Hierarchy of management that covers different levels of management
Antecedents and consequences of brand loyalty: An empirical study
1. thus reducing the probability of switch-
ing brands.6
Finally, brand loyalty has
been identified as a major determinant
of brand equity.7
The concept of brand loyalty has
not, however, been uniquely defined
and operationalised in the marketing
literature. For example, brand loyalty
has been defined as a repeat purchase,8
preference9
and commitment,10
and as
retention and allegiance.11
These
diverse definitions of brand loyalty are
in part due to the fact that loyalty is a
very complex construct.12
Further-
more, there exist various aspects of
brand loyalty (such as behavioural and
attitudinal brand loyalty). If these
aspects were to be integrated, however,
then one could come up with a
INTRODUCTION
There is no doubt, among academics
and practitioners alike, that the con-
cept of brand loyalty is of strategic
importance for companies in order
to obtain a sustainable competitive
advantage. This is due to a num-
ber of reasons. First, brand-loyal con-
sumers are less expensive, since they
reduce the marketing costs of doing
business.1–3
Secondly, brand extensions
are less risky for brands that exhibit
high loyalty.4
Thirdly, brand loyalty has
been shown to be associated with
higher rates of return on investment
through increases in market share.5
Fourthly, brand-loyal consumers have
fewer reasons to engage in an extended
information search among alternatives,
䉷 HENRY STEWART PUBLICATIONS 1479-1803 BRAND MANAGEMENT VOL. 11, NO. 4, 283–306 APRIL 2004 283
Spiros Gounaris
Assistant Professor of Marketing,
Department of Marketing and
Communications,
Athens University of Economics
and Business, Patission 76,
Athens 10434, Greece
Tel: ⫹32 10 8203 445
Fax: ⫹32 10 8211 269
E-mail: sgounar@aueb.gr
Antecedents and consequences of
brand loyalty: An empirical study
Received (in revised form): 5th November, 2003
SPIROS GOUNARIS
has a PhD from Athens University of Economics and Business, and is an assistant professor of marketing at the
Department of Marketing and Communication at Athens University of Economics and Business. His research
interests pivot around consumer behaviour, satisfaction and loyalty, service quality, tourist marketing,
business-to-business marketing and market orientation development. His work has been published in many
journals.
VLASIS STATHAKOPOULOS
has a PhD from the University of Arizona, and is an associate professor of marketing at the Department of
Marketing and Communication at Athens University of Economics and Business. His work has been published in
a number of journals and been included in various research proceedings. His research interests include marketing
management and strategy, sales management, services marketing, and consumer satisfaction and loyalty.
Abstract
The authors consider the relationships among characteristics associated with the consumer (risk
aversion and variety seeking), the brand (brand reputation and availability of substitute products),
the social environment (social group influences and peers’ recommendations), four types of loyalty
(premium loyalty, inertia loyalty, covetous loyalty and no loyalty), and four consumer-related
behaviour types (word-of-mouth communication, buy alternative brand, go to different store and buy
nothing). To test the hypothesised relationships a survey of Greek consumers was conducted. The
findings provide general support for the postulated linkages among the above variables. Implications
for marketing practice and directions for future research are discussed.
2. the different types of brand loyalty.
After discussing the findings and their
managerial implications, this paper
concludes with study limitations and
directions for future research.
TYPES OF LOYALTY
The authors’ review of past literature
suggests that brand loyalty has been
viewed from three different, albeit
complementary, perspectives, namely:
the behavioural, the attitudinal and the
reasoned action perspectives.
More specifically, the behavioural
perspective has conceptualised brand
loyalty in terms of repeated pur-
chases (for example, Cunningham19
and Kahn et al.20
). In fact, several
models have been proposed in the
literature in order to study brand
loyalty from the behavioural perspec-
tive, the Dirichlet model being one of
the most prominent.21–23
These ap-
proaches model the consumers’ faithful
enactment of a promise to consistently
purchase only one brand, although
they fail to model the reason(s) behind
this behaviour.
One possible insight could be
found in the attitudinal perspective in
conceptualising loyalty. According to
this perspective, brand loyalty con-
sists of a strong internal disposition
towards a brand leading to repeated
purchases.24–26
As such, the attitudinal
approach conceives brand loyalty based
on stated preferences, commitment, or
purchase intentions. One would ex-
pect attitudinal and behavioural brand
loyalty to be positively correlated,
although not perfectly, otherwise there
would be little need for different
concepts.27
Thus an increase in at-
titudinal brand loyalty should lead to an
increase in behavioural brand loyalty.
more accurate definition and thus
operationalisation of brand loyalty.
Hence the first objective of this paper
is to conceive a better definition
of brand loyalty and validate its
operationalisation.
Furthermore, until now there have
been few studies that have examined
the antecedents of brand loyalty (for
example, Dick and Basu,13
Ha14
and
Hog et al.15
). Hence the second
objective of this study is to add to this
stream of research by empirically
examining the role of context in
shaping the development of brand
loyalty. Finally, a third objective of this
manuscript is to empirically examine
the effects of brand loyalty on con-
sumers’ behaviour. For instance there is
empirical evidence that demonstrates
that loyalty is not necessarily reflected
upon the systematic purchase of a
single brand.16
In fact, researchers have
long questioned whether the systematic
purchase of a single brand is the result
of increased levels of loyalty to this
brand or whether it is the outcome of
loyalty to a store which carries a
limited number of brands for a given
product category.17
Moreover, empiri-
cal research has demonstrated that
brand loyalty does not result only in a
specific purchase pattern. For instance,
it can also bring about positive word-
of-mouth communication, which is
not necessarily tied with the purchase
of the brand to which the consumer
feels loyal.18
The rest of the paper is organised as
follows. First, the different types of
brand loyalty are discussed. Next, the
authors advance a conceptual model
and associated research hypotheses.
Then a description is given of an
empirical study designed to test the
hypotheses and compare the effects of
284 䉷 HENRY STEWART PUBLICATIONS 1479-1803 BRAND MANAGEMENT VOL. 11, NO. 4, 283–306 APRIL 2004
GOUNARIS AND STATHAKOPOULOS
3. This then implies that the attitudinal
perspective is of limited value in grasp-
ing the notion of loyalty. If, however,
there are changes in the marketplace, as
is often the case (for example, a new or
improved product is introduced, and
there is increased perceived risk), the
consumer is likely to engage in a
decision-making process, breaking the
cycle of habitual purchases.
The cycle of purchases may or may
not break, however, if the consumer
holds strong positive sentiments and
identifies with the brand. Including the
attitudinal perspective in conceptualis-
ing loyalty is useful, since it allows the
decision-making process occurring in
the consumer’s mind during the pur-
chase to be more realistically described.
It is the cognitive activities that one
describes with this perspective.
Similarly, the cycle of purchases may
or may not break because of pres-
sures exercised by the consumer’s
social environment. Thus embodying
the reasoned action approach in the
proposed conceptualisation recognises
the fact that there are some situa-
tions where consumers’ behaviour is
not fully under their control, but is
influenced by the expectations of
relevant others.
Therefore, one could conceive
brand loyalty as comprising three
dimensions. Each of them determines
the type of loyalty a consumer will
exhibit towards a brand. For instance,
a consumer who is unfavourable to the
purchase of a certain brand may still
purchase the brand. This loyalty
behaviour is likely to be converted into
a behaviour of switching the brand
when the consumer is no longer forced
to keep purchasing the brand. Thus the
following four generic types of brand
loyalty can be identified: ‘no loyalty’,
Another possible explanation can,
however, be derived from the theory
of reasoned action. According to this
perspective, the consumer’s behaviour
may be influenced by social pressures,
thus explaining how a consumer’s
brand attitude may be unfavourable,
while the consumer repeats the pur-
chases of the particular brand. In such
a case, the consumer’s brand loyalty
would be superficial.28
Recognising the
above difficulties in defining and ex-
plaining brand loyalty, Ha29
proposed
the theory of reasoned action to
explain brand loyalty. According to the
reasoned action paradigm — based on
the theory of reasoned action, intro-
duced by Fishbein30
— brand loyalty is
conceived as a notion that is dependent
on normative influences (such as in-
fluences deriving from social peers).
These influences, in turn, are reflected
in the behavioural consequences of
loyalty.31,32
According to this view, one
may hold a favourable attitude towards
a brand but still not purchase it because
of not being able to afford it, a partner
disliking the brand, or for many other
reasons.33,34
Such an individual, al-
though having never actually pur-
chased the brand, promotes it in public,
recommends it, and compels others to
buy it. This situation is similar to the
theoretical discussion by Oliver35
of the
loyalty phases, and particularly the
cognitive phase, where loyalty is based
merely on ‘brand belief’ and not on
brand experience.
For the purposes of this research, a
conceptualisation of loyalty is adopted
that attempts to combine all three
approaches to brand loyalty in ex-
plaining purchasing behaviour. Thus,
incorporating the behaviour paradigm
suggests that repeat purchases are often
the outcome of habitual behaviour.
䉷 HENRY STEWART PUBLICATIONS 1479-1803 BRAND MANAGEMENT VOL. 11, NO. 4, 283–306 APRIL 2004 285
ANTECEDENTS AND CONSEQUENCES OF BRAND LOYALTY: AN EMPIRICAL STUDY
4. self-perception and personality. The
consumer trusts it and is willing to
recommend it to peers, friends
or relatives, although, for reasons
beyond the consumer’s control, the
purchase itself may never occur. In
such cases, the consumer is strongly
discouraged to be loyal to a certain
brand by social influences. For
instance, a young, newly appointed
lecturer in a business school might
covet a Mercedes, but not purchase
it because he cannot afford it or
because he might not wish to
publicise his economic status. The
lecturer may, however, still recom-
mend the brand.
— Inertia loyalty: An individual, al-
though purchasing the brand, does
so out of habit, convenience or for
some other reason, but not as a
consequence of emotional attach-
ment to the brand or a real
social motive. Inertia loyalty is
characterised by a habitual attach-
‘inertia loyalty’, ‘premium loyalty’ and
‘covetous loyalty’ (see Figure 1).
The four types of brand loyalty are
characterised as follows:
— No loyalty: No purchase at all, and
a complete lack of attachment to
the brand. Also no social influences
to be even cognitively loyal to a
brand.
— Covetous loyalty: No purchase but,
unlike the case of ‘no loyalty’, the
individual exhibits a very high level
of relative attachment to the brand
as well as a strong positive predis-
position towards the brand, which is
developed from the social en-
vironment. This condition arises
from perceived human characteris-
tics which a consumer identifies in
a specific brand.36
The individual
comes to like the brand and
thus emotional attachment with the
brand increases. The brand becomes
an extension of the consumer’s own
286 䉷 HENRY STEWART PUBLICATIONS 1479-1803 BRAND MANAGEMENT VOL. 11, NO. 4, 283–306 APRIL 2004
High
Low
High
Low
High
COVETOUS
INERTIA
NO
LOYALTY
PREMIUM
Purchasing behaviour
Emotional
attachment
Social influences
GOUNARIS AND STATHAKOPOULOS
Figure 1 A conceptualisation of loyalty based on purchasing behaviour, emotional attachment and social influences
5. vinced that the selected brand is in
some way the best brand to buy.43
This conviction arises from both
personal and social motives. Varia-
tions in the price of their favourite
brand may affect the quantity of the
brand they purchase, but not the
brand they choose to buy, since
these consumers are committed to
the brand.
CONCEPTUAL MODEL AND RESEARCH
HYPOTHESES
Figure 2 presents the conceptual model
of brand loyalty that guides this
research. The model shows potential
drivers of brand loyalty. These drivers
are classified in three basic categories:
consumer drivers, brand drivers and
social drivers. By focusing on potential
drivers, it may be possible to manage
brand loyalty better. In addition, the
model used in this study focuses on
consumers’ behavioural responses to
brand loyalty — word-of-mouth com-
munication, buying alternative brand,
going to different point of sale (store)
and buying nothing.
Consumer drivers
Both normative and empirical studies
have substantiated the importance of
the individual’s characteristics in decid-
ing to purchase a specific brand.44
Two
such characteristics are examined in
this study: risk aversion and variety
seeking. Although many characteristics
of consumers may have an impact on
the decision to purchase a specific
brand, this study chooses to focus on
these two specific attributes, which
both relate to how consumers handle
risks. Loyalty has been described as a
means of handling the risk associated
ment that is to a large ex-
tent unemotional and convenience
driven. The consumer may sys-
tematically choose the specific
brand over other brands, but this
choice involves little emotional
involvement, little personal invest-
ment, and no brand commitment.37
Hence this is a very fragile
relationship that may be easily
terminated by a rival product
capable of breaking the con-
sumer’s habitual behavioural pat-
tern. Oliver38
terms this type of
loyalty ‘phantom loyalty’, while
Day39
and Dick and Basu40
call it
‘spurious loyalty’.
— Premium loyalty: An individual
exhibits a high degree of relative
attachment to the brand, a high
instance of repeat purchases, and
appears to be highly influenced by
social pressure. Premium loyalty is
characterised by the greatest degree
of consumer attachment to the
brand, and in this case the con-
sumer purposefully seeks to pur-
chase the particular brand, while
attempting to overcome obstacles.
This is similar to the descrip-
tion by Oliver41
of ‘action loyalty’
— ‘commitment to the action
of re-buying’. Premium loyalty
propels individuals to suffer various
sacrifices in order to acquire their
favoured brand. Football fans are a
good, although extreme, example of
people showing this type of loyalty.
They may see their team losing one
game after the other, and yet be
willing to travel with the team or
watch its games on television.
Consumers who exhibit ‘premium
loyalty’ have been won over by the
brand alternative through the value
it provides to them42
or are con-
䉷 HENRY STEWART PUBLICATIONS 1479-1803 BRAND MANAGEMENT VOL. 11, NO. 4, 283–306 APRIL 2004 287
ANTECEDENTS AND CONSEQUENCES OF BRAND LOYALTY: AN EMPIRICAL STUDY
6. of rational market behaviour — in
order to reduce perceived risk. The
risk element may be either a functional
risk or a social acceptance risk. In fact
the perceived risk can be so intense
that individuals become reluctant to
proceed with the action. Instead, they
procrastinate until they have reduced
the perceived complexity or the uncer-
tainty associated with the situation.53,54
Hence the consumer’s need to control
the risk will be a significant positive
factor in the formation of brand loyalty.
On these grounds, this paper inves-
tigates the following hypothesis:
H1: Risk aversion will relate to the type of
brand loyalty the individual develops
towards a specific brand.
Variety seeking
On the other hand, uncertainty of the
outcome of a purchase and the risk
associated with a certain decision
provides stimulation to the consumer.55
If the stimulation obtained is within
the bounds of the optimal stimulation
with the decision to purchase a specific
brand.45
Risk aversion
Individuals are often confronted with
situations that differ in the degree of
uncertainty or complexity they present
to them.46
Typically, decisions linked to
highly valued goals47
such as the
purchase of a new car and/or deci-
sions on high-involvement product
categories48
encompass greater risk for
the individual buyer. Such decisions
may evoke negative emotions that the
buyer attempts to deal with.49
Emo-
tions accentuate the risk associated
with the purchase of a specific brand,
leading to a greater search,50
which, in
turn, may lead to lower levels of brand
loyalty. On the other hand, emo-
tions may lead consumers to ex-
hibit avoidance behaviour51
and/or
greater dependence on previously held
choices, which result in higher levels of
brand loyalty. Furthermore, Sheth and
Parvatiyar52
argued that consumers
become brand loyal — a manifestation
288 䉷 HENRY STEWART PUBLICATIONS 1479-1803 BRAND MANAGEMENT VOL. 11, NO. 4, 283–306 APRIL 2004
CONSUMER DRIVERS
Risk aversion
Variety seeking
BRAND DRIVERS
Brand reputation
Availability of substitute
brands
SOCIAL DRIVERS
Social group influences
Peers’ recommendation
Type of
brand
loyalty
Buy nothing
Buy alternative
brand
Word-of-mouth
communication
Visit other store
GOUNARIS AND STATHAKOPOULOS
Figure 2 Conceptual model
7. of the transaction and the cost if he
decides to acquire the product. Hence
to understand how product characteris-
tics contribute to brand loyalty, it is
necessary to comprehend what creates
and determines the value individuals
derive from a brand. The following
two brand characteristics are examined
in this study: brand reputation and
availability of substitute brands.
Brand reputation
Although not part of the physical
product itself, the reputation of a
brand’s name has been described as an
extrinsic cue, that is, an attribute
related to the product.61
A reputable
brand name conveys a strong indica-
tion of the product’s quality and equity
that is not necessarily related to detailed
knowledge of the intrinsic — technical
— specification of the product.62
Therefore, the choice between alterna-
tive brands within a single product class
is facilitated, since brands are differen-
tiated easily by their consumers. As
Oliver63
suggests, loyalty is not merely
about product superiority and satisfying
customers. Loyalty is about having
customers who can become deter-
mined defenders of the brand. If
the firm cannot develop, support
and maintain brand uniqueness and
perceived brand equity, then it is
not possible to expect loyalty to
develop.64
Thus having a brand with a strong
reputation will be a significant positive
factor in the formation of brand loyalty,
since the brand’s reputation strengthens
its perceived equity.65,66
Moreover, the
reputation of the brand strengthens the
habitual behaviour of consumers by
rewarding their choice and making the
brand more desirable and alluring.67
As
level, then it is desirable and the
consumer actively seeks to attain it.
However, if the optimal stimulation
level is exceeded, it becomes too
intensive, leading consumers to try to
reduce the complexities that are as-
sociated with such a condition and, as
pointed out earlier, they attempt to
routinise the decision-making process
and its outcomes. In fact, Sheth and
Parvatiyar56
pointed out that routinisa-
tion and variety-seeking behaviour
become cyclical over time, but the
cycles are asymmetrical in favour of the
longer duration of routinised be-
haviour.
Routinisation, although initially
helpful, may, however, lead an
individual to feelings of monotony and
boredom, which may lead to ex-
perimentation with new brands.57
Moreover, it appears that the level of
variety-seeking behaviour depends on
the intrinsic need of consumers to seek
variety (personal differences) and
on the product category level of
involvement.58
It is within this
framework that studies report a break
in the link between satisfaction and
loyalty.59
Indeed, as Homburg and
Giering60
have demonstrated, variety
seeking is one of the key consumer
characteristics which moderates the
relationship between perceived quality
and satisfaction with the loyalty to a
specific brand. Within this framework,
this paper hypothesises that:
H2: Variety-seeking behaviour will relate to
the type of brand loyalty an individual
develops towards a specific brand.
Brand drivers
An individual’s intention to purchase a
product reflects a search for value out
䉷 HENRY STEWART PUBLICATIONS 1479-1803 BRAND MANAGEMENT VOL. 11, NO. 4, 283–306 APRIL 2004 289
ANTECEDENTS AND CONSEQUENCES OF BRAND LOYALTY: AN EMPIRICAL STUDY
8. do so in the absence of any attractive
alternative — as is the case when
no substitute brands are available73
— the relationship tends to last
only for as long as there is no
alternative.74
Research suggests that
customers in such constrained situa-
tions attempt to restore their freedom
to choose.75
According to resource-
dependence theory,76
consumers may
attempt to break free from constrained
relationships by identifying acceptable
substitutes. Hence the perception of
similar substitutes may be expected to
influence negatively the creation of
relational ties to the brand within the
specific category, and it might therefore
be considered as a deterrent to the
formation of brand loyalty. On these
grounds the following hypothesis is
investigated:
H4: The availability of substitute products
will influence the type of brand loyalty
an individual develops towards a spec-
ific brand.
Social drivers
Finally, when studying the antecedents
of loyalty, one should not neglect the
social norms which may influence con-
sumers’ behaviour patterns. Consumers
do not take decisions isolated from
social influences. Rather, they are sub-
jected to heavy social control over the
attitudes they have and the behaviour
they develop.77
Social group influences
One strong type of such social in-
fluence is that derived from reference
groups — the social groups that have
a direct or indirect influence on the
person’s attitude or behaviour.78
In the
a result, reputable brands enjoy higher
loyalty due to their higher market
share.68
This higher market share is
attributed to the fact that higher-share
brands are not only bought by more
consumers, but they are also bought
more frequently. In other words, high-
share brands benefit both from greater
market penetration and higher pur-
chase frequency. This is the well-
known double-jeopardy phenomenon,
an ‘empirical law’ that researchers have
observed and modelled for nearly 30
years.69
On these grounds, the follow-
ing hypothesis is investigated:
H3: Brand reputation will relate to the
brand loyalty type an individual
develops towards a specific brand.
Availability of substitute brands
Brand reputation is subjected to the
shopper’s perception of both the range
of competing products and brands
as well as the class of substitute
products.
When a product class comprises
several brands which are perceived by
consumers to be similar to each
other, discriminating among them is
hard. Consequently, individuals have
no reason to show loyalty towards one
or another. In fact, the more alike the
brands are perceived to be, the less
likely loyalty is to emerge.70
Rather,
consumers are prone to make their
purchases from a predetermined set of
alternative products without showing a
particular preference to any specific
brand from this set.71,72
Thus the
availability of substitute products is
expected to affect brand loyalty sig-
nificantly.
Moreover, when customers stay in a
relationship because they are forced to
290 䉷 HENRY STEWART PUBLICATIONS 1479-1803 BRAND MANAGEMENT VOL. 11, NO. 4, 283–306 APRIL 2004
GOUNARIS AND STATHAKOPOULOS
9. loyalty, since the desire for the brand
may be affected by group preference.87
On these grounds, this paper inves-
tigates the following hypothesis:
H5: Social group influences will relate to
the type of brand loyalty an individual
develops towards a specific brand.
Peers’ recommendation
Another strong source of social
influence is the recommendations and
suggestions made by the individual’s
peers. Hite and Hite88
found that a
party’s reputation could lead to posi-
tive expectations about the party
which, in turn, leads other parties to
develop reciprocity and loyalty for the
reputable party. When it comes to
brand names, their reputation reflects
the opinion of others that a specific
brand possesses or does not possess
certain characteristics.89
While adver-
tising and/or public relations help
brands to demonstrate their qualities,
peers are among the most influential
sources of information used by con-
sumers in shaping their opinion
concerning a brand’s qualities.90
Peers
exercise both normative (conformist)
and identificational influences on con-
sumers. Informative influences help to
guide consumers in product, brand
and store searches,91
whereas norma-
tive influences direct and control
evaluations, choices and loyalties.92
Thus peers’ recommendations are
expected to significantly affect brand
loyalty.93
Following the reasoned-action
paradigm, Bearden and Etzel94
suggest,
however, that the recommendations of
peers may not necessarily convert into
actions (ie purchase). Under certain
circumstances, they could merely
context of the present work, two types
of social influence are considered: so-
cial group influences and peers’ recom-
mendations.
A group becomes a reference one
when an individual identifies with it so
much that he takes on many of the
values, attitudes and/or behaviour of its
members.79
The power of the influence
of a reference group is dependent on
the individual’s susceptibility to this
influence, the strength of his involve-
ment with the group and the degree
of product conspicuousness.80
Powerful
reference groups may easily change the
behaviour of their members, or their
aspirant members, and align it more
with the norms and standards that the
group considers to be acceptable.81
Hence the individual’s loyalty
towards a product is also dependent on
the acceptance of his preference for a
certain product by the social group the
individual refers to, particularly when
the conditions under which individuals
feel coerced to give in to the group’s
norms are met. By adapting their
attitudes and behaviour, consumers
fulfil their aspirations and at the same
time reduce the perceived risk of
making a decision.82
Besides, recent
empirical studies have attested to the
impact of social stimuli (or normative
information) on loyalty.83,84
For instance, Mascarenhas and
Higby,85
in their study of how
youngsters choose a brand, indicated
that parents’ consistent choice of a
particular brand influences children to
perceive the brand as ‘good’, and thus
become loyal to it. Furthermore, Hog
et al.86
found that families and peer
groups led young consumers to form a
more positive image of a brand. Hence
group social influences are expected to
have a strong positive impact on brand
䉷 HENRY STEWART PUBLICATIONS 1479-1803 BRAND MANAGEMENT VOL. 11, NO. 4, 283–306 APRIL 2004 291
ANTECEDENTS AND CONSEQUENCES OF BRAND LOYALTY: AN EMPIRICAL STUDY
10. On these grounds this paper inves-
tigates the following hypothesis:
H7: The type of brand loyalty will depend
on the occurrence or not of word-of-
mouth communication between con-
sumers.
Buy alternative brand
An interesting situation arises when
a consumer is loyal to a specific
brand, but the brand is unavailable
when required at a particular store.98
How likely is it that the individual
will betray the brand and purchase
another?
Oliver99
has shown that consumers,
when faced with uncertainty about
how to handle a decision and about its
outcomes (concerning, for example,
specifying relative uncertainties, what
information to seek, or how to assess
consequences), tend to delay the actual
decision. This is in line with the
empirical findings of Greenleaf and
Lehmann,100
demonstrating that such
procedural uncertainty causes con-
sumers to delay a decision. Hence, for
instance, when consumers are deprived
of the brand towards which they have
developed a feeling of loyalty, they
may delay their purchase until either
‘their’ brand is available again or they
have managed to handle the new
situation.
On the other hand, some con-
sumers might find delaying the pur-
chase too ‘costly’ and thus decide
to switch brands. Many consumers
adapt their brand preferences accord-
ing to the time when they prefer
to shop.101
Therefore they would
rather stay in one store and switch
brands. Such behaviour is in line with
the consequences of developing the
influence the consumer’s emotional
attachment to the brand. Consider, for
instance, a young teenager who
develops a high attachment to Sony’s
Playstation II after it was
recommended by a friend, but still feels
reluctant to purchase it because he
perceives that his parents would
disapprove. Nonetheless, following the
conceptualisation of loyalty outlined,
the teenager in question is (covetously)
loyal to the brand. On these grounds,
this paper investigates the following
hypothesis:
H6: Peers’ recommendation will relate to
the type of brand loyalty an individual
develops towards a specific brand.
Consequences of loyalty
Scholars studying the notion of brand
loyalty have discussed a number of
behavioural consequences. In the
context of the present work, four
alternative consequences of loyalty are
examined, namely: word-of-mouth
communication, buy alternative brand,
go to different point of sale (store) and
buy nothing.
Word-of-mouth communication
Perhaps the single most expected be-
havioural outcome of loyalty is brand
recommendation. Consumers become
loyal as a result of the satisfaction
they experience with their purchase.95
Satisfied consumers who share their
experiences with other individuals are
the best advocators of any company or
its products.96
In fact, as Oliver97
sug-
gests, in certain cases it is the sharing of
the experience regarding the brand that
ultimately provides the satisfaction and
not the brand itself.
292 䉷 HENRY STEWART PUBLICATIONS 1479-1803 BRAND MANAGEMENT VOL. 11, NO. 4, 283–306 APRIL 2004
GOUNARIS AND STATHAKOPOULOS
11. available), and in doing this there is less
risk associated with the decision. The
individual has developed the highest
level of loyalty — ‘action loyalty’.
Betraying the brand will be like
betraying himself.104
Based on the
above discussion, this paper investigates
the following hypothesis:
H10: The type of brand loyalty will relate
to whether the consumer decides to
buy nothing if the brand is unavail-
able.
METHODOLOGY
Data collection and sample
The sample for the study was drawn
from the area of Athens, Greece.
Trained personnel conducted inter-
views, based on a questionnaire, in
order to increase the validity and
reliability of the responses.
The sample consisted of 850 con-
sumers of whisky who were randomly
approached in the street and shop-
ping malls. Although the sample was
clearly chosen for convenience, the
interviews were conducted at different
locations and on different days, as well
as at uniform intervals, in order to
reduce location-, date- and time-re-
lated response bias.
The decision to focus the study on
whisky buyers was based on three
factors. First, there is a remarkable
variety of alternative whisky brands in
the Greek market, a fact which
gives many options to the shopper.
Moreover, whisky is bought quite
often by the majority of the adult
Greek population. Consequently, it
could reasonably be expected that
signals of loyalty, where identified in
terms of behaviour, emotional attach-
‘spurious’ type of loyalty as sug-
gested by Dick and Basu.102
On these
grounds, this paper investigates the
following hypothesis:
H8: The type of brand loyalty will relate
to the purchase or not of alternative
brands.
Go to a different store
In the absence of the desired brand,
loyal consumers may choose to go to
a different point of sale to seek the
brand. Once more, this behaviour
depends on the perceived risk as-
sociated with the decision of purchas-
ing an alternative brand, but perhaps
also with the emotional disappoint-
ment of not finding the brand with
which the consumer has an emotional
attachment.103
It can be expected that
the decision of whether to go to a
different store to find the desired brand
will be determined by the type of
brand loyalty. Based on the above
discussion, this paper investigates the
following hypothesis:
H9: The type of brand loyalty will affect
the decision to go to a different store.
Buy nothing
The decision to buy nothing if the
preferred brand is unavailable is by
definition a strong indicator of
premium loyalty. To make this
decision, the individual has to go
through the same cognitive/conative
process to decide on an alternative as
he did originally to choose the
preferred product, and this creates both
cognitive and emotional discomfort.
The result is that the decision to buy
is postponed (for when the product is
䉷 HENRY STEWART PUBLICATIONS 1479-1803 BRAND MANAGEMENT VOL. 11, NO. 4, 283–306 APRIL 2004 293
ANTECEDENTS AND CONSEQUENCES OF BRAND LOYALTY: AN EMPIRICAL STUDY
12. This process resulted in a battery of 21
items worded to capture the different
types of loyalty. Next, a mini pilot
survey was conducted among a ran-
domly selected sample of 250 stu-
dents. Exploratory factor analysis was
employed to refine their answers on
loyalty by deleting items with high
loadings on multiple factors. This
process resulted in a four-factor solu-
tion, namely premium loyalty (com-
prised of three items), covetous loyalty
(three items), inertia loyalty (four items)
and no loyalty (five items). These 15
items were employed in the analysis of
the results, which is reported in this
manuscript.
Dimensionality and psychometric attributes
of the loyalty scale
In the main study the psychometric
attributes of the scales employed to
measure loyalty were assessed using
confirmatory factor analysis (CFA).
This was preferred to exploratory
factor analysis (EFA), a method about
which various concerns have been
raised. For instance, according to
Mulaik,106
EFA may ‘find optimal
knowledge’ (p. 265). Mulaik107
made it
clear that ‘there is no rationally
optimal way to extract knowledge
from experience without making cer-
tain prior assumptions’ (p. 265). In
addition, the factor structures yielded
by an exploratory factor analysis are
determined by the mechanics of the
method and are dependent on specific
theories and the mechanics of extrac-
tion and rotation procedures. This,
too, can result in inaccurate results.
Mulaik108
also made it clear that
exploratory techniques do not provide
any way of indicating when something
is wrong with one’s assumptions,
ment or social stimuli, would be
authentic and not the consequence of
a constrained choice. Finally, the large
number of alternative whisky brands
that can be found in the Greek market
differ markedly in terms of their
positioning strategy (product, price,
promotion and distribution) in Greece.
The same is also true for the purchas-
ing occasion. Whisky is bought in
Greece for private or public (in-home
or on-premises) consumption, as well
as for offering as a gift. In addition,
whisky can be bought legally from
off-licences, supermarkets and con-
venience stores. Therefore, the authors
expected that there would be enough
scope for all types of loyalty to
develop.
Measures
Although attempts were made to use
existing measures, they were not avail-
able for several constructs or were
otherwise limited in their applicability
to the context of this study. It was
therefore necessary to adapt the current
measures or develop new ones (as dis-
cussed subsequently).
Development of the loyalty scale
The review of the current litera-
ture did not reveal any empirically
validated scales with respect to the
different types of loyalty. Thus new
scales were developed to measure
the loyalty construct. In doing so,
the scale-development instructions sug-
gested by Churchill105
were followed.
More specifically, once the domain of
the construct (brand loyalty) was speci-
fied, the preliminary set of items was
developed through discussion with
consumers and group brand managers.
294 䉷 HENRY STEWART PUBLICATIONS 1479-1803 BRAND MANAGEMENT VOL. 11, NO. 4, 283–306 APRIL 2004
GOUNARIS AND STATHAKOPOULOS
Tải bản FULL (24 trang): https://bit.ly/3u7fzrO
Dự phòng: fb.com/TaiHo123doc.net
13. cal foundation.112
In addition, CFA
offers the researcher a more viable
method for evaluating the validity of a
construct. The researcher is able to
explicitly test hypotheses concerning
the factor structure of the data due to
having the predetermined model spec-
ifying the number and composition
of the factors. Confirmatory methods,
after specifying the a priori factors, seek
to optimally match the observed and
theoretical factor structures for a given
data set in order to determine the
‘goodness of fit’ of the predetermined
factor model.
Figure 3 depicts the results of CFA
while Table 1 summarises the fit
indices using the chi-square test, the
goodness of fit index (GFI), Bentler’s
comparative fit index (CFI) and the
root mean square error of approxima-
tion (RMSEA) for the four scales
employed in this study to measure the
different types of loyalty. The RMSEA
because the technique is designed
to fit the data regardless of other
considerations. Rather than justifying
the ‘knowledge’ produced, exploratory
factor analysis suggests hypotheses, but
does not justify knowledge. Finally,
another problem lies in the interpreta-
tion of the results. The interpretation
of factors measured by a few vari-
ables is frequently complicated.109,110
Stevens111
suggested that the difficulty
in interpretation often comes about
because the researcher lacks prior
knowledge and therefore has no basis
on which to make an interpretation.
CFA on the other hand is a theory-
testing model as opposed to a theory-
generating method like EFA. In CFA,
the researcher begins with a hypothesis
prior to the analysis. This model speci-
fies which variables will be correlated
with which factors and which factors
are correlated. The hypothesis is based
on a strong theoretical and/or empiri-
䉷 HENRY STEWART PUBLICATIONS 1479-1803 BRAND MANAGEMENT VOL. 11, NO. 4, 283–306 APRIL 2004 295
e1 e2 e3
C1 C2 C3
e6
e5
e4
P1 P2 P3
COVETOUS PREMIUM
e9
e8
e7
I1 I2 I3
INERTIA
I4
e10
NO LOYALTY
N2 N3 N4 N5
e15
N1
e11 e12 e13 e14
0.94
0.89
0.93
0.89 0.79
0.83
0.93
0.85 0.75 0.69 0.68
0.78
0.82 0.88
0.84
0.88 0.83 0.87
0.80
0.83 0.79
0.84
0.79 0.79 0.85
0.88 0.81 0.85 0.81 0.82
0.60
0.69
0.62
0.61
0.69
0.62
ANTECEDENTS AND CONSEQUENCES OF BRAND LOYALTY: AN EMPIRICAL STUDY
Figure 3 The loyalty measurement: Confirmatory factor analysis
Tải bản FULL (24 trang): https://bit.ly/3u7fzrO
Dự phòng: fb.com/TaiHo123doc.net
14. Table 1, the four-factor solution was
superior to that of the single factor.
Convergent and discriminant
validity for the loyalty scale were
evaluated by calculating the average
variance extracted (AVE) for each
factor. Convergent validity is
established if the shared variance
accounts for 0.50 or more of total
variance. Discriminant validity is
evident when the AVE for each
construct is greater than the squared
correlation between that construct and
any other construct in the model.117
The results presented in Table 2
confirm both the convergent and
discriminant validity of the four
scales.
Internal consistency was assessed
by means of the Cronbach’s alpha
coefficient.118
An alpha value of 0.60 or
above for new scales, and 0.70 or above
for established scales, is considered to be
measures the lack of fit and takes
parsimony into account by assessing the
discrepancy per degree of freedom
between the population covariance
matrix and the fitted matrix. That is, it
penalises for overfitting. The ad-
justed goodness-of-fit index (AGFI)
is not reported as its usefulness is
questionable.113
The measurement model was first
tested for the adequacy of a four-factor
solution. As can be seen in Table 1,
although the overall chi-square test
was statistically significant (2
⫽ 122,
df 83) the GFI of 0.98 and a low
value (below 0.10) of the root mean
square residual (RMSEA ⫽ 0.024) sug-
gested a good model fit.114
Besides, the
2
statistic is known to be strongly
dependent on sample size, and thus
its appropriateness has been strongly
questioned.115,116
A single-factor model
was also tested, but, as can be seen in
296 䉷 HENRY STEWART PUBLICATIONS 1479-1803 BRAND MANAGEMENT VOL. 11, NO. 4, 283–306 APRIL 2004
GOUNARIS AND STATHAKOPOULOS
Table 1 Fit statistics
Loyalty measurement
Four factors One factor
Chi square (2
)
Degrees of freedom (df)
GFI*
CFI*
RMSEA**
122
83
0.980
0.983
0.024
970
89
0.795
0.638
0.108
* CFI and GFI values close to 1 indicate a good fit.
** The lower the RMSEA values, the better the model; values below 0.1 suggest adequate fit.
Table 2 Reliability and validity assessment
Dimensions of loyalty
PREM CVTS INRT N-LTY
Rel AVE (Corr)2
Rel AVE (Corr)2
Rel AVE (Corr)2
Rel AVE (Corr)2
Conv Disc
0.82 0.72 0.71 0.94 0.85 0.78 0.88 0.60 0.59 0.91 0.69 0.59 YES YES
PREM, premium loyalty; CNTS, covetous loyalty; INRT, inertia loyalty; N-LTY, no loyalty; Rel, Cronbach’s alpha
coefficient; AVE, average variance extracted ⫽ 兺 (standard loadings)2
/兺(standard loadings)2
⫹ 兺ij; Conv,
convergent validity (AVE > 0.50); Disc, discriminant validity ⫽ AVE/(Corr2
) > 1; (Corr)2
, highest (Corr)2
between
factor of interest and remaining factors.
3849518