The current issue and full text archive of this journal is available at

http://www.emerald-library.com/ft

Developing a comprehensive Comprehensive

picture of service
picture of service failure failure
Mark Colgate
Department of Marketing, University of Auckland, New Zealand, and 215
Melissa Norris
Interlock, Auckland, New Zealand Received November 1999
Revised August 2000
Keywords Service levels, Services, Loyalty, Banking Accepted January 2001

Abstract Much attention has been paid recently to the concept of service failure. In light of
this, the paper develops a model of the potential outcomes from service failure. Results are
utilised from interviews with business banking customers who have recently encountered a
service failure. The results show that service recovery is only one of the reasons a customer may
stay or exit a service organisation after a service failure. Other factors, such as barriers to exit
and loyalty, are just as prevalent in the decision-making process. A model of service failure,
which was generated through the results of the research, is presented in the discussion section
of the paper.

Introduction
Much research has focused on how to improve a service provider's service
quality, and how to attract and maintain customers (particularly with the use of
service recovery techniques). A major factor that affects a customer's
perception of a service provider's quality is the number, and extent, of the
problems they encounter, and how these problems are handled by the provider.
As services are inherently variable in how they are conducted, it is to be
expected that problems will occur (Hart et al., 1990, p. 255). A term that is used
synonymously for a problem that a customer has with a service is ``service
failure'' (Spreng et al., 1995).
Customers react in two ways when they encounter a service failure. They
either remain with the service provider or leave (i.e. exit). To expand this
further, when customers encounter a service failure they can either not
complain, or complain and give the service provider an opportunity to
rectify the problem. If they choose not to complain they may remain with the
provider, despite their dissatisfaction, or exit. Alternatively, if they
complain they may also choose to stay or they may exit, and this is
influenced by how the situation is handled by the service provider (i.e. the
service recovery).
In the case where a service provider is notified that a problem has occurred
(i.e. a customer complains), it is necessary for service recovery to be conducted.
Service recovery includes all actions taken by a service provider in order to try
to resolve the problem a customer has with their organisation (Gronroos, 1990).
There is a strong focus in the literature on managerial ways to improve service International Journal of Service
recovery. For example, much research looks at what aspects of service recovery Industry Management,
Vol. 12 No. 3, 2001, pp. 215-233.
encourage a customer to stay (Bell and Ridge, 1992; Zemke, 1994). # MCB University Press, 0956-4233
IJSIM In fact, a thorough investigation of the literature reveals that research into
12,3 the service recovery process has reached a relatively sophisticated level
(Boshoff and Leong, 1998; Tax et al., 1998). In fact the plethora of literature,
both managerial and academic, on service recovery leads us to believe that
service recovery is the most important factor in driving a customer to exit or
stay after a service failure. This may not necessarily be the case, and further
216 investigations into the outcomes of service failure are required.
In light of this, the aim of this paper is to generate a model of how customers
react to service failure by gaining a more detailed understanding of what drives
a customer to remain or exit after they have encountered a service failure with
their service provider ± in the same service context. This is important because
by analysing customers in the same service context (i.e. same industry, same
segment of customers) we obtain more comparative results. The extant
literature does not investigate this area fully, but instead tends to focus on one
dimension of this research field (e.g. service recovery or exit). This leads us to
the research question: Why do some customers exit their service provider after
a service failure while others stay?
In order to answer this question the paper is broken down as follows. First,
the literature on all aspects of service failure is examined. Second, propositions
for the study are developed based on the literature review. Third, the results of
interviews with customers who have stayed after a service failure and those
customers who have exited are discussed. Next, a model based on the results of
the interviews is developed which aims to present a more complete picture of
the outcomes of service failure. Finally, the implications of this research, both
academic and managerial, are presented.

Service failure
Throughout the literature the differing reactions customers may have to
service failure have been discussed. These include loyalty, voice (e.g.
complaining to the service provider, or another party) and exit. These are
discussed in turn.

Loyalty
Loyalty is developed over a period of time from a consistent record of meeting,
and sometimes even exceeding, customer expectations (Teich, 1997). Loyal
customers are beneficial to service providers because they, ``often cost less to
service, spend more as their time with the firm lengthens, and provide a good
source for new business'' (Levesque and McDougall, 1993, p. 40). They usually
lead to lower marketing costs, more efficient operations and higher profits
(Mittal and Lassar, 1998).
However, loyalty comes in several differing forms and these include: pure
loyalty, latent loyalty and spurious loyalty (Dick and Basu, 1994). Pure
loyalty is patronising a company often, as well as thinking highly of it. This
is the kind of loyalty most people associate with the term. Latent loyalty is
when a customer has a low repeat patronage but thinks highly of the
organisation. Spurious loyalty, however, is the opposite of this. It occurs Comprehensive
when a customer has a high repeat patronage but a relatively low attitude to picture of service
the company. Hirschman (1970) captures this notion in the definition of failure
loyalty in the exit, voice and loyalty model ± it is remaining with a company
despite dissatisfaction. It is the notion of spurious loyalty that this paper is
most interested in, as it has been neglected by managers and academics
alike. 217
Initially, the choice by a spuriously loyal customer to remain with a
service provider does not immediately threaten profits, but with continued
dissatisfaction they may spread negative word of mouth or exit (Hirschman,
1970). As Colgate et al. (1996) noted, it is not always the case that customer
defection is the inverse to loyalty ± companies need to understand why a
customer still patronises them, as it may not necessarily be a positive,
conscious choice. Customers may stay with an organisation even if they are
dissatisfied because they perceive they have no choice (Hocutt et al., 1997;
Holmlund and Kock, 1996; Mittal and Lassar, 1998; Zeithaml et al., 1996).
Hence, a loyal customer may not necessarily be a satisfied customer.
The findings from the study conducted by Levesque and McDougall (1993,
p. 52) suggested that, ``even when a problem is not solved, approximately half
of the respondents would remain with the firm''. This may be due to switching
costs, lack of perceived differentiation of alternatives, locational constraints on
choice, time or money constraints, habit or inertia (Bitner, 1990; Ennew and
Binks, 1996). Hirschman (1970) also proposes the idea that dissatisfied
customers may choose to remain with a service provider in the belief that the
likelihood of an improvement outweighs the cost of searching for another
supplier.
Loyalty is one of the responses a customer has to a service failure, the
literature suggests. However, we should not believe that loyalty is one
dimensional. Customers may stay with a service provider because they have a
high opinion of them and see the service failure as an aberration. Customers
may also stay after a service failure because they are spuriously loyal. That
is, they feel trapped, are apathetic or there are no alternatives so they do not
leave.

Voice
The terms voice and exit were introduced by Hirschman (1970), when he
suggested that dissatisfaction can provoke two active negative responses; voice
and exit. Day and Landon (1977) extended the notion of ``voice'' further by
explaining that voice can be complaining to the service provider, complaining
to acquaintances (negative word of mouth), or complaining to third parties in
order to help seek redress.
Any problem a customer has with their service provider will arouse some
emotion whether they complain about it to the provider or not. By complaining
to the service provider, customers give them a chance to rectify the problem.
However, a customer must have sufficiently strong emotions about a situation
IJSIM before they make an effort to complain. High transaction costs are another
12,3 principal reason why consumers choose not to complain.
These [costs] include the time, cost, and effort involved, uncertainty about how to complain,
and beliefs that redress will not be provided (Day and Landon, 1977).

Service providers must provide open lines of communication if they want

218 customers to complain, then they can try and solve their problem(s). If a
customer voices by any other means (e.g. complaining to a business colleague)
or exits, service providers will find it difficult to find out why those customers
are unhappy (Levesque and McDougall, 1993).
Customer complaints (or lack of) has been identified, therefore, by the
literature as an integral part of service failure. A customer who does not
complain may exit and give the organisation no chance to ``save'' the customer,
or stay and ``fester'' ± neither is desirable. A complaining customer, however,
gives the company an opportunity and should be seen as such.

Stay or exit?
At the end of the day a customer's final decision after a service failure is to
reside or leave. Exit is ``the voluntary termination of an exchange relationship''
(Singh, 1990, p. 2). In order to exit in response to a service failure, it requires the
customer to have motivation and expend effort (Singh, 1990; Stewart, 1998).
Customer exit is important because it can cost the seller a great deal of revenue
through decreased income from the customer's future revenue, higher costs in
attracting new customers, loss of free advertising through word of mouth, and
decreased employee retention (Reichheld and Sasser, 1990; Keaveney, 1995;
Zeithaml and Bitner, 1996).
The benefits of customer retention have also been emphasized. For example,
Reichheld and Sasser (1990) found that by retaining 5 per cent more customers,
a service provider can increase profits by almost 100 per cent.
Many factors influence a customer's reaction when they face a problem with
their service provider. This paper tries to uncover what these factors are and
how they interact, leading to the decision by the customer to leave or stay.

Service recovery
Of the above reactions to service failure, voice, in respect of complaining to
the service provider, creates the potential for the customer and the
organisation to resolve the problem that led to the service failure (the service
recovery process). Gronroos (1990) defined service recovery processes as
``those activities in which a company engages to address a customer
complaint regarding a perceived service failure''(p. 7) . It encompasses ``all the
actions taken to get a disappointed customer back to a state of satisfaction''
(Bell, 1994, p. 49).
There are several reasons posited throughout the literature that explain the
increasing trend of companies to invest in service recovery efforts (Holmlund
and Kock, 1996). If a service recovery situation is not regarded as important or
handled well, customer satisfaction can be eroded and ultimately the service Comprehensive
provider's profitability may diminish (Bailey, 1994). Results from Boshoff picture of service
(1997) support this, and he showed that poor service delivery cannot be failure
compensated for by great service recovery, although service recovery can help
significantly in limiting the poor service delivery's harmful impact.

Models on the service recovery process 219

There have been advances made in the literature in terms of modelling
customer exit and service recovery. Two models have been identified that
directly relate to this paper. Stewart's (1998) customer exit process model, and
Boshoff's (1997) service recovery model.
Stewart (1998) developed a comprehensive model of the customer exit
process. It is designed to help management understand the process. She notes
that customers usually engage in customer complaints prior to exit and go
through a whole series of emotions. The availability of alternatives and the
relationship the customer has with the bank are also considered (as well as the
complaint handling) before exit is invoked.
Stewart (1998) suggests that by the time a customer has decided to exit an
organisation it is too late to recover them. This is often because they are leaving
due to the unresponsiveness of the organisation, so when the organisation does
respond when they have decided to leave, this is viewed as ironic.
Stewart's model, although most insightful, focuses solely on a customer's
decision to exit, and it needs expansion when applied to the service failure
process as a whole.
Boshoff (1997) introduced a model of service recovery, not just about the
customer exit process. He attempted to bring together all the aspects of service
recovery identified by earlier authors. The model incorporates Hirschman's
(1970) exit, voice and loyalty, Gilbert et al.'s (1992) confirmation/
disconfirmation theory, and Zemke's (1994) loyalty and word of mouth theory.
The model also depicts, to some extent, the different types of reactions
customers have to dissatisfaction. These are:
. Those who have seriously considered leaving the service organisation
and do not complain but value loyalty, so they stay.
. Those who also do not complain, but just exit the service organisation.
. Those customers who complain and, depending on the outcome of their
complaint, they either stay due to satisfaction or loyalty, or leave.
Despite the importance of these models, other research presented suggests they
are incomplete. There are some permutations that seem to exist that are not
included in the models. This research, therefore attempts to tie all these
disparate areas of research together ± only then can the outcomes of service
failure be fully understood.
IJSIM Service failure outcomes
12,3 A customer may stay after they have experienced a problem because they have
complained and they are satisfied with the service recovery process. This is the
ideal situation for service providers who plan for good service recovery (Zemke,
1993). However, many times customers stay even though they are dissatisfied
with the service recovery they have received. One of these situations stems
220 from spurious loyalty.
This spurious loyalty may be due to their perception that the barriers to exit
are high, and/or their perception that the quality of alternative providers is no
better than their current provider. These two factors combine to create a body
of ``prospective switchers'' ± customers who stay even though they are
dissatisfied, but given a change in the rules may leave immediately (e.g. more
alternatives come along).
The other outcome, when a customer experiences a service failure, is to
leave. This may be as a result of complaining and experiencing the service
provider's recovery efforts. These recovery efforts may be poor, causing the
customer to be even more dissatisfied and leave (Sheth and Parvatiyar, 1995).
Alternatively, the customer could be satisfied with the recovery they have
received but still exit. This could be due to a perceived lack of barriers to exit
by the customer, a sense of apathy as to which provider they use, or that the
recovery could not make up for the incident(s) they have experienced.
A further reason customers may leave, even though they are satisfied with
the recovery they have received, may be due to variety seeking (Rust and
Zahorik, 1993). Hence, service recovery may not be the main reason why a
customer remains or exits, it appears to be one of several reasons.
Similarly, service providers do not always get the opportunity to conduct
service recovery, as they are not always aware that a problem has occurred
(TARP, 1995). The customer responses here are the same as those after service
recovery efforts; that is, a customer may choose to remain or exit.
Hence, customers may stay after they have experienced a problem even
though they have not complained, and have not given the service provider an
opportunity to conduct service recovery. This could be because of the loyalty
they have towards their service provider, or because of spurious loyalty. Also, a
customer may leave without complaining, not giving the service provider an
opportunity to solve the problem and ease the customer's dissatisfaction.
In light of the literature review, six propositions are presented:
P1: There will be times when customers exit without complaining, thus
offering the service provider no chance to conduct service recovery.
P2: There will be times when customers exit after they have complained
because have received poor service recovery.
P3: There will be times when customers exit after they have complained
even though they have received good service recovery.
P4: There will be times when customers stay with a service provider even
though they encountered a service failure and did not complain.
 P5: There will be times when customers stay with a service provider when Comprehensive
they complained and received good service recovery. picture of service
P6: There will be times when customers stay with a service provider even failure
though they complained and received poor service recovery.
These can been seen in Figure 1.
221
Methodology
Qualitative research is the most appropriate method for these propositions, as
research into the service failure process is inconclusive (hence the propositions)
and additional outcomes of service failure may exist given the absence of
research that specifically looks at this area. Customers also need to be given the
opportunity to tell their story and explain the situation from their point of view
(Perrien and Ricard, 1995). Quantitative research does not allow customers to
express their true feelings about a situation, so it may not be possible to
understand what has really happened without the use of qualitative research
(Perrien et al., 1994).
The domain of this research is small business customers in the banking
industry. One reason for choosing this industry is because banking has a
strong relationship/loyalty aspect to it. This makes the customer's decision to
exit a harder, more thought-involving process (Ennew and Binks, 1996).
Second, banking relationships are particularly important to small business
because of their importance in the provision of finance:
The nature of the overall banking relationship is arguably of particular importance in the
context of small businesses because of the problems of asymmetric information and the
potential for credit restraints as a result of information asymmetries (Stiglitz and Weiss, 1981,
p. 398).

This highlights the importance of banking relationships for small businesses,

and this suggests that the decision to remain or leave after encountering a
problem will be well considered.
Small business customers are also a relatively unresearched group of
customers, making the findings from this research more important. Finally, the

Figure 1.
Propositions in context
of service failure
IJSIM banking industry is characterised by many service failures (Anonymous, 1996)
12,3 and so this would make it easier to collect data.
At this point it is important to understand what drives satisfaction in the
small business-bank relationship, in order to comprehend which events may
lead to service failure. According to Winstanley (1997, p. 38), in small business
relationships in banking there are four key factors driving customer
222 satisfaction:
(1) satisfaction with borrowing; (2) satisfaction with the relationship manager; (3) satisfaction
with the day-to-day service; and (4) satisfaction with value for money.

Hence, problems in any of these areas could lead to dissatisfaction. Winstanley

(1997) found that if the company is a borrower, the most important driver is
satisfaction with borrowing, whereas if the company is not a borrower, the
most important driver is satisfaction with the relationship manager. So for
borrowers, there are four components of satisfaction with borrowing, and these
are ``proactive, policy, structure and price'' (Winstanley, 1997, p. 40). The key
drivers under these components are ``availability of credit, anticipation of credit
needs, and flexibility in structuring loans'' (Winstanley, 1997, p. 40). For non-
borrowers, a customer ``wants a relationship manager who understands their
business needs, acts as an advocate for them at the bank, and resolves
problems'' (Winstanley, 1997, p. 41). The absence of these satisfaction drivers
may lead to service failures.
There are two different groups of respondents questioned in this study:
those who left their bank after a service failure, and those who remained with
their bank after seriously considering leaving (a service failure had to have
occurred). Stewart's (1998) definition was adopted (with a slight revision) for
this study, so that to leave meant that ``a customer's main account was no
longer with that bank'' (i.e. the customer had shifted the main account to
another provider). A main account is defined as, ``the account where most of
your transactions take place''. Someone who has remained is defined as a
customer who experienced a service failure and seriously considered leaving,
but who had finally decided not to move their main current account. Hence,
most of their transactions still occur with that service provider, and the
problem did not adversely affect the amount of business they had with that
provider.
Most of the studies conducted to date about service failure (or aspects of it),
tend to ask behavioural intentions questions in terms of what they expect will
happen in a service failure situation (e.g. Boshoff and Leong, 1998). This style
of research has been proven to be a less than accurate reflection of what
customers actually do when faced with the problem (cf. Zeithaml et al., 1996).
Therefore, this study investigates those customers who have actually remained
or exited based on their own situation, so that it reflects more accurately what
customers do.
A sample size of 20 respondents was selected, comprising ten customers
who had left their service provider recently after experiencing a service failure,
and ten customers who had remained with their service provider after recently Comprehensive
experiencing a service failure and seriously considering leaving. picture of service
The research was collected via a series of face-to-face, in-depth interviews, failure
which had been pre-arranged over the telephone. Approximately one in 16
companies contacted had encountered a service failure(s) large enough to make
them seriously consider leaving or actually leave their bank. Out of those
companies, one in four refused to participate in the research. Hence, a total of 223
435 small businesses were contacted by telephone, and if there was no reply or
the contact was busy, up to three call backs were made. A stratified random
sample was used and they were selected from a small business database
purchased from a third party.
All interviews were conducted at the offices of the business and were
recorded. Sample questions from the ``remain'' and ``exit'' interviews can be seen
in the Appendix. The results from these interviews were transcribed, and then
content analysed by using a number of methods, including checklist matrices,
``eyeballing'' and searches in NUDIST.
A checklist matrix is ``a format for analysing field data on a major variable
or general domain of interest'' (Miles and Huberman, 1994, p. 105). The aim of
this type of analysis is to note patterns or themes, and they are especially useful
when exploring new domains (as is the case in this study). The checklist
technique is a useful way to analyse the within-group and between-group
similarities and differences.
Another method of analysis that was used is a technique commonly referred
to as ``eyeballing''. This procedure requires transcriptions of the interviews to
be read and examined several times, with key phrases then highlighted and
assigned codes. The goal in selecting these phrases is to identify key thoughts,
ideas and perceptions from each respondent.
The final content analysis method used was the computer programme
NUDIST. This uses inductive techniques to decode qualitative data. The
programme was used to search for key words and phrases common to several
respondents as identified during eyeballing. For example, when searching for
emotions the respondents had mentioned during their interviews, the words
``annoyed'' and ``frustrated'' were commonly expressed. NUDIST was used to
find all the uses of these words and their derivatives.

Results
The ``remain'' group
The majority of the respondents in the ``remain'' group were emotional about
the service failure they had experienced, and this led most (80 per cent) of them
to complain. There was a general feeling of dissatisfaction with the service
recovery efforts made by the banks after the service failures, as out of the eight
respondents who gave their bank an opportunity to conduct service recovery,
there were only three who were satisfied with the recovery. These remaining
three were still not entirely satisfied after they had received the recovery, but
IJSIM were willing to acknowledge that the bank had made a concerted effort to
12,3 rectify the situation.
More than half of the respondents in the ``remain'' group felt that the
relationship they had with their bank was not close and this would have
contributed to the low sense of loyalty this group felt. This suggests that these
respondents remained for reasons other than loyalty.
224 The four respondents who had relatively close relationships with their bank
tended to have been with that bank for a long period of time (i.e. for more than
ten years) and they expressed moderate to strong levels of loyalty. One
interviewee went as far to say ``Basically, except for your shareholders, your
bank is maybe your second most important client''.
The majority of respondents in this group perceived high barriers to exit,
with seven out of the ten perceiving this, and an additional two respondents
perceived moderate barriers to exit. The barriers to exit that these customers
mentioned included: the amount of time and hassle it would take to change,
having to get new cheque books and transfer funds, and having to learn a new
cash management system. The amount of paperwork involved, issues with
existing loans and guarantees, and a good history and rapport with their
current bank also seemed to be factors. Typical quotes from interviewees on
this subject included: ``You've got to give financial information and history . . .
to other bankers out there, and there is confidentiality involved, and there
would be lots of costs involved in changing bank''. ``If we were to change it
would cause some upheaval'' and ``With a lot of our product coming in via a
letter of credit or agent's bills, it would have been a big exercise to change
banks in itself''.
The perception of high barriers to exit was accentuated by the overall
feeling by most of these respondents that the alternative banks were of the
same (or a similar) quality as their own. Quotes which illustrated this were:
``Are they going to be truly any better? Most of them are fairly equal in their
competition and charge''. ``A bank's a bank. They all do the same thing'' and
``We wouldn't really mind which bank we use. They all sort of supply the
same service''.
This is a compelling reason why these respondents had not switched to
another supplier, as they felt that they would be no better off by switching.
Three respondents used the quote ``the devil you know is better than the devil
you don't''. This would infer that they would prefer to remain with this supplier
rather than switch to an unknown supplier, as there is risk and uncertainty
involved.

The ``exit'' group

The majority of the respondents in the ``exit'' group complained to their bank
about the failure. Only three out of the eight respondents who complained to
their bank about the service failure(s) were satisfied with the recovery efforts of
their bank. Notably, two respondents in this group, who gave their bank a
chance to recover, received no recovery at all. When one of the respondents
wrote a letter complaining, they said they got ``no response ± not even a word''. Comprehensive
This lack of recovery will have contributed to their decision to exit. picture of service
Similarly, some of the service failures were so bad that even good recovery failure
did not change their mind regarding their decision to switch. For example, one
interviewee stated that ``They [the bank] came back with a counter proposal
and apology and they didn't want to lose us, but by then it was too late. We
didn't really feel that they were going to do anything that was any different 225
from what they had done in the past''.
Nine of the ten respondents in this group felt that the relationship that they
had had with their bank was not close. This lack of closeness would have been
a driving factor contributing to their decision to exit their bank. Comments
regarding these respondents' feelings about the closeness of their relationship
included: ``It was always an uncomfortable relationship'' and ``The business
manager that we dealt with was very old fashioned. He was very, very staid
from the old school, and it didn't fit our business at that point''.
Many of the comments from interviewees implied that the lack of closeness
for most of this group can be attributed to the fact that they did have a business
banker or a bank manager at the time (i.e. one person with responsibility) that
acted as the agent between the bank and the business.
Seven of the respondents in the ``exit'' group perceived the barriers to exit to
be low, with eight of the respondents indicating that alternative banks were of
a marginally or clearly better quality. This was certainly a difference from the
remain group. Quotes which illustrated the perception of better alternatives
were ``You have such a huge choice you can go to anyone now and they'll take
your business'' and ``I had seriously started making inquiries with other
banking facility providers and I was very impressed with the services that
were available''.

Discussion
The literature implies that there is a direct relationship between service
recovery and how customers react ± if a service recovery is satisfactory a
customer will remain, but if it is unsatisfactory a customer will exit. However,
this is not always the case as the results show ± clearly there are many other
factors that impact on a customer's decision. For example, interviews revealed
that the loyalty they feel towards their provider and the perception they have of
the barriers to exit are also important.
It appears that there are three major factors that influence a customer's
decision to remain or exit after a service failure. These are satisfaction with
recovery (if the customer complained), loyalty and barriers to exit, and each of
these is discussed below.

Satisfaction with recovery

A number of customers in the ``exit'' group left despite satisfaction with service
recovery and a number of customers in the ``remain'' group stayed despite
IJSIM dissatisfaction, indicating that service recovery does not always make a
12,3 difference.
The reason this occurs is best illustrated when dissatisfaction with service
occurs ± the results show that if a customer is dissatisfied with recovery they
will exit, except when other factors, such as loyalty and barriers to exit exist,
enticing the dissatisfied customer to stay. Great service recovery does not
226 always compensate for service failures, but it can limit its harmful impact, as
Boshoff (1997) suggests.
Hence, when dissatisfaction with service recovery efforts is combined with
factors such as loyalty and high barriers to exit, a customer may well remain
rather than exit. However, if loyalty and barriers to exit are low, a dissatisfied
customer will exit. Other factors, therefore, seem to be more important in
driving a customer's decision to remain or to exit rather than satisfactory
recovery.
Similarly, Stewart (1998) suggests that by the time a customer has decided to
exit an organisation it is too late to recover them, so even if recovery efforts are
satisfactory, the customer may exit because they have already made their
decision. The results in this research support the notion that a customer may
receive satisfactory recovery but still exit. Three of the respondents in the ``exit''
group exited despite satisfactory recovery, implying that they had already
decided to exit before the recovery was provided.

Loyalty
The results from this study found if a customer feels a strong sense of loyalty to
their bank then they will remain with their service provider after a service
failure although this is not always the case. This research discovered that when
good service recovery did not occur, then loyalty to the service provider seemed
to moderate the customers' decision to exit or stay. This loyalty may result
from the customers' reliance on the relationship they have formed with the
service provider (Gwinner et al., 1998), it may be a reflection of the period of
time the relationship has existed, or it may be that they feel their provider is
giving them the ``best deal''.
However, the results also showed that some previously loyal customers still
left. The service failure is often strong enough to force these customers to leave
despite the strength of the relationship. The loyal customers often felt, in these
cases, that the bank was not reciprocating their sense of loyalty. Hence, they
left because they had a feeling of almost betrayal at the way they had been
treated by the bank, especially after many years of a relationship with that
bank.

Perceived barriers to exit

The results from this study support the theory in the literature that if a
customer who is dissatisfied, is not loyal, and has a perception that the barriers
to exit are low, he/she will exit rather than remain. All customers who remained
either perceived that there were at least some barriers to exit, or that they felt at Comprehensive
least a moderate sense of loyalty towards their bank. picture of service
A further dimension of barriers to exit is a customer's perception of failure
alternative providers. The results illustrated that a far greater number of
customers in the ``exit'' group perceived alternative banks to be of a better
quality than their current bank than in the ``remain'' group. Customers who
remained tended to perceive all banks as the same or slightly better. 227
In terms of the propositions, there is evidence to suggest that all are
supported, albeit weakly, given the nature of the data. Interestingly though,
new categories seem to exist that were not part of the propositions. There are
some customers who stay after poor service recovery because of perceived
barriers to exit, while some may stay because of loyalty. Similarly, some
customers may not complain after a service failure and stay because of loyalty
and some because of perceived barriers to exit. Finally, some customers may
exit after good service recovery because of low levels of loyalty (i.e. variety
seeking behaviour) and some may switch because of a perception of low
barriers to exit. All of these categories can be built into a model of customer
reactions to service failure which appears in Figure 2.
Most of the research into customer retention and customer exit investigates
the processes separately, not linking the two processes together. By building on
the models of Stewart (1998) and Boshoff (1997), and by incorporating the
relevant literature and the summary of the results above, this model shows the
different ways customers react to service failure with their service provider.

Explanation of the model

There are four concepts on the model that encompass the reactions of exiting
(boxes one to four). The first of these incorporates the notion that customers
exit without complaining because they are so dissatisfied. This stems from the
research that says customers can have two active negative reactions to
dissatisfaction, voice and exit (Hirschman, 1970; Richins, 1987). In this instance,
customers have chosen not to voice and exited instead.
The next three concepts under the exit reaction encompass the idea that if
customers are dissatisfied with a provider they usually complain in the hope
that their problem will be resolved (i.e. they hope to receive satisfactory service
recovery efforts). Even if a customer is satisfied with recovery efforts, they may
still exit. They may do so because they perceive the barriers to exit to be low,
meaning that a change to an alternative provider is not perceived to be difficult.
We call this low inertia.
The other reason a customer may leave, despite satisfactory recovery, is due
to a low sense of loyalty to the provider. Sometimes even if customers are
satisfied with the complaint handling it does not necessarily mean they will
stay loyal to the service provider. ``A satisfied customer may be willing (or even
eager) to patronise alternative suppliers hoping to receive even more satisfying
results'' (Rust and Zahorik, 1993, p. 199). This implies that the factors that make
IJSIM
12,3

228

Figure 2.
Model of customer
reactions to service
failure
a customer loyal are different from those that make them satisfied (Mittal and Comprehensive
Lassar, 1998). picture of service
The fourth concept accounts for when customers are dissatisfied with the failure
service recovery from their provider, they may exit. Spreng et al. (1995) support
this concept by stating that this may be because it is the second time the service
provider has failed to meet a customer's expectations.
The other reaction a customer may have to a service failure, after seriously 229
considering leaving, is to remain. The next five concepts (boxes five to nine)
capture the different iterations of remaining. If a customer complains, they may
remain because they are satisfied with the recovery they have received. The
loyalty the service recovery creates is contributing to their decision to remain
and this is depicted in the ninth box.
Concepts five through to eight are characterised by a customer remaining
despite dissatisfaction. This could be dissatisfaction with the service recovery
efforts of their provider after a complaint was made (boxes five and seven), or
dissatisfaction with the provider which has not been addressed because the
customer has not complained (boxes six and eight). Concepts five and six have
spurious loyalty aspects to them because of the perception that the barriers to
exit are high (Dick and Basu, 1994; Ennew and Binks, 1996), labeled as high
inertia. Concepts seven and eight have loyalty aspects to them, as customers
remain (despite dissatisfaction), but still think highly of the organisation. That
is, they are loyal and the incident they encountered was not harmful enough to
destroy their commitment and make them exit.

Implications
The aim of this paper was to investigate service failure more fully and provide
a more complete picture of its outcomes, which is represented in Figure 2. The
purpose of this was to draw the attention of managers and academics alike to
the importance of understanding fully what drives certain service failure
outcomes. A customer's perceptions of barriers to exit, perceived value of the
relationship, perception of service recovery, and feelings of loyalty to the
provider will all affect their decision to exit or stay.
Managers, therefore, need to understand the role barriers to exit plays in
their industry, the extent to which loyalty can be built to isolate the effects of
service failures, and the power of service recovery. This suggests that even
before customers encounter a service failure, providers should develop (and
endeavor to sustain) close relationships with their customers, as relationship
strength and loyalty are driving factors when customers are deciding whether
to remain or exit.
This is not to suggest that service recovery is not important, far from it.
However, given that marketers need to increasingly justify their investments in
service quality improvements (Rust et al., 1994), it is essential for managers to
realise exactly how powerful service recovery can be and that there are other
ways of insulating a company from service failures.
IJSIM For academics, the results indicate that investigating the ``mix'' of factors
12,3 that drive the outcomes of service failure is important. This research suggests
that perceived barriers to exit, loyalty and service recovery all play a role in
service failure outcome decisions. The discovery that more research is required
within this area (particularly as regards spurious loyalty) is an important
implication of this paper.
230 It is important to note the limitations of this study. First, the size of the
interview sample may be viewed as a limitation in this study. With only ten
respondents in each group (i.e. 20 in total), the findings from their interviews
are not generalisable. However, the aim of the research was not to generate
generalisable results but to provide evidence that service failure is a multi-
faceted construct that is inherently complex. Second, the results were collected
from only one industry, the banking industry. This may be particularly
important given that the banking industry is characterised by high levels of
customer inertia (Colgate, 1997) and the results, therefore, may be unique to this
industry.

Conclusion
Research into services is growing rapidly; however, there are still unexplained
phenomena that need to be investigated. A literature review of previous
research, both theoretical and empirical, identified that further knowledge was
required into how customers react to service failures. An examination of this
has identified that the three major factors that affect a customer's decision to
remain or exit are the customer's satisfaction with recovery efforts, their sense
of loyalty to the provider, and their perception of the barriers to exit. Service
recovery may also not be as important in driving exit and retention in a service
failure situation as was initially thought.
By using these factors the study proposed a model that encapsulates all the
major reasons as to why a customer may stay or leave a service organisation
after a service failure. The model has been developed, as the literature has not
yet adequately covered all customer reactions to service failure. The aim of
the model is to develop a more in-depth explanation of how and why
customers behave as they do, after experiencing a service failure. With this
knowledge, service providers should ultimately be able to deliver a more
effective service.

References
Anonymous (1996), ``Bank satisfaction down'', The Daily Telegraph, 2 October, p. 34.
Bailey, D. (1994), ``Service recovery: A ten-stage approach in the training of front-line staff'',
Training & Management Development Methods, Vol. 8 No. 4, pp. 4.17-4.21.
Bell, C.R. (1994), ``Turning disappointment into customer delight'', Editor & Publisher, Vol. 127
No. 32, pp. 38-48.
Bell, C.R. and Ridge, K. (1992), ``Service recovery for trainers'', Training & Development, Vol. 46
No. 5, pp. 58-63.
Bitner, M.J. (1990), ``Evaluating service encounters: the effects of physical surroundings and Comprehensive
employee responses'', Journal of Marketing, Vol. 52 No. 2, pp. 69-82.
Boshoff, C. (1997), ``An experimental study of service recovery options'', International Journal of
picture of service
Service Industry Management, Vol. 8 No. 2, pp. 110-30. failure
Boshoff, C. and Leong, J. (1998), ``Empowerment, attribution and apologising as dimensions of
service recovery: an experimental study'', International Journal of Service Industry
Management, Vol. 9 No. 1, pp. 24-47.
231
Colgate, M. (1997), ``Personal bankers and relationship marketing: a New Zealand case study'',
Journal of Financial Services Marketing, Vol. 2 No. 1, pp. 84-96.
Colgate, M., Stewart, K. and Kinsella, R. (1996), ``Customer defection: a study of the student
market in Ireland'', International Journal of Bank Marketing, Vol. 14 No. 3, pp. 23-9.
Day, R.L. and Landon, E.L. Jr (1977), ``Toward a theory of consumer complaining behaviour'', in
Woodside, A.G., Sheth, J.N. and Bennett, P.D. (Eds), Consumer and Industrial Buying
Behavior, North-Holland, New York, NY, pp. 425-37.
Dick, A.S. and Basu, K. (1994), ``Customer loyalty: toward an integrated conceptual framework'',
Journal of the Academy of Marketing Science, Vol. 22 No. 2, pp. 99-113.
Ennew, C.T. and Binks, M.R. (1996), ``The impact of service quality and service characteristics on
customer retention: small businesses and banks in the UK'', British Journal of
Management, Vol. 7, pp. 219-30.
Gilbert, F.W., Lumpkin, J.R. and Dant, R.P. (1992), ``Adaptation and customer expectations of
health care options'', Journal of Health Care Marketing, Vol. 12 No. 3 pp. 46-55.
GroÈnroos, C. (1990), ``Relationship marketing approach to the marketing function in service
contexts: the marketing and organizational behavior influence'', Journal of Business
Research, Vol. 20 No. 1, pp. 3-12.
Gwinner, K.P., Gremler, D.D. and Bitner, M.J. (1998), ``Relational benefits in services industries:
the customer's perspective'', Journal of the Academy of Marketing Science, Spring,
pp. 101-14.
Hart, C., Heskett, J. and Sasser, E. (1990), ``The profitable art of service recovery'', in Christopher
et al. (Eds), Turning Sows' Ears into Silk Purses, Dryden, New York, NY.
Hirschman, A.O. (1970), Exit, Voice and Loyalty: Responses to Decline in Firms, Organizations,
and States, Harvard University Press, Cambridge, MA.
Hocutt, M.A., Mowen, J.C. and Chakraborty, G. (1997), A Model of Relationship Dissolution:
Antecedents and Consequences of a Buyer-Seller Relationship, American Marketing
Association, Chicago, IL.
Holmlund, M. and Kock, S. (1996), ``Relationship marketing: the importance of customer-
perceived service quality in retail banking'', Service Industries Journal, Vol. 16 No. 3,
pp. 287-304.
Keaveney, S.M. (1995), ``Customer switching behaviour in service industries: an exploratory
study'', Journal of Marketing, Vol. 59 No. 2, pp. 71-82.
Levesque, T.J. and McDougall, G.H.G. (1993), ``Managing customer satisfaction: the nature of
service problems and customer exit, voice and loyalty'', Asia Pacific Journal of Quality
Management, Vol. 2 No. 2, pp. 40-58.
Miles, M.B. and Huberman, A.M. (1994), Qualitative Data Analysis: An Expanded Source Book,
Sage Publications, New York, NY.
Mittal, B. and Lassar, W.M. (1998), ``Why do customers switch? The dynamics of satisfaction
versus loyalty'', The Journal of Services Marketing, Vol. 12 No. 3, pp. 177-94.
Perrien, J. and Ricard, L. (1995), ``The meaning of a marketing relationship: a pilot study'',
Industrial Marketing Management, Vol. 24, pp. 37-43.
IJSIM Perrien, J., Lalonde, M. and Filiatrault, P. (1994), ``Divorce in a relationship: the case of
commercial banking'', Journal of Professional Services Marketing, Vol. 10 No. 2,
12,3 pp. 157-68.
Reichheld, F.F. and Sasser, W.E. Jr (1990), ``Zero defections: quality comes to services'', Harvard
Business Review, September-October, pp. 105-11.
Richins, M.L. (1987), ``A multivariate analysis of response to dissatisfaction'', Journal of the
Academy of Marketing Science, Vol. 15 No. 3, pp. 24-31.
232
Rust, R.T. and Zahorik, A.J. (1993), ``Customer satisfaction, customer retention, and market
share'', Journal of Retailing, Vol. 69 No. 2, pp. 193-215.
Rust, R.T., Zahorik A.J. and Keiningham, T.L. (1994), Return on Quality, Probus Publishing
Company, Chicago, IL.
Sheth, J.N. and Parvatiyar, A. (1995), ``Relationship marketing in consumer markets: antecedents
and consequences'', Journal of the Academy of Marketing Science, Vol. 23 No. 4, pp. 255-71.
Singh, J. (1990), ``Voice, exit, and negative word-of-mouth behaviors: an investigation across three
service categories'', Journal of the Academy of Marketing Sciences, Vol. 18, pp. 1-15.
Spreng, R.A., Harrell, G.D. and Mackoy, R.D. (1995), ``Service recovery: impact on satisfaction and
intentions'', Journal of Services Marketing, Vol. 9 No. 1, pp. 15-23.
Stewart, K. (1998), ``An exploration of customer exit in retail banking'', International Journal of
Bank Marketing, Vol. 16 No. 1, pp. 6-14.
Stiglitz, J.E. and Weiss, A. (1981), ``Credit rationing in markets with imperfect information'',
American Economic Review, Vol. 71 No. 3, pp. 393-409.
TARP (1995), Study of Consumer Complaint Behaviour in Australia, The Society of Consumer
Behaviour Professionals.
Tax, S.S., Brown, S.W. and Chandrashekaran, M. (1998), ``Customer evaluations of service
complaint experiences: implications for relationship marketing'', Journal of Marketing,
Vol. 62 (April), pp. 60-76.
Teich, I. (1997), ``Holding on to customers: the bottom-line benefits of relationship building'', Bank
Marketing, Vol. 29 No. 2, pp. 12-13.
Winstanley, M. (1997), ``What drives customer satisfaction in commercial banking?'', Commercial
Lending Review, Vol. 12 No. 3, pp. 36-42.
Zeithaml, V.A and Bitner, M.J. (1996), Services Marketing, International Edition, McGraw-Hill,
Singapore.
Zeithaml, V.A., Berry, L.L. and Parasuraman, A. (1996), ``The behavioural consequences of
service quality'', Journal of Marketing, Vol. 60, April, pp. 31-46.
Zemke, R. (1993), ``The art of service recovery: fixing broken customers ± and keeping them on
your side'', in Scheuing, E. and Christopher, W. (Eds), The Service Quality Handbook,
AMACOM Books, New York, NY, pp. 463-76.
Zemke, R. (1994), ``Service recovery'', Executive Excellence, Vol. 11 No. 9, pp. 17-18.

Appendix. Sample questions

Remain interview
How long have you been with your main bank?
So you seriously considered leaving your bank in the last year ± when was that?
What caused you to consider leaving your bank?
Was there one main reason that caused you to consider leaving or was it a progression of
problems?
What factors stopped you from changing banks?
In hindsight, are you happy with the decision you made to remain with this bank? Comprehensive
Hypothetically, if a bank removed most of the cost, effort and hassle from moving, would you
move? picture of service
failure
Exit interview
What caused you to leave your bank?
Was there one main reason that caused you to leave or was it a progression of problems?
Did you seriously consider leaving for a period of time before you actually left? 233
Why did you leave at that particular time?
How long did you stay with your bank after initially becoming dissatisfied?
What could the bank have done differently to make you stay?
In hindsight, would you make the same decision again to switch from (old bank) to (new bank)?
(Are you happy with the decision you made?)