Global-local consumer identities as drivers of global digital brand usage

Purpose – The purpose of this paper is to represent the first empirical attempt to explore global-local consumer identities as drivers of global digital brand usage. Specifically, this study considers a unique category of digital products, social networking sites (SNS), and develops a set of hypotheses to assess the mechanism through which location-based identities influence the actual usage of global SNS (Facebook and Instagram). Moreover, cross-country variations are investigated under the lens of developed vs developing countries. Design/methodology/approach – Cross-country surveys in a developed (Austria) and a developing country (Thailand) were conducted. Data collected from 425 young adults were analyzed using SEM techniques in order to test a set of hypotheses. Findings – Results show that in Thailand, users with a global identity enjoy participating in global SNS more than their counterparts in Austria. In addition, consumers with a local identity in Thailand demonstrate less pleasure when participating in global SNS than their counterparts in Austria, and consequently are less inclined to use global SNS. Practical implications – Findings provide digital marketers with useful insights into important strategic decisions regarding the selection and potential adaptation of global digital brands according to the country context. Originality/value – This research is the first to extend the location-based identity research in the context of global digital brands, explain how global-local identities predict SNS usage through an engagement mechanism and investigate cross-country variations of this mechanism.


Introduction
Research on conceptualizing consumers' disposition toward global brands is burgeoning, with a number of researchers focusing on the effect of global-local consumer identities (e.g.Zhang and Khare, 2009).A global identity implies that a person identifies with all of humankind and consequently feels like a citizen of the world (Cannon and Yaprak, 2002).Consumers with a global identity have been found to have a more positive view toward globalization, to perceive people around the globe as similar to one another, and to show more interest in global events (Guo, 2013).A local identity derives from the overall awareness of belonging to a community that shares similar national values and cultural norms (Thompson, 2001).In an increasingly globalized marketplace, global and local identities co-exist, with consumers identifying simultaneously with both the global and their local community, and adopting whichever identity is more salient at a given time, based on contextual factors (Erez et al., 2013;Steenkamp and de Jong, 2010;Zhang and Khare, 2009).Past studies suggest that a positive disposition toward global brands is boosted by a global identity, since consumers with a more accessible global identity are more likely to hold positive attitudes toward global brands (Alden et al., 2006;Steenkamp and de Jong, 2010;Zhang and Khare, 2009).
Global brands are brands that have global awareness, availability, acceptance and desirability (Özsomer and Altaras, 2008).Prior research has highlighted that global brands hold an enhanced identity-expressiveness function, which makes them highly symbolic goods (Strizhakova and Colter, 2013;Xie et al., 2015).Hence, since our material possessions are viewed as major components of our extended self (Belk, 1988), consumers use global brands in order to signal the group to which they affiliate (Zhang and Khare, 2009), or to construct identity and promote their self-image (Strizhakova et al., 2011).Importantly, global brands, in contrast to local brands, are more appealing for publicly-visible goods because they are higher in aspirational value and associated with status, modernity, cosmopolitan sophistication and technology (Batra et al., 2000;Dimofte et al., 2008;Strizhakova et al., 2008;Zhao and Belk, 2008).Empirical findings further corroborate this notion by indicating that the impact of the global effect on brand preference varies across several product categories (Davvetas and Diamantopoulos, 2016).Furthermore, although extant research has greatly improved our understanding of the role of global-local identities in the consumption of global brands, existing studies on this topic have exclusively concentrated around material (physical) products, ignoring how digitalization has transformed the international marketing landscape (Oxley and Yeung, 2001;Prasad et al., 2001;Samiee, 1998).
This digital transformation has revolutionized consumers' engagement processes, since the nature of their physical interaction with physical products differs from online interaction with digital products in several ways.Prior research reports that consumers perceive physical goods as having greater social identity-signal capacities and legacy potentials relative to digital goods (Giles et al., 2007).Also, due to their greater permanence and tangibility, physical products provide greater ease of establishing attachment or association with the self (Belk, 2013).Digital products, being consumed online, lack some factors associated with the consumption of physical products, such as tangibility and visibility, which are desirable properties for consumers who aim to signal identification through consumption of identity-relevant products (Davvetas and Diamantopoulos, 2016).
At this stage, an important distinction should be drawn between digital technologies as consumption objects and possessions (e.g.participating in social networking sites or listening to streamed music) and consumption of goods and services through digital means (e.g.online shopping).Digital products refer to the former, where digital technologies are consumption objects themselves (e.g.search engines).Although material goods embody a system of meanings through which consumers express themselves and communicate with others (Dittmar, 1992), little is known as to whether this also holds for dematerialized digital goods.Moreover, studies have widely documented that certain attributes borne by global brands, such as higher quality and social prestige, comprise key benefits, particularly among consumers in developing countries (Batra et al., 2000;Steenkamp et al., 2003).Compared to their counterparts in developed markets, consumers in emerging markets have a stronger desire for global elements to construct their identity (Cleveland and Laroche, 2007;Strizhakova et al., 2008).Thus, in developing countries, consumers are more prone to using self-identity signals embedded in global brands (Strizhakova and Coulter, 2013;Zhou et al., 2008).
The present paper provides novel insights into the role of location-based identities in shaping preferences for global brands in the context of digital consumption among consumers from developed and developing countries.In this respect, we aim to contribute to existing international marketing literature by investigating the engagement mechanism through which global and local identities lead to consumption of global digital brands.In particular, our study considers social networking sites (SNS hereafter) as a representative category for digital products, and aims to investigate the engagement mechanism through which location-based identities drive preference for global digital brands, and, more specifically, the actual usage of two of the most popular global SNS, Facebook and Instagram.Through an empirical study of 425 global SNS users from Austria and Thailand, representing a developed and a developing country, respectively, we also aim to investigate potential cross-country variations of the above-mentioned mechanism.
To summarize, this study contributes to relevant literature by extending location-based identity research, through investigating the unexplored case of consumers' disposition toward global digital brands.For this purpose, it provides a path of theoretically-driven relationships, which aims to explain how global-local identities predict usage of global SNS.Finally, we investigate potential cross-country variations of the above-mentioned mechanism, which will extend our knowledge on how location-based identities predict global consumption in developing and developed countries.Findings are particularly relevant for digital brand managers interested in boosting consumers' preferences toward their global brand, especially when operating in developing markets.Furthermore, our research offers managerial advice to marketers of global SNS on how to adapt their marketing efforts in order to differentiate the positioning of their products.
We begin with a theoretical discussion which aims to clarify the concept of global digital brands.Then, drawing from the social identity theory (SIT) and literature around SNS usage, a number of conceptual relationships are developed.Hypotheses are presented, followed by the method and results.The research concludes with a general discussion, implications for managers of global digital brands, limitations and directions for further research.

Theoretical background
The global digital brand Central to our research is the notion of digital products and their differences from physical products, which provide the context of our study.To begin with, a conceptual distinction should be drawn between digital technologies as consumption objects and possessions and consumption of goods and services through digital means.We focus our attention on the first case, where digital technologies (or digital products) are consumption objects themselves (e.g.streamed music).Social networking sites, being virtual platforms where people of similar interests may gather to communicate, share and discuss ideas, represent an interesting cluster of digital products.Importantly, digital products occupy a liminal product category between the material and imaginary worlds (Denegri-Knott and Molesworth, 2010).As a result, digital consumption differs from material consumption as the object of consumption lacks material substance.As such, this product category has some inherent characteristics, such as limited options for social exhibition and thus social signaling, and limited visibility aspects.Since both social signaling and visibility represent product properties that can explain preferences for global brands (Davvetas and Diamantopoulos, 2016), their limited presence might influence their impact.Nevertheless, motivations for acquiring digital products resemble those for acquiring material consumer goods, including gaining status and prestige and expressing identity (Wang et al., 2014).Although digital possessions may not have the heft and gravitas of physical possessions, they can still play a key role in self-expression (Belk, 2014).Thus, examining whether the mechanism of expressing identities through consumption applies to this distinct product category is of great relevance.

IMR 36,5
Since the literature clearly suggests that consumers use knowledge at the brand level to process information and make purchase decisions (Low and Lamb, 2000), the global digital brand is the focus of this research endeavor.We base our definition of this on Steenkamp et al.'s (2003) operationalization of perceived brand globalness, which is built on the basis of consumer perceptions of a brand's worldwide availability, acceptance and desirability.Thus global digital brands are digital brands having "global awareness, availability, acceptance and desirability, and are often found under the same name with consistent positioning, image, personality, look, and feel in major markets enabled by standardized and centrally coordinated marketing strategies and programs" (Özsomer and Altaras, 2008, p. 1).A more recent view of perceived brand globalness by Steenkamp (2014) highlights three key components of what makes a brand global: use of a standardized marketing strategy worldwide, availability in multiple world regions and awareness outside the brand's customer base.
For a better understanding of the differences between digital products and digital brands, as well as what a global digital brand is, Figure 1 provides a hierarchical categorization of the digital product, based on the product-category schema-based approach.Product-category schemata can be thought of as providing a basic segmentation of the marketplace (Halkias, 2015).Such schemata aim to divide the market into distinct groups with homogeneous content, based on a taxonomic categorization (Blanchard et al., 2012).Hence, in Figure 1, the digital product is classified in a hierarchical fashion, starting from the superordinate level, representing the broadest level of organization, and proceeding to the finest level of categorization, namely, digital brands.Consistent with the relevant literature (Halkias, 2015;Hoyer and McInnis, 2008), at the superordinate level, products are different from each other, but still have a few common features and associations, such as the fact that they are consumed online.The subordinate level offers finer discrimination, where one can find several clusters of digital products (e.g.SNS, search engines, applications, etc.).Within the level below, products become branded and are differentiated from one another based on whether they are perceived as global or local.This distinction is based on the definition of perceived brand globalness by Steenkamp (2014) or of perceived brand localness, which incorporates consumer perceptions about a brand's symbolic association with a given local culture and its role as a key player in the local market (Swoboda et al., 2012).Finally, at the lowest level, one can find examples of global digital brands such as Facebook or Instagram and local digital brands such as Line or Weibo.Since we aim to investigate whether location-based identities influence preferences for global digital brands, we select global SNS as the context of our research.
Hereafter, we explain how location-based identities are expected to influence actual usage of global SNS.

Global digital brand usage
Global-local identity and global SNS hedonic motivation Identity theory was originally formulated by Stryker (1968) as a micro-sociological theory to explain individuals' role-related behaviors (Hogg et al., 1995).The theory maintains that individuals have distinct components of self-identity for each role position they occupy in society (Terry et al., 1999).Taking one step further, the SIT describes the relationship between individuals and the social group they feel affiliated to in order to provide explanations as to how memberships can influence individuals' behavior (e.g.Tajfel and Turner, 1986).In this respect, SIT explains in-group favoritism, or favorable evaluations or preferential treatment of people perceived to belong to the same in-group.This in-group favoritism is a result of the intrinsic need for positive social identity and the need to positively differentiate the in-group from out-groups.It is that need for positive distinctiveness that triggers a sequential process of social categorization, social identification and social-group comparison, which leads to in-group favoritism (Tajfel and Turner, 1986).Hence, from an SIT perspective (Tajfel, 1978), in-group or social identities reflect group membership, which defines an individual's self-concept, confers on them a relevant group identity and shapes their cognition and behavior (Tajfel and Turner, 2001).
Drawing on the SIT, several studies have sought to explain the role of an individual's group identification in shaping attitudinal responses and favoritism toward global or local brands (Bartsch et al., 2016;Cleveland et al., 2011).
Relevant literature acknowledges that people derive a sense of identity from their membership of spatial groups, such as a specific nation (e.g.Blank and Schmidt, 2003) or the entire world (e.g.Zhang and Khare, 2009).This phenomenon, known as location-based identification, was extensively examined by Tu et al. (2012, p. 36), who formally defined global and local identity: a local identity "consists of mental representations in which consumers have faith in and respect for local traditions and customs, recognize the uniqueness of local communities, and are interested in local events," whereas a global identity "consists of mental representations in which consumers believe in the positive effects of globalization, recognize the commonalities rather than dissimilarities among people around the world, and are interested in global events." Global and local identities may overlap, as conceptually one person can have both.They are not exact opposites, but can co-exist, achieving saliency in different contexts (Erez et al., 2013), and tend to be understood as separate and independent processes (Reed et al., 2012;Steenkamp and de Jong, 2010).In line with this view, past studies have suggested that cultural identity is often framed as a tension or competing choice between global and local identity, but there is increasing recognition that both are intertwined in mediated and nuanced conversations with each other (Strizhakova et al., 2012;Dong and Tian, 2009;Varman and Belk, 2009;Zhao and Belk, 2008).However, most people tend to lean toward whichever is stronger as implied by situational factors, as maintaining both equally may lead them to uncomfortable identity confusion (Zhang and Khare, 2009;Arnett, 2002).Therefore, when multiple identities are present, individuals shift among them, seeking to apply those that, depending on the relevant context they find themselves in, are perceived to be congruent (e.g.Fetscherin, 2010;Reed et al., 2012).This mechanism fulfills the innate need for self-consistency (Xie et al., 2015).Which particular identity becomes salient at any given point varies based on specific situational and/or individual circumstances (Forehand et al., 2002).For example, in the context of evaluating foreign products, a global identity is more likely to become salient since by its very nature it implies embracing the interplay of different foreign elements (Micevski et al., 2018;Cannon and Yaprak, 2002).
In the digital context, self-identity, among other factors, can explain an orientation toward technology usage (Westjohn et al., 2009).Thorbjørnsen et al. (2007) argue that digital communication services constitute an interesting arena for investigating identity-expressiveness and self-presentation effects on other constructs.Moreover, past studies have shown that 706 IMR 36,5 self-identity has a positive effect on intention to use digital products (Nysveen et al., 2005;Thorbjørnsen et al., 2007).Likewise, Brandtzaeg and Heim (2009) propose self-identity as one of the main drivers of SNS usage.
At the same time, global brands are seen as sources of symbolic values such as status, prestige, social approval, excitement and modernity (Halkias et al., 2016;O ̈zsomer, 2012;Steenkamp et al., 2003).In this context, recent studies support the view of an identity-based function of global brands, whereby consumers view global brands as a medium to express their modern self-image and global identity (Strizhakova and Coulter, 2013;Xie et al., 2015).In fact, location-based identification puts forward a congruence effect such that global identity typically leads to preference for global brands (Zhang and Khare, 2009;Zeugner-Roth et al., 2015).Identity theorists concur that positive emotions result from meeting one's identity-enactment expectations (Burke and Stets, 2009) because identities carry with them a motivation to act in identity-consistent ways (Oyserman, 2009).Coleman and Williams (2013) suggest that individuals will prefer to experience emotions that are consistent with a social identity, as well as engage in identity-consistent emotion regulation processes.
Hedonic motivation refers to the hedonic fulfillment such as experiencing fun, amusement, fantasy and sensory stimulation that is related to shopping experience (Babin et al., 1994).In line with previous research, this study suggests that, in the process of achieving congruence with their global identity when individuals recognize that they can enact their identity in a fulfilling way (i.e. by consuming a global brand or by becoming active participants of a global SNS such as Facebook or Instagram), they are expected to be driven by a hedonic motivation that will enable them to obtain pleasure and enjoyment from this specific enactment.Therefore, location-based identities are considered drivers of SNS hedonic motivation, as users seek consistency between their identity and emotions (Burke and Stets, 2009;Oyserman, 2009;Reed, 2004).Users driven by a global identity in particular are expected to display increased hedonic motivation when participating in global SNS.Thus, we posit that: H1.(a) Global and (b) local identities are both drivers of global SNS hedonic motivation, with global identity enhancing hedonic motivation more than local identity.

From hedonic motivation to perceived global SNS value and usage
In general, perceived value refers to the value that consumers perceive by using a product/service (Bettman et al., 1998).Of course, value perception differs according to personal values, needs and preferences (Pura, 2005) and thus prior research has defined perceived value as a multi-dimensional construct (Yu et al., 2013).Sheth et al. (1991) introduced a framework viewing perceived value as having social, emotional, functional, epistemic and conditional dimensions, whereas Babin et al. (1994) regarded the concept of perceived value as consisting of hedonic value among other components.By adopting Ducoffe's (1995, p. 1) definition of perceived advertising value, this study conceptualizes perceived SNS value as "a subjective evaluation of the relative worth or utility of Social Networking Sites to users."This conceptualization is also consistent with a stream of researchers who equate value with an overall assessment of subjective worth considering all relevant evaluative criteria (i.e.Schechter, 1984;cited in Zeithaml, 1988, p. 13).
Previous research has used the motivation theory to explain perceived value in the field of information technology (Davis et al., 1992).In the context of SNS specifically, hedonic motivation is a distinct motive which appears frequently across studies (e.g.Chi, 2011;Heinonen, 2011;Park et al., 2009;Zhang and Mao, 2016).It refers to actions committed because of a general perceived value of interest (Davis et al., 1992), and focuses on consumers' emotional needs, taking into consideration non-functional benefits derived from the experience, such as happiness, fantasy, sensuality and enjoyment (Xu et al., 2012).Moreover, research around SNS indicates that a perceived value of enjoyment is positively related to SNS usage (Lin and Bhattacherjee, 2008;Moon and Kim, 2001;Van der Heijden, 2004).Recent gratification research further corroborates these findings by suggesting that SNS users aim to satisfy a set of needs, including entertainment, which allows them to formulate general value perceptions about the specific SNS platform (e.g.Ko et al., 2005;Whiting and Williams, 2013;Dunne et al., 2010).Therefore, we hypothesize that: H2.Global SNS hedonic motivation enhances perceived global SNS value.
Perceived value of enjoyment, derived from hedonic motivation, is also incorporated in discussions about the technology acceptance model theory, and linked positively to SNS usage (Davis et al., 1992;Hong and Tam, 2006;Van der Heijden, 2004;Venkatesh, 2000).In particular, perceived enjoyment is an important factor of predicting the intention to use a pleasure-oriented information system such as SNS, as individuals' motivation to use SNS can be expected to increase when they experience more intense enjoyment (Kang and Lee, 2010;Sledgianowski and Kulviwat, 2009).Enjoyment and fun are also motives for mobile services usage (Pagani, 2004;Nysveen et al., 2005), as well as drivers of usage in an entertainment context (Van der Heijden, 2004).In the context of SNS particularly, both aforementioned perceived values are found to positively influence usage (Yu et al., 2013).Finally, gratification research shows that SNS perceived values are fundamental drivers of SNS usage (e.g.Huang et al., 2014;Kim et al., 2011).Bringing this notion to the global SNS context, we hypothesize that: H3.Perceived global SNS value predicts actual global SNS usage.

Cross-country variations
Global identity signifies the main outcome of globalization, and is more prominent in developing than developed countries (Arnett, 2002).In emerging marketplaces such as Brazil and India, consumers identify themselves as "global citizens" more often than those in developed marketplaces such as the USA and Germany (Holt et al., 2004;Guo, 2013).In general, consumers in emerging markets seek to affirm their affinity with members of the advanced world (Batra et al., 2000;Oszomer, 2012), which explains their higher preference toward global products.For example, Steenkamp and de Jong (2010) suggest that Chinese consumers favor global brands more than US consumers.In developing countries, consumers show similar preferences as they believe that they offer them social status, social conformity and wealth expression (Guo, 2013;Wang and Yang, 2008).In less developed countries, consumers associate global brands with aspirational lifestyles of more advanced economies (Alden et al., 1999;Dimofte et al., 2008).These symbolic attributes of global brands primarily influence tastes in developing countries, which are characterized by low-level economic development and high social mobility (Batra et al., 2000).Using a global brand is a way of displaying competence with regard to alien cultures and is an important motive behind global identity in many developing countries (Batra et al., 2000).Therefore, in emerging economies, global-oriented consumers are even more likely to prefer global brands because their self-identity is reflected in them; and in emerging markets particularly, this need is stronger (Cleveland and Laroche, 2007;Strizhakova et al., 2011).We suggest that the magnitude of the relationship between global 708 IMR 36,5 identity and hedonic motivation will be stronger for users from developing than from developed economies, i.e.: H4.The relationship between global identity and global SNS hedonic motivation is stronger among users in developing countries than among users in developed countries.
Emerging markets are becoming home markets for local brands, which are gaining success in global markets in their own right (Xie et al., 2015) Moreover, it is generally accepted that the production and control of global culture resides in the developed countries of the West (USA especially) and the flow of global culture is mostly from Western countries to developing ones (Batra et al., 2000).Therefore, consumers from developed countries should perceive global products, effectively symbols of global culture (Alden et al., 1999), as having stronger cultural links with their local community and closer proximity to their local market.On the contrary, consumers from developing countries are expected to perceive global brands as more distant, with less local-community relatedness.
Local identity represents a connection to the local community (Zhang and Khare, 2009).As previously described, the closer a brand's identity is perceived to be to self-identity (Chernev et al., 2011) and group identity (Brewer and Gardner, 1996), the more attractive it should become.Overall, in developed compared to developing countries, local identity is expected to be more strongly associated with global brand preferences because of the brands' closer relatedness to and origin in the local market.Thus, in the context of our study, we propose: H5.The relationship between local identity and global SNS hedonic motivation is stronger among users in developed countries than among users in developing countries.
Figure 2 illustrates the conceptual framework of the study.

Data collection and sample
An online survey was conducted in Austria and Thailand for the purposes of our study.
The selected countries represent a good prototype of a West European developed country (Austria) and a Southeast Asian developing country (Thailand).To assess economic development, we relied on gross national income per capita converted to US$ for 2017.We retrieved data from the World Data Bank (2017): Austria (GNI/capita: US$45,440) and Thailand (GNI/capita: US$5,960).Moreover, both regions, West Europe and Southeast Asia, exhibit high SNS penetration rates of 54 and 55 percent, respectively, higher than the global SNS penetration rate of 42 percent (We are Social, 2018).In Austria, as in most European countries, a variety of global SNS compete for market share, with Facebook consistently occupying the top position in active online reach, followed by Instagram and Twitter (Ofcom, 2018).As for Thailand, the more turbulent political and socio-economic context of Asia is reflected in its country-specific markets.With regards to SNS in particular, Asia's market composition is quite different from that in Europe, with intense competition between global and regional companies.Thailand is among the top 15 countries in terms of number of SNS users (Statista, 2018a), making it a highly attractive market.
Both global SNS companies (Facebook, Instagram, Twitter, etc.) and regional SNS companies operate in Thailand quite successfully.For instance, Thailand ranks among the top ten countries as regards active Facebook users (We are Social, 2018).At the same time, though, it is the second most successful market for Line ( Japanese SNS), after Japan.WeChat (Chinese SNS) is also increasingly gaining popularity in Thailand ever since it entered the market.Therefore, these countries provide a fruitful ground for our study.
For our online survey, two versions of a questionnaire were designed: one for Facebook users and another for Instagram users.Facebook and Instagram are leading SNS, with 2,167m and 800m active users worldwide, respectively (Statista, 2018a, b).In 2017, Facebook had 3.8m users in Austria and 49m users in Thailand.Similarly, Instagram had 1.7m users in Austria and 13.6m users in Thailand.Their ubiquity makes them good representatives of global digital brands.Moreover, they are available and successful within both Austria and Thailand.The questionnaire was initially developed in English and then translated and back-translated by native speakers and bilinguals into the native languages of the participants, i.e.German and Thai, as in previous research (Makri and Schlegelmilch, 2017).In order to establish the conceptual equivalence of our measures (Douglas and Craig, 1983), inconsistencies with the original versions were discussed among translators and resolved, to minimize idiomatic issues.
A convenience sample of students was used in both counties.Bachelor students from two business schools in Vienna and Bangkok who attended a course offered by one of the researchers were asked to forward the online questionnaire via e-mail to their peers as part of the course's assignments.A total sample of 425 (n AT ¼ 247, n TH ¼ 178) young adults was used to test our hypotheses.In Austria, 56 percent of the respondents were female, and respondents were on average 26 years old (M age ¼ 26.13, SD ¼ 11.11).Of the Austrian sample, 81 percent was composed of unemployed students and the vast majority of the sample (89 percent) were students holding a high school diploma or a bachelor degree.Austrian users had on average 324 friends/followers in Facebook/Instagram, with an average of 33 percent being international acquaintances.Apart from Facebook and Instagram, the most frequently used SNS among Austrian users were Snapchat (53 percent), Twitter (23 percent) and YouTube (12 percent).In Thailand, 36 percent of the respondents were female, and respondents were, on average, 21 years old (M age ¼ 21.35, SD ¼ 3.21).Of the Thai sample, 87 percent was composed of unemployed students and the vast majority of the sample (96 percent) were students holding a high school diploma or a bachelor degree.Thai users had on average 554 friends/followers in Facebook/Instagram, with an average of 21 percent international.Apart from Facebook and Instagram, the most frequently used SNS among Thai users were Line (54 percent) and Twitter (32 percent).

Measures
Appendix includes all measures used in our survey which were based on well-established existing scales.In particular, global and local identity was measured by Tu et al.'s (2012) scale.Hedonic motivation was measured by Babin et al.'s (1994) scale, and perceived SNS value with a scale adapted from Ducoffe (1995).Actual usage was measured by the average time spent on the specific SNS daily.All measures used a seven-point scale format (from 1 ¼ strongly disagree to 7 ¼ strongly agree).We also included ethnocentrism, measured with Shimp and Sharma's (1987) CETSCALE, number of friends/followers and SNS type as control variables.SNS type was operationalized with a dummy variable (0 ¼ Instagram, 1 ¼ Facebook), in order to account for potential differences caused by the different platforms.The items per construct used for further analysis are presented in the Appendix.

Measurement validation
Confirmatory factor analysis (CFA) was first conducted to assess the measurement properties of the models, using Amos 23.Following Anderson and Gerbing (1988), reliability and validity of the scales were investigated by examining the fit of three measurement models, one for the whole sample (n ¼ 425) and one for each country separately (Austria and Thailand).For purification purposes, a number of items that displayed low factor loadings (o0.5) and therefore did not exhibit adequate psychometric properties in all models were dropped.The Appendix features the items per construct retained for further analysis.
The purified measurement models were then further evaluated.
As shown in Table I, fit statistics indicate a close fit to the data for all three models.Moreover, all factor loadings of the remaining items were large (ranging from 0.567 to 0.990) and significant ( po 0.01), providing evidence of convergent validity.We also estimated construct reliability using composite reliability scores and average variance extracted (AVEs).All constructs have composite reliabilities scores exceeding 0.70 (ranging from 0.83 to 0.97), and the AVE for all constructs was greater than 0.50 (ranging from 0.62 to 0.92), satisfying the recommended thresholds (Bagozzi and Yi, 1988;Fornell and Larcker, 1981).Table II presents descriptive statistics, reliability scores and intercorrelations for the study constructs.Finally, we assessed discriminant validity using the most restrictive test provided by Fornell and Larcker (1981).This method compares the square root of the AVE with the correlation of latent constructs.As Table II illustrates, the square root of each construct's AVE has a greater value than the correlations with other latent constructs, providing strong evidence of discriminant validity.Thus, all measurement scales possess good levels of reliability and validity.

Invariance analysis
Invariance analysis with multi-group structural equation modeling was further performed following Byrne et al. (1989).To ensure invariance, the pattern of factor loadings for each observed measure should first be tested for equivalence across the groups ( Jöreskog, 1971).As a result of the measurement validation process, we determined a baseline model which fits the data adequately for each group (i.e.country) separately.
The next process is to estimate the same parameters that were estimated in the baseline model in a multi-group model.For this purpose, as an initial step, no equality constraints were imposed on any of the parameters.As Table IV shows, the fit of the unconstrained baseline model, also known as configural model, is satisfactory (w 2 160 ð Þ ¼ 309:4, p o0.00; χ 2 /df ¼ 1.93; CFI ¼ 0.96; RMSEA ¼ 0.047).This is evidence that the number of factors and the pattern of their structure are similar across the Austrian and the Thai sample, thus indicating the existence of configural invariance.
However, configural invariance is not a sufficient condition for establishing measurement equivalence (Byrne et al., 1989).When testing for measurement invariance, the focus turns to

712
IMR 36,5 the extent to which parameters in the measurement components of the model are equivalent across the two groups.This process requires formal comparison between the baseline model and a full metric invariance model, in which equality constraints are assigned on all factor loadings.Table III presents the results of this comparison also.Given that the χ 2 difference between the baseline and full metric invariance models is insignificant (Dw 2 10 ð Þ ¼ 17:5, pW0.05), we assume that full measurement invariance is supported.The full invariance model will be used in subsequent cross-country analyses.

Common method bias
Common method bias (CMB) is a potential problem in studies that rely on a single informant and collect data at one point in time.This bias results "from any artefactual covariance between the predictor and criterion variable produced by the fact that the respondent providing the measure of this variables is the same" (Podsakoff et al., 2003, p. 881).We aimed to control for CMB by employing Harman's single factor test and including all measurement items in an exploratory factor analysis.We repeated this process for both the Austrian and Thai samples.In both cases, the unrotated factor solution extracted nine factors (with 27.53 percent being the most variance explained by any one factor in Austria and 31.54percent by any one factor in Thailand), thus providing no evidence for CMB.
In addition, we employed the marker variable approach proposed by Lindell and Whitney (2001) for both measurement models (Austria and Thailand).Specifically, we introduced social desirability as a latent marker variable in our measurement models which, from a conceptual point of view, was unrelated to the constructs analyzed in our model.Social desirability was captured by three items: "I like to gossip at times," "I have never deliberately said something that hurt someone's feelings" and "I am always willing to admit it when I make a mistake" (measured on a seven-point scale anchored at strongly disagree/ strongly agree) (Strahan and Gerbasi, 1972).We compared the significance of the structural parameters of the models before and after the introduction of the marker variable.The significance of the resulting coefficients did not change, indicating that CMB was not a problem in the analysis.Therefore, we can safely assume that CMB cannot explain the observed associations among our study constructs.

Hypotheses testing
In order to test our research hypotheses, structural equation modeling was performed in Amos 23.A full model (including both Austrian and Thai respondents) was tested to investigate hypotheses H1a-H3.Because the number of Facebook/Instagram friends/ followers may influence the time spent on the relative SNS, we include this variable as control in our model.Furthermore, SNS type was included as a control variable to account for the potential influence stemming from structural differences between the two platforms (i.e.Facebook and Instagram).Moreover, consumer ethnocentrism is one of the most established constructs used to explain consumer preference for local/regional brands over foreign ones (Shimp and Sharma, 1987;Steenkamp et al., 2003).As such, it may be

713
Global digital brand usage negatively associated with the usage of global products and is thus also included as a control variable in our model.
The model exhibited acceptable levels of fit (full model: w 2 128 ð Þ ¼ 326:4, p o 0.00; χ 2 /df ¼ 2.55; CFI ¼ 0.94; RMSEA ¼ 0.060).Additionally, standardized coefficients and corresponding t-values reported in Table IV provide evidence for H2 and H3 for the whole sample.Furthermore, both global and local identity are significant drivers of hedonic motivation.For H1a-H1b, a formal comparison of our model with a model in which the two coefficients were set equal showed that there was no improvement in overall fit (Dw 2 1 ð Þ ¼ 0:9, p ¼ 0.34).However, a closer examination within the county subsamples indicated that although in Thailand global identity predicted hedonic motivation better than local identity (Dw 2 1 ð Þ ¼ 4:8, p ¼ 0.02), in Austria, both identities predicted hedonic motivation equally (Dw 2 1 ð Þ ¼ 1:4, p ¼ 0.24), weakening the overall effect.Overall, H1a-H1b was partially confirmed.
Subsequently, cross-country comparisons were performed to investigate differences in the structural parameters between users in Austria and Thailand [1].A two-group model was tested in order to formally test for H4 and H5.The model exhibited acceptable levels of fit (two-group model: w 2 256 ð Þ ¼ 584:2, p o 0.00; χ 2 /df ¼ 2.28; CFI ¼ 0.91; RMSEA ¼ 0.055).As Table IV shows, a positive and significant relationship between global identity and hedonic value is observed for both Austrian and Thai users.The relationship between global identity and hedonic value was further examined by comparing a model in which this path was constrained to be equal between groups to a model where the parameter was freely estimated between groups.A significant difference was found (Dw 2 1 ð Þ ¼ 4:2, p o 0.05; β AT ¼ 0.17; β THAI ¼ 0.54), indicating the positive association between global identity and hedonic value is stronger among Thai than among Austrian users.Thus, H4 was supported.
H5 was tested in a similar way.As shown in Table IV, the relationship between local identity and hedonic motivation was not significant in Thailand.This finding can be interpreted by considering the nature of the Thai SNS market.In general, Asia's often more turbulent political and socio-economic context is reflected in its diverse country-specific markets.This diversity is also mirrored in the SNS marketplace, with many regional

714
IMR 36,5 platforms gaining popularity (i.e.Japanese "Line," Chinese "Weibo" or "WeChat," etc.).Particularly in Thailand, both global and regional SNS platforms operate with success, whereas in Austria, as in the majority of European countries, global SNS platforms largely monopolize the market.In the presence of a variety of regional SNS substitutes (Line, WeChat, etc.), the market is naturally polarized, with Thai users who identify themselves on a local level being expected to obtain pleasure from engagement with their regional SNS rather than the global ones.This identification might be particularly strong among consumers brought up in collectivistic societies such as Thailand, and therefore the relationship between local identity and hedonic motivation to use a global product may be further weakened.Although the coefficient of the relationship between local identity and hedonic value was found not significant at the 95 percent confidence level in Thailand, it is still descriptively lower than in Austria.Indeed, a formal comparison of this relationship's magnitude between Austria and Thailand confirms a significant difference.A model in which the path between local identity and hedonic attitude was constrained to be equal between groups was compared to a model in which the corresponding parameter was freely estimated between groups.The χ 2 difference between the two models was significant (Dw 2 1 ð Þ ¼ 5:7, po0.01; β AT ¼ 0.28; β THAI ¼ −0.03), providing support for H5.Therefore, the relationship between local identity and hedonic value is stronger among Austrian than among Thai users.

Discussion and implications
Evidence exists that a large share of consumers around the world reveal a preference toward global brands (Steenkamp et al., 2010), with researchers in the field providing theoretical explanations as to why.For instance, global brands have been shown to be associated with high perceived quality and prestige, especially in developing countries (Steenkamp et al., 2003).Other studies go further and claim that global brands have the ability to provide consumers with a desired identity (Dimofte et al., 2008), with a number of researchers focusing particularly on the role of global-local consumer identities (Tu et al., 2012;Verlegh, 2007;Zhang and Khare, 2009).Researchers of the field claim that global and local identity can co-exist within an individual, with one of the two becoming occasionally salient and triggering favorable attitudes toward brands that are identity-congruent (Zhang and Khare, 2009).Despite the growing academic focus on consumers' location-based identities and their impact on consumer decision making, research on this topic has exclusively focused on physical products.Consequently, evidence is lacking regarding the effect of consumers' global-local identities on brand preferences in the digital context, thus overlooking how the digitalization and dematerialization of our digital possessions has transformed the international marketing landscape.To address this notable research gap, this study investigates the role of location-based identities in predicting usage of global SNS.Facebook and Instagram, being globally available and desirable SNS, are profound symbols of global consumer culture.As such, they have the ability to promote and diffuse global values and lifestyles among their users worldwide.Through their participation in global SNS, consumers from all around the world acquire knowledge, skills and behaviors that are characteristic of the global consumer culture and are encouraged to express their global identity (e.g.Cleveland and Laroche, 2007).That said, extending relevant research on digital products and digital offerings can provide significant insights for both academics and digital managers.

Theoretical implications
Given the digital transformation in today's competitive global marketplace, the contribution of our research is threefold.First, drawing on identity theory, our paper contributes to the existing literature with regard to how global-local identity is influencing consumers' disposition toward global brands, by being the first to consider the case of global digital brands.Therefore, our study offers new insights into the evolving nature of global consumption and international marketing in the age of digitalization.Second, by considering Facebook and Instagram as prominent global digital brands, our findings provide empirical evidence on how global and local consumer identities influence the usage of global SNS in Austria and Thailand.In particular, our results suggest that global-local consumer identities have an effect on the usage of global digital brands, through a path of relationships indicating how location-based identities influence users' perceived hedonic motivation, which, in turn, enhances their perceived SNS value and leads to increased usage.Third, our study advances the stream of research which investigates the role of global-local consumer identities in formulating consumers' positive disposition toward global products under the lens of emerging vs developed markets (Alden et al., 2013;Guo, 2013;Strizhakova and Coulter, 2013;Swoboda et al., 2012).In emerging countries, global brands are perceived as a kind of passport to global citizenship, and thus consumers with a global identity are more prone to both quality and self-identity signals embedded in global brands, and hence more likely to prefer them (Strizhakova et al., 2008).Our findings corroborate these claims by providing empirical evidence that in the developing country (Thailand), users with an accessible global identity demonstrate higher hedonic motivation to participate in global SNS than their counterparts in the developed country (Austria).Thai users think of their participation in global SNS as a passport to global citizenship, which enables them to achieve congruence with their global identity and thus enhance the pleasure obtained from using Facebook and Instagram.This underlying mechanism is also significant for Austrian users, but significantly weaker.
Extant research suggests that global-local identity produces an assimilation effect such that consumers with high global identity tend to prefer global products, whereas those with high local identity lean toward local products (Zhang and Khare, 2009).However, it is widely documented that global and local identity are not exact opposites, can co-exist within an individual, and can become salient in different contexts (Erez et al., 2013;Reed et al., 2012;Steenkamp and de Jong, 2010).Our analysis further corroborates this notion, as global and local identity are positively correlated in both our samples, as clearly evidenced in Table II.Also, according to our findings, both global and local identities are significant drivers of global SNS preference.In cases where both identities are equally strong, a person tends to decide on one particular identity, which leads to one identity being usually more accessible (Arnett, 2002;Zhang and Khare, 2009).In other words, according to the context and their accessible identity, consumers seek to enact identity-congruent behaviors to maintain consistency in their feelings and actions (Westjohn et al., 2012) and to avoid potential tensions ( Josiassen, 2011).
This offers an explanation as to why no significant association between local identity and hedonic motivation was found among users of global SNS in Thailand.Thai users with an accessible local identity, who identify themselves with their local community, are expected to show preference for the regional digital products available in their market and therefore show weak motivation to use global SNS.In Austria, however, in the absence of regional or local competition which would polarize the market and consequently initiate the assimilation or congruence effect described above, this relationship was significant.In addition, for consumers from developed countries, given that the association between global products and developed economies is quite established, global products are conceptually closer to their local in-group (Halkias et al., 2016).Therefore, the link between local identity and hedonic motivation is stronger among global SNS users in Austria than among their counterparts in Thailand.
To summarize, our findings show that the impact of global identity on hedonic motivation is significantly higher among users of global SNS in the developing country (Thailand) than in the developed country (Austria).On the contrary, the impact of local identity on hedonic motivation is significantly higher among users in the developed country (Austria) than in the developing country (Thailand).In general, hedonic motivation enhances the perceived value of global SNS, which, in turn, increases the actual usage.

IMR 36,5
Practical implications Our findings are particularly interesting for digital marketers interested in boosting consumer preferences toward their global offering, especially when these marketers find themselves operating in developing markets or in markets where local/regional substitutes feature strongly.First, our findings show that global identity has a positive effect on the hedonic motivation of individuals in emerging markets, which, in turn, affects positively perceived global SNS value and usage.Therefore, managers should build or improve the cosmopolitan nature of their brands in order to appeal to the global identity of their target audience.In this way, consumers will feel members of a global brand-club which adds prestige and modernity to their self-identity.In practice, global events (e.g.Google products trainings) that promote such characteristics can boost the global identity expression of the individual, positively influence perceived value and increase product usage.
Second, our results suggest that hedonic motivation of global digital brands increases actual usage through the enhancement of perceived value, especially in developing markets.Thus, managers should promote the digital brand value derived from usage.Fun, fantasy, excitement and entertainment constitute an individual's basic hedonic needs and therefore should be attached to any digital brand.For instance, Netflix and Spotify have recently successfully entered developing markets, including Thailand; both brands meet such hedonic consumer needs worldwide, and, at the same time, their patrons feel like members of a global entertainment community.
Third, research has emphasized that global brands are positioned as a way of expressing one's global belongingness (Alden et al., 1999(Alden et al., , 2006;;Steenkamp et al., 2003); it has been argued that consumers who believe in global citizenship embrace global brands as a way of expressing engagement with the world (Strizhakova et al., 2008).In this respect, marketers should try to craft winning strategies by finding ways to reinforce the global character of their digital brands, thereby technically making that global identity more accessible for existing or potential consumers.For instance, they might use global symbols, landmarks or even international language throughout their marketing efforts in order to activate and boost the accessibility of global identity and hence connection to the world.Thus, for marketers of global SNS, our research offers a strong argument for fostering the salience of global identification, particularly for operations in developing countries/markets.
Finally, when it comes to global SNS, users increasingly express themselves to others in ways that can potentially attract large audiences across the world.Therefore, global SNS, apart from being digital products themselves, are also important tools for companies and marketers that aim to reach big and global audiences in order to increase awareness of their products/brands.As such, the managerial implications of our findings are relevant not only to SNS platform managers, but also to companies that use these platforms as means of promoting their offerings.In this respect, our results may have multiplicative effects, since they concern a significant number of companies worldwide.

Limitations, future research directions and conclusions
Our research endeavor has limitations that offer avenues for future studies.First, our study, being the first to investigate the effect of location-based identities on global digital brands, focuses on the very specific product category of SNS.In order to mitigate any potential confounding effect arising from this specificity, future research should aim to replicate findings for other categories of digital products, such as video-streaming platforms, search engines and internet browsers.Second, we should also note that the student nature of our sample may not have been representative of the population at large.While one would expect the relationships uncovered in the present research to apply to the general population as well, some caution should be taken with regard to the generalizability of the findings.Third, the cross-sectional survey-based method used to conduct our study imposes some inherent limitations to our results.Different methods and samples in future studies should aim to verify the generalizability of the present study.For instance, experimentally priming global and local identities could examine the sequencing and causality of relationships more confidently.Moreover, more countries should be included for more insightful cross-country research.
Our study only takes into consideration global digital brands as represented by global SNS, and offers insights on the role that identities play in shaping attitudes and behaviors toward Facebook and Instagram.Due to the lack of local SNS alternatives, results are unable to delineate users' attitudes and behaviors toward global and local digital brands.Thus, future studies should attempt to collect comparable data from users of local or regional SNS, such as the Chinese RenRen or Japanese Mixi in order to offer complete empirical support for the occurrence of the assimilation effect, which suggests that consumers with high global identity tend to prefer global products, whereas consumers with local identity tend to prefer local products (Zhang and Khare, 2009).To this end, researchers should take into account potential discrepancies between local and regional SNS.In cases where regional is not equivalent to local, as for instance RenRen for Thai users, the anticipated effects of location-based identities on product preference might not be replicable.
Future research would also benefit by investigating a number of potential moderators which might alter the magnitude of the relationships between identities and product preferences in the context of digital product offerings.For instance, international marketing literature identifies xenocentrism (e.g.Balabanis and Diamantopoulos, 2016) or cosmopolitanism (Cleveland et al., 2011) as individual traits which can influence consumers' behaviors.The inclusion of these constructs, among others, within models that examine the links between location-based identities and digital product preferences can offer fruitful research directions.Another notable shortcoming of our research which future studies should aim to remedy is the exclusion of several factors that might interact with our constructs and thus act as control variables in our conceptual model, such as individual measures of collectivistic-individualistic value or the use of private vs public SNS profiles.
To conclude, insights offered by the present study expand our understanding of how location-based identities influence consumers' disposition toward global brands in the context of digital products.Results showing that global and local identity lead to enhanced global SNS usage through an increase of the enjoyment obtained by participation provide practical strategic insights for global SNS marketing managers.Empirical evidence of cross-country variations of this mechanism allows managers to adapt their marketing strategies according to the country-market they target.
Figure 2. Conceptual framework . Unlike emerging markets, developed markets not only provide home markets for local products, but, at the same time, represent typical sources of global brands.According to the BrandZ's (2017) top 100 most valuable global brands list, the vast majority of global brands are from developed economies, with brands from North America (USA primarily), UK and Continental Europe leading the list.In fact, almost all global SNS have been founded in the USA (e.g.Facebook in Boston, Instagram in California).Similarly, other global digital brands have emerged from economically well-developed countries (e.g.Spotify, founded in Sweden).Since global brands typically originate from developed countries, the link between global brands and developed countries in the mind of consumers can be expected to strengthen.