank loyal. When the decision-mak-
patronage”. However, behavioural measures, ing and evaluative processes are not explicit
such as repeat purchasing or visiting and very limited, the consumer will not
sequence, have been criticised for a lack of a become committed to the bank and cannot be
conceptual basis and for having a narrow, i.e. bank loyal. Consumers whose patronage is
outcome-focused, view of what is in fact a not based on bank loyalty may exhibit an
dynamic process (Day, 1969). For instance, a attachment to bank attributes and can easily
low degree of repeat purchasing of a particu- be lured away by competitors through, for
lar service may very well be the result of instance, pricing strategies. In our view, the
situational factors such as non-availability, level of consumer commitment can differ
variety seeking and lack of provider prefer- considerably. Therefore, we assume there is a
ence. Therefore, the behavioural approach to continuum of bank loyalty. At one end of the
loyalty may not yield a comprehensive continuum, one finds true bank loyalty; the
insight into the underlying reasons for loy- repeat visiting behaviour based on a maxi-
alty, instead it is a consumer’s disposition in mum amount of commitment. At the other
terms of preferences or intentions that plays end of the continuum, one finds spurious
an important role in determining loyalty bank loyalty; the repeat visiting of the bank
(Bloemer and Kasper, 1995; Jain et al., 1987).
not based on any commitment at all. In this
Furthermore, repeat purchasing behaviour
way, bank commitment enables us to define a
may not even be based on a preferential dispo-
degree of bank loyalty.
sition but on various bonds that act as switch-
As mentioned above, service quality and
ing barriers to consumers (Liljander and
satisfaction have both been advanced as
Strandvik, 1995). We define bank loyalty as:
antecedents of service loyalty. Therefore, in
the biased (i.e. non random) behavioural
the next section we will zoom in on these
response (i.e. revisit), expressed over time, by
concepts and the relationship between them.
some decision-making unit with respect to
one bank out of a set of banks, which is a
function of psychological (decision-making
Service quality and satisfaction
and evaluative) processes resulting in brand
commitment. In most models of client evaluations of retail
This definition is based on Jacoby and banking services the focus has been on a
Chestnut (1978). The critical part of our defin- comparative judgement of expectations
ition of bank loyalty is bank commitment. In versus perceived performance resulting in
theories of interorganisational relationships the two major evaluative judgements of
[ 277 ]
Josée Bloemer, Ko de Ruyter perceived service quality and client satisfac-
and Pascal Peeters tion (Murphy, 1996; Smith, 1992). Both con- The relationship between
Investigating drivers of bank cepts have been frequently used and mea-
image, quality, satisfaction and
loyalty: the complex
sured in the retail banking services area
loyalty
relationship between image,
service quality and (Lewis, 1993; Lewis and Mitchell, 1990; Smith, It has been suggested recently that retail
satisfaction 1992). However, it has been argued that prob- banking has been suffering from an identity
International Journal of lems of definition, delineation and conceptu- crisis and that image research studies should
Bank Marketing alisation concerning these evaluative judge- be undertaken with the objective of providing
16/7 [1998] 276–286 information that is as strategically important
ments still exist (De Ruyter et al., 1997). Much
of the confusion arises from the fact that both as financial performance data (Worcester,
forms of evaluative judgements are based on 1997). A favourable image is viewed as a criti-
comparable underlying constructs. Clients cal aspect of a company’s ability to maintain
form expectations prior to their encounter its market position, as image has been related
with a bank (employee), they develop percep- to core aspects of organisational success such
tions during the service delivery process and as customer patronage (Granbois, 1981; Kor-
gaonkar et al., 1985).
subsequently they compare their perceptions
Many conceptualisations of image have
to their expectations in evaluating the out-
been advanced in the past (Doyle and Fen-
come of the service encounter. While service
wick, 1974; James et al., 1976; Kunkel and
quality and satisfaction are concepts that
Berry, 1968). Image has been treated as a
have a number of similar characteristics,
“gestalt”, reflecting a customer’s overall
they have points of differentiation as well, as
impression. Keaveney and Hunt (1992) have
becomes clear from major advances in the argued that the image of a retail institution is
services marketing literature (Patterson and formed along the lines of category-based
Johnson, 1993). In the first place, it is processing theory, i.e. when a customer
frequently argued that in order to form a encounters a bank, he or she will form a
satisfaction judgement, consumers must have mental picture as to whether the bank
experienced a service, whereas perceived matches any other categories of banks experi-
service quality is generally viewed as being enced in the past. According to the category-
not necessarily experience-based. Second, it based processing paradigm, it is proposed
has been argued that the two concepts are that incoming information, as well as cus-
determined by different antecedents. Evi- tomer evaluation of attributes, will be judged
dence exists regarding a number of cognitive relative to the bank image.
and affective processes (equity, attributions, The exact relationship between image and
cost/benefit analyses) that influence satisfac- loyalty has remained a matter of debate.
tion. The number of antecedents to service Sirgy and Samli (1989), for instance, report a
quality is regarded as more limited (Oliver, direct positive relationship between image
1993). Clearly the two types of evaluative and loyalty. On the other hand, it has been
judgement are not perceived as isomorphic demonstrated that the link between image
and increasingly treated as separate con- and loyalty is mediated by customer evalua-
structs in research on services. There is a tive judgements such as quality perceptions.
growing consensus on the sequential order of Moreover, in the product literature there is
service quality and satisfaction. The latter is ample evidence that image significantly
affects perceptions of quality (Darden and
increasingly regarded as the superordinate
Schwinghammer, 1985; Render and O’Connor,
construct based on conceptual work by Oliver
1976; Stafford and Enis, 1969). Therefore, we
(1993) and Rust and Oliver (1994), and empiri-
propose that as customer evaluative judge-
cal evidence provided by Cronin and Taylor
ments such as perceived service quality and
(1992) and De Ruyter et al. (1997). Cronin and
satisfaction are established in a process of
Taylor (1992) undertook an empirical test of
inference making of expectations, image will
the reciprocity between satisfaction and precede customer evaluations, rather than
quality across several service industries. these evaluations being components of image
Using structural equation modelling, they (Hildebrandt, 1988; Mazursky and Jacoby,
found that service quality can be seen as a 1986). That is, image determines the nature of
determinant of satisfaction which in turn consumer expectations which, in turn, are a
influences purchase intentions. Lately, how- decisive influence on the formation of quality
ever, it has been suggested that, in addition to perceptions.
service quality and satisfaction, image is also Based on our review of the literature we
an important determinant of customer formulate the following hypotheses:
patronage. We will discuss image in the next H1: Image will have a direct positive effect
section. on loyalty.
[ 278 ]
Josée Bloemer, Ko de Ruyter H2: Image will have an indirect positive the bank, in terms of a scale varying from 1 to
and Pascal Peeters effect on loyalty via satisfaction (i.e. a 10”).
Investigating drivers of bank mediator effect). A commitment scale, together with the
loyalty: the complex H3: Image will have an indirect positive chance of visiting the same bank again, was
relationship between image,
service quality and effect on loyalty via quality (i.e. a mediator used to determine bank loyalty (Bloemer and
satisfaction effect). Kasper, 1995). Customers had to rate the
International Journal of H4: Quality will have an indirect positive intention to visit the same bank next time
Bank Marketing effect on loyalty via satisfaction (i.e. a they needed to visit a bank for personalised
16/7 [1998] 276–286 mediator effect). service, in contrast with a visit to the auto-
H5: Satisfaction will have a direct positive matic teller machine – repeat visiting behav-
effect on loyalty. iour (RPB) – ranging from 0 per cent to 100
per cent). Bank commitment (COM) was
In the next section we will report the results
measured with a scale containing four four-
of an empirical study designed to test these
point Likert-scale items. The Cronbach alpha
hypotheses.
for this scale was 0.76. This operationalisa-
tion concerns the behavioural as well as the
commitment aspect of loyalty. Therefore, it is
An empirical study in line with our definition of bank loyalty
Data collection which stresses the two aspects. Non-Dutch
An empirical study was conducted among items were translated into Dutch via a proce-
customers of a major bank in The Nether- dure of double-back translation by a qualified
lands in 1996. Nationwide, 2,500 existing pri- translator (Brislin, 1980).
vate customers were interviewed by phone
with regard to their image of the bank, their
quality perception of the bank, their satisfac- Results
tion with the bank and their loyalty towards
In Table I, we present an overview of the cor-
the bank. The sample was randomly drawn
relations between the variables: image, qual-
and found to be representative for the bank’s
ity, satisfaction and loyalty[1].
private customers by checking the customer
Table I shows that all the correlation coeffi-
database. Moreover, the descriptive data gath-
cients are significant (p < 0.001). Further-
ered in the study were compared with demo-
more, there is a clear positive relationship
graphic variables that were available from
between image and quality (r = 0.59), image
previous market research studies within this
and satisfaction (r = 0.44), image and loyalty
bank.
(r = 0.53), quality and satisfaction (r = 0.55),
quality and loyalty (r = 0.66) and satisfaction
Questionnaire development
and loyalty (r = 0.59).
The design of the questionnaire was primar-
Next, we used multivariate regression
ily based on multiple-item measurement
analysis to gain additional insight into the
scales taken from previous research. The
data and to test our hypotheses.
items were adapted to the specific character-
We expect a positive direct relationship
istics of our research setting. The image (IM)
between the image of a bank and the loyalty
of the bank was measured with a scale con-
towards that bank. For this purpose we spec-
taining 17 four-point Likert-scale items, rang-
ify a model to test the relationship between
ing from 1 = completely disagree, to 4 = com-
image and loyalty, and the effects of quality
pletely agree. The scale was developed on the
and satisfaction on loyalty:
basis of a qualitative research study
LOY = b0 + b1 * SAT + b2 * QUA +
conducted prior to the quantitative data col-
b3 * IM + ε1 (Model 1.1)
lection phase. This scale was also pre-tested
and found to be valid and reliable on the basis where SAT = satisfaction; QUA = quality
of our study. The Cronbach alpha for this perceptions; IM = image; and LOY = loyalty of
scale was 0.83. the customers towards the bank. The results
With regards to perceived service quality, of this analysis are shown in Table II.
items were based on the service quality liter- Table II shows that both quality perception
ature (e.g. Parasuraman et al., 1988) on the and satisfaction have a positive impact on
results of a qualitative study for the bank loyalty (beta = 0.38 and beta = 0.42 respec-
setting. The quality perception (QUA) of the tively). Image has no significant influence on
bank was measured with a scale containing loyalty. The explained variance of loyalty by
19 four-point Likert-scale items. The Cron- quality and satisfaction is 53 per cent.
bach alpha for this scale was 0.73. We expect an indirect positive effect
In order to measure satisfaction, customers between the image of a bank and the loyalty
were asked to state their satisfaction (SAT) towards that bank through satisfaction (i.e. a
with the bank (“how satisfied are you with mediator effect). For this purpose we specify
[ 279 ]
Josée Bloemer, Ko de Ruyter towards that bank through quality (i.e. a
Table I
and Pascal Peeters mediator effect). For this purpose we specify
Investigating drivers of bank Pearson correlations between the central
a model to test the relationship between
loyalty: the complex variables
image and quality.
relationship between image,
service quality and IM QUA SAT LOY QUA = b0 + b1 * IM + ε1 (Model 1.3)
satisfaction IM 0.59a 0.44a 0.53a where QUA = quality perceptions; and IM =
International Journal of QUA 0.55a 0.66a the image of the bank. The results of this
Bank Marketing
16/7 [1998] 276–286 SAT 0.59a analysis are shown in Table IV.
Notes: IM = image; QUA = quality perceptions; SAT = Table IV shows that the image of the bank
overall satisfaction; LOY = loyalty; a = one-tailed has a clear positive influence on the quality
significance < 0.001 perception (beta = 0.59). The explained vari-
ance is 34 per cent.
Furthermore, we expected that quality has
an indirect positive effect on loyalty via
Table II satisfaction and that satisfaction will have a
Results of regressions analysis based on direct positive effect on loyalty. Based on the
model 1.1a analysis presented above, we conclude that
b beta p indeed quality has an indirect effect via satis-
faction (beta = 0.55) and that satisfaction has
Intercept –1.48 0.00
a direct effect on loyalty (beta = 0.42).
SAT 0.44 0.42 0.00
The results of the former analyses in terms
QUA 1.30 0.38 0.00
of the significant beta coefficients are
IMb
depicted in Figure 1. Figure 1 shows image
R2 0.53
does not have a direct positive effect on loy-
Note: aLOY = bo + b1 * SAT + b2 * QUA + b3 * IM + ε1;
b = not significant (p > 0.05); IM = image; QUA = quality alty. Therefore, H1 has to be rejected. We also
found that no indirect effect of image through
perceptions; SAT = overall satisfaction; LOY = loyalty
satisfaction could be detected. We therefore
have to reject H2 too. In addition, we found
a model to test the relationship between that quality has a direct and an indirect
image and satisfaction, we also incorporated impact on loyalty. The indirect influence is
the effects of quality. via satisfaction. Altogether this implies that
SAT = b0 + b1 * QUA + b2 * IM + ε1 (Model although image has no indirect influence on
1.2) loyalty directly via satisfaction, it neverthe-
less has an impact on satisfaction via quality.
where SAT = satisfaction; QUA = quality
Next, Figure 1 shows the mediator effect of
perceptions; and IM = the image of the bank.
the quality perception in the relationship
The results of this analysis are shown in
between image and loyalty. Therefore H3 can
Table III.
be accepted. Likewise, H4 and H5 can be
Table III shows that only quality perception
accepted on the basis of our analysis.
has a significant effect on the loyalty of the
As we developed a number of multi-dimen-
customers towards this bank (beta = 0.55).
sional scales to measure the constructs intro-
The explained variance of the loyalty of the
duced above, we decided to perform a number
customer by the quality perception is 30 per
of additional analyses in order to gain a
cent. Image seems to have no impact on loy-
deeper and more comprehensive insight into
alty here.
the relationships between our variables.
We expect an indirect positive effect
These analyses were used to examine the
between the image of a bank and the loyalty
dimensional distinctiveness of both the
image and quality constructs. In fact, we used
Table III exploratory factor analysis to determine the
Results of regressions analysis based on
model 1.2a
Table IV
b beta p Results of regressions analysis based on
Intercept 0.92 0.00 model 1.3a
QUA 1.85 0.55 0.00
IMb b beta p
R2 0.30 Intercept 0.44 0.00
Notes: aSAT = b0 + b1 * QUA + b2 * IM + ε1; QUA 0.47 0.59 0.00
b = not significant (p > 0.05); R2 0.34
IM = image; QUA = quality perceptions; SAT = overall Notes: aQUA = b0 + b1 * IM + ε1;
satisfaction IM = image; QUA = quality perceptions

[ 280 ]
Josée Bloemer, Ko de Ruyter different dimensions of image and quality. In driven, position in the market, society-driven
and Pascal Peeters each of the factor analyses, the number of and prices.
Investigating drivers of bank factors retained were those with eigenvalues The results of the factor analysis for the
loyalty: the complex greater than 1. The results of those factor quality items are shown in Table VI.
relationship between image,
service quality and analysis for the image items are shown in The overall pattern of rotated factor load-
satisfaction Table V. ings suggested a seven-dimensional solution,
International Journal of The overall pattern of rotated factor load- accounting for 58.7 per cent of variance
Bank Marketing ings suggested a six-dimensional solution, extracted. The factors found could be called
16/7 [1998] 276–286 accounting for 63.8 per cent of variance reliability, empathy, efficiency, interest rates,
extracted. We labelled the factors as follow- procedures, expertise and access to money.
ing: customer contacts, advice, relationship Next, we used again multivariate regres-
sion analysis to gain additional insight into
the data in terms of the importance of the
Figure 1 different image and quality dimensions in
Model based on ordinary multivariate least squares analyses relation to loyalty. According to the outcome
of the factor analysis we specify a new model
to test the relationship between loyalty, satis-
faction, quality and image dimensions. On
QUA β = .38 the basis of factor analyses we are now able to
β = .59
zoom in on the relationships between deter-
minants of loyalty.
β = .55 LOY
IM LOY = b0 + b1 * SAT + b2 * QF1... b8 * QF7
+ b9 * IF1+...+ b14 * IF6 + ε1 (Model 2.1)

SAT β = .42 where SAT = satisfaction; QF1 = factor scores


quality perceptions first factor etc.; QI1 =
image factor scores first factor and LOY =
loyalty of the customers towards the bank.
The results of this analysis are shown in
Table VII.
Table VII shows that satisfaction (beta =
Table V 0.38), reliability (beta = 0.36), efficiency (beta
Result factor analysisa on image items = 0.18) and position in the market (beta = 0.30)
have a significant positive impact on loyalty.
Item Description Factor loadingsb
The total explained variance of the model
Customer contacts (IFI) rises significantly from 53 per cent to 58 per
8 Little or no mistakes 0.76436 cent, as compared to the original model that
5 Customer treated respectfully 0.75255 did not include the different factors for qual-
7 Time arrangement 0.64440 ity and image.
10 Time and attention to customer 0.58235
According to the outcome of the factor
Advice (IF2)
analysis we also specify a model to test the
13 Advice on investment funds 0.82671
relationship between image and satisfaction.
4 Advice on insurances 0.68525
Here we incorporated the effects of quality:
6 Expertise in advising 0.66151
SAT = b0 + b1 * QF1... b7 * QF7+ b8 * IF1
Relationship driven (IF3)
+...+ b13 * IF6 + ε1 (Model 2.2)
15 Taking other than business aspects
into account 0.77410 where SAT = satisfaction; QF1 = factor scores
16 Efforts for local community 0.67089 quality perceptions first factor etc.; and QI1 =
17 Own interest not on the first plan 0.62373 image factor scores first factor of the bank.
Position in the market (IF4) The results of this analysis are shown in
1 New products or services 0.74323 Table VIII.
9 Attractive advertisements 0.61260 Table VIII shows that only reliability and
11 Strongest financial institutions of empathy have a significant positive impact on
the world 0.51652 satisfaction (beta = 0. 39 and beta = 0.32
2 Modern 0.49383 respectively, explained variance is 25 per
Society-driven (IF5) cent).
12 Financial institution for everyone 0.81116 We also expected an indirect positive effect
14 Strong commitment to society 0.60782 between the image factor scores of a bank and
Prices (IF6) the loyalty towards that bank through quality
3 Reasonable prices 0.74857 (i.e. a mediator effect). For this purpose we
Total explained variance 63.8% specify a model to test the relationship
Notes: a = Using principal axis factoring and varimax rotation; b = Factor loadings > 0.4 between image and quality.

[ 281 ]
Josée Bloemer, Ko de Ruyter QF1 = b0 + b1 * IF1 + ... + b6 * IF6 + ε1 are: reliability (beta = 0.51), satisfaction (beta
and Pascal Peeters (Model 2.3) = 0.38), position in the market (beta = 0.30),
Investigating drivers of bank QF2 = b0 + b1 * IF1 + ... + b6 * IF6 + ε1 efficiency (beta = 0.18), customer contacts
loyalty: the complex
(Model 2.4) (beta = 0.16), empathy (beta = 0.12) and soci-
relationship between image,
service quality and QF3 = b0 + b1 * IF1 + ... + b6 * IF6 + ε1 ety-driven (beta = 0.02) (taking into account
satisfaction (Model 2.5) both direct and indirect effects).
International Journal of where QF1 = factor scores quality percep-
Bank Marketing
16/7 [1998] 276–286 tions first factor etc.; and IF1 = factor scores Table VII
image of the bank first factor etc. The results Results of regressions analysis with factor
of this analysis are shown in Table IX. scores according to model 2.1a
Table IX shows that customer contacts have
a significant positive influence on reliability b beta p
(beta = 0.32; explained significant variance 11 Intercept 0.54 0.05
per cent); that customer contacts and society- SAT 0.38 0.38 0.00
driven have a significant positive influence QF1 0.36 0.36 0.00
on empathy (beta = 0.20 and beta = 0.18 QF2 not significant (p > 0.05)
respectively, explained significant variance QF3 0.14 0.18 0.01
seven per cent) and that position in the mar- QF4 not significant (p > 0.05)
ket has a significant positive impact on effi- QF5 not significant (p > 0.05)
ciency (beta = 0.16, explained significant QF6 not significant (p > 0.05)
variance three per cent). QF7 not significant (p > 0.05)
The results of the former analyses in terms IF1 not significant (p > 0.05)
of the significant beta coefficients are IF2 not significant (p > 0.05)
depicted in Figure 2. IF3 not significant (p > 0.05)
Figure 2 shows that, from most to least IF4 0.27 0.30 0.00
important, the major determinants of loyalty IF5 not significant (p > 0.05)
IF6 not significant (p > 0.05)
R2 0.58
Table VI Notes: aLOY = bo + b1 * SAT + b2 * QF1 + ...+ b8 * QF7
Result factor analysisa on quality items + b9 * IF1 + ... + b14 * IF6 + ε1;
IF1 ... IF6 = factor scores image; QF1 ... QF6 = factor
Item Description Factor loadingsb scores of quality perceptions; SAT = overall satisfaction;
Reliability (QFI) LOY = loyalty

10 Accuracy of employees 0.69945


7 Expertise of employees 0.59682
6 Handling complaints 0.57773
16 Personalised consulting 0.57432 Table VIII
14 Privacy at the office 0.40249 Results of regressions analysis with factor
18 Proactive suggestions scores according to model 2.2a
Empathy (QF2)
17 Attention of employees 0.74234 b beta p
9 Kindness of employees 0.73697 Intercept 3.51 0.00
13 Efforts for the customer 0.58296 QF1 0.39 0.39 0.00
11 Recognition by employees 0.40747 QF2 0.32 0.32 0.00
Efficiency (QF3) QF3 not significant (p > 0.05)
15 Queuing time at the office 0.79880 QF4 not significant (p > 0.05)
8 Speed of handling at the office 0.75316 QF5 not significant (p > 0.05)
Interest rates (QF4) QF6 not significant (p > 0.05)
4 Level of savings rates 0.75083 QF7 not significant (p > 0.05)
5 Level of mortgage rates 0.74595 IF1 not significant (p > 0.05)
Procedures (QF5) IF2 not significant (p > 0.05)
1 Sending invoices 0.76423 IF3 not significant (p > 0.05)
3 Cost of using an account 0.66828 IF4 not significant (p > 0.05)
Expertise (QF6) IF5 not significant (p > 0.05)
19 Expertise on investment funds 0.86090 IF6 not significant (p > 0.05)
Access to money (QF7) R2 0.25
2 Hours of opening 0.73091 Notes: aSAT = b0 + b1 * QF1 + ... + b7 * QF7 + b8 * IF1
12 Trouble-free cash dispenser –0.55281 + ... + b13 * IF6 + ε1;
Total explained variance 58.7% IF1 ... IF6 = factor scores image; QF1 ... QF6 = factor
Notes: a = Using principal axis factoring and varimax rotation; b = Factor loadings > 0.4 scores of quality perceptions; SAT = overall satisfaction

[ 282 ]
Josée Bloemer, Ko de Ruyter Table IX we are able to obtain a more detailed insight
and Pascal Peeters Results of regressions analysis with factor into the relationship between image and
Investigating drivers of bank quality on the one hand and loyalty on the
loyalty: the complex scores according to models 2.3a, 2.4b and 2.5c
other. In essence, our conclusion is that all
relationship between image,
service quality and b beta p three constructs (i.e. image, quality and satis-
satisfaction Model 2.3a faction) exert an influence on customer loy-
International Journal of Intercept not significant (p > 0.05) alty with banks. Furthermore, we encoun-
Bank Marketing IF1 0.32 0.32 0.00 tered a relatively strong relationship between
16/7 [1998] 276–286 IF2 not significant (p > 0.05) reliability and satisfaction. This points to the
IF3 not significant (p > 0.05) relative importance of the reliability factor in
IF4 not significant (p > 0.05) retail banking. Finally, we found that factors
IF5 not significant (p > 0.05) like society-driven, empathy and customer
IF6 not significant (p > 0.05) contacts have an indirect effect on loyalty via
R2 0.11 satisfaction and quality. It should also be
Model 2.4b noted that customer contacts have a direct
Intercept not significant (p > 0.05) impact on empathy.
IF1 0.18 0.20 0.01 A number of theoretical shortcomings
IF2 not significant (p > 0.05) follow from our research, suggesting a num-
IF3 not significant (p > 0.05) ber of issues that merit further research.
IF4 not significant (p > 0.05) First of all, it seems important to validate the
IF5 0.15 0.18 0.02 distinction between image and quality in
IF6 not significant (p > 0.05) other financial services settings (e.g. insur-
R2 0.07 ance, corporate banking) and to verify
Model 2.5c whether the multi-dimensional approach
Intercept not significant (p > 0.05) taken here is valid for these settings. Second,
IF1 not significant (p > 0.05) this study was limited to loyalty as an expres-
IF2 not significant (p > 0.05) sion of consumer preference. Future research
IF3 not significant (p > 0.05) should focus on other attitudinal and behav-
IF4 0.18 0.16 0.03 ioural outcomes, such as word-of-mouth com-
IF5 not significant (p > 0.05) munication, the compositions of the evoked
IF6 not significant (p > 0.05) set and information search behaviour, and
R2 0.03 investigate whether dimensions of satisfac-
tion also has an effect on these consequences.
Notes: Additional research is needed to investigate
aQF1 = b + b * IF1 + ... + b * IF6 + ε ; whether a distinction can be made between
0 1 6 1
bQF2 = b + b * IF1 + ... + b * IF6 + ε ; cognitive and affective dimensions of satisfac-
0 1 6 1
cQF3 = b + b * IF1 + ... + b * IF6 + ε; tion. Third, the literature on customer-firm
0 1 6
IF1 ... IF6 = factor scores image; relationships has suggested various types of
QF1 ... QF6 = factor scores of quality perceptions commitment, such as affective, calculative
and moral commitment (Allen and Meyer,
1990; Kumar et al., 1994). The obvious implica-
tion would be to investigate whether the type
Theoretical implications of commitment to a bank can further nuance
An important implication of our results is the satisfaction-image-quality-loyalty rela-
that analysis on the level of the various tionship. Fourth, as our study replicates and
underlying dimensions of elusive and diffi- extends findings from the consumer product
cult to operationalise constructs like image literature to bank image, bank quality and
and quality provides additional insight into bank satisfaction and loyalty, additional
the relationship among them. Our analysis research is required to test our model in busi-
shows that, although we did not find a direct ness-to-business relationships. Obviously, the
relationship between the overall image con- external validity of our findings needs addi-
struct and loyalty, one distinct dimension tional attention in terms of the replication of
(position in the market) has a direct positive our study in a similar research setting. Fifth,
effect on customer loyalty. Likewise, in although we found significant relationships,
addition to the direct positive effect of the it should be taken into account that the levels
overall quality construct, we are now able to of variance explained are relatively modest.
nuance that finding as it becomes clear that Further research is needed to gain additional
specifically reliability and efficiency could be insight into the explanation of bank loyalty.
viewed as important determinants of loyalty. Finally, all constructs were measured at one
Therefore, our initial rejection of H1 could be point in time, thus essentially from a static
revoked on the basis of the multi-dimensional perspective. It may be worthwhile to study
approach that was taken subsequently. Thus, bank loyalty over time, in order to be able to
[ 283 ]
Josée Bloemer, Ko de Ruyter Figure 2
and Pascal Peeters Model based on factor analyses in combination with ordinary multivariate least squares analyses
Investigating drivers of bank
loyalty: the complex
relationship between image,
service quality and
satisfaction β = .18 β = .32
Society
International Journal of driven
Empathy Satisfaction β = .38
Bank Marketing
16/7 [1998] 276–286
Loyalty
β = .20 β = .39

β = .32
Customer
Reliability β = .36
contacts

β = .18

β = .30
Efficiency

β = .16

Position in
the market

take into account the dynamics in consumer evaluate the bank in terms of the relative
patronage behaviour. Should such an position it has within the market place. After
approach be taken, then measures of actual all, retail banking is very much a service
behaviour and bank objective performance depending on credence properties. Therefore,
(e.g. switching behaviour, vulnerability to the use of corporate advertising creating the
price competition, turnover, relative market perception of a strong financial institution,
share), in addition to perceptual gauges, with innovative products and services and
could be taken into account. modern facilities, seems important for the
establishment of customer loyalty in retail
banking.
Managerial implications A third managerial implication is that
satisfaction is not the sole determinant of
In terms of the practical relevance of our customer loyalty in retail banking. Many
research, a number of managerial implica- banks have a customer satisfaction measure-
tions may be derived. First of all, reliability ment program, providing customer feedback.
seems to be the most important factor influ- Our study shows that although there is a
encing customer loyalty with banks. Looking direct positive relationship between satisfac-
at the individual indicators of this quality tion and loyalty, other determinants play an
dimension, it follows that banks should invest important role too. Just focusing on satisfac-
in monitoring employees in order to make a tion may result in overlooking other impor-
trustworthy impression on the customers, tant drivers of customer loyalty.
both in the case of the general service Fourth, efficiency, i.e. queuing time and
encounter, as well as in the handling of cus- speed of handling, has a direct influence on
tomer complaints. In the dialogue with bank loyalty. Customers are not willing to spend
customers, management and employees their valuable time waiting for services.
should strive to find out what customers Banks that take this into account promote
expect in terms of accuracy, expertise, com- loyalty among their customers. Although
plaint handling and proactive suggestions. customer contacts, empathy and society-
This implies an extensive and continuous driven seem to be of lower order importance,
training program. they still have an indirect effect on customer
Following from the importance of the relia- loyalty and should be handled with
bility dimension, customers apparently look “customer” care. Finally, caution should be
for “external cues” in order to be able to taken with regards to the impact of image,

[ 284 ]
Josée Bloemer, Ko de Ruyter which seems to be a construct that has a rela- Dick, A.S. and Basu, K. (1994), “Customer loyalty:
and Pascal Peeters tively low direct impact on bank loyalty, espe- toward an integrated conceptual framework”,
Investigating drivers of bank cially when one considers the influence of the Journal of the Academy of Marketing Science,
loyalty: the complex individual dimensions of image. Vol. 22, pp. 99-113.
relationship between image, Doyle, P. and Fenwick, I. (1974), “Shopping habits
service quality and
satisfaction in grocery chains”, Journal of Retailing, Vol.
Note 50, pp. 39-52.
International Journal of
Dwyer, R.F., Schurr, P.H. and Oh, S. (1987), “Devel-
Bank Marketing
1. Additional information on the data set may be oping buyer-seller relationships”, Journal of
16/7 [1998] 276–286
obtained from the authors. Marketing, Vol. 51, pp. 11-27.
Granbois, D. (1981), “An integrated view of the
store choice/patronage process”, in
References Monroe, K.B. (Ed.), Advances in Consumer
Research, Vol. 8, Association for Consumer
Allen, N.J. and Meyer, J.P. (1990), “The measure-
ment and antecedents of affective, continu- Research, Ann Arbor, MI, pp. 693-5.
ance, and normative commitment to the orga- Gremler, D.D. and Brown, S.W. (1996), “Service
nization”, Journal of Occupational loyalty; its nature, importance and implica-
Psychology, Vol. 63, pp. 1-18. tions”, in Edvardsson, B., Brown, S.W., John-
Berry, L.L. (1983), “Relationship marketing”, in ston, R. and Scheuing, E. (Eds), QUIS V:
Berry, L.L., Shostack, G.L. and Upah, G.D. Advancing Service Quality: A Global Perspec-
(Eds), Perspectives on Services Marketing, tive, ISQA, New York, NY, pp. 171-81.
American Marketing Association, Chicago, Guiltinan, J.P. (1989), “A classification of switch-
IL. ing costs with implications for relationship
Bloemer, J.M.M. and Kasper, J.D.P. (1995), “The marketing”, in Childers, T.L. and Bagozzi, R.P.
complex relationship between consumer (Eds), Proceedings of AMA Winter Educators’
satisfaction and brand loyalty”, Journal of Conference: Marketing Theory and Practice,
Economic Psychology, Vol. 16, pp. 311-29. AMA, Chicago, IL, pp. 216-20.
Brislin, R. W. (1980), “Translation and content Hildebrandt, L. (1988), “Store image and the pre-
analysis of oral and written materials”, in diction of performance on retailing”, Journal
Triandis, H.C. and Berry, J.W. (Eds), Handbook of Business Research, Vol. 17, pp. 91-100.
of Cross-Cultural Psychology: Methodology, Jacoby, J.W. and Chestnut, R.W. (1978), Brand
Volume 2, Allyn & Bacon, Boston, MA. Loyalty Measurement and Management,
Cronin, J.J. and Taylor, S.A. (1992), “Measuring John Wiley & Sons, New York, NY.
service quality: a re-examination and exten- Jain, A.K., Pinson, C. and Malhotra, N.K. (1987),
sion”, Journal of Marketing, Vol. 56, pp. 55-68. “Customer loyalty as a construct in the mar-
Crosby, L.A., Evans, K.R. and Cowles, D. (1990), keting of bank services”, International Jour-
“Relationship quality in services selling: an nal of Bank Marketing, Vol. 5 No. 3, pp. 48-72.
interpersonal influence perspective”, Journal James, D.L., Durand, R.M. and Dreves, R.A. (1976),
of Marketing, Vol. 54, pp. 68-81. “The use of a multi-attributes attitudes model
Czepiel, J.A. (1990), “Service encounters and in a store image study”, Journal of Retailing,
service relationships: implications for Vol. 52, pp. 23-32.
research”, Journal of Business Research, Vol. Keaveney, S.M. (1995), “Customer switching
20, pp. 13-21. behaviour in service industries: an
Czepiel, J.A. and Gilmore, R. (1987), “Exploring exploratory study”, Journal of Marketing,
the concept of loyalty in services”, in Vol. 59, pp. 71-82.
Czepiel, J. A., Congram, C.A. and Shanahan, J. Keaveney, S.M. and Hunt, K.A. (1992), “Conceptu-
(Eds), The Services Marketing Challenge: Inte- alization and operationalization of retail store
grating for Competitive Advantage, AMA, image: a case of rival middle-level theories”,
Chicago, IL, pp. 91-4. Journal of the Academy of Marketing Science,
Darden, W.R. and Schwinghammer, J.K.L. (1985), Vol. 20 No. 2, pp. 165-75.
“The influence of social characteristics on Kiesler, C.A. (1968), “Commitment”, in
perceived quality in patronage choice behav- Abelson, P.A. et al. (Eds), Theories of Cognitive
ior”, in Jacoby, J. and Olson, J. (Eds), Perceived Consistency: a Source Book, Rand Mc Nally,
Quality: How Consumers View Stores and Mer- Chicago, IL.
chandise, Lexington Books, Lexington, MA, Klemperer, P. (1987), “Markets with consumer
pp. 161-72. switching costs”, The Quarterly Journal of
Day, G.S. (1969), “A two-dimensional concept of Economics, Vol. 102, pp. 375-94.
brand loyalty”, Journal of Advertising Korgaonkar, P.K., Lund, D. and Price, B. (1985), “A
Research, Vol. 9, pp. 29-36. structural equations approach toward exami-
De Ruyter, J.C., Bloemer, J.M.A. and Peters, P. nation of store attitude and store patronage
(1997), “Merging service quality and service behavior”, Journal of Retailing, Vol. 61, Sum-
satisfaction: an empirical test of an integrative mer, pp. 39-60.
framework”, Journal of Economic Psychology, Kumar, N., Hubbard, J.D. and Stern, L.W. (1994),
Vol. 18, pp. 387-406. The Nature and Consequences of Marketing

[ 285 ]
Josée Bloemer, Ko de Ruyter Channel Intermediary Commitment, No. 94-115, Patterson, P.G. and Johnson, L.W. (1993), “Discon-
and Pascal Peeters Marketing Science Institute, Cambridge. firmation of expectations and the gap model of
Investigating drivers of bank Kunkel, J.H. and Berry, L.L. (1968), “A behavioral service quality: an integrated paradigm”,
loyalty: the complex conception of retail image”, Journal of Market- Journal of Consumer Satisfaction, Dissatisfac-
relationship between image, ing, Vol. 32, pp. 21-7. tion and Complaining Behavior, Vol. 6, pp. 90-9.
service quality and
Lastovicka, J.L. and Gardner, D.M. (1977), “Com- Render, B. and O’Connor, T.S. (1976), “The influ-
satisfaction
ponents of involvement”, in Maloney, J.C. and ence of price, store name and brand name on
International Journal of
Silverman, B. (Eds), Attitude Research Plays perception of product quality”, Journal of the
Bank Marketing
16/7 [1998] 276–286 for High Stakes, American Marketing Associa- Academy of Marketing Science, Vol. 4, Fall,
tion, Chicago, IL. pp. 722-30.
Levesque, T. and McDougall, G.H.C. (1996), “Deter- Rust, R.T. and Oliver, R.L. (1994), “Service quality:
minants of customer satisfaction in retail insights and managerial implications from the
banking”, International Journal of Bank Mar- frontier”, in Rust, R.T. and Oliver, R.L. (Eds),
keting, Vol. 14 No. 7, pp. 12-20. Service Quality: New Directions in Theory and
Lewis, B.R. (1993), “Service quality: recent devel- Practice, Sage, London, pp. 1-19.
opments in financial services”, International Shemwell, D.J., Cronin, J.J. and Bullard, W.R.
Journal of Bank Marketing, Vol. 11 No. 6, pp. (1994), “Relational exchange in services: an
19-25. empirical investigation of ongoing customer
Lewis, B.R. and Mitchell, V-W. (1990), “Defining service-provider relationships”, International
and measuring the quality of customer ser- Journal of Service Industry Management, Vol. 5,
vice”, Marketing Intelligence and Planning, pp. 57-68.
Vol. 8 No. 6, pp. 11-17. Sirgy, M.J. and Samli, A.C. (1989), “A path analytic
Liljander, V. and Strandvik, T. (1995), “The nature
model of store loyalty involving self-concept,
of customer relationships in services”, in
store image, geographic loyalty, and socioeco-
Swartz, T.A., Bowen, D. and Brown, S.W. (Eds),
nomic status”, Journal of the Academy of Mar-
Advances in Services Marketing Management,
keting Science, Vol. 13 No. 3, pp. 265-91.
Vol. 4, pp. 141-67.
Smith, A.M. (1992), “The consumers’ evaluation of
Mazursky, D. and Jacoby, J. (1986), “Exploring the
service quality: some methodological issues”,
development of store images”, Journal of
in Whitelock, J. (Ed.), Proceedings of the 1992
Retailing, Vol. 62, pp. 145-65.
Annual Conference, MEG, University of Salford,
Meidan, A. (1996), Marketing Financial Services,
Marketing in the New Europe and B eyond, pp.
MacMillan Press, Houndmills.
633-48.
Moorman, C., Zaltman, G. and Despande, R.
Snyder, D.R. (1986), “Service loyalty and its mea-
(1992), “Relationships between providers and
surement: a preliminary investigation”, in
users of marketing research: the dynamics of
Venkatesan, M., Schmalensee, D.M. and Mar-
trust within and between organizations”,
Journal of Marketing Research, Vol. 29, shall, C. (Eds), Creativity in Service Marketing:
pp. 314-29. What’s New, What Works, What’s Developing,
Morgan, R.M. and Hunt, S.D. (1994), “The commit- AMA, Chicago, IL, pp. 44-8.
ment-trust theory of relationship marketing”, Stafford, J.E. and Enis, B.M. (1969), “The price-
Journal of Marketing, Vol. 58, pp. 20-38. quality relationship: an extension”, Journal of
Murphy, J.A. (1996), “Retail banking”, in Buttle, F. Marketing Research, Vol. 6, November, pp. 456-8.
(Ed.), Relationship Marketing, Theory and Surprenant, C.F. and Solomon, M.R. (1987), “Pre-
Practice, Paul Chapman, London, pp. 74-90. dictability and personalization in the service
Oliver, R.L. (1993), “A conceptual model of service encounter”, Journal of Marketing, Vol. 51,
quality and service satisfaction: compatible pp. 86-96.
goals, different concepts”, in Swartz, A.T., Worcester, R.M. (1997), “Managing the image of
Bowen D.E. and Brown, S.W. (Eds), Advances in your bank: the glue that binds”, International
Services Marketing Management, Vol. 2, JAI Journal of Bank Marketing, Vol. 15 No. 5,
Press, Greenwich, CT, pp. 68-85. pp. 146-52.
Osman, M.Z. (1993), “A conceptual model of retail Yavas, U. and Shemwell, D.J. (1996), “Bank image:
image influences on loyalty patronage behav- exposition and illustration of correspondence
ior”, The International Review of Retail, Distrib- analysis”, International Journal of Bank Mar-
ution and Consumer Research, Vol. 31, pp. 149-66. keting, Vol. 14 No. 1, pp. 15-21.
Parasuraman, A., Zeithaml, V. and Berry, L. (1988), Zeithaml, V.A. (1981), “How consumer evaluation
“SERVQUAL: a multiple-item scale for measur- processes differ between goods and services”,
ing consumer perceptions of service quality”, in Donnelly, J.H. and George, W.R. (Eds), Mar-
Journal of Retailing, Vol. 64, Spring, pp. 12-40. keting of Services, AMA, Chicago, IL.

[ 286 ]