72 Journal of International Marketing The Interplay Between Global and Local Brands: A Closer Look at Perceived Brand Globalness and Local Iconness… [623775]

72 Journal of International Marketing The Interplay Between Global and
Local Brands: A Closer Look at
Perceived Brand Globalness and
Local Iconness
Ay șegül Özsomer
ABSTRACT
By studying consumer samples in an emerging market, Turkey, and two mature markets, Singapore and Denmark, the
author tests the chain of relationships that drive consumers’ likelihood of purchasing the global brand in the presence of
a local brand in a linear structural relations framework. The results indicate that perceived brand globalness is posi –
tively related to local iconness in an emerging market, but the relationship is negative in advanced markets. Developing
local iconness helps build the perception of prestige in all three markets. Furthermore, local iconness is positively
related to local brand quality perceptions in the culturally grounded categories of food in an emerging market, whereas
in nonfood categories, local iconness has no connection to quality. In terms of cross-effects, as expected, the perceived
quality of the local brand is negatively associated with global brand purchase likelihood in all markets and categories
studied. In contrast, local brand prestige dampens global brand purchase likelihood for older consumers in an emerging
market. The article concludes with implications for global and local brand managers.
Keywords: survey, structural equations modeling, signaling theory, associative network memory model, global
brand management
In many product categories, consumers are increas –
ingly faced with a choice between global and local
brands. In today’s global marketplace, it is important
for marketers to understand how consumers make this
choice and why some consumers prefer global brands to
local brands. For example, since launching its “Path to
Growth” strategy in 2000, Unilever has reduced the
number of brands from 1600 to 400 leading brands and
fewer than 250 tail brands (Unilever 2007). This enables
Unilever to concentrate resources on a portfolio of lead –
ing global brands with strong growth potential that best meet the needs and aspirations of people around the
world. At approximately the same time, Procter &
Gamble also pruned its brand portfolio in favor of
global brands (Pitcher 1999), and in 2003, Heinz
declared its intention to focus on a smaller number of
“power brands” and selling many of its local brands.
These and many other companies are betting their
futures on global brands: brands that consumers can
find under the same name in multiple countries with
generally standardized and centrally coordinated mar –
keting strategies (Özsomer and Altaras 2008;
Steenkamp, Batra, and Alden, 2003; Yip 1995). While
most researchers agree that wide availability and recog –
nition are the key features of global brands, universal
relevance, global image, standardization, esteem, high Ay șegül Özsomer is Associate Professor of Marketing, Ko ç University
(e-mail: [anonimizat]). The author thanks Asli Berispek,
Soohow Wong, and Huseyin Erol for their assistance in the project.
The author also thanks Rick Bagozzi, Skander Essagier, Tarcan
Kumkale, and Zeynep Gürhan-Canlı and participants at Manchester
University, CIMAR 2007 for helpful comments. This research was
funded by the Dean’s Research Fund at the College of Administrative
Sciences and Economics, Koç University. Journal of International Marketing
©2012, American Marketing Association
Vol. 20, No. 2, 2012, pp. 72 –95
ISSN 1069-0031X (print) 1547-7215 (electronic)

The Interplay Between Global and Local Brands 73 quality, and social responsibility have been suggested as
well (e.g., Dimofte, Johansson, and Ronkainen 2008;
Holt, Quelch, and Taylor 2004; Johansson and
Ronkainen 2005; Strizhakova, Coulter, and Price 2008).
In many markets, global brands compete with strong
local counterparts. Local brands are developed for and
tailored to the unique needs and desires of local mar –
kets. Despite being described as “only available in a spe –
cific geographical region” (Dimofte, Johansson, and
Ronkainen 2008, p. 118), local brands have their own
strengths, such as perceptions of uniqueness, originality,
and pride of representing the local market. Local brands
have traditionally benefited from a high level of aware –
ness and close relationships with consumers in their
countries. As Dimofte, Johansson, and Ronkainen
(2008) reveal, local brands not only reflect but also help
define the character of the local market. As such, some
of these brands are perceived as local icons in their mar –
kets insofar as they are associated with symbols of the
local culture, heritage, and country.
The opportunities and threats of increasing globaliza –
tion have created a sense of urgency for companies to
succeed in not only mature but also emerging inter –
national markets (Burgess and Steenkamp 2006). In
striving to succeed in emerging markets, multinational
corporations (MNCs) are bringing their global brands
to emerging-market consumers because most of the eco –
nomic growth in the next decades will come from these
markets. However, cross-national studies reveal signifi –
cant differences between consumers from emerging and
mature markets. For example, Alden, Steenkamp, and
Batra (1999) indicate that globally positioned brands
might be more attractive than local brands, especially
for emerging-markets consumers. Similarly, Batra et al.
(2000) find that emerging-market consumers prefer
nonlocal (i.e., foreign) brands to local brands, aspiring
to a greater global community and, in the process,
downgrading their own brands compared with global
brands. Given the influx of global brands, this research
focuses on the interplay between local and global brands
and, in particular, the antecedents of successful local
brands and the consequences for global brands. How
can local brands defend themselves against their global
counterparts? Are there differences across mature and
emerging markets in the usefulness of strategies that
local brands can use? These questions guide this
research.
Building on Steenkamp, Batra, and Alden’s (2003)
work, this study suggests “local iconness” as a possible counter strategy available to local brands when compet –
ing against global brands. Holt (2004, p. 4) defines
iconic brands as consumer brands that carry “consensus
expressions of particular values held dear by some mem –
bers of a society.” Icons succeed because they forge a
deep connection with the culture (Holt 2003). Brands
that become local icons achieve dominant positions
partly due to their high levels of group symbolism or
associations with a group and its culture (Torelli, Tat
Keh, and Chiu 2008). Similarly, a major strength of
local brands is their closer connection to national iden –
tity, local culture, and heritage (Ger 1999). In this arti –
cle, “local iconness” is defined as the degree to which a
brand symbolizes the values, needs, and aspirations of
the members of the local country. Perceptions of a brand
being a good symbol of the country, embodying the
local culture, and representing what the country is all
about are included in the conceptual domain of local
iconness. In this conceptualization, the emphasis is on
local associations, and the cultural reference group is the
local country culture. Thus, this research investigates
the extent to which local brand iconness can help sway
global preferences in an emerging and mature country
setting.
The literature suggests that the interplay between local
and global brands could depend on the product category
and consumer segment. Local brands may be more
attractive for products consumed in the household, such
as foods. Global brands, however, continue to be more
appealing for publicly visible goods because they are
higher in aspirational value and are associated with sta –
tus, modernity, cosmopolitan sophistication, and tech –
nology (Batra et al. 2000; Dimofte, Johansson, and
Ronkainen 2008; Strizhakova, Coulter, and Price 2008;
Zhou and Belk 2004). Food and drink categories have a
stronger cultural grounding because local tastes and a
stronger affection for tradition give local brands the
ability to challenge the marketing efforts of global
brands (Schuh 2007). Studies on multidomestic versus
global industry distinction also support this expectation
(Porter 1986; Prahalad and Doz 1987; Yip 1995). A
global industry is characterized by the presence of cus –
tomers with homogeneous needs. Thus, marketing is
transferable across countries, giving global brands an
edge in their competition against local brands. Similarly,
teenagers and younger adults are more attracted to
global brands (Hassan and Katsanis 1994; Strizhakova,
Coulter, and Price 2008), endangering the feasibility of
local iconness as an alternative strategy. Thus, this study
examines the moderating role of product category (food
and drinks vs. nonfood categories) and consumer age to

74 Journal of International Marketing enhance understanding of the competition between
global and local brands.
This study integrates two theoretical approaches, signal –
ing theory and associative network memory model, to
explicate the underlying relationship between the local
brand and its global counterpart. This is accomplished
by investigating local iconness more closely as a counter
strategy available to local brands and determining its
relationship to other local brand equity characteristics
as well as the brand equity of the global brand simulta –
neously. By studying consumer samples in an emerging
market (i.e., Turkey) and two mature markets (i.e., Sin –
gapore and Denmark), the chain of relations that drive
consumers’ likelihood of purchasing the global brand in
the presence of a local brand is tested in a linear struc –
tural relations framework. The results indicate that
global expansion (perceived brand globalness) is a feasi –
ble defense strategy for a local brand because of its posi –
tive association with local iconness in an emerging mar –
ket, but the relationship is negative in mature markets.
In emerging markets, going away from the home raises
iconness, whereas in mature markets, it lowers iconness.
Furthermore, developing local iconness helps build the
perception of prestige in all three markets.
In terms of moderators, local iconness is positively
related to local brand quality in the culturally grounded
categories of food in an emerging market, whereas in
nonfood categories it has no connection to quality. In
terms of cross-effects, the perceived quality of the local
brand is negatively associated with global brand pur –
chase likelihood (hereinafter, GBPL) in all markets and
categories studied. When consumers have a viable local
alternative, the preference for a global alternative dimin –
ishes. Furthermore, local brand prestige dampens GBPL
for older consumers in an emerging market.
THEORETICAL BACKGROUND AND
HYPOTHESES DEVELOPMENT
Previous research has regularly used both signaling
theory (Erdem and Swait 1998, 2004) and the associa –
tive network memory model (Anderson 1983; Brewer
1988; Keller 1993, 2003) as frameworks for explaining
consumer-based brand evaluations and consumer brand
choice (e.g., Aaker 1991; Erdem and Swait 1998, 2004).
Aaker’s (1991) conceptualization and Keller’s (1993)
framework of associative network model focuses on
consumers’ brand associations. Aaker suggests that
brand associations such as perceived quality, prestige, and other proprietary brand assets (e.g., patents) influ –
ence brand evaluations directly and preference indi –
rectly. Recent research on global branding suggests that
brand globalness is a distinct brand attribute that is con –
sidered and weighted in the brand evaluation process
much like other attributes (Dimofte, Johansson, and
Ronkainen 2008; Steenkamp, Batra, and Alden 2003).
That is, as a brand association, globalness can add or
subtract value depending on country (i.e., emerging vs.
mature), category (food and drinks vs. nonfood), and
consumer characteristics (consumer age).
Signaling theory, which is based in information econom –
ics, asserts that when imperfect and asymmetric infor –
mation characterizes a market, firms may use signals
(i.e., manipulate attributes or activities) to convey infor –
mation about their characteristics (Spence 1974).
According to Erdem and Swait (1998, p.131), “when
consumers are uncertain about product attributes, firms
may use brands to inform consumers about product posi –
tions.” When a brand positions itself as a high-quality,
high-prestige, and global brand, it is signaling the prom –
ised brand position.
On the basis of previous research (Dimofte, Johansson,
and Ronkainen 2008; Holt, Quelch, and Taylor 2004;
Steenkamp, Batra, and Alden 2003), a global brand can
signal five underlying characteristics that form its
promised position: (1) wide availability, recognition,
and geographical reach; (2) aspirations of achievement,
excitement, and happiness; (3) convenience and low
risk, implying a perceived quality association; (4) envi –
ronmental and ethical responsibility; and (5) uniformity
and standardization (i.e., being the same everywhere,
with no particular adaptation to local markets). This
last dimension is where local brands signal different
positions. Previous research (Dimofte, Johansson, and
Ronkainen 2008; Kapferer 2005) has indicated that
adaptation to local preferences and culture is a key
characteristic signaled by local brands. Local brands
also signal uniqueness, originality (derived from local
adaptation), pride of representing the local area and
culture, and high awareness and availability in their
local markets.
With asymmetric and imperfect information, brand
globalness and local iconness can act as signals that
firms use to convey brand positions of high quality and
prestige. Perceived brand globalness may serve as a sig –
nal because it embodies a firm’s (past and present) mar –
keting strategies in other markets. The perception that
the brand is available in other markets not only provides

The Interplay Between Global and Local Brands 75 direct product availability, recognition, and reach infor –
mation, but also conveys indirect information on prod –
uct attributes about which consumers are imperfectly
informed. Similarly, availability in other markets may
function as a quality (and prestige) signal by guiding
consumer inferences about demand- or supply-related
quality information. Perceived globalness may reflect a
high worldwide demand for superior quality and/or the
accumulated expertise that comes with producing larger
quantities for many markets. Indeed, global brands
often advertise their worldwide availability and accept –
ance (Alden, Steenkamp, and Batra 1999) as a signal to
convey information about their quality. Similarly, if con –
sumers in other parts of the world use the brand, con –
sumers may perceive the brand as signaling higher pres –
tige. In Holt, Quelch, and Taylor’s study (2004, p. 74),
a respondent directly articulated this association: “The
more people who buy (a) brand … the better quality it
is.” Steenkamp, Batra, and Alden (2003) find support
for a positive perceived brand globalness and prestige
association. Thus, as a brand attribute, brand globalness
functions as a signal of and holds positive associations
with quality and prestige.
As mentioned previously, brand globalness by itself,
independent of any effects associated with quality and
prestige, may enhance brand preference. This associa –
tion is called the “belongingness pathway” by
Steenkamp, Batra and Alden (2003); the “global myth”
effect by Holt, Quelch, and Taylor (2004); and the
“aspiration component” of a global brand by Dimofte,
Johansson, and Ronkainen (2008). The argument is that
global brands offer consumers the opportunity to
acquire and demonstrate participation in an aspired-to
global consumer culture (Alden, Steenkamp, and Batra
1999, 2006). Consumers use global brands to create an
imagined identity that they share with like-minded
people (Hannerz 1990; Holt, Quelch, and Taylor 2004).
Consumers’ ownership and use of global brands signal
information to other consumers of membership in a
worldwide consumer segment (Dawar and Parker
1994), with global brands functioning as a passport to
global citizenship (Strizhakova, Coulter, and Price
2008). In Holt, Quelch, and Taylor’s (2004, p. 71)
study, an Argentinean consumer stated, “Global brands
make us feel like citizens of the world, and … they
somehow give us an identity.” In the same study, a Costa
Rican expressed the global myth effect: “Local brands
show what we are; global brands show what we want to
be” (p. 71). Thus, perceived brand globalness alone sig –
nals additional information for consumers and may
function as a distinct brand attribute (e.g., belonging to a global consumer segment), beyond those captured by
quality and prestige.
This signaling practice is not limited only to global
brands. Many local brands with a foreign presence use
their availability in other markets in their communica –
tions with local consumers. Indeed, perceived globalness
as a signal of quality would be particularly useful for
brands in emerging countries (e.g., India, Turkey), in
which the quality of local brands may vary widely in a
given product category (Maxwell 2001). 1A strong local
consumer electronics brand in Turkey, Beko, has been
using this route successfully for almost a decade, with
the slogan “Beko—A World Brand.” The competitor,
Vestel, published full-page advertisements in leading
local newspapers highlighting exports to 103 countries
(Vestel 2006). The signal from a local brand about its
global availability, acceptance, and reach informs local
iconness, quality, and prestige perceptions of consumers
in the local market. In an effort to defend themselves
from global brands and to capture a unique balance,
local brands may also signal their local symbolism and
closer associations with local country culture to build
local iconness as a point of difference in their competi –
tion against global brands. That is, signals of global
presence may be combined with signals of local iconness
in brand management strategies for local brands, par –
ticularly in emerging markets.
Local Iconness
The homogenization argument of globalization’s conse –
quences proclaims the dominance of global marketing
strategies of multinationals and their global brands
(Alden, Steenkamp, and Batra 2006; Holton 2000).
However, others suggest that global interconnection and
interdependence do not necessarily mean conformity
and homogeneity (Ger 1999; Holton 2000; Pieterse
1995) and that local culture remains a central influence
on consumer behavior (Samli 1995). According to Ger
(1999), local brand managers can achieve competitive
success by using local cultural capital, heritage, and tar –
geting and positioning strategies that reflect a deeper
understanding of local identity, culture, tradition, tastes,
and needs. Forging a connection with the local country
and culture may be one way of building perceptions of
local iconness for local brands.
The relationship between perceived brand globalness
and local iconness captures the marriage of global suc –
cess and local (cultural, country) connections. The
inter penetration of these attributes and signals may be

76 Journal of International Marketing mutually reinforcing in emerging markets, but these sig –
nals may be in conflict in mature markets. Recall that
the main signals of a local brand are its uniqueness,
originality, and local cultural orientation; thus, when
consumers from mature markets learn that a local brand
is expanding abroad, they may fear loss of commitment
to home—and with it, not only a loss of purity and the
brand’s iconness, but also a sense of anger or betrayal
(i.e., “Who do they think they are?”) In contrast, in
emerging markets, seeing that a local brand is “doing
well abroad” will probably stimulate pride and respect
in homegrown talent. 2
In general, emerging market consumers’ aims are not to
differentiate themselves from other consumers but to
affirm their resemblances to the members of the
advanced world (Batra et al. 2000) and their presence in
world markets through the availability of their local
brands in other markets. In contrast, consumers from
mature markets may want their local icons to be original,
unique, and available only for them in their local mar –
kets. While emerging-market consumers may be striving
for more validation and similarity to those in the
advanced world, consumers from advanced markets may
be driven by a need for uniqueness (Brewer 1991). Thus,
in emerging countries, going away from home raises
iconness, whereas in mature markets it lowers iconness.
Furthermore, factual and spatiotemporal connection to
history or a particular location captures a type of
authenticity of marketing offerings (Grayson and Mar –
tinec 2004) relevant to local iconness. For mature mar –
ket consumers, when their local brands are perceived to
be global and sold overseas, they may lose this type of
authenticity (connection to a particular location). The
quest for iconness may be dampened when its degree of
originality and connection to a local origin is weakened
by perceptions of global presence. Therefore, mature
country consumers are expected not to approve the
incorporation of globalness perceptions into their local
icons. They want their local icons to be pure, not mixed
with brand elements that would be normally associated
with global brands:
H1: Perceived brand globalness of the local brand
is positively related to perceptions of local
iconness for consumers from emerging mar –
kets; the relationship is negative for mature
country consumers.
The Local Iconness–Prestige Relationship . Studies have
provided some evidence of a positive relationship between the perceived globalness of a brand and its per –
ceptions of prestige (Steenkamp, Batra, and Alden
2003). This effect is particularly strong in emerging
markets (e.g., Batra et al. 2000). In their efforts to
defend themselves against global brands, local brands
can also benefit from higher prestige perceptions. How –
ever, the sources for creating prestige would be different
for local brands. Inspired by Ger (1999) and Steenkamp,
Batra, and Alden (2003), local iconness is suggested as a
unique and differentiating source of prestige for local
brands in all types of markets. By tapping into local cul –
tural capital, symbolism, and heritage; a deeper under –
standing of the local culture, tastes, and needs; and
therefore locally responsive targeting and positioning,
local iconness can help local brands create and signal
prestige. Although recently, in response to enhanced
competition, many global brands have tried to commu –
nicate being insiders, such local anchoring and symbol –
ism are available to local brands to a greater extent and
with more believability and sincerity. The positive rela –
tionship of local iconness to prestige has been supported
by Steenkamp, Batra, and Alden (2003) in Korean and
U.S. samples for brands that vary on globalness percep –
tions. However, the authors do not investigate these
relationships in local–global brand pairs, as the current
study does. Thus, the following is proposed:
H2: Local iconness is positively related to con –
sumer perceptions of local brand prestige.
The Local Iconness–Quality Relationship. The relation –
ship of local iconness to quality perceptions may be more
nuanced. Local iconness captures a brand’s associations
with local culture that create brand value. On the one
hand, local icons are expected to have a better assess –
ment of what the consumer wants in terms of quality,
because “high quality” may be interpreted differently
across markets (Ger 1999). On the other hand, the sym –
bolism of uniqueness and authenticity may be contrary
to quality perceptions. Indeed, Steenkamp, Batra, and
Alden (2003) do not find a significant relationship
between local iconness and quality. Thus, a main effects
hypothesis is not developed. This relationship is expected
to be moderated by product category (food vs. nonfood);
this hypothesis is developed subsequently.
Local–Global Brand Relationships (Cross-
Effects)
According to the associative network memory model,
brands in a category, regardless of whether they are
global or local, are related to one another. Activating

The Interplay Between Global and Local Brands 77 associations related to the local (global) brand should
facilitate retrieval of associations related to the global
(local) brand in the same category (Alba and Hutchin –
son 1989; Meyers-Levy and Tybout 1989). This is
because the cognitive schemas of the local and global
brand also share an evaluative domain (Fiske and
Pavelchak 1986): They are satisfying the same need
(e.g., food). Therefore, they are substitutes for one
another, and their cognitive networks are related. Fur –
thermore, the literature on competitive positioning sug –
gests that perceptions of brands are affected by the per –
ceptions of their competitors in the market (Carpenter
and Nakamoto 1989), and the global brand and its local
competitor are compared with each other. Therefore,
quality and prestige perceptions of global and local
brand pairs in this study are expected to be correlated. 3
In the competition between global brands and their local
alternatives, other cross-effects occur that need to be
empirically analyzed. First, just as previous research has
found that consumer perceptions of the quality and
prestige of the global brand are positively associated
with global brand purchase likelihood (Holt, Quelch,
and Taylor 2004; Steenkamp, Batra, and Alden 2003),
the quality and prestige of the local alternative is
expected to demonstrate the opposite association on
purchase likelihood of the global brand. The perceived
quality of the local brand increases the purchase likeli –
hood of the local brand, which in turn decreases the
purchase likelihood of the global alternative. Similarly,
the prestige of the local brand is expected to enhance the
purchase likelihood of the local brand, thus, indirectly
dampening the purchase likelihood of the global brand.
Thus:
H3: Perceived quality of the local brand is nega –
tively related to the purchase likelihood of the
global brand.
H4: Perceived prestige of the local brand is nega –
tively related to the purchase likelihood of the
global brand.
Figure 1 presents the hypotheses and the chain of rela –
tionships that drive consumers’ likelihood of purchasing
the global brand in the presence of a local brand alter –
native. In line with previous research (e.g., Erdem and
Swait 1998; Steenkamp, Batra, and Alden 2003), con –
sumer’s likelihood of purchasing the brand is used as a
measure of brand value or utility. To capture the net
effect on global brand value or utility that the local
alternative exerts, consumer’s likelihood of purchasing the local brand is subtracted from the likelihood of pur –
chasing the same category global alternative, leading to
the focal measure, GBPL.
Moderating Role of Product Category and
Consumer Age
There is some recent evidence that the relationship
between local and global brands may be moderated by
the product category. In Zhou and Belk’s (2004) study,
global brands were associated with status, modernity,
cosmopolitan sophistication, and technology, whereas
local brands were associated with local cultural values.
Food has a foundational role in defining local culture
and tradition. Therefore, strong local symbolism and cul –
tural connections may be more desirable and easy to
establish for culture-bound categories such as food and
drinks. The global versus multidomestic industry distinc –
tion also supports this expectation (Porter 1986; Praha –
lad and Doz 1987; Yip 1995). A multidomestic industry
exists when customer needs are heterogeneous, making
marketing nontransferable across countries. The unique –
ness of local tastes in multidomestic industries gives local
brands an edge in learning, understanding, and adapting
to local consumers and making their marketing and
brand building more believable and authentic. Kobrin
(1991) identifies food and drinks as low on global inte –
gration, suggesting a multidomestic industry structure.
Özsomer, Bodur, and Cavusgil (1991) provide some
indirect empirical evidence in support of the multi –
domestic nature by showing the standardization of the
marketing programs to be lower in food and drink than
in nonfood categories. Furthermore, the market pene –
tration of global brands is lower in food and drinks in
Central and Eastern European markets (Schuh 2007).
Being representative of the country and the local culture
and based on the strong local cultural symbolism, local
iconness is expected to be associated with perceived
quality only in categories like food, for which the
uniqueness is more in line with local expectations,
tastes, habits, needs, and definitions of quality.
H5: Local iconness is positively related to per –
ceived brand quality of the local brand
(PBQuality L) in food categories; the relation –
ship is nonsignificant in nonfood categories.
The literature also suggests that teens and young adults
are affected by globalization to a greater extent than
their older counterparts (Walker 1996) and that global
brands have a greater appeal among teenagers (Hassan

78 Journal of International Marketing and Katsanis 1994). In the competition between global
and local brands, younger consumers may have a
greater preference for global brands because they
enhance their social acceptability and signal their mem –
bership in a global consumer segment (e.g., Strizhakova,
Coulter, and Price 2008). In contrast, older consumers
may value the uniqueness and authenticity that local
brands offer. Using this logic, this study examines the
moderating role of a consumer’s age. Specifically, the cross-effects of the local brand prestige (Prestige L) and
quality (Quality L) on GBPL are expected to be stronger
for older consumers. Whereas younger consumers may
downplay the brand equity of the local brand in their
evaluations of the global counterpart, older consumers
may put more weight on these associations.
H6: The negative relationship between Prestige L
and GBPL is stronger for older consumers. Figure 1. A Model of the Interplay Between Global and Local Brands
*Significant at ≥ 90% level. (All others significant at ≥ 95% level.)
Notes: The L and G subscripts refer to the brand equity associations of the local and global brands, respectively. n.s. = statistically not significant. H1
H2
H3 H4 .43
.23
.31 .35
n.s.
.41 .14
–.28
–.43
n.s.
n.s.
.17 .28
.34
.31
.16
.53
.20* .92
.62
n.s. .23
15
n.s.
.64
.59
.66 –.64
–.42
–.71 –.12
n.s.
n.s. PB
GlobalnessL Local
Iconness
PrestigeL PB
QualityL
PB
GlobalnessG PrestigeG
PB
QualityG
GBPL Path Parameter
Legend
1. Turkey
2. Singapore
3. Denmark .07*
.27
.13*
Global–Local
Difference Global Brand Local Brand

The Interplay Between Global and Local Brands 79 H7: The negative relationship between Quality L
and GBPL is stronger for older consumers.
SAMPLES, MEASURES, AND EMPIRICAL
RESULTS
The hypotheses are tested using consumers from Turkey,
Singapore, and Denmark, which represent different
positions relative to each other according to Hofstede’s
(1980) dimensions of individualism (37, 20, and 74,
respectively), uncertainty avoidance (85, 8, and 23,
respectively), power distance (66, 74, and 18, respec –
tively), and masculinity (45, 48, and 16, respectively).
As such, these countries represent a relatively high vari –
ability on major cultural dimensions and increase the
generalizability of the results.
In terms of sample selection, Sekaran (1983) identifies
two primary ways to achieve sample comparability:
drawing nationally representative samples or selecting
matched samples on the basis of some characteristics of
interest. Budgetary constraints prevented representative
sampling. Therefore, an attempt was made to match
samples by collecting data in shopping malls catering to
middle-income clientele in their respective countries.
Middle-income consumers are expected to be reasonably
familiar with global (and local) brands. With the help of
local researchers, four malls were identified in Istanbul,
Turkey; two in Singapore; and two in Aarhus, Denmark.
The resulting sample was comparable in terms of age dis –
tribution (92%, 83%, and 95% of respondents were in
the 18- to 45-year age range), gender distribution (48%,
48%, and 50% males), average daily television watching
hours (2.7, 2.8, and 2.1), and education (90%, 73%, and
97.5% had a high school education or more) for Turkey,
Singapore, and Denmark, respectively. 4
In Turkey, eight product categories, including durables and
nondurables, involving 16 brands were used. Categories
and brands were identified in three steps. Every year,
ACNielsen ranks brands in a nationwide “Brand Aware –
ness” study, in which respondents identify the most
known brands through unaided recall. The best-known
global (local) brands were identified and matched with
their best-known local (global) counterparts to represent
both types of brands in a pairwise comparison (ACNielsen
2006). In all product categories, the global and local
brands were in the top three in terms of nationwide aware –
ness in their respective categories. It should be noted that,
first, the globalness (localness) perceptions are relative and
represent various degrees of globalness (localness) rather than absolute categorizations. Second, because global
brands are generally associated with higher aspirational
value (e.g., Dimofte, Johansson, and Ronkainen 2008;
Holt, Quelch, and Taylor 2004), brands from both utili –
tarian and hedonic categories (implying a higher aspira –
tional value for hedonic brands) are included. Using
Ratchford’s (1987) measures, undergraduate students
rated the categories that emerged on involvement, utili –
tarian, and hedonic dimensions separately.
Third, given the focal antecedent status of the local brand
in the model, specifically that the perceived globalness of
the local brand is tested for its relationship to local icon –
ness, perceived quality, and prestige, it is important to
establish that the local brand is actually perceived as local
and not global by consumers. To this end, four focus
groups were conducted with the help of a professional
research company. After providing relevant definitions,
focus group members were asked to identify brands that
they perceived as global and local in any category. Local
brands that were identified as being high on globalness
were eliminated (Beko, Migros, and Ülker). Focus group
members were also asked the following question: “If a
brand were to carry the Turkish flag, which brand would
it be?” Arçelik, Cola Turca, Mavi Jeans, and Vestel were
among the brands identified as such.
Next, focus group results were integrated; to maximize
variance on involvement and utilitarian/hedonic dimen –
sions, the categories (brands) selected included cola drinks/
low involvement–hedonic (Coca-Cola, Cola Turca), facial
cream/high involvement–hedonic (Nivea, Arko), color
television sets/high involvement–utilitarian (Sony, Vestel),
refrigerators/high involvement–utilitarian (Bosch, Arçe –
lik), toothpaste/low involvement–utilitarian (Colgate,
Ipana), laundry detergent/low involvement–utilitarian
(Ariel, Alo), jeans/high involvement–hedonic (Levi’s, Mavi),
and yogurt/low involvement–utilitarian (Danone, Pınar).
Respondents rated each brand separately on the mani –
fest items. Each respondent evaluated two global–local
brand pairs (a total of four brands). On average, respon –
dents completed the survey in 25 minutes. Each respon –
dent received 5 Turkish Lira (approximately US$3.50)
for completing the survey. A total of 480 global–local
brand pair evaluations were collected.
In Singapore, three product categories involving six
brands were used. Categories (brands) were identified
with a procedure similar to that used in Turkey. However,
involvement and hedonic/utilitarian dimensions were
assessed using a focus group. The categories (brands)

80 Journal of International Marketing studied included beer/high involvement–hedonic (Carls –
berg, Tiger), consumer banking/high involvement–
utilitarian (HSBC, DBS Bank), supermarkets/high
involvement–utilitarian (Carrefour, NTUC Fairprice).
There were two service categories because Singapore, an
island of few natural resources, has few product brands,
and the country is better known for its services. Respon –
dents were paid S$2 for their participation in the study
(because the number 2 is regarded as lucky in Asian cul –
ture). The money was put in a red packet, a color consid –
ered lucky since the survey was conducted during the fes –
tive season, the Lunar New Year. A total of 202 global–
local brand pair evaluations were completed.
In Denmark, similar procedures to those used in Singa –
pore identified the categories (brands) as follows:
beer/high involvement–hedonic (Heineken, Ceres),
jeans/high involvement–hedonic (Levi’s, Jack Jones),
yogurt/low involvement–utilitarian (Danone, Arla),
sports clothing/high involvement–hedonic (Nike, Hum –
mel), and supermarket/high involvement–utilitarian
(Aldi, Netto). A pretest revealed that potential respon –
dents did not want a token in exchange of participating
in the survey. Instead, the academic nature of the study
was emphasized, and respondents were encouraged to
“help” by participating in the study. One hundred sixty-
four global–local brand pair evaluations were completed.
Categories and brands were rotated completely across
participants in all three markets. Categories differed
somewhat between countries because well-known
global and local brand pairs were not available in all
three countries. Measures were developed to accommo –
date such brand-level differences (a difference measure,
GBPL was used). The original English questionnaire was
translated into Turkish by the researchers. The resulting
Turkish version was back-translated into English by a
bilingual language instructor. Discrepancies were
resolved in face-to-face discussions. Furthermore, some
minor changes were made based on the feedback from
pretests with four respondents in each country. Pretests
in Singapore and Denmark revealed that respondents
were comfortable with the English questionnaire,
enabling the use of the original English version for the
Singaporean and Danish legs of the data collection.
Measures
A unique feature of this study is that all brand associa –
tions (e.g., perceived brand globalness, local iconness)
are measured separately for both the local and global
brands, enabling us to empirically partial out the effects of local brand equity associations on global brand
evaluations. In terms of measurement, previously used
measures and scales (Steenkamp, Batra, and Alden 2003)
were used (for all scales used, their sources and their psy –
chometric properties, see the Appendix). All items were
measured on seven-point scales. Perceived brand global –
ness was initially measured with four items capturing the
extent to which consumers believed the brand was mar –
keted and used in other countries. Prestige was measured
with a single item. Local iconness was measured with
three items that captured the extent to which consumers
thought the local brand was a symbol of the local coun –
try and represented what the country was all about.
Local iconness of the global brand was measured as a
validity check (for construct means and standard devia –
tions, see Table 1). Two items were used to capture brand
purchase likelihood. A measure of GBPL was created by
subtracting responses for local brand purchase likelihood
(BPL L) from corresponding responses for GBPL (BPL G)
items (GBPL = BPL G– BPL L). Thus, positive values of
GBPL imply a likelihood of purchasing the global brand,
while negative values imply a preference for the local
brand in the pairwise local–global comparisons. This
type of difference measure also takes into account brand-
level fixed effects, eliminating the need to have brand
dummies in the models.
Examination of Data Pooling
To decide whether separate models should be estimated
for each sample, the possibility of pooling data across
countries by means of multigroup EQS analyses was
investigated. Two nested models were investigated: (1) a
model in which all structural paths were set equal across
the country samples (equal model) and (2) a model in
which all structural paths were set free across the three
country samples (free model). A chi-square difference
test revealed that the free model had a significantly bet –
ter fit than the equal model. Therefore, the data were
not pooled across countries.
Overall Measurement Model Validation
The quality of the measurement model was assessed by
investigating unidimensionality, convergent validity,
reliability, and discriminant validity separately for
Turkey, Singapore, and Denmark. Through exploratory
factor analysis, evidence for the unidimensionality of
each construct was supported by appropriate items that
loaded at least .60 on their respective hypothesized com –
ponent and loaded no greater than .30 on other compo –
nents. In the confirmatory factor analysis itself, two

The Interplay Between Global and Local Brands 81 poorly loading items were dropped from both the local
and global measures of perceived brand globalness in all
three samples (Anderson and Gerbing 1988), resulting
in a two-item measurement. The purified complete
measurement model demonstrated a good fit for Turkey,
Singapore, and Denmark (Turkey: 2= 133.33, d.f. =
64, p= .000; comparative fit index [CFI] = .98; non –
normed fit index [NNFI] = .96; root mean square error
of approximation [RMSEA] = .05; and standardized root
mean square residual [SRMR] = .04; Singapore: 2=
102.56, d.f. = 64, p= .002; CFI = .97; NNFI = 94;
RMSEA = .05; and SRMR = .05; Denmark: 2= 124.10,
d.f. = 64, p= .00; CFI = .95; NNFI = .92; RMSEA = .07;
and SRMR = .04). The overall goodness of fit also sup –
ports unidimensionality (Steenkamp and Van Trijp
1991). Furthermore, the loadings of items on their
respective factors were all significant and positive, stan –
dardized factor loadings were all greater than .5 in all
three samples, and all t-test values were highly signifi –
cant, providing support for convergent validity.
For discriminant validity, as a basic test, it was first
determined whether correlations among the latent con –
structs were significantly different from one. In all
samples, construct correlations met this criterion (for
correlations, see Table 2), indicating that the con –
structs are fairly well discriminated. Then a series of
nested confirmatory factor models were compared in
which, for every pair of the constructs, a one- and two-
factor model were tested to determine which would fit
best (Anderson 1987; Bollen 1989). If the two-factor
model fits significantly better than the one-factor
model, discriminant validity of the two factors is sup -ported (Bagozzi, Yi, and Phillips 1991). Chi-square dif –
ferences were significant for all model comparisons
(p< .05) in all three samples. Third, Fornell and Lar –
cker’s (1981) test provides evidence for discriminant
validity between each possible pair of latent constructs
in all samples.
Cross-National Measurement Validation
To cross-nationally investigate the interrelationships
between constructs in a nomological net, configural and
metric invariance using multigroup confirmatory factor
analyses (Mullen 1995; Steenkamp and Baumgartner
1998) was also tested. This was necessary because the
scale intervals of the latent constructs must be compara –
ble across Turkey, Singapore, and Denmark. Configural
invariance of the eight-factor model was supported, as
the three-group confirmatory factor analysis model fit
was very good ( 2= 359.9, d.f. = 192, p= .00; CFI =
.97; NNFI =.95; and RMSEA = .03). Following the
nested sequential procedures that Bagozzi and Foxall
(1996) and Steenkamp and Baumgartner (1998) sug –
gest, metric invariance was assessed by comparing
nested complete measurement models in terms of the
difference in chi-square relative to degrees of freedom,
RMSEA, and CFI. In the first model (the base model),
all factor loadings, error variances, and all factor vari –
ances/covariances were allowed to be free across the
three samples. (One marker item was selected, and the
same marker item was used in each sample). In the sec –
ond model (the equal loading model), the factor load –
ings (apart from the marker item) were constrained to
be equal across the three samples. A model in which all Table 1. Means and Standard Deviations of Key Constructs
Turkey Singapore Denmark
LocalGlobalLocalGlobalLocalGlobal
Local iconness 4.67 3.19 4.92 3.18 4.47 1.70
(1.51) (1.51) (1.41) (1.42) (1.59) (1.06)
PBGlobalness 3.95 5.11 3.73 5.24 3.70 6.26
(1.71) (1.73) (1.72) (1.33) (1.63) (1.44)
PBQuality 5.16 5.82 4.94 5.00 4.92 5.24
(1.47) (1.21) (1.30) (1.22) (1.21) (1.26)
Prestige 4.54 5.06 4.28 4.44 3.91 4.65
(1.92) (2.00) (1.68) (1.71) (1.48) (1.73)
Notes: All ratings are based on seven-point scales.

82 Journal of International Marketing Table 2. Means, Standard Deviations, and Predicted Correlations for Model Constructs
A: Turkish and Singaporean Samplesa
Measures 123456789 10
M 3.95 5.16 3.32 4.54 5.11 5.82 5.06 .34 5.47 5.37
SD 1.71 1.47 1.51 1.92 1.73 1.21 2.00 2.10 1.19 1.30
1. PBGlobalness L 3.73 1.72 .48 .24 .37 –.11 .02 –.00 –.30 .00 .00
2. PBQuality L 4.94 1.30 .33 .25 .47 .04 .09 .12 –.57 .39 .10
3. Local iconness 3.08 1.41 –.44 .04 .38 .34 .21 .32 –.15 .37 .10
4. Prestige L 4.28 1.68 .01 .39 .27 .22 .07 .33 –.35 .42 .11
5. PBGlobalness G 5.24 1.33 .43 .13 .21 –.08 .30 .63 .09 .21 .41
6. PBQuality G 5.00 1.23 .24 .35 .11 .21 .50 .39 .41 .16 .43
7. Prestige G 4.44 1.71 –.01 –.06 .20 .15 .20 .17 .10 .24 .37
8. GBPL –.33 1.69 .18 –.21 –.09 –.08 .27 .36 .18 –.30 .36
9. Familiarity L 5.57 1.22 .00 .31 .41 .24 .36 .26 .11 –.16 .26
10. Familiarity G 5.23 1.14 .00 .16 .22 .13 .66 .36 .15 .27 .53
B: The Danish Sample (n = 164)
Measures 123456789 10
M
SD
1. PBGlobalness L 3.70 1.63
2. PBQuality L 4.92 1.21 .13
3. Local iconness 3.53 1.59 –.28 .25
4. Prestige L 3.91 1.48 .16 .43 .27
5. PBGlobalness G 6.26 1.14 –.10 .04 .01 .04
6. PBQuality G 5.24 1.26 .04 .21 .15 .18 .37
7. Prestige G 4.65 1.73 –.01 .16 .15 .08 .18 .59
8. GBPL –.38 2.15 –.04 –.41 –.14 –.16 .33 .43 .31
9. Familiarity L 5.58 1.14 .00 .32 .40 .16 –.05 –.02 –.04 –.45
10. Familiarity G 5.20 1.28 .00 –.05 –.06 –.02 .49 .45 .32 .46 –.14
aThe matrix’s upper triangle corresponds to the Turkish sample (n = 480), while the lower triangle corresponds to the Singaporean sample (n = 202).
factor loadings were constrained to be equal yielded a
significant decrease in model fit ( 2= 397.4, d.f. = 206,
p= .00; CFI = .96; NNFI =.94; and RMSEA = .04). Fol –
lowing Steenkamp and Baumgartner’s (1998) approach,
partial metric invariance was then tested by releasing 2
of the 14 constraints. The resulting insignificant chi-
square difference at p= .01 (  2= 24.57, d.f. = 12)
together with an insubstantial decrease in the CFI (.004)
and a .01 improvement in RMSEA suggest retaining this constrained model. In general, equality of factor load –
ings is supported in that only 2 of the 14 constraints on
the three-group model were released.
These findings support full invariance of the factor pat –
tern (configural invariance) and partial invariance of
the factor loadings (metric invariance) of the global and
local brand-level constructs in the model. The measure –
ment models for Turkey, Singapore, and Denmark are

The Interplay Between Global and Local Brands 83 clean, with evidence of unidimensionality, convergent
validity, discriminant validity, and configural and met –
ric invariance. Thus, the structural model was evalu –
ated next.
RESULTS
Structural Model Evaluation
To test the hypotheses, structural equation modeling
was used. Figure 1 presents the results for all three mar –
kets. To enhance connection with previous research,
unhypothesized path results are also presented in the
same figure. Table 3 reports the parameter estimates for
the hypothesized paths and goodness-of-fit indicators of
the structural equation system for each sample. Separate
measures of local and global brand familiarity were
added to each structural equation as covariates to con –
trol for possible confounds. The model fits the data very
well (Turkey: 2= 218.58, d.f. = 98, p= .00; CFI = .97;
NNFI = .95; RMSEA = .05; and SRMR = .06; Singa –
pore: 2= 155.27, d.f. = 98, p= .00; CFI = .95; NNFI =
.93; RMSEA = .05; and SRMR = .06; Denmark: 2=
185.19, d.f. = 98, p= .00; CFI = .94; NNFI = .91;
RMSEA = .07; and SRMR = .06). Approximately 60% of the variance of the focal endogenous construct,
GBPL, is accounted for by the model for Turkey and
Denmark, and 40% is accounted for in the sample from
Singapore (Bagozzi and Yi 1988). Thus, overall, the
model is a good representation of the data.
Local Iconness Hypotheses . H 1and H 2explicate the
associations among local iconness and related con –
structs. Consistent with H 1, perceived brand globalness
of the local brand is positively associated with local
iconness only in the emerging market, Turkey. In the
mature markets of Singapore and Denmark, this rela –
tionship is negative and significant. There is also uni –
form support for H 2, revealing that local iconness is
positively and significantly related to the prestige of the
local brand in both emerging and mature markets.
Cross-Brand-Level Hypotheses . Two interesting cross-
brand effects are revealed in this study: (1) the negative
relationship of local brand quality on GBPL (H 3) and
(2) the negative association between local brand pres –
tige and GBPL (H 4). H 3is uniformly supported in all
three samples, in that GBPL is negatively related to the
perceived quality of the leading local brand in the same
category. The mechanism behind this effect is as fol –
Table 3. Structural Parameter Estimates and Goodness-of-fit Indexes (Full Samples)
Turkey Singapore Denmark
Hypothesized Path Estimate T-Value Estimate T-Value Estimate T-Value
Local Iconness
H1: PBGlobalness LÆLocal iconness .14 1.98** –.28 3.07** –.43 2.89**
H2: Local iconness ÆPrestige L .28 3.93** .34 2.57** .31 3.48**
Cross-Brand Level
H3: PBQuality LÆGBPL –.64 –8.03** –.42 –3.66** –.71 –5.22**
H4: Prestige LÆGBPL –.12 –2.57* .00 .02 –.09 –1.03
Goodness-of-Fit Measures
2(98 d.f.) 218.58 155.27 185.19
CFI/NNFI/RMSEA/SRMR .97/.95/.05/.06 .95/.93/.05/.06 .94/.91/.07/.06
R2 PBQ L .31 .20 .18
R2PBQ G .22 .33 .24
R2GBPL .58 .38 .60
N 480 202 164
*p< .10.
** p< .05.

84 Journal of International Marketing lows: Perceived quality of local brand increases the pur –
chase likelihood of the local brand, which in turn
decreases GBPL. The results show that when consumers
have a high-quality local brand, their likelihood of pur –
chasing the global alternative decreases. This result
provides strong empirical evidence for the cross-
country validation of this cross-brand relationship,
which is noteworthy given that the countries studied
differ considerably on demographic, economic, and cul –
tural dimensions. H 4is supported only in Turkey: The
prestige of the local brand increases the purchase likeli –
hood of the local brand, which in turn dampens the
purchase likelihood of the global brand in an emerging
market setting.
To better understand the mechanisms through which the
effects of the local brand operate on GBPL, the indirect
effects of PBGlobalness Land local iconness on GBPL
are considered. The indirect effects of PBGlobalness Lare
uniformly significant in all three samples and the indi –
rect effect of local iconness is significant in Denmark. In
addition to the direct negative effect of PBQuality Lon
GBPL, the local brand exerts an indirect negative effect
on GBPL, triggered by its perceived globalness. (Signifi –
cant indirect effects for Turkey, Singapore, and Den –
mark at p< .10 are –.30, –.06, and –.18, respectively.)
Furthermore, local iconness matters in Denmark: It indi –
rectly dampens the purchase likelihood of the global
brand further (–.13, significant at p< .10).
Global Brand–Level Relationships. Although not the
main focus of this study, in general, the current find –
ings support the results of Steenkamp, Batra and
Alden (2003) across three different countries, thus
enhancing the generalizability of their results. Per –
ceived brand globalness is positively associated with
both perceived brand quality and prestige (barely
missing significance in Denmark). Moreover, no sup –
port is found in any of the three countries for a direct
relationship between PBGlobalness Gand GBPL, when
perceptions of global brand quality and prestige are
controlled for and when local brand equity associa –
tions are included. Thus, this path is omitted from the
final model (Figure 1). However, the uniformly posi –
tive and significant indirect effects of PBGlobalness G
on GBPL (at p< .10, Turkey, Singapore, and Denmark
are .10, .32, and .13, respectively) are in line with pre –
vious findings. Increased brand globalness increases
quality perceptions, which in turn increase purchase
likelihood. The premise that perceived brand global –
ness increases purchase likelihood still holds, albeit
along an indirect path. To enhance confidence in the validity of the postulated
model, this study turns to three model specifications. First,
rival structural models are tested. Second, effect sizes for
the significant paths are tested to determine whether they
are equal across the samples. Third, moderator tests on
some of the hypothesized relationships are performed.
Rival Models
Bagozzi and Yi (1988) recommend that researchers com –
pare rival models and not just test the performance of a
proposed model. Therefore, three pertinent rival models
were tested. The first specifies the reverse causal direction
between perceived brand globalness and perceived brand
quality in that perceived quality is considered an
antecedent (not outcome) of perceived globalness for
both the local and global brand sections of the structural
model studied here. It could be argued that a global (or
local) brand is considered to have higher levels of per –
ceived brand globalness and prestige if its perceived
quality is high. With respect to the overall fit, the rival
model has a higher chi-square and lower CFI and NNFI
than the hypothesized model in all three samples (Turkey:
2= 411.15, d.f. = 98, p= .00; CFI = .91; NNFI = .87;
RMSEA = .09; and SRMR = .10; Singapore: 2=
199.43, d.f. = 98, p= .00; CFI = .91; NNFI = .88;
RMSEA = .07; and SRMR = .10; Denmark: 2= 229.89,
d.f. = 98, p= .00; CFI = .90; NNFI = .86; RMSEA = .09;
and SRMR = .09). In addition, because the two models
have the same degrees of freedom, their chi-square sta –
tistic can be compared directly, revealing that the rival
model’s fit is 88%, 28%, and 4% weaker for Turkey,
Singapore, and Denmark, respectively. Because this
model is a worse fit in all three samples, it must be dis –
carded in favor of the original model.
In the second rival model, the causal ordering between
local iconness and prestige is reversed, in that prestige
is an antecedent (not outcome) of local iconness. As a
reviewer suggested, a local brand could be perceived
to have a higher level of local iconness when its pres –
tige is high. To reflect cross-brand associations of the
same brand equity attributes the Prestige LÆPrestige G
path was preserved in this rival model. While demon –
strating good fit, this model has a higher chi-square
and lower CFI and NNFI than the hypothesized model
in all three samples (e.g., Turkey: 2= 246.67, d.f. =
98, p= .00; CFI = .95; NNFI = .94; RMSEA = .06; and
SRMR = .08). Specifically, the increase in chi-square
( 2= 28.09), the decline in the CFI ( CFI = –.02), and
the increase in SRMR ( SRMR = .02) supported pre –
serving the original model for further analysis.

The Interplay Between Global and Local Brands 85 However, the causal ordering between local iconness
and prestige deserves further analysis, preferably with
longitudinal data.
The third rival model challenged the direct effects of
local brand equity constructs on GBPL. When first
introducing the model, theoretical rationale for expect –
ing direct effects of the perceived quality and prestige
of the local brand on the purchase likelihood of the
global brand (H 3and H 4) was provided. For purposes
of contrast, the competing model eliminates these two
cross-brand effects. With respect to overall fit, this sec –
ond rival model performed relatively poorly (Turkey:
2= 342.45, d.f. =100, p= .00; CFI = .93; NNFI = .90;
RMSEA = .07; and SRMR = .10; Singapore: 2=
170.47, d.f. =100, p= .00; CFI = .94; NNFI = .92;
RMSEA = .06; and SRMR = .07; Denmark: 2=
221.38, d.f. =100, p= .00; CFI = .91; NNFI = .88;
RMSEA = .09; and SRMR = .08). Because this model
is a nested version of the theoretical model, chi-square
difference tests indicated that for Turkey, Singapore,
and Denmark, these direct paths improved model fit
significantly and that the original model with these
cross-brand effects was superior.
To further refine the main results, the extent to which
the paths reported significant in all three samples were
equal in strength was evaluated. A multiple groups
approach was used to answer this question. Five paths
were reported to be significant in all three samples:
PBGlobalness LÆPBQuality L, Local Iconness ÆPres –
tige L, PBGlobalness GÆPBQuality G, and PBQualityL
(–) ÆGBPL, and PBQuality GÆGBPL. A correspon –
ding series of chi-square difference tests were per –
formed on these paths, one path at a time. A model in
which the path of interest was set equal across the
three country samples (equal model) was compared
with one in which the same path was set free across the
three samples (free model). The chi-square differences
range from 1.34 to 3.27 ( d.f. = 2). The insignificant
decline in model fit reveals that, indeed, paths that are
significant in Turkey, Singapore, and Denmark do not
differ statistically in strength across countries (e.g., the
negative effect of local brand quality on GBPL has the
same strength in Turkey, Singapore, and Denmark). As
such, the postulated model shows some consistency
across countries. Overall, the testing of these rival
models and effect sizes point to (1) the relative robust –
ness and validity of the postulated model, (2) the exis –
tence of cross-brand effects of the local brand on
GBPL, and (3) cross-country consistency of some por –
tions of the model. Moderating Effects of Product Category and
Consumer Age
Moderating effects are tested through multigroup analy –
ses, splitting the samples in each country into subsam –
ples on the two moderating variables of interest. Table 4
displays the results for separate structural model estima –
tions in terms of chi-square and degrees of freedom. 5
Regarding H 5, the formal moderation test revealed that
product category moderates the relationship of local
iconness to local brand quality in Turkey. (The relation –
ship is insignificant in the full sample.) The within-
country path coefficients are positive and significant in
food categories in Turkey and Denmark (food cate –
gories: Turkey: = .35, t = 2.33; Singapore: = .05, t =
.25; Denmark: = .25, t = 2.96) but consistently insig –
nifi cant for nonfood categories (nonfood categories:
Turkey: = –.07, t = –.37; Singapore: = –.01, t = –.07;
Denmark: = .14, t = 1.18). For Turkish and Danish
consumers, the data suggest that investing in local icon –
ness enhances quality perceptions in food categories.
To test the moderating influences of consumer age,
each country sample was divided into subsamples of
younger (18–25 years of age) and older consumers
(26–60 years of age). The results show that age signifi –
cantly moderates the relationship between local brand
prestige and GBPL in Turkey. For older consumers, the
prestige of the local brand has a stronger dampening
effect on the purchase likelihood of the same category
global brand. The within-country path coefficients are
consistently larger in the older than the younger sub –
samples in all three countries (for older consumers:
Turkey: = –.21, t = –3.43; Singapore: = –.13, t = –1.11;
Denmark:  = –.28, t= –1.90; for younger consumers:
Turkey: = –.02, t = –.35; Singapore: = .07, t = .95;
Denmark: = –.05, t = –.41 ). The data suggest that for
older Turkish and, to some extent, Danish consumers,
investing in local brand prestige generates a higher
payoff in terms of dampening the purchase likelihood
of the global brand. Age does not moderate the rela –
tionship between PBQuality L and GBPL (H 3), confirm –
ing the stability and robustness of this cross-brand
effect across countries and consumer segments.
DISCUSSION AND IMPLICATIONS
This study extends previous research by analyzing both
global and local brands simultaneously in an integrative
model. The design of this study and the findings from
the empirical analysis based on consumer data from an

86 Journal of International Marketing emerging market and two mature markets contribute to
the global branding literature in several ways.
First, the study includes local iconness (e.g., Steenkamp,
Batra, and Alden 2003) as a brand signal and a critical
brand attribute for local brands and investigate its rela –
tionship to brand globalness perceptions. It shows that
perceived brand globalness elevates brands to local icon –
ness status in emerging markets but not in mature mar –
kets (H 1). The findings suggest evidence of a hybridiza –
tion process (Pieterse 1995) of local brands in an
emerging market. The seemingly contradictory percep –
tions of globalness and local iconness are mutually rein –
forcing in consumer evaluations of local brands in emerg –
ing markets. This relationship capturing the marriage or
melange of global presence and strong local (cultural,
country) connections in emerging markets suggests that
the globalization process is not simple or unidirectional,
as the homogenization argument suggests (e.g., Levitt
1983). Instead, through a hybridization process, the local
brand drives strength from associations that would nor -mally be contrary to localness. Some elements of being
global are integrated into the elements of being local, cre –
ating the many shades of gray in the hybridized world of
successful local brands in emerging markets.
This study finds evidence that consumers in advanced
countries do not approve the marriage of the diverse
associations of globalness perceptions and local icon –
ness. The availability of their local icons in other mar –
kets as well leads these consumers to downgrade their
local iconness perceptions. The loss of spatiotemporal
authenticity (Grayson and Martinec 2004) or the higher
need for uniqueness (Brewer 1991) could explain the
dampening of these consumers’ iconness perceptions.
Second, this study finds that local iconness is a driver of
prestige for local brands in both emerging and advanced
markets (H 2). In line with Ger’s (1999) expectation, by
building strong local cultural connections and symbolism,
local iconness creates unique perceived value associated
with prestige perceptions in all markets studied. Third, this Table 4. Moderating Influences
Turkey Singapore Denmark
Moderator: Product Category (Food vs. Nonfood)
Equal Model
d.f. 208 208 208
² 381.77 299.53 356.87
H5: Local iconness ÆPBQuality L: free
d.f. 207 207 207
² 372.97 298.83 354.84
 ² 8.80* .70 2.03
Moderator: Consumer Age
Equal Model
d.f. 208 208 208
² 368.48 319.48 329.18
H6: Prestige LÆGBPL: free
d.f. 207 207 207
² 363.76 317.07 328.67
 ² 4.72* 2.41 0.51
H7: PBQuality LÆGBPL: free
d.f. 207 207 207
² 368.26 319.44 329.14
 ² .22 .04 .04
*p< .05.

The Interplay Between Global and Local Brands 87 study identifies the cross-brand effects of the local brand
on same-category, comparable global brand evaluations. It
shows that GBPL is negatively related to the quality per –
ceptions of its local counterpart. Thus, quality of the local
brand is the key association in dampening GBPL, and
local brand globalness enhances local brand quality per –
ceptions. These relationships hold even after controlling
for familiarity, prestige, quality, and perceived globalness
of the global brand and of its local counterpart. 6
Fourth, by integrating the associative network memory
model (Anderson 1983; Keller 1993, 2003) and signal –
ing theory (Erdem and Swait 1998, 2004), this study
develops a theoretical logic that brand globalness is a
firm signal and a critical brand attribute for both global
and local brands. The relevance of this conceptualiza –
tion is demonstrated by the empirical evidence that per –
ceived brand globalness is beneficial for both the global
and local brands: It is positively associated with quality
and prestige perceptions.
Implications for Local Brand Managers
Because local brands are increasingly faced with competi –
tion from global brands, this study has several implica –
tions for local brand managers to defend themselves
against global brands. First, in emerging markets, local
brands that are perceived as global enhance their local
iconness perceptions. 7Seeing that a local brand is “doing
well abroad” stimulates feelings of success and pride and
respect in homegrown talent. Global expansion is one
defense strategy local brands can use—defense by attach –
ing. 8Therefore, it is recommended that local brand man –
agers invest in building presence in foreign markets and
using this foreign availability and success (particularly in
advanced countries) in their communications with local
consumers. Indeed, the leading Turkish jeans brand, Mavi,
ran advertisements specifying that its brand is sold in the
United States (among many other countries), the home of
the original Levi’s jeans. In 2005, Asia Pacific Breweries,
the owner of Singapore’s Tiger beer brand, “unleashed the
new Tiger,” in which the new look exuded a stronger uni –
versal appeal. The words “World-Acclaimed Beer” were
prominently displayed on Tiger’s bottle and can labels
(APW 2005). In the new logo, the signature Tiger icon
strode forth powerfully, “symbolising that Tiger has
evolved from a local beer to a Singapore icon.”
Some other ways of building perceptions of globalness
include indicating conspicuously on the packaging selec –
tive foreign countries (favorable in the mental maps of tar –
geted consumers) where the products are available and providing usage instructions in different languages. Global
availability could also be the underpinning of print or tele –
vision commercials. International quality endorsement
marks, such as the ISO 9000 family of quality standards
and European Commission quality standards, are also
useful. Public relations departments can court the foreign
media by giving them access to their key executives for
opinions and comments about current issues.
However, calling attention to global expansion seems to
work only for emerging markets, not for mature mar –
kets. For example, in launching the new look in 2005,
Tiger beer managers tried to preserve the brand’s sense
of familiarity through its brewing tradition, Singa –
porean origin, and exotic Asian appeal (APW 2005).
NTUC Fairprice, the largest supermarket chain in Singa –
pore and one of the brands in this study, uses “Singa –
pore’s very own” as its brand slogan (Wikipedia 2012).
The positive relationship between local iconness and
local brand prestige, uniformly supported in all three
markets, deserves local brand managers’ attention. In
support of Steenkamp, Batra, and Alden’s (2003)
empirical results and embracing Ger’s (1999) emphasis
on local culture, local brands that invest in and commu –
nicate their brands as icons of local culture can enhance
their brand’s prestige perceptions. For example, one of
the Turkish brands in this study, Mavi jeans, launched a
communications campaign in 2007 (Mavi 2007) in
which jeans were wrapped like turbans on models’
heads and several local accessories were used.
Such local symbolism and cultural connections builds
local iconness, which enhances the prestige of the brand.
In turn, prestigious local brands increase purchase like –
lihood of the local brand and indirectly dampen GBPL
in an emerging market. Thus, the competition between
global and local brands remains a factor. Brands high in
local iconness, whether iconness is driven from per –
ceived globalness or not, enhance their prestige percep –
tions in both emerging and industrialized markets.
In food categories, there are further returns to local
iconness perceptions. In both Turkey and Denmark,
local icons are also perceived to be of higher quality.
Definitions and expectations of quality may be influ –
enced more by cultural, traditional, and habitual needs,
giving local icons an edge in food categories. Local icons
may be more tuned to such local definitions and expec –
tations of quality in food. For example, founded in
1856, Denmark’s Ceres Beer, one of the local brands in
this study, preserves its high alcohol content as a way to

88 Journal of International Marketing keep its connection to the Danish brewery tradition and
origin. (In Denmark, traditionally, beers were brewed
with special ingredients [some using lard], giving the
beer a unique taste and higher alcohol content [Glaser
2003.]) For managers of local food brands, developing
local iconness is a path worth investing in because the
returns are positive in terms of quality perceptions.
Although not formally hypothesized, observing the same
brand equity associations for the local and global brand
(e.g., PBQuality LÆPBQuality G), it is evident that percep –
tions of local brand quality tend to be associated with per –
ceptions of global brand quality at a higher level of inten –
sity in nonfood categories (nonfood, PBQuality LÆ
PBQuality G: Turkey: = .14, t = 3.04; Singapore: = .39,
t = 3.91; Denmark: = .12, t = 1.30), whereas in food cate –
gories the associations tend to be weaker (food, Turkey: 
= –.02, t = –.28; Singapore: = .16, t = 1.88; Denmark:
= .10, t = 1.01). According to the associative network
model, in food categories, the other brand may be less rele –
vant, and therefore its attributes may not be salient. That
is, because food is more culturally grounded, with local
tastes, traditions, and habits playing a greater role (Schuh
2007), consumers may consider the global brand alterna –
tive less relevant in satisfying social, cultural, and tradi –
tional needs (weaker interplay). In nonfood categories,
because the global brand is considered more relevant, its
attributes come to the forefront, creating stronger associa –
tions (stronger interplay). This is an intriguing observation
in need of further research and validation.
Finally, local brands that are perceived to be global are
also perceived to be of higher quality. The consumers may
be using the following inference: If the company is selling
this brand in other world markets, it must be committed
to this brand, and it cannot risk hurting its brand name
by producing low-quality products. Other consumers
around the world must be buying this brand because it is
of higher quality. Thus, though not explicitly tested, the
perceived globalness of a local brand may enhance its
credibility (Erdem and Swait 1998). The positive associa –
tion between the perceived globalness and quality of the
local brand is critical in that perceived brand globalness
increases perceived quality, which in turn increases local
brand purchase likelihood and thus reduces the purchase
likelihood of the global alternative. This study provides
the first empirical validation of this chain of effects.
In terms of the effects on a comparable global brand, in
emerging markets and especially for older consumers, the
prestige of their local brand dampens GBPL. This effect
exists also in Denmark among older consumers. Thus, local brand managers can invest in enhancing the prestige
perceptions of their brands, particularly in their targeting
and communication strategies with older consumers.
Implications for Global Brand Managers
As many multinational companies are pruning their
brand portfolios in favor of global brands and as they
are eliminating many “successful” local brands, it is
critical to understand the effects of same-category local
brands on consumer global brand preference. How
should global managers act if local brands build on
iconness or perceptions of globalness? In a recent study
on U.S. consumers, Dimofte, Johansson, and Ronkainen
(2008, p. 129) find that the self-reported indifference to
or outright dislike of global brands is not representative
of consumer’s true attitudes. Indirect ways of evaluating
affective responses to brand globalness show that
“everyone feels good about global brands and what they
convey to their users.” Therefore, global brand man –
agers should be cautious in building an “insider” status.
Some of the affective “aspirational” aura (Dimofte,
Johansson, and Ronkainen 2008), the happiness, and
the excitement associated with global brands may be
lost as they cultivate insider associations and become
“one of us.” Furthermore, the insider positioning may
not be authentic, credible, and believable coming from a
global brand. Local companies can “out-localize” a
global brand (Ger 1999).
Global firms can respond to attacks from independently
owned, nimble local brands by preserving local brands
in their brand portfolios. Thus, keeping a portfolio of
local brands, whether gained through past acquisitions
or built in-house in previous decades when the indus –
tries were multidomestic, gives multinational corpora –
tions the strategic flexibility to respond to nimble local
brand owners by building local iconness of their local
brands. Indeed, many multinational corporations keep
the acquired local brands in their portfolios and com –
municate deeper connections to the local country and
culture, in many cases preserving or building the local
iconness of these brands.
Considering the results of this study along with those of
previous findings on global branding (e.g., Dimofte,
Johansson, and Ronkainen 2008; Strizhakova, Coulter,
and Price 2008) and on marketing standardization (e.g.,
Özsomer and Simonin 2004; Zou and Cavusgil 2002),
global brands should emphasize their global reach, their
universal availability and recognition, and their uniform
and standardized but high-quality features, reducing the

The Interplay Between Global and Local Brands 89 perceived risk and in turn increasing the value and
utility for their target consumers. The strong positive
associations between global brand quality and purchase
likelihood and the lack of significant associations
between global brand prestige and purchase likelihood
in all three markets, taken together, reveal that investing
in creating and communicating quality rather than pres –
tige is a surer route to consumer preference for global
brands. Indeed, the strong effects of quality prevail even
after controlling for the quality and prestige effects of
the local brand. Thus, global brands should heavily
invest in perceptions of quality to enhance and protect
their competitive positions (Holt, Quelch, and Taylor
2004; Steenkamp; Batra and, Alden 2003).
Global brand managers might also nurture other attrib –
utes beyond quality and familiarity to create differential
advantage. Some of these points of difference might be
greater attention to social responsibility in the local
market (Dimofte, Johansson, and Ronkainen 2008;
Holt, Quelch, and Taylor 2004) and authenticity per –
ceptions (Grayson and Martinec 2004). However, the
focus groups conducted for this study during the brand
selection process revealed greater suspicion and cyni –
cism regarding the social responsibility activities of
global brands. Furthermore, in line with previous
research, the social responsibility expectations were also
greater from global brands. Global brands are subjected
to greater expectations and tougher metrics when evalu –
ated in terms of their social responsibility, making such
signals difficult to execute and manage.
While local brands can more easily and convincingly
build authenticity perceptions with unique associations
to their local identity and culture (Ger 1999), a handful
of global brands, such as Apple, have succeeded in
building and maintaining high authenticity perceptions
(for further details, see Belk and Tumbat 2005) even
though they are available throughout most of the world.
While social responsibility and authenticity of global
brands are fruitful areas for further research, these fac –
tors are beyond the scope of this study.
LIMITATIONS AND FURTHER RESEARCH
There are many avenues of further research to pursue
and limitations of this study to address. First, both per –
ceived brand globalness and local iconness need to be
conceptualized and measured more extensively. These
constructs are likely to be multidimensional. Further –
more, the relationship between perceived brand global -ness and local iconness deserves further scrutiny. For
example, how would the relationship be different if
globalness perceptions were driven by success in other
advanced markets versus success in emerging markets?
Would the relationship be different for emerging and
advanced market consumers? The relationship might be
expected to be positive for advanced-country consumers
if globalness is driven from success in other advanced
markets. Second, consumer need for uniqueness (Brewer
1991) and authenticity (Grayson and Martinec 2004)
can be explicitly measured and included to provide a
better understanding of why the relationship between
perceived globalness and iconness is negative in
advanced countries.
Third, the design of this study necessitated identification
of well-known global and local brand pairs. As a result,
categories differed somewhat between countries because
the same global brands were either not well-known or
well-known local counterparts were not available in all
three countries. Using a reviewer’s suggestion to control
for product category differences (in addition to brand-
level differences, controlled for with the difference
measure, GBPL), the hypothesized model was run,
dividing GBPL by the purchase likelihood of the local
brand ([BPL G– BPL L]/BPL L). This cleaner measure cap –
tured the percentage shift of the purchase likelihood of
the global brand relative to its local counterpart and
took care of product category–level differences. The
substantial results did not change significantly. Thus, the
original analysis is reported (thanks to a reviewer for
providing this suggestion).
Fourth, although in most cases the local and global
brands are top brands in their categories, products
might still differ in pricing, which might account for dif –
ferences in perceived quality, prestige, and purchase like –
lihood. For a subset of the brands, there is global–local
price information. Their price levels are very compara –
ble, and T-tests showed no significant differences. More –
over, the main model was run on these brands control –
ling for price, and the substantial results did not change
significantly—significant paths were still significant.
However, the lack of price information for all brand pairs
is a limitation of the current study in need of attention in
further research. A related point is differences in the exe –
cution of brand communication. This research did not
control for communication execution variability, which
might influence perceptions of brand equity elements.
Fifth, by design, this research did not include any luxury
brands but rather positioned the local as an alternative

90 Journal of International Marketing to the “normal” global brand (Ger 1999). The relation –
ship between globalness and local iconness as well as the
interplay between the global and local brands might be
different when luxury brands are involved. Similarly, in
times of economic distress, consumers may be expected
to return to local concerns with renewed interest in local
brands (Quelch and Jocz 2009). Testing the role of local
iconness and globalness perceptions in times of eco –
nomic crises could yield fruitful insights for global and
local brand managers.
Sixth, previous research identified ethnocentrism as a
moderator of the relationship between perceived brand
globalness and purchase likelihood (Steenkamp, Batra,
and Alden 2003) and as a mediator of the relationship
between global consumption orientation and global
brand attitudes (Alden, Steenkamp, and Batra 2006).
Other consumer characteristics such as susceptibility
to normative influence (Batra et al. 2000; Alden, Steen –
kamp, and Batra 2006), materialism (Ahuvia and
Wong 2002), and cosmopolitanism (Cannon and
Yaprak 2002), might moderate the interplay between
the global brand and its local counterpart. Specifically,
ethnocentrism and susceptibility to normative influ –
ence might moderate the local iconness–Prestige Lrela –
tionship. Another area for further research would be to
integrate Hofstede’s (1980) dimensions of national cul –
ture into the model by measuring them at the con –
sumer level (in line with Erdem, Swait, and Valenzuela
2006) to investigate culture’s influence on global
(local) brand purchase likelihood. Because the model
presented here is already complex, such possibilities
were not investigated.
Seventh, this study compares a relatively large emerging
market with two small, mature markets. While Denmark
has strong local brands in many categories, the availabil –
ity of local brands may be more limited in Singapore.
This might be one reason some of the paths were not sig –
nificant in Singapore (e.g., local iconness ÆPBQuality L
in food categories). Future studies should include larger
mature markets (e.g., Japan, the United Kingdom, South
Korea) for a better test of model generalizability.
This study focuses on the interplay between local and
global brand pairs. Asymmetrical relationships are iden –
tified between perceived brand globalness and local
iconness in emerging and advanced markets. Evidence is
found of a hybridization process in emerging markets, in
which the global and local elements of a local brand’s
equity are mutually reinforcing. With this research, I
hope to attract greater interest in the interplay between global and local brands. Indeed, the competition
between same-category comparable local and global
brands will be a fruitful and managerially relevant area
of research for many years to come.
NOTES
1. Quality variation depends on the ability to implement
reliable, low-defect manufacturing. Not all local
manufacturers in emerging markets have access to
such manufacturing sophistication, research and
development, and automation levels. Thus, quality
variation is expected to be higher for local brands
reflected in perceptual measures as well.
2. Thanks to a reviewer for this insight.
3. The cross-brand correlations among quality and pres –
tige control for category main effects and enable a
closer investigation of how category differences (food
vs. nonfood) strengthen or weaken these associations.
4. Assessing comparability through matched samples
was critical, particularly in Turkey, given lower per
capita gross domestic product rates at purchasing
power parity ($7,900 in Turkey vs. $29,900 in Sin –
gapore and $33,400 in Denmark; CIA Factbook
2006).
5. Considering product category a moderator, in the
equal models, all paths of the structural model across
food and nonfood subsamples were set equal. In the
free models, all equality constraints remained in place,
except the path that was potentially affected by the
moderator variable. A significant decrease in chi-
square from the equal model to a model in which one
relationship is free implies that the moderator variable
has a significant influence on that relationship.
6. In all markets, brand familiarity plays a critical role.
Brand familiarity with the global brand increases the
likelihood of purchasing the global brand, whereas
brand familiarity with the local brand has the oppo –
site effect on the purchase likelihood of the global
brand.
7. This relationship is stable across food and drinks and
nonfood categories according to chi-square difference
tests.
8. Thanks to a reviewer for providing these insights.

The Interplay Between Global and Local Brands 91 Appendix. Measures, Composite Reliabilities, and Average Variance Extracted
Constructs Item Source
Perceived Brand Globalness
PBG L: CR T= .62, AVE T= .51, CR S= .57, AVE S= .42, CR D= .72, AVE D= .65
PBG G: CR T= .72, AVE T= .62, CR S= .60, AVE S= .47, CR D= .71, AVE D= .61
PBG Land PBG GaTo me this is a global brand/To me this is a local brand. Steenkamp, Batra, and Alden (2003)
I don’t think consumers overseas buy this brand/I do think Same as above
consumers overseas buy this brand.
This brand is sold only in (Turkey, Singapore, Denmark)/This Same as above
brand is sold all over the world. b
Local Iconness
CR T= .83, AVE T= .64, CR S= .87, AVE S= .68, CR D= .89, AVE D= .72
I associate this brand with things that are (Turkish, Singaporean, Steenkamp, Batra, and Alden (2003)
Danish)/I do not associate this brand with things that are (Turkish,
Singaporean, Danish).
To me, this brand represents what (Turkey, Singapore, Denmark) Same as above
is all about/to me this brand does not represent what (Turkey,
Singapore, Denmark) is all about.
To me, this brand is not a very good symbol of (Turkey, Singapore, Same as above
Denmark)/to me, this brand is a very good symbol of (Turkey,
Singapore, Denmark).
Perceived Brand Prestige
Prestige Land This is a very prestigious brand/This is not a very prestigious Batra et al. (2000)
Prestige Ga brand.
Perceived Brand Quality
PBQ L: CR T= .95, AVE T= .91, CR S= .85, AVE S= .74, CR D= .91, AVE D= .84
PBQ G: CR T= .89, AVE T= .81, CR S= .86, AVE S= .76, CR D= .90, AVE D= .82
PBQ Land PBQ GaThis brand is very low on overall quality/This brand is very high Steenkamp, Batra, and Alden (2003)
on overall quality.
This is a brand of inferior quality/This is a brand of superior Same as above
quality To me this brand is a very good symbol of (Turkey,
Singapore, Denmark).
Brand Familiarity c
Familiarity Land This brand is very familiar to me/This brand is very unfamiliar Same as above
Familiarity Gato me.
Everybody here has heard of this brand/Almost nobody here has Same as above
heard of this brand.
I’m not at all knowledgeable about this brand/I’m very knowledge- Same as above
able about this brand.
I have never seen ads for it in (Turkish, Singaporean, Danish) Same as above
magazines, radio, or TV/I have seen many ads for it in (Turkish,
Singaporean, Danish) magazines, radio, or TV.

92 Journal of International Marketing Global Brand Purchase Likelihood d
GBP L: CR T = .92; AVE T = .86; CR S = .82; AVE S = .72; CR D = .94; AVE D = .88
BPL Land BPL GaI would not buy it (assuming it was available)/I would certainly Same as above
buy it (assuming it was available).
I’m not at all likely to buy it (if available)/I’m very likely to buy it
(if available).
Consumer Age
Please check the box for each that represents your age: Newly developed
___ Below 18
___ 18–25
___ 26–35
___ 36–45
___ 46–55
___ 56 and above
Category
Food and drinks versus other Coded by researcher
aThese questions were answered separately for the local and the global brand by the respondent, generating the separate constructs in the model in Figure 1. All
measures are seven-point, unless otherwise specified.
bThese items were deleted in the final analysis.
cAn index of familiarity was developed separately for the local and global brand by averaging responses to the four familiarity items. The corresponding index was
used in structural model equations.
dResponses for local brands on the two items were subtracted from the corresponding global brand items to create the two GBPL measures. Reported CR and AVEs
are for these difference measures.
Notes: CR = composite reliability, AVE = average variance extracted. T = Turkey, S = Singapore, and D = Denmark. Appendix. Continued
Constructs Item Source
REFERENCES
Aaker, David A. (1991), Managing Brand Equity . New York:
The Free Press.
Ahuvia, A.C. and Nancy Wong (2002), “Personality and Values
Based Materialism: Their Relationships and Origins,” Journal
of Consumer Psychology , 12 (4), 389–402.
Alba, Joseph W. and J.W. Hutchinson (1989), “Dimensions of
Consumer Expertise,” Journal of Consumer Research , 13 (4),
411–54.
Alden, Dana L., Jan-Benedict E.M. Steenkamp, and Rajeev
Batra (1999), “Brand Positioning Through Advertising in
Asia, North America, and Europe: The Role of Global Con –
sumer Culture,” Journal of Marketing , 63 (January),
75–87. ———, ———, and ——— (2006), “Consumer Attitudes
Toward Marketplace Globalization: Structure, Antecedents,
and Consequences,” International Journal of Research in
Marketing , 23 (3), 227–39.
Anderson, J.R. (1983), The Architecture of Cognition . Cam –
bridge, MA: Harvard University Press.
Anderson, James C. (1987), “An Approach for Confirmatory
Measurement and Structural Equation Modelling of Organi –
zational Properties,” Management Science , 33 (4), 525–41.
——— and D.W. Gerbing (1988), “Structural Equations Mod –
elling in Practice: A Review and Recommended Two-Step
Approach,” Psychological Bulletin , 103 (3), 411–23.
ACNielsen Brand Research (2006), “Türkiye’nin En Bilinen
Markaları Raporu, (The Most Known Brands in Turkey

The Interplay Between Global and Local Brands 93 Report),” (accessed June 20, 2007), [available at www.
acnielsen.com/products/reports/].
APW (2005), “Tiger Beer Unveils a Fresh New Look,” (accessed
January 3, 2011), [available at http://www.apb.com.sg/tiger-
beer-unveils-fresh-new-look-.html].
Bagozzi, Richard P. and G.R. Foxall (1996), “Construct Valida –
tion of a Measure of Adaptive-Innovative Cognitive Styles in
Consumption,” International Journal of Research in Market –
ing , 13 (3), 201–213.
——— and Youjae Yi (1988), “On the Evaluation of Structural
Equation Models,” Journal of the Academy of Marketing Sci –
ence , 16 (1), 77–94.
———, ———, and Lynn W. Phillips (1991), “Assessing Con –
struct Validity in Organizational Research,” Administrative
Science Quarterly , 36 (3), 421–59.
Batra, R., V. Ramaswamy, Dana L. Alden, Jan-Benedict E.M.
Steenkamp, and S. Ramachander (2000), “Effects of Brand
Local/Nonlocal Origin on Consumer Attitudes in Develop –
ing Countries,” Journal of Consumer Psychology , 9 (2),
83–95.
Belk, W. Russ and G. Tumbat (2005), “Cult of Macintosh,”
Consumption, Markets and Culture , 8 (3), 205–217.
Bollen, Kenneth A. (1989), Structural Equations with Latent
Variables . New York: John Wiley & Sons.
Brewer, Marilynn B. (1988), “A Dual Process Model of Impres –
sion Formation,” in Advances in Social Cognition , Vol. 1, T.K.
Srull and R.S. Wyer Jr., eds. Hillsdale, NJ: Lawrence Erlbaum
Associates, 1–36.
——— (1991), “The Social Self: On Being the Same and Differ –
ent at the Same Time.” Personality and Social Psychology Bul –
letin , 17 (October), 475–82.
Burgess, S.M. and Jan-Benedict E.M. Steenkamp (2006), “Mar –
keting Renaissance: How Research in Emerging Markets
Advances Marketing Science and Practice,” International
Journal of Research in Marketing , 23 (4), 337–56.
Cannon, Hugh M. and Atilla Yaprak (2002), “ Will the Real-
World Citizen Please Stand Up! The Many Faces of Cosmo –
politan Consumer Behavior,” Journal of International Mar –
keting , 10 (4), 30–52.
Carpenter, Gregory and Kent Nakamoto (1989), “Consumer
Preference Formation and Pioneering Advantage,” Journal of
Marketing Research , 26 (August), 285–98.
CIA Factbook (2006), “Country Comparison,” (accessed Feb –
ruary 7, 2012), [available at http://www.cia.gov/cia/library/
publications/the-world-factbook/index.html]. Dawar, Niraj and Philip Parker (1994), “Marketing Universals:
Consumers’ Use of Brand Name, Price, Physical Appearance
and Retailer Reputation as Signals of Product Quality,” Jour –
nal of Marketing , 58 (April), 81–96.
Dimofte, Claudio V., Johny K. Johansson, and Ilkka A.
Ronkainen (2008), “Cognitive and Affective Reactions of U.S.
Consumers to Global Brands,” Journal of International Mar –
keting , 16 (4), 113–35.
Erdem, Tülin and Jeffrey Swait (1998), “Brand Equity as a Sig –
naling Phenomenon,” Journal of Consumer Psychology , 7
(April), 131–57.
——— and ——— (2004), “Brand Credibility, Brand Consider –
ation, and Choice,” Journal of Consumer Research , 31 (June),
191–98.
———, ———, and A. Valenzuela (2006), “Brands as Signals:
A Cross-Country Validation Study,” Journal of Marketing , 70
(January), 34–49.
Fiske, S.T. and M.A. Pavelchak (1986), “Category-Based Ver –
sus Piecemeal-Based Affective Responses: Developments in
Schema-Triggered Affect,” in Handbook of Motivation and
Cognition: Foundations of Social Behavior , R.M. Sor –
rentino and E.T. Higgins, eds. New York: Guilford Press,
167–203.
Fornell, Claus and David F. Larcker (1981), “Evaluating Struc –
tural Equation Models with Unobservable Variables and
Measurement Error,” Journal of Marketing Research , 18
(February), 39–50.
Ger, Güliz (1999), “Localizing in the Global Village: Local
Firms Competing in Global Markets,” California Manage –
ment Review , 41 (4), 64–83.
———, ———, and Dana-Nicoleta Lascu (1993), “The Devel –
opment of Consumer Desire in Marketizing and Developing
Economies: The Cases of Romania and Turkey,” in Advances
in Consumer Research , Vol. 20, Leigh McAlister and Michael
L. Rothschild, eds. Provo, UT: Association for Consumer
Research, 102–107.
Glaser, Gregg (2003), “Denmark’s Other Breweries: To Survive
a Consolidating Market, Denmark’s Regional Breweries
Merged into the ‘Bruggerigruppen,’” Modern Brewery Age ,
(March 31), (accessed on January 3, 2012), [available at
http://findarticles.com/p/articles/mi_m3469/is_13_54/ai_1000
74938].
Grayson, Kent and R. Martinec (2004), “Consumer Perceptions
of Iconicity and Indexicality and Their Influence on Assess –
ments of Authentic Market Offerings,” Journal of Consumer
Research , 31 (September), 296–312.
Hannerz, Ulf (1990), “Cosmopolitans and Locals in World Cul –
ture,” Theory, Culture and Society , 7 (2), 237–51.

94 Journal of International Marketing Hassan, Salah S. and L.P. Katsanis (1994), “Global Market Seg –
mentation Strategies and Trends,” in Globalization of Con –
sumer Markets: Structures and Strategies , Salah S. Hassan and
Erdener Kaynak, eds. New York: International Business Press,
47–62.
Hofstede, Geert (1980), Culture’s Consequences: International
Differences in Work-Related Values . Beverly Hills, CA: Sage
Publications.
Holt, Douglas B. (2003) “What Becomes an Icon Most?” Har –
vard Business Review , 81 (March), 43–49.
——— (2004), How Brands Become Icons: The Principles of
Cultural Branding . Boston: Harvard Business School Press.
———, Jack A. Quelch, and Earl L. Taylor (2004), “How
Global Brands Compete,” Harvard Business Review , 82 (Sep –
tember), 68–75.
Holton, Robert (2000), “Globalization’s Cultural Conse –
quences,” Annals of the American Academy of Political and
Social Science ,570 (July), 140–52.
Johansson, Johny K. and Ilkka A. Ronkainen (2005), “The Esteem
of Global Brands,” Brand Management , 12 (5), 339–54.
Kapferer, Jean-Noél (2005), “The Post-Global Brand,” Brand
Management , 12 (5), 319–24.
Keller, Kevin L. (1993), “Conceptualizing, Measuring, and
Managing Customer-Based Brand Equity,” Journal of Market –
ing , 57 (January), 1–22.
——— (2003), “Brand Synthesis: The Multi-Dimensionalities
of Brand Knowledge,” Journal of Consumer Research , 29 (4),
595–600.
Kobrin, S.J. (1991), “An Empirical Analysis of the Determi –
nants of Global Integration,” Strategic Management Journal ,
12 (S1), 17–31.
Levitt, Theodore (1983), “Globalization of Markets,” Harvard
Business Review , 61 (May/June), 92–102.
Mavi (2007), (accessed on June 20, 2007), [available at www.
mavi.com/advertising.asp].
Maxwell, S. (2001), “An Expanded Price/Brand Effect Model:
A Demonstration of Heterogeneity in Global Consumption,”
International Marketing Review , 18 (3), 325–44.
Meyers-Levy, J. and Alice M. Tybout (1989), “Schema Con –
gruity as a Basis for Product Evaluation,” Journal of Con –
sumer Research , 16 (1), 39–54.
Mullen, Michael R. (1995), “Diagnosing Measurement Equiva –
lence in Cross-National Research,” Journal of International
Business Studies , 26 (3), 573–96. Özsomer, Ay șegül and Selin Altaras (2008), “Global Brand Pur –
chase Likelihood: A Critical Synthesis and an Integrated Con –
ceptual Framework,” Journal of International Marketing , 16
(4), 1–28.
———, Muzaffer Bodur, and Tamer Cavusgil (1991), “Mar –
keting Standardization by Multinationals in an Emerging
Market,” European Journal of Marketing , 25 (12),
50–68.
——— and Bernard Simonin (2004), “Marketing Program
Standardization: A Cross Country Exploration,” Inter –
national Journal of Research in Marketing ,21 (4), 397–419.
Pieterse, Jan N. (1995), “Globalization as Hybridization,” in
Global Modernities , M. Featherstone, Scott Lash, and Roland
Robertson, eds. London: Sage Publications, 45–68.
Pitcher, G. (1999), “Consumer Brands Cull Makes Room to
Grow,” Marketing Week , 22 (40), 23.
Porter, Michael E. (1986), “Competition in Global Industries: A
Conceptual Framework,” in Competition in Global Indus –
tries , Chap. 1, Michael E. Porter, ed. Boston: Harvard Busi –
ness School Press, 15–60.
Prahalad, C.K. and Yves L. Doz (1987), The Multinational Mis –
sion . New York: The Free Press.
Quelch, John A. and Katherine E. Jocz (2009), “How to Mar –
ket in a Downturn,” Harvard Business Review , 87 (4), 52–62.
Ratchford, Brian T. (1987), “New Insights About the FCB
Grid,” Journal of Advertising Research , 27 (4), 24–38.
Samli, Coskun A. (1995), International Consumer Behavior: Its
Impact on Marketing Strategy Development . Westport, CT:
Quorum Books.
Schuh, A. (2007), “Brand Strategies of Western MNCs as Driv –
ers of Globalization in Central and Eastern Europe,” Euro –
pean Journal of Marketing , 41 (3/4), 274–91.
Sekaran, Uma (1983), “Methodological and Theoretical Issues
and Advancements in Cross-Cultural Research,” Journal of
International Business Studies , 14 (2), 61–73.
Spence, M. (1974), Market Signaling: Information Transfer in
Hiring and Related Screening Processes . Cambridge, MA:
Harvard University Press.
Steenkamp, Jan-Benedict E.M., Rajeev Batra, and Dana L.
Alden (2003), “How Perceived Brand Globalness Creates
Brand Value,” Journal of International Business Studies , 34
(1), 53–65.
——— and Hans Baumgartner (1998), “Assessing Measure –
ment Invariance in Cross-National Consumer Research,”
Journal of Consumer Research , 25 (June), 78–90.

The Interplay Between Global and Local Brands 95 ——— and M.J. De Jong (2010), “A Global Investigation into
the Constellation of Consumer Attitudes Toward Global and
Local Products,” Journal of Marketing , 74 (November),
18–40.
——— and H.C.M. van Trijp (1991), “The Use of LISREL in
Validating Marketing Constructs,” International Journal of
Research in Marketing , 8 (4), 283–300.
Strizhakova, Y., R.A. Coulter, and Linda L. Price (2008),
“Branded Products as a Passport to Global Citizenship: Per –
spectives from Developed and Developing Countries,” Journal
of International Marketing ,16 (4), 57–85.
Torelli, Carlos J., H. Tat Keh, and C.Y. Chiu (2008), “Cultural
Symbolism of Brands,” in Contemporary Branding Issues: A
Research Perspective , B. Loken, R. Ahluwalia, and M.J.
Houston, eds. London: Psychology Press.
Unilever (2007), “Unilever Investor Seminar,” (accessed
January 3, 2012), [available at http://www.unilever.
com/images/ir_1.2_growth_strategy_rivers_speech_tcm13-
86705.pdf]. Vestel (2006), “Vestel’s Distribution Network,” (accessed Feb –
ruary 7, 2012), [available at http://www.zorlu.com.tr/en/
GRUP/ves_dagitim.asp].
Walker, C. (1996), “Can TV Save the Planet?” American Demo –
graphics , 18 (May), 42–49.
Wikipedia (2012), “NTUC Fairprice,” (accessed January 3,
2012), [available at http://en.wikipedia.org/wiki/NTUC_Fair –
Price].
Yip, George S. (1995), Total Global Strategy: Managing for
Worldwide Competitive Advantage . Upper Saddle River, NJ:
Prentice Hall.
Zhou, Nan and Russell Belk (2004), “ Chinese Consumer Read –
ings of Global and Local Advertising Appeals,” Journal of
Advertising , 33 (4), 63–76.
Zou, Shaoming and Tamer S. Cavusgil (2002), “The GMS: A
Broad Conceptualization of Global Marketing Strategy and
Its Effect on Firm Performance,” Journal of Marketing , 66
(October), 40–57.

Copyright of Journal of International Marketing is the property of American Marketing Association and its
content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's
express written permission. However, users may print, download, or email articles for individual use.