An executive summary for

managers and executive Communication message

readers can be found at the
end of this article strategies for brand extensions
Joo Young Kim
Doctoral Candidate and Instructor of Advertising, College of
Journalism and Communications, University of Florida, Gainesville,
Florida, USA

Keywords Brand extensions, Marketing strategy, Brand identity,

Communication management
Abstract This study searches for the communication message strategies for two distinct
brand extension types: close and remote brand extensions. An experiment is conducted
with four cells which were exposed to different communication strategies for five
extension types. Communication strategies used were brand essence cue, extension
attribute cue, extension dissonance reducer cue, and some combinations of the named
cues. Results show that different communication strategies are necessary for extension
situations that differ in distance from parent brand territories.

Introduction and purpose of the research

Marketing researchers
The decision to extend a brand or create a new brand is often problematic.
Many disadvantages of brand extensions have been suggested in the
marketing literature such that brand extension can confuse consumers,
encounter retailer resistance, hurt the parent brand's image or sales, and
forgo the chance to develop a new brand (Keller, 1998). In spite of these
disadvantages, many companies have introduced a myriad brand extensions
throughout the world, because it is seen as cost effective and makes efficient
use of existing marketing resources.
As brand extension has been a widely accepted product developing strategy,
marketing researchers have attempted to explain the reasons behind
consumers' acceptance of brand extensions. The consensus to date is that the
perceived fit between brand extension and the parent brand is the essential
factor to understand the success of brand extensions. There are several
conceptual definitions of perceived fit such as:
. the similarity or feature overlap between the parent brand and extension
category (Aaker and Keller, 1990; Boush and Loken, 1991; Romeo,
1990);
. both category similarity and brand concept consistency (Park et al.,
1991);
. the unique brand-specific associations in the extension (Broniarczyk and
Alba, 1994); and
. the connection of any salient and relevant parent brand association (e.g.
category, brand concept, brand-specific associations) to the extensions
(Bridges et al., 2000).
In spite of such prolific research in the brand extension field, few studies
(e.g. Aaker and Keller, 1990; Bridges et al., 2000) have dealt with marketing
communication strategies for various brand extension situations. Bridges

The Emerald Research Register for this journal is available at

http://www.emeraldinsight.com/researchregister
The current issue and full text archive of this journal is available at
http://www.emeraldinsight.com/1061-0421.htm

462 JOURNAL OF PRODUCT & BRAND MANAGEMENT, VOL. 12 NO. 7 2003, pp. 462-476, # MCB UP LIMITED, 1061-0421, DOI 10.1108/10610420310506029

Tauber's typology
JOURNAL OF PRODUCT & BRAND MANAGEMENT, VOL. 12 NO. 7 2003
et al. (2000) studied different communication effects on different extension
settings in which salience and relevance of the brand extensions were
manipulated, where the salience of associations was assumed to depend on
the dominant parent brand associations, and where the relevance was
assumed to depend on the parent brand-to-extension category relationship.
While they argued that relational or elaborational communication strategy
would help consumers build explanatory links between the parent brand and
extensions, their results were theoretically, rather than empirically, useful.
They did not clearly suggest a useful form for message strategy
recommendation. As brand extensions are often initially perceived as
unusual (i.e. low perceived-fit) by consumers, marketers need to know about
and use the proper message strategies in order to improve perceived fits.
Thus, the need for research in practical message strategies is paramount. In
consensus with such demand, this study is intended to investigate appropriate
empirical communication strategies for several extension types, largely
categorized as close and remote brand extensions.
The current study first identifies the categorization scheme of brand
extension types using Tauber's (1988) typology. Second, the concept of
parent brand essence, which this study views as an important determinant of
the perceived-fit of brand extensions, is discussed. Third, extension
dissonance reducer, which is considered a critical influencer of the
perceived-fit change, is introduced. Using these categorization schema and
their important concepts, an experiment will be designed and conducted in
order to identify the message strategies that work best for the specific
extension types.
Classification of brand extensions
There are two broadly classified extension types. One is a line extension: a
new product is offered within a current product category. The second is a
category extension: a new product is introduced in a different new product
category from that currently served by the parent brand (Keller, 1998). There
is another type of extension classification such as close/remote extensions,
which is based on the distance of extension from the parent brand's territory
(Kapferer, 1998). The present study considers the brand territory as meaning
any area (whether it is product-related or non-product-related) with which a
brand is uniquely and/or strongly associated. This is consistent with Bridges
et al.'s (2000) study that suggested that the connection of any salient and
relevant parent brand association (e.g. category, brand concept,
brand-specific associations) to the extensions can be used by consumers in
forming the perception of fit (Bridges et al., 2000). In regard to the specific
extension strategy classifications, Tauber's (1988) seven general strategies of
extension have served as useful guides in identifying the extension cases:
(1) introduce the same product in a different form;
(2) introduce products that contain the brand's distinctive taste, ingredient,
or componet;
(3) introduce companion products for the brand;
(4) introduce products relevant to customer franchise of the brand;
(5) introduce products that capitalize on the firm's perceived expertise;
(6) introduce products that reflect the brand's distinctive benefit, attribute,
or feature owned; and
463
 (7) introduce products that capitalize on the distinctive image or prestige of
the brand.
Strategy classification To conduct the extension communication study, this research synthesized the
past findings and classifications of brand extensions and created an
integrated brand extension strategy classification. First, using Tauber's
classification, the current study conceptually created four simpler strategy
classifications (feature extension: Tauber's strategy Nos. 1, 2, 6; know-how
extension: No. 5; companion extension: No. 3; image extension: Nos. 4, 7)
(Table I). These four were further simplified into two: extensions of product-
related associations (feature, know-how, companion extension), and
non-product related associations (image extension), which is consistent with
Keller's view that brand extension is about capitalizing on an original
brand's customer-based brand equity, which is based on brand knowledge,
mainly composed of product-related and non-product-related attributes
(Keller, 1998).

Brand essence fit

Consumer behavior literature has shown that consumers' evaluation of brand
extensions, in large part depends upon the perceived fit between a parent
brand and the extension. Although a consensus exists regarding the
importance of the perceived fit in brand extensions, there are several
conceptual definitions of perceived fit such as:
. the similarity or feature overlap between the parent brand and extension
category (Aaker and Keller, 1990; Boush and Loken, 1991; Romeo,
1990);
. both category similarity and brand concept consistency (Park et al.,
1991);
. the unique brand-specific associations in the extension (Broniarczyk and
Alba, 1994); and
. the connection of any salient and relevant parent brand association (e.g.
category, brand concept, brand-specific associations) to the extensions
(Bridges et al., 2000).
Basis of fit Although past research provides valuable insights into understanding brand
extensions, one thing that most past studies lack is the study of the process of
determining the ``basis of fit'' consumers use to evaluate brand extensions,
and specific extension information which need to be more emphasized for
different types of extensions. These issues are related to a major part on
which consumers focus when they evaluate brand extensions. In this

Product-related associations extension Non-product-related associations extension

Product feature extension Image extension
New products that contain the original New products that capitalize on the
brand's distinctive features such as original brand's personality, user and
taste, ingredient, or component usage imagery (e.g. Porsche sunglasses)
(e.g. Hershey's chocolate milk)
Know-how extension (e.g. Honda lawnmowers)
Usage extension
Companion products: supplement or
complement (e.g. Duracell Durabeam flashlights)

Table I. Classification of brand extensions

464 JOURNAL OF PRODUCT & BRAND MANAGEMENT, VOL. 12 NO. 7 2003

 research, that part is assumed to be ``brand essence'', defined as the deep
meaning of a brand in the consumers' mind and life (Upshaw, 1995). This
study assumes that every well-established brand possesses a defined set of
essence. The essence of a brand would be the core of brand identity, and
typically coming from a flagship product of the brand. A flagship product is
defined as the product that consumers most closely associate with the
original brand (John et al., 1998), such as shampoo with Vidal Sassoon.
Influenced by the attributes of a flagship product, a set of brand essence
would be formed in the minds of consumers as the ``-ness'' of the brand
(Upshaw, 1995), such as ``coke-ness.'' Brand essence can be thought of as
the root association set of a brand that contains both product-related and
non-product-related attributes. For example, Volvo's brand essence may
include ``safety'' and ``car'', which are product-related attributes, while its
brand essence also can include other brand essence such as ``prestige'',
which is a non-product-related attribute.
Attributes In summary, brand essence should be able to help us understand the basis of
fit because it is the core of brand identity and can consist of both
product-related and non-product-related attributes. Accordingly, this
research's assumption is that consumers evaluate brand extensions primarily
by evaluating brand essence consistency rather than just product-similarity or
concept-consistency, no matter what kind of extension it is.
In addition, consumers would feel the degree of extension's distance (i.e.
closeness and remoteness) by estimating the perceived fit of the original
brand essence in extensions. Original brand's brand essence would be a
required part of both close and remote extensions. However, perhaps it is
more critical in remote extensions because very remote extensions would
more likely keep only the original brand's brand essence, while most of the
other parent brand associations do not exist. For this reason, it might be
better for companies to explicitly ensure their consumers that the original
brand essences are in the brand extensions, by communicating and reminding
the appropriate essences. Based on these assumptions, first two hypotheses
are stated as follows:
H1. Consumers evaluate a brand extension's closeness and remoteness by
evaluating the perceived fit of the original brand essence with the
extension (category).
H2. Consumers' attitude toward brand extensions (especially, remote
extensions) can be improved when original brand's brand essence is
cued in extension communication contexts.

Attribute dissonance
Some dissonance Consumers often experience some dissonance between the original brand's
attributes and the extended brand's attributes. For example, the softness
attribute of facial tissue would be difficult to transfer into bread with the
same meaning because the feeling found in facial tissue will not be soft
inside the mouth. Also, the softness attribute of bread will result in a
dissonance, i.e. the feeling of being greasy on the face, not soft like tissue.
In psychology, dissonance has been generally defined as psychological
discomfort (Festinger, 1957; Carlsmith and Aronson, 1963; Menasco and
Hawkins, 1978; Elliot and Devine, 1994). Festinger (1957) described a
person as being in a dissonant state if two elements in his/her cognition, that
is, in his knowledge of himself, behavior, feelings, desires, or his/her

JOURNAL OF PRODUCT & BRAND MANAGEMENT, VOL. 12 NO. 7 2003 465

 knowledge of the world, are inconsistent. Dissonance can occur when people
try to form an opinion or make a decision when cognition and opinions pull
or direct them in different directions (Sweeney et al., 2000). Thus, in the
brand extension situations, consumers would have two conflicting or
commingled directions of approach (e.g. ``one of my favorite brand names'')
or withdraw (e.g. ``unfitting extension category with the brand name'' or
``product-attribute concept seems consistent but felt-meaning is very
different such as the case of `softness' between the facial tissue and the
bread''). Therefore, reducing such dissonance would bring more extension
acceptance in consumers' minds. Back to the facial tissue and bread
example. If the facial tissue company wishes to extend its brand into a bread
category, it would be better to attempt to address the dissonance by stating
something like ``feel the tasty softness in your mouth'', not just saying ``feel
the softness in your mouth''. Although it appears to be minor, the extra word
``tasty'', which is referred to as a ``dissonance reducer'' in this research,
might at least address the softness attribute dissonance situation by assuring
consumers that the bread is very soft in mouth. This generates the third
hypothesis:
H3. A consumer's attitude toward a brand extension can be improved by
adding attribute dissonance reducers in the extension communication
context.

Communication message study

Four groups An extension positioning study conducted by Aaker and Keller (1990) tried
to solve the question: ``How are consumer evaluations affected when
different types of information are provided in the extension context?'' Their
research measured attitudes toward four hypothetical brand extensions by
differentiating communication contexts through four groups:
(1) Group 1 ± no quality cues or attribute elaborations.
(2) Group 2 ± quality cues only.
(3) Group 3 ± attribute elaborations only.
(4) Group 4 ± both cues and elaborations.
They found that the most effective advertising strategy for an extension is
one that emphasizes information about the extension rather than reminders
about the original brand.
Salience and relevance In addition, as introduced previously, Bridge et al. (2000) studied different
communication effects on different extension settings in which salience and
relevance of the brand extensions were manipulated, where the salience of
associations was assumed to depend upon the dominant parent brand
associations, and where the relevance was assumed to depend on the parent
brand-to-extension category relationship. Their study found that low fit
perceptions, which lack either salience or relevance, could improve with
relational or elaborational communication strategies, which alter the salience
and relevance of the associations that come to mind when evaluating
extensions.
However, these research results do not give clear information to solve the
present study's questions and hypotheses because they:
. tested original brands' quality cues, but did not discuss brand
essence;

466 JOURNAL OF PRODUCT & BRAND MANAGEMENT, VOL. 12 NO. 7 2003


Experiment method
. did not consider the extension's closeness and remoteness;
. did not classify the extension types (although Bridge et al. classified
extensions into two that lack either salience or relevance of parent brand
associations in extensions, this classification does not fit with our study
purpose because their classification and communication strategies are
largely based on ``lacks'', but not much on ``remains (i.e. brand
essence)''; and
. did not approach the communication messages as means of both brand
essence reminder and dissonance reducer.
Consequently the present research studies how both brand essence reminder
and dissonance reducer in the communication contexts can affect consumers'
attitude toward different extension situations (i.e. close or remote extensions)
using Aaker and Keller's (1990) experiment method.

Method
There are four different types of studies in this research: three pretests and an
experiment. The three pretests are the investigation of the brand essence; the
examination of the closeness/remoteness perception of brand essence and
extension category; and a focus group that revealed a set of attribute
dissonance. In selecting a test brand, one relatively neutral and well-known
brand (Colgate) was chosen to minimize variations from mixing several
functional and prestige brands. Although using fictitious brand names might
provide more experimental control, this study utilizes a real brand name to
avoid misinterpreting category-level effects as brand-level effects
(Broniarczyk and Alba, 1994).

Pretest 1
Brand essence The brand association test was conducted using 30 graduate students.
Operationally viewed as the brand essence, Colgate's salient brand
associations were collected by asking the students to give five
top-of-the-mind words associated with Colgate (see Table II). The five most
mentioned associations were used as tests of brand essence. Although tube
and red tied in with ``fresh and clean'', they were eliminated from the test
because they were attributes that are more superficial. However, it is
important to note that these superficial attributes would be important as
visual consistency elements in extensions. Based on the five associations (i.e.
brand essence), four hypothetical brand extensions (based on Table I) were
created and tested in the following experiment:
(1) product feature extension (chewing gum);
(2) usage extension (electric toothbrush);

Associations Percentage of respondents (n = 30)

a
Toothpaste 78
Whitea 43
Teetha 43
Fresha 29
Cleana 25
Red 23
Tube 23
Note: a Colgate's brand essence

Table II. Top seven brand associations of Colgate

JOURNAL OF PRODUCT & BRAND MANAGEMENT, VOL. 12 NO. 7 2003 467

 (3) medium-level image extension (aftershave); and
(4) very remote image extension (men's underwear).
The order of its distance of extension was from close to remote. The distance
of extension will be verified in pretest 2.

Pretest 2
Generic categories To verify the perceived distance between brand essence and extension
categories, a second pretest was conducted. The survey studied how closely the
subjects felt to those four generic extension categories via the five Colgate brand
associations. For example, one question asked a perceived imagery distance
between ``aftershave'' and ``clean'' without the brand name Colgate. The means
of the five words' distance scores toward the four categories are in Table III.
The result suggests a high possibility that the distance of extension would be
judged through examination of distance between parent brand essences and
extension categories as we hypothesize. This result also implies that our
extension distance manipulation was successful.

Pretest 3
In order to study the effects of attribute dissonance reducers, dissonant
attributes were investigated in a focus group of ten graduate students. In this
pretest, most dissonant attributes the focus group had were negative aspects
of toothpaste ± feeling found in extensions as well as the company's
expertise. The participants freely discussed what kind of message cue in the
communication context might reduce their negative feelings associated with
the extensions. The dissonance ``attributes'' and ``reducers to be tested'' in
Table IV reflect the qualitative findings from this focus-group study.

Electric Chewing Men's

toothbrush gum Aftershave underwear
Toothpaste 6.47 (1.07) 4.07 (1.82) 3.07 (2.07) 1.13 (0.35)
White 4.60 (2.34) 3.33 (1.40) 2.23 (1.77) 5.67 (1.75)
Teeth 6.90 (0.55) 5.87 (1.31) 1.43 (1.04) 1.33 (0.80)
Clean 6.23 (1.30) 3.37 (1.45) 5.57 (1.19) 4.80 (1.77)
Fresh 5.13 (1.72) 5.97 (1.16) 5.97 (1.35) 3.60 (2.04)
Average 5.86 4.52 3.63 3.30
Notes: Scale of the closeness/remoteness score: 1-7 (1 = remotest, 7 = closest);
numbers in parentheses: standard deviations

Table III. Distance scores between Colgate's brand essence and extension
categories
Dissonance
reducers to be
Extensions Brand essence Dissonant attributes tested
Colgate chewing Teeth, freshness, Toothpaste taste Great flavor
gum white
Colgate electric Teeth, cleanness, No skill, non-durable A new technology
toothbrush toothpaste
Colgate Freshness, No skill, toothpaste feeling, Elegant fragrance
aftershave cleanness low symbolic association
Colgate men's Freshness, No skill, not stylish Newly patented
underwear cleanness fabric

Table IV. Dissonant attributes and reducers

468 JOURNAL OF PRODUCT & BRAND MANAGEMENT, VOL. 12 NO. 7 2003


Communication contexts
The experiment
Based on the results of the three pretests, the experimental brand
extension communication study was conducted. To compare different
effects of communication, five different communication contexts were
employed: no-cue, brand essence cue, brand extension attribute cue,
brand extension attribute cue + attribute dissonance reducer, and all cues.
The dissonance reducers in cue No. 4, which were used in the group
3 and 4 (Table V), were hypothesized to reduce anticipated
dissonances. Therefore, subjects in the group 3 and 4 saw mixed two cues as
one cue.
Subjects were assigned to one of the four groups, and each group
saw the same brand extensions as no-cue context on the first page of the
questionnaire, while watching different communication cue contexts in
the following pages. The first question asked respondents their overall
attitude toward the original brand (Colgate) using a one to seven scale in
order to compare them to the extensions. After the initial attitude
question, the no-cue tests were conducted to see how subjects' attitude
changed when they saw specific cues. The attitudes toward different
extensions and communication cues were gathered from 120 students (56
undergraduate students and 64 graduate students; 30 for each survey) by
averaging two measures: perceived quality and likelihood of trying the
extension (Aaker and Keller, 1990). Tables V and VI shows each group
and cues manipulated.

Group no. Cues

Group 1 No-cues + (a) brand essence cues
Group 2 No-cues + (b) extension attribute cues
Group 3 No-cues + (c) extension attribute cues + attribute dissonance reducers
Group 4 No-cues + (a, c) all cues

Table V. Experiment groups and cues

No-cues
Colgate chewing gum
Colgate electronic toothbrush
Colgate aftershave
Colgate men's underwear
Note: Attribute dissonance reducers are in italics
Cues
(a) For your oral care
(b) In peppermint flavor with anti-tartar formula
(c) In great peppermint flavor with anti-tartar formula
(a) From the leader of quality dental care products
(b) Great plaque remover
(c) Great plaque remover with a new technology
(a) From the maker of freshness
(b) Soothes and refreshes your skin
(c) Soothes and refreshes your skin with elegant
fragrance
(a) From the maker of clean and fresh personal care
products
(b) Great style in various fresh colors
(c) Great style in various fresh colors with newly
patented fabric which allows for maximum
breathability

Table VI. Description of message cues

JOURNAL OF PRODUCT & BRAND MANAGEMENT, VOL. 12 NO. 7 2003 469

 Results and managerial implications
Table VII shows the pattern of changes in attitudes toward extensions with
different communication message strategies. Each of the specific cue
attitudes was compared with each no-cue attitude in each same group to see
the within-group attitude changes. However, the mean of all group's attitude
for the original brand was calculated and shown as the overall original brand
attitude (mean = 4.9, STD = 0.7). There is no significant difference among
groups. Differences in means were examined for each group with no-cue
message and specific cue messages, in order to test the within-group attitude
changing effects of specific communication cues.
Six findings There are six main research findings as follows:
(1) Finding No. 1. Subjects evaluated the brand extensions' closeness and
remoteness by evaluating the perceived fit of the original brand essence
with the extensions. H1 is indeed supported. This research hypothesized
that consumers evaluate a brand extension's closeness and remoteness by
evaluating the perceived fit of the original brand essence with the
extensions. As we see, subjects' overall average attitudes toward no-cue
message ranged from highest (4.3; chewing gum ± product feature
extension) to lowest (2.4; underwear ± image extension) while the
original brand attitude was 4.9. This result is consistent with the results
of the pretest 2 (brand essence and closeness/remoteness correlation
study). The exception is the electric toothbrush. A reason for this may be
that one of the five words that pretest 2 used (i.e. toothpaste) was much
more closely associated with electric toothbrush than chewing gum.
Figure 1 shows the comparison of attitude means between pretest 2

Attribute
No-cue cue +
(control Essence Attribute dissonance
groups) cue only cue only reducer All cue
Chewing gum 4.3 (0.68)** 5.0 (1.53) 5.1 (1.36) 5.1 (1.46) 4.9 (1.29)
Electric toothbrush 4.2 (0.58) 4.4 (1.63) 4.8 (1.36) 4.6 (1.24) 4.4 (1.46)
Aftershave 3.4 (0.78) 3.6 (1.44) 3.6 (1.53) 3.5 (1.47) 3.6 (1.61)
Underwear 2.4 (0.59) 2.5 (1.21) 2.6 (1.41) 2.8 (1.19) 3.0 (1.45)
Sample size 120 30 30 30 30
Notes: * Extension attitudes are the average of perceived extension quality (1 = inferior,
7 = superior) and intention to buy the extension (1 = not at all likely, 7 = very likely).;
** Numbers in parenthesis: standard deviations of each attitude

Table VII. Brand extension attitudes* in different cues

Figure 1. Mean comparison between pretest 2 and experiment

470 JOURNAL OF PRODUCT & BRAND MANAGEMENT, VOL. 12 NO. 7 2003

 (perceived distance between brand essences and extended categories: a
higher score suggests the closer distance) and the experiment (attitude
toward extensions). Although attitude and distance perception are
different in construct and were measured with different methods (but in
same scale points), it would be interesting to see how they achieved such
similar results. This comparison implicitly supports this research's
hypothesis that consumers evaluate extensions by focusing on brand
essence.
Brand information (2) Finding No. 2. The effect of using ``only brand essence cue'' was positive
in most extensions, though not sufficient (partial support of H2).
Although brand essence cues were not significantly effective in the
electric toothbrush extension case (t = 1.65, p > 0.1), they were
significantly effective in other extension cases (chewing gum, after-
shave, underwear: all t > 2.30, all p < 0.03). However, those effects were
not sufficient because other cues generated higher attitudes in three
extension cases, though the closest extension (i.e. chewing gum)
achieved significantly high attitude (mean = 5.0, Diff. = 0.7, t = 2.30,
p < 0.05). These results imply that consumers may want more
information on farther extensions, rather than just brand essence
information. However, in close extensions, it seemed that essence
information could be sufficient to communicate.
(3) Finding No. 3. The combined effects of brand essence cues and other
cues were not significant. Although all cue effects were significant
((chewing gum: t = 3.52, p = 0.001), (electric toothbrush: t = 3.04,
p = 0.005), (underwear: t = 2.37, p = 0.025)), there were no significant
attitude differences between groups 3 and 4 (±0.7 < all t < 0.65, 0.50 < all
p < 0.71). Since the only cue difference between group 3 and 4 was the
brand essence cue, this result implies that there are no significant
combined effects of brand essence and other cues. Although brand
essence is theorized as critical in the process of brand extension
evaluations, this implies that the essence does not need to be actually
communicated with other attribute cues because consumers may have
already found, assumed, and linked the essence to extensions by
presented extension attributes. Accordingly, consumers might not want
to be over-communicated. However, as discussed, brand essence cues
would be effective when they are used alone, especially in close
extensions.
Attribute cues (4) Finding No. 4. In close-extension, it seems to be best to communicate
only extension attribute cues. In this study, the close extensions were
the chewing gum and electric toothbrush. As we see, among the
attitudes of these extensions, the highest attitudes were achieved when
only extension attribute cues were given ((chewing gum: t = 3.33, p =
0.002), (electric toothbrush: t = 2.85, p = 0.008)). This suggests that
consumers believe that it is enough to know extension attribute cues, and
those attributes are properly communicated. If other information beyond
that is given, it seems that they tend to regard it as just over-
communication, which results in more negative rather than positive
effects.
(5) Finding No. 5. The attribute dissonance reduction hypothesis (H3) was
not supported. Theoretically, in brand extension, attribute dissonance
must be reduced. Although group 3 showed a significant attitude change
when they saw dissonance reducers with extension attribute cues, there

JOURNAL OF PRODUCT & BRAND MANAGEMENT, VOL. 12 NO. 7 2003 471

 were no significant attitude differences between groups 2 and 3 (±0.59
< all t < 0.66, all p > 0.56). In addition, although the effect of including
dissonance reducers seemed to be more positive than excluding in the
underwear extension case (Table VII), it was only marginally significant
(t = 0.66, p = 0.512). Accordingly, these results imply that companies
might not need actually to induce attribute dissonance reduction. It
seems that consumers can easily overcome or overlook dissonance by
seeing extension attribute cues. Cooper and Fazio (1984) support this
result, suggesting that dissonance reduction (e.g. attitude change) would
emerge only when individuals felt personally responsible for producing
foreseeable negative consequences. Since the experiment did not require
any responsibility for the consequences of self-report of their attitudes,
respondents might not be motivated to reduce dissonance. Furthermore,
Harmon-Jones (2000) suggests that the dissonance reduction may not
occur and people may simply live with the negative affect. Moreover,
individuals may differ in their tolerance for dissonance (Festinger 1957).
Accordingly, dissonance reduction would seem to depend on consumers'
involvement with the purchase situation and their individual level of
tolerance for dissonance. Therefore, dissonance reducers would need to
be very carefully used for low involving and high dissonance-tolerable
product purchase situations because consumers would be able to easily
overcome the dissonance, otherwise they would probably be more
stimulated to experiencing dissonance and inducing more negative
associations.
Relevant cue (6) Finding No. 6. In very remote extensions, it seemed to be better to use all
relevant cues. Consumers seem to need a large amount of information in
evaluating remote extensions. As seen in Table VII, when subjects saw
all relevant cues, their attitude toward the remotest extension (i.e. men's
underwear) went up highest to 3.0 (t = 2.37, p = 0.025) from the same
group no-cue's 2.4. However, although using all cues seems to be also
effective in the medium-level remote extension (i.e. aftershave), it was
not significant (t = 0.40, p = 0.69). In addition, although using all cues
was effective in close extensions (all t > 3.0, p = 0.005), it could not
generate sufficiently high attitudes. These results imply that for
extensions that are very far removed, consumers may require the greatest
amount of information to provide an improvement in attitude, but for
extensions that are little far removed, there should be careful
considerations in using all cues.

Discussions
This paper has served to contribute to our understanding of brand extension
classification and the process of extension evaluation. In addition, this
research has contributed to an understanding of the core communication
strategies of brand extensions by suggesting several important considerations
resulting from the survey. The experiment showed that brand essence should
be carefully used (or not used) in communication contexts, and dissonance
reducers would not be much effective to use for the situation of
low-involving product purchase.
There are four limitations. First, the survey could not measure the pure effect
of dissonance reducers since the cues were used with other extension
attribute cues. Second, the survey simply tested attitudes toward the contents
of literal contexts. Since today's communications are integrated in various
channels and forms, consumers can have different attitudes toward

472 JOURNAL OF PRODUCT & BRAND MANAGEMENT, VOL. 12 NO. 7 2003

 In close extensions
Brand essence cue is effective to use
(2, 3)
Attribute dissonance reducers would be
ineffective to use for low-involving
products (5)
It would be best to focus on only
extension attribute cues (4)
Do not ]ust use many cues; avoid
over-communication (4)
Note: Numbers in parenthesis are the relevant finding numbers of the experiment
In remote extensions
Brand essence cue is effective to use
(2, 3)
Attribute dissonance reducers would be
ineffective to use for low-involving
products (5)
It might be best to focus on all relevant
cues if it is very remote extension (6)

Table VIII. Communication message recommendations

extensions depending on the amount and quality of different communication

methods. Third, this research only studied a brand, which has one flagship
product (toothpaste). Therefore, other extension cases of brands that have
several flagship products or categories (e.g. Yamaha, Sony, 3M, etc.) need
more extensive research. Finally, sample characteristics (i.e. college and
graduate students) and size may limit the generalizability of the result.
Close and remote The present experiment offers several practical considerations of
extensions communication strategies for brand extensions. Those considerations can be
summarized into two criteria ± recommendations for close extensions and
remote extensions (Table VIII). Using these recommendations, and
understanding the proposed model of extension classification, companies
would be able to design more thorough brand extension communication
strategies.

References
Aaker, D.A. and Keller, K.L. (1990), ``Consumer evaluations of brand extensions'', Journal of
Marketing, Vol. 54, January, p. 38.
Boush, D. and Loken, B. (1991), ``A process-tracing study of brand extension evaluations'',
Journal of Marketing Research, Vol. 28, February, pp. 16-28.
Bridges, S., Keller, K.L. and Sood, S. (2000), ``Communication strategies for brand extensions:
enhancing perceived fit by establishing explanatory links'', Journal of Advertising, Vol. 29,
pp. 1-11.
Broniarczyk, S.M. and Alba, J.W. (1994), ``The importance of the brand in brand extensions'',
Journal of Marketing Reseach, Vol. 31, May, pp. 214-28.
Carlsmith, J.M. and Aronson, E. (1963), ``Some hedonic consequences of the confirmation and
disconfirmation of expectations'', Journal of Abnormal and Social Psychology, Vol. 66,
pp. 151-6.
Cooper, J. and Fazio, R.H. (1984), ``A new look at dissonance theory'', in Berkowitz, L. (Ed.),
Advances in Experimental Social Psychology, Vol. 17, Academic Press, San Diego, CA,
pp. 229-66.
Elliot, A.J. and Devine, P.G. (1994), ``On the motivational nature of cognitive dissonance:
dissonance as psychological discomfort'', Journal of Personality and Social Psychology,
Vol. 67, pp. 382-94.
Festinger, L. (1957), A Theory of Cognitive Dissonance, Stanford University Press, Stanford,
CA.
Harmon-Jones, E. (2000), ``A cognitive dissonance theory perspective on the role of emotion
in the maintenance and change of beliefs and attitudes'', in Frijda, N.H., Manstead, A.R.S.
and Bem, S. (Eds), The Effects of Emotions upon the Formation and Strength of Beliefs,
Cambridge University Press, Cambridge, pp. 185-211.
John, D.R., Loken, B. and Joiner, C. (1998), ``The negative impact of extensions: can flagship
products be diluted?'', Journal of Marketing, Vol. 62, p. 19.

JOURNAL OF PRODUCT & BRAND MANAGEMENT, VOL. 12 NO. 7 2003 473

 Kapferer, J. (1998), Strategic Brand Management: Creating and Sustaining Brand Equity
Long-Term, Kogan Page, London.
Keller, K.L. (1998), Strategic Brand Management: Building, Measuring, and Managing Brand
Equity, Prentice-Hall, Upper Saddle River, NJ.
Menasco, M.B. and Hawkins, D.I. (1978), ``A field test of the relationship between cognitive
dissonance and state anxiety'', Journal of Marketing Research, Vol. 15, November,
pp. 650-5.
Park, C.W., Milberg, S. and Lawson, R. (1991), ``Evaluation of brand extensions: the role of
product feature similarity and brand concept consistency'', Journal of Consumer
Research, Vol. 18, September, pp. 185-93.
Romeo, J.B. (1990), ``The effect of negative information on the evaluation of brand extension
and the family brand'', in Holman, R.H. and Solomonin, M.R. (Eds), Advances in
Consumer Research, Association for Consumer Research, Provo, UT, pp. 399-406.
Sweeney, J.C., Hausknecht, D. and Soutar, G.N. (2000), ``Measuring cognitive dissonance: a
multidimensional scale'', Psychology and Marketing, Vol. 17 No. 5, pp. 369-85.
Tauber, E.M. (1988), ``Brand leverage: strategy for growth in a cost-control world'', Journal of
Advertising Research, August/September, pp. 26-30.
Upshaw, L.B. (1995), Building Brand Identity: A Strategy for Success in a Hostile
Marketplace, John Wiley & Sons, New York, NY.
&

474 JOURNAL OF PRODUCT & BRAND MANAGEMENT, VOL. 12 NO. 7 2003

 This summary has been
provided to allow managers
and executives a rapid
appreciation of the content
of this article. Those with a
particular interest in the
topic covered may then read
the article in toto to take
advantage of the more
comprehensive description
of the research undertaken
and its results to get the full
benefit of the material
present
460
Executive summary and implications for managers and
executives
``Strange'' ads can rejuvenate a tired old brand
Every now and then creative teams in advertising agencies have a rush of
blood to the head and produce off-the-wall treatments for well known
brands. And sometimes the client bites. When UK beer brand Boddington's
relaunched as ``the cream of Manchester'' by apeing face cream and
perfume ads the brand took off (and the agency raked in the awards). Levi's
broke their tradition of sexy male models and old pop songs when they
introduced ``Flat Eric'' a cartoon character. It worked.
Us folk in advertising are always looking to push the boundaries for brand
owners but we have not been doing so in a considered way. Rather we have
been acting on ``gut feel'' and instinct. What Lange and Dahlen do here is to
provide something of a rationale for dramatic changes in advertising
treatment.
Should we always stick with the knitting?
Conventional wisdom ± backed up with some research ± has always argued
for consistency, gradual change and the protection of brand position and
personality. The brand is too important an asset for the firm to take huge
risks by indulging the eccentric ideas of advertising creative teams.
Off-the-wall advertising can only serve to indulge the advertiser and is likely
to damage the brand.
Lange and Dahlen set out to challenge this position by arguing that ``. . .
some brands may benefit from advertising breaking the consistency, using
`strange' ads that are incongruent with the associations consumers hold with
the brand.'' And, say Lange and Dahlen, it is the familiar brands ± the big
brands ± that stand to gain most from this radical approach.
Lange and Dahlen's argument is built round what we already know about
information incongruency having positive effects as a result of increased
memorability. After years of familiar, even boring, advertising from a big
brand, the ``strange'' ad acts to reinvigorate consumers by stimulating
curiosity, interest and involvement. The new and unusual ad grabs the
consumer's attention leading to a situation of ``. . . more cognitive processing
and careful elaboration which in turn makes the ad/information more
persuasive and the response more favourable.''
What about the risks of breaking the mould?
Surely, despite the added stimulus of the ``strange'' ad, there are still huge
risks? We cannot assume that the consumer response will be positive ± the
existing advertising is not broken, why change it? The flip side of this
argument, which Lange and Dahlen note, is that the boring old advertising
approach can lead to over-familiarity, boredom and apathy. Hardly the kind
of response we want to get from consumers!
However, the less familiar the consumer is with the brand, the greater the
risks associated with ``strange'' advertising. What Lange and Dahlen show
when they test their idea, is that the brands that gain from significant
changes in advertising treatment are the bigger, more familiar brands.
Smaller, less familiar brands do not get the same advantages from a radical
shift in advertising treatment ± indeed such a shift may damage the brand.
It is clear that adopting a radical, different advertising approach represents
a considerable risk even for the most powerful brand. Lange and Dahlen
JOURNAL OF PRODUCT & BRAND MANAGEMENT, VOL. 12 NO. 7 2003
 report that advertising incongruent with the brand is less readily recalled by
the consumer. As a result, normal approaches to ad testing and screening
are likely to rule out the radical treatment since the most widely used tests
involve assessment of advertising recall.
What Lange and Dahlen argue is that the lower recall numbers are
acceptable for the bigger brand because, as their research shows, the
``strange'' ad increases brand memorability. And, in the final analysis, it is
the brand we want the consumer to recall not just the ad. However, to
achieve this response, the consumer has to be familiar with the brand in the
first place.

Familiarity breeds contempt

If the old advertising is working, why change? Brand awareness figures are
high, the advertising retains a high level of recall and the consumer
continues to buy. What Lange and Dahlen suggest is that familiarity with the
brand does not translate automatically into liking for the brand. Moreover,
consumers will attribute their familiarity to seeing a lot of very familiar ads
and, if pressed, will probably express boredom and perhaps irritation with
the old advertising treatment.
As Lange and Dahlen point out the ``strange'' ad for a familiar brand ``. . .
made it easier for consumers to remember the brand and harder to remember
the ad, thus making the brand more salient with a decreased risk of
advertising wear-out.'' The ``strange'' ad makes consumers work a little
harder in order to link the incongruent treatment with the established brand
``schema'' in their minds.
While familiarity may breed contempt, unfamiliarity presents a different
problem. The unfamiliar brand does not have an established ``schema'' in
the mind of the consumer, so faced with an incongruous treatment that
consumer has no reference point. As a result, the consumer is likely not only
to fail to recall the ad itself but also to fail to recognise the brand being
advertised. For the smaller brand, radical changes to advertising treatment
are clearly a very bad idea ± whatever your advertising agency says, stick to
the tried and tested approach in order to build the levels of brand awareness.
Without this awareness, unusual, radical ± ``strange'' ± advertising simply
does not work.
The work here presents advertisers with a further option in the development
of advertising strategy and, perhaps more importantly, it provides
advertising planners with a basis on which to research and test radical
advertising approaches. Many planners have argued against relying on
recall testing since it always produces more of the same advertising (if I test
an ad I have run against an ad I have not run, the former will almost always
get a high level of recall). In addition to ad recognition, we should also be
looking to assess brand recognition and memorability.
Finally, advertisers need to recognise that many consumers see advertising
as part of the entertainment they are exposed to. A rationale for ``strange''
advertising is that it sustains consumer interest and, over time, levels of ad
awareness will rise as consumer becomes familiar with the new approach.

(A preÂcis of the article ``Let's be strange: brand familiarity and ad-brand

incongruency''. Supplied by Marketing Consultants for Emerald.)

JOURNAL OF PRODUCT & BRAND MANAGEMENT, VOL. 12 NO. 7 2003 461