Do brand warmth and brand competence add value to consumers? A stereotyping perspective

Contributing to the literature on brand stereotyping, we draw on the Stereotype Content Model to investigate whether the content of the brand stereotype (in terms of warmth and competence) impacts consumers' perceptions of functional, emotional and social value. In doing so, we explicitly account for the brand's level of perceived globalness (PBG) and localness (PBL) as known in ﬂ uences on both stereotype content and value perceptions. Across two studies, we ﬁ nd that brand warmth consistently and positively impacts functional and emotional value, whereas brand competence enhances functional value. The impact of the stereotyping dimensions on value is subsequently re ﬂ ected in increased purchase intentions and higher brand ownership. Surprisingly, none of the latter outcomes is a ﬀ ected by social value. Our ﬁ ndings corroborate previous research showing that PBG and PBL are important drivers of brand stereotype content, but also reveal that brand warmth has a stronger impact on behavioral outcomes than brand competence.


Introduction
Brand stereotypes represent consumers' oversimplified and generalized beliefs about brands as intentional agents (Greenwald & Banaji, 1995;Kervyn, Fiske, & Malone, 2012), that is, brands are perceived by consumers as capable of expressing intentions (i.e. warmth) as well as able to enact these intentions (i.e. competence). An example of a brand that relies on its stereotype content in its branding efforts is the German cosmetic brand Nivea. Nivea is stereotyped as high in both warmth and competence by its consumers (Ivens, Leischnig, Muller, & Valta, 2015), which implies that the brand is perceived to have good intentions (e.g. care for the nature in terms of packaging and recycling; Nivea, 2019) and to be able to enact them (e.g. Nivea is one of the leading companies in the field of skin care, with over 130 years of experience; Nivea, 2019). According to the Brands as Intentional Agents Framework (BIAF), "brands differ in how well (or ill) intentioned brands seem to be, as well as on how able they are perceived to be" (Kervyn et al., 2012, p. 9). For instance, Johnson & Johnson is seen as a brand with good intentions and high ability (i.e. high warmthhigh competence), thus representing a popular brand among consumers, whereas Marlboro's good intentions are questionable and its perceived ability mediocre (i.e. low warmthlow competence), making it a stereotypical troubled brand . Other brands may be associated with "mixed" stereotype content whereby high scores on the warmth dimension may be accompanied with low scores on the competence dimension or vice versa (Fiske, Cuddy, Glick, & Xu, 2002). For instance, Rolls Royce is typically seen as a highly competent, but cold brand .
Previous research shows that brand stereotyping results in more favorable brand evaluations, brand-related emotions, increased purchase intentions and brand ownership (e.g. Aaker, Garbinsky, & Vohs, 2012;Ivens et al., 2015). Despite the rich body of research on brand stereotypes (see Appendix A for a comprehensive summary), the question of how stereotypical judgements of warmth and competence influence consumer's perceptions of value (Sweeney & Soutar, 2001) and, through them, behavioral outcomes (e.g. purchase intentions and brand ownership) has yet to be addressed. Perceived value is defined as the "consumer's overall assessment of the utility of a product (or service) based on perceptions of what is received and what is given" (Zeithaml, 1988, p. 14). The construct of perceived value is crucial to marketing theory and thus, to a better understanding of consumer behavior (Gallarza, Gil-Saura & Holbrook, 2011), because every purchase involves a tradeoff of benefits received against the sacrifice incurred to obtain these benefits (Woodall, 2003). The central importance of the value construct is also reflected in a rich body of research, which shows that value perceptions impact consumers' loyalty, attraction and brand repurchase, among others (Holbrook, 1994;Yang & Peterson, 2004;Zeithaml, 1988). Interestingly, while prior research has shown that different dimensions of perceived value are influenced by consumers' perceptions of brand globalness (PBG) and localness (PBL) (Swoboda, Pennemann, & Taube, 2012;Swoboda & Hirschmann, 2016) and while both PBG and PBL have been found to impact brand warmth and competence (Davvetas & Halkias, 2019;Kolbl, Arslanagic-Kalajdzic, & Diamantopoulos, 2019), the direct link between brand stereotype content and different perceived value dimensions has not been assessed 1 .
Against this background, we draw from stereotyping literature and, in particular, the Stereotype Content Model (SCM; Fiske et al., 2002) as well as from signaling theory (Erdem & Swait, 1998;Kirmani & Rao, 2000) to investigate the links between brand stereotype warmth and competence and dimensions of perceived brand value, while controlling for the level of PBG and PBL. We further consider behavioral outcomes of perceived value and explicitly compare the relative impact of brand stereotype warmth and competence on both, value dimensions and outcomes (i.e. purchase intentions and brand ownership).
Our intended contribution is four-fold. First, we contribute to brand stereotyping literature by revealing how distinct dimensions of brand stereotype content influence different dimensions of perceived value (Sweeney & Soutar, 2001). Given that brand stereotypes represent a shortcut to consumers' cognitive evaluations of brands , it is essential to understand how consumers' stereotypical beliefs influence their assessment of perceived brand value, not least because perceived value represents "the imperative focus for both practitioners and researchers" (Gallarza et al., 2011, p. 179). Our study is, to the best of our knowledge, the first to throw light on this issue. In this context, our multidimensional representation of value accounts for the functional benefits offered by the brand (i.e. functional value), captures relevant consumer-brand emotional bonds (i.e. emotional value), and also accounts for any social signaling related to the brand (i.e. social value) (Sheth et al., 1991). Importantly, by taking PBG and PBL into account we are able to highlight the unique contribution of warmth and competence on explaining perceived brand value, over and above the influence of PBG and PBL.
Second, we offer a deeper understanding of the stereotype transfer mechanism by explicitly assessing the relative importance of brand competence and warmth across three perceived value dimensions (i.e. functional, emotional, social). In line with signaling theory (Kirmani & Rao, 2000), we conceptualize brand stereotype competence and warmth as consumer judgements which enable informational transfer relevant for consumers' assessment of perceived value. Specifically, we reveal whether competence (i.e. ability) has a stronger impact on functional value than on emotional and social value and, conversely, whether warmth (i.e. intentions) is more diagnostic for emotional and social value than for functional value. By delineating the value-building roles of warmth and competence, we contribute to the body of literature regarding the diagnosticity of the warmth and competence dimensions.
Third, we provide a reassessment of the roles of PBG and PBL as antecedents of brand stereotype content by clarifying the extent to which PBG and PBL impact both warmth and competence (see Kolbl et al., 2019) or whether competence is influenced only by PBG while warmth only by PBL (see Davvetas & Halkias, 2019). We thus reconcile and formally test two competing views on PBG and PBL as drivers of brand warmth and competence (i.e. the "exclusive" and the "inclusive" approaches) and highlight their relative explanatory power. We do so in an enriched empirical setting comprising both global and local brands, from both a developed and a developing country.
Fourth, we contribute to international marketing literature by revealing the extent to which the observed effectsrelating to both the antecedents and the outcomes of brand stereotypesare stable or contingent on the market setting involved (i.e. developed vs. developing). This has important implications for practitioners regarding how to best leverage the stereotype content dimensions in order to boost consumers' perceptions of brand value and, ultimately, encourage purchase intentions and brand ownership.

Stereotypes and the stereotype content model (SCM)
Stereotyping research has its roots in social cognition and was firstly introduced to characterize different social groups (Fiske et al., 2002). A comprehensive body of research in social psychology has identified two basic stereotyping dimensions, which are able to grasp around 80% of the variance in perceptions of different social behaviors . These two dimensions are warmth and competence Fiske et al., 2002), but have also been labeled as communion and agency (Abele, 2003;Bakan, 1966;Wojciszke & Abele, 2008); morality and competence (Wojciszke, 2005); or as intentions and ability .
The common denominator among the above labels is that warmth (i.e. communion, morality) reflects the other's intentions (Fiske et al., 2002) and "pertains to functioning in social relations" (Wojciszke & Abele, 2008, p. 1139, whereas the competence (i.e. agency, ability) describes the other's ability to enact these intentions (Fiske et al., 2002) and "refers to task functioning and goal achievement" (Wojciszke & Abele, 2008, p. 1139. Warmth is therefore captured with descriptors such as kind, warm, and good-natured, whereas competence includes items such as capable, competent, and efficient. Warmth and competence are the two dimensions of the Stereotype Content Model (SCM; Fiske et al., 2002), the most widely used model in stereotyping research that has been applied to different topics of interest and social groups, for example people, animals, companies, and brands (Cuddy et al., 2008;Sevillano & Fiske, 2016;Aaker, Vohs, & Mogilner, 2010;. A substantial body of social psychology research shows that stereotyping impacts perceptions, evaluations and behavior (e.g. Kunda & Spencer, 2003).
Despite the fact that stereotyping was firstly used as a tool to better understand the phenomenon of social judgement and perception in social psychology, the warmth and competence dimensions have lately also been applied in marketing and consumer research on topics as diverse as country-of-origin (COO; e.g., Chen, Mathur, & Maheswaran, 2014), branding (e.g.,  and environmentally responsible consumption (e.g. Antonetti & Maklan, 2016). In fact, stereotypical associations apply to "every stimulus object that is ascribed to the stereotypical category" (Halkias, Davvetas, & Diamantopoulos, 2016, p. 3642), including brands. The application of a stereotyping perspective to branding is elaborated in the next section.

Brand stereotypes
The seminal work of Fournier (1998) on consumer-brand relationships has laid new foundations on how people relate to brands. Consumers started being recognized "not as a passive object of marketing transactions but as an active, contributing member of the relationship dyad" (Fournier, 1998, p. 344). The consumer-brand relationship bond is enabled by anthropomorphism, which refers to the need, want, or desire to humanize everything, in order to simplify exchanges with the nonmaterial world (Fournier, 1998;Guthrie & Guthrie, 1995). Thus, with the help of anthropomorphism, consumers started seeing brands as their relationship partners (Fournier, 1998).
Based on the idea that people relate to brands similarly as they do to people (Fournier 1998;2009), researchers applied the social perception model from social psychology to brand perception in marketing . The brand's intentions (i.e. warmth) and ability (i.e. competence) are two fundamental dimensions within the BIAF , which is theoretically consistent with the SCM (Fiske et al., 2002), with an adaptation to suit brand perception 2 . For example, most people believe that Audi is "typically perceived as offering good quality and safe vehicles" (Warranty Direct, 2017). This means that the brand is stereotyped as being competent through showing its ability to produce good quality and safe vehicles. However, it is not only the competence that consumers assess when stereotyping a certain brand. The big CO 2 "Dieselgate" scandal in the automotive industry in 2017, where Audi, BMW, Daimler, Volkswagen, and Porsche were involved was seen as a "monstrous antitrust case" (Forbes, 2017) and revealed these brands clearly had bad intentions towards their consumers and, according to the SCM, they would be stereotyped as "cold".

Brand stereotypes vs. brand personality
Brand stereotypes must be conceptually distinguished from the brand personality construct. Brand personality is defined as "the set of human characteristics associated with a brand" (Aaker, 1997, p. 347). Brand stereotypes, on the other hand, represent the consumers' oversimplified and generalized beliefs about brands as intentional agents (Greenwald & Banaji, 1995;. Researchers investigating brand stereotypes recognize that "there are clear links between our brand perception model and Aaker's (1997) brand personality scale" (Kervyn et al., 2012, p. 171) and, indeed, brand personality has been modeled as an antecedent of brand stereotypes (Ivens et al., 2015), meaning that the attributes of brands, which sum up into brand personality, predict the content of brand stereotypes, as reflected in warmth and competence. In what follows, we briefly highlight the points of similarity and difference between these two constructs.
Brand personality and brand stereotypes share the notion of anthropomorphism as a common denominator, since both constructs have been adapted to marketing from (social) psychology through assigning human characteristics to brands. Importantly, however, brand personality and brand stereotypes have a different target focus. As explained by Kervyn et al. (2012, p. 171) "personality scales make sense when focusing on one or on a small number of brands, to provide a more detailed description of their actual attributes. Social perception models on the other hand allow researchers to measure perception of a larger number of social objects, thus creating a whole landscape in which the images of all the relevant objects can be located and compared." Thus, brand personality deals with individual perception and represents the fit between self-versus brand perception, whereas the brand stereotype construct deals with social perception. This view is congruent with that found in psychology, which speaks of differentiation between "personality scales (what a person is) and social perception (how a person seems)" (Kervyn et al., 2012, p. 171). Therefore, brand stereotypes represent a generalized and shared opinion of society and move away from the individual, or subjective perception, which is common to brand personality. Second, while brand personality is associated with human characteristics of the brand, it does not consider brands as intentional agents (i.e. their perceived intentions and ability), which is fundamental to the BIAF (Fournier & Alvarez, 2012).
There is also an important difference in how brand personality and brand stereotypes are measured. When evaluating brand personality, respondents are asked about their individual perception of how well the items represent or describe a certain brand (e.g. Coca Cola is: exciting) (Aaker, 1997). In contrast, when evaluating brand stereotypes, respondents are asked about a common consensus of a society (e.g. Most people in Germany think that Coca Cola is: competent) (Fiske et al., 2002;.

Research model and hypotheses
Our research model showing the study constructs and the hypothesized links between them is summarized in Fig. 1. The focal relationship in our model is the link between the brand stereotype content and perceived value. PBG and PBL are exogenous variables modeled as antecedents to both stereotype content and perceived value dimensions; their function as a common cause also ensures that any observed link between warmth/competence and perceived value is not spurious. Purchase intentions and brand ownership are respectively modeled as direct and indirect consequences of perceived value and represent the behavioral outcomes of interest. Finally, brand familiarity is used as a control variable on the latter. We elaborate on the model relationships and associated hypotheses below.
3.1. PBG and PBL as antecedents to brand stereotype content PBG and PBL represent two separate, yet mutually reinforcing constructs that shape consumer perceptions, attitudes and behaviors (Halkias et al., 2016;Kolbl et al., 2019;Özsomer, 2012;Xie, Batra, & Peng, 2015). PBG is mainly achieved through brand's (a) visual representation, (b) marketing communication, and (c) market presence (Swoboda et al., 2012), and builds on the premise of global consumer culture positioning (GCCP; Alden, Steenkamp, & Batra, 2006). On the other hand, PBL does not necessarily imply that the brand is only present in a single country, but in contrast to PBG, emphasizes the localness aspect of a brand through its symbolic or even iconic role for a local culture (Özsomer, 2012). Domestic, as well as global brands can decide to emphasize different combinations of localness and globalness associations (Riefler, 2012) in order to either depict their closeness with the local community, or to show their strength through the brand's global presence. Indeed, cases where a domestic brand internationalizes into a strong global player, but still emphasizes their origin (i.e. Apple stating being designed in California), or where a foreign global brand connects to a local culture (i.e. Nike promoting sneakers designed by locally-based designers) are becoming increasingly common (Sichtmann, Davvetas, & Diamantopoulos, 2019). Research evidence shows that, in addition to impacting brand quality and prestige (Steenkamp, Batra, & Alden 2003), PBG and PBL boost a brand's identity expressiveness (Xie et al., 2015) and serve as strong attitudinal (Halkias et al., 2016) and brand identification drivers (Sichtmann et al., 2019).
Particularly relevant for current purposes are recent findings which show that PBG and PBL influence brand stereotype content (Davvetas & Halkias, 2019;Kolbl et al., 2019). However, the studies concerned differ with regards to the nature of the influence involved. Specifically, Davvetas and Halkias (2019, p. 12) argue that "the extent to which an individual brand is perceived to be globally available will foster judgements of brand competence while the degree to which an individual brand is perceived to be embedded in the local culture will foster judgements of brand warmth". This implies that PBG and PBL influence different dimensions of brand stereotype content and precludes the presence of cross-links (i.e. from PBG to warmth or from PBL to competence). In contrast, Kolbl et al. (2019) hypothesize and empirically support such cross-links but also highlight that these may depend on the country setting involved (e.g. while the PBL → competence cross-link was consistently significant in both a developed and a developing country context, the PBG → warmth cross-link was only observed in the latter context).
Bearing the above in mind, we revisit the two perspectives in our 2 Warmth in the SCM is equivalent to "intentions" in BIAF, whereas competence in SCM corresponds to "ability" in BIAF.
model and subject them to a formal comparison. Following Kolbl et al.'s (2019) conceptual rationale, we draw from signaling theory (Erdem & Swait, 1998;Kirmani & Rao, 2000) and treat PBG and PBL as signals which help to shape a brand's stereotypical content. Regarding PBG, while we concur with Davvetas and Halkias (2019) that globalness perceptions are more prone to invoke the competence dimension, there is substantial evidence demonstrating that brands perceived as global invoke sophistication, excitement, social approval and belongingness to the global consumer culture (Strizhakova, Coulter, & Price, 2008;Schuiling & Kapferer, 2004), which prompt the brand's good intentions and therefore, to a certain extent, also invoke the warmth dimension. Similarly, regarding PBL, we agree with Davvetas and Halkias (2019) that localness perceptions are expected to especially impact the warmth dimension due to their authenticity, originality and tailor making to consumers' tastes (Ger, 1999;Schuiling & Kapferer, 2004;Van Ittersum & Wong, 2010). However, brands that invoke localness associations are also adapted to local needs and are perceived to be trustworthy (Ger, 1999;Schuiling & Kapferer, 2004), thus, also triggering the competence dimension. In short, both PBG and PBL can be expected to impact stereotypical judgements of competence as well as warmth, although to different extents. We thus hypothesize that: H1. PBG is positively associated with brand stereotype (a) competence, and (b) warmth, but with (c) a stronger influence on competence than on warmth.
H2. PBL is positively associated with brand stereotype (a) competence, and (b) warmth, but with (c) a stronger influence on warmth than on competence.

Brand stereotypes, perceived value and consumer response
Perceived value has been recognized as one of the most important and essential parts of the relationship-marketing paradigm, which focuses on the relationship between consumers and companies/brands (Grönroos, 1994;Payne & Holt, 2001;Ravald & Grönroos, 1996). The importance of the construct for marketing has been demonstrated by intense interest among researchers throughout the years (more than 196 empirical studies until 2013) (Vieira, 2013); by being viewed as a research priority (Marketing Science Institute, 2006); by scholars observing that value creation "must be the reason for the firm's existence and certainly for its success" (Slater, 1997, p. 166); and by recognizing that "the creation of customer value has become a strategic imperative in building and sustaining a competitive advantage" (Gallarza et al., 2011, p. 427 as in Wang, Lo, & Yang, 2004. In line with the theory of consumption values (Sheth, Newman & Gross, 1991), which seeks to explain why consumers buy/use one product/brand over another , consumer choices are regarded as a function of multiple 'consumption value' dimensions (Sheth et al., 1991). Specifically, perceived value represents the assessment of the total worthiness of a purchase, and is comprised of functional, emotional and social dimensions (Sweeney & Soutar, 2001). Functional value, which is traditionally assumed to be the primary driver of consumer choice, is defined as "the perceived utility acquired from an alternative's capacity for functional, utilitarian, or physical performance" (Sheth et al., 1991, p. 160). Emotional value refers to "the perceived utility acquired from an alternative's capacity to arouse feelings or affective states" (Sheth et al., 1991, p. 161); importantly, even tangible and seemingly utilitarian products are frequently associated with emotional responses, one example being luxury fashion apparel brands (Li, Li, & Kambele, 2012). Lastly, social value reflects "the perceived utility acquired from an alternative's association with one or more specific social groups" (Sheth et al., 1991, p. 161).
Consumers rely on different cues emitted by the brand to weight the benefits against the sacrifices and to develop their value perception. These cues need to be credible enough to enable informational transfer and reduce the uncertainty that arises from imperfect and asymmetric information that is available to consumers for most of the products (Erdem, Swait, & Valenzuela, 2006;Spence, 1974). By stereotyping the brand's ability (i.e. competence) and intentions (i.e. warmth), consumers are making cognitive assessments of brands that influence their perceptions and attitudes . Consistent with signaling literature (Erdem & Swait, 1998;Kirmani & Rao, 2000), we thus expect that consumers' assessments of brand stereotype warmth and competence enables informational transfer relevant for the consumer evaluation of potential benefits/sacrifices (i.e. perceived value) since they embody relevant information on brand's ability and intentions that can reduce uncertainty. We hence argue that higher brand competence is seen as advantageous (the brand is able to solve the consumer's problem) thus increasing perceived brand benefits (such as better workmanship) while at the same time reducing the need to consider other brand options, thus decreasing perceived sacrifices (such as search costs). Similarly, higher brand warmth assessment (i.e. the brand is warm and good-natured) can positively impact consumer value perceptions by increasing perceived benefits (such as enjoyment) and decreasing perceived sacrifices (such as social disapproval).
In light of the above, we expect that brand stereotype content will positively impact consumers perceptions of value since favorable consumer judgements on the stereotype dimensions result in a more favorable evaluation of the brand. For example, the very successful global brand of Warby Parker eyeglasses states its mission as "to offer designer eyewear at a revolutionary price, while leading the way for socially conscious businesses" (Warby Parker, 2019). The competence that is signaled by the brand (i.e. offering designer eyewear at revolutionary price) can serve as a driver of functional value, whereas its warmth (i.e. leading the way for socially conscious business) signals the brand's good nature and thus enhances perceived social and emotional value. Bearing in mind, that previous research shows that the dimensions of perceived value are also impacted by a brand's PBG and PBL (Swoboda et al., 2012;Swoboda & Hirschmann, 2016), we propose a direct link between the stereotype dimensions and perceived value while controlling for the impact of PBG/PBL on value. Specifically, we hypothesize that: H3. Brand stereotype (a) competence, and (b) warmth will positively impact consumers' perceived value over and above any impact of PBG and PBL.
To advance explanation on how competence and warmth function as value-enhancers, their relative importance vis-à-vis functional, emotional and social perceived value dimensions needs to be considered. According to the Stereotype Content Model (Fiske et al., 2002; and BIAF , brands differ substantially in how well intentioned and how able they are perceived to be. Brands assessed as highly competent are also viewed as more capable and credible Fiske et al., 2002). For example, the slogan for the Adidas brand ("Impossible is nothing", www.adidas-group.com), reflects competence through the athletic performance that the brand enables. Similarly, the New York Times' slogan "All the news that is fit to print" (www.nytimes.com), emphasizes the credible news that it carries.
Highly competent brands can convey value-for-money benefits which then translate to perceived functional value, without much relation to perceived social or emotional value (Sheth et al., 1991). For example, the Dollar Shave Club's tagline "Shave time. Shave money." (www.dollarshaveclub.com) refers explicitly to functional value only, that is, cost and convenience benefits for consumers. Consequently, the relative influence of brand stereotype competence is expected to differ across perceived value dimensions, eliciting more perceived functional value and trading off perceived emotional and social value (Sheth et al., 1991). Therefore: H4. Brand stereotype competence will have a stronger impact on functional value than on emotional value and social value.
Brands also differ in how warm they are perceived to be Fiske et al., 2002). Brands stereotyped as highly warm are often seen as well-intentioned, friendly and kind (Kervyn, Bergsieker, & Fiske, 2012;. This invokes more favorable affective states as well as stronger sense of belonging to a particular social group; for example, L'Oréal's "Because you're worth it" (www.lorealparisusa.com). Warmth depicts intentions and intentions often signal belongingness and social approval (i.e. offering the promise of joy and relaxation, giving the "ticket" to the certain social circle). For example, the insurance brand Cigna is perceived as high on warmth (Aaker et al., 2012) and communicates cooperative and prosocial cues that increase perceptions of emotional and social value, but not necessarily functional value since the ability and usefulness of the brand are not communicated. Given that consumers trade off among value dimensions for the ones that are most salient, we argue that brands judged high on warmth will arouse positive feelings or associations with desirable social groups, while performance-related aspects will not be as notable. We thus hypothesize that: H5. Brand stereotype warmth will have a stronger impact on emotional and social value than on functional value.
Turning attention to the behavioral outcomes shown in Fig. 1, prior research indicates that perceptions of value directly influence consumers' willingness to buy (e.g. Dodds, Monroe, & Grewal, 1991); thus, the positive relationship between perceived value and purchase intentions is well recognized in previous studies (e.g. Chen & Chang, 2012;Johnson, Herman, & Huber, 2006;Sánchez-Fernández & Iniesta-Bonillo, 2007). Following the logic of the increase in perceived benefits and decrease in perceived costs, higher functional, emotional and social value should lead to stronger purchase intentions, and, ultimately, result in higher brand ownership (Naseem, Verma, & Yaprak, 2015; Salehzadeh & Pool, 2017); we assess these previously established relationships in our research model, but do not formally hypothesize them.
To control for potential confounding effects on our results, we also include brand familiarity in our model as a control variable. While brand familiarity is not of theoretical interest in our study, there are at least three reasons why it might impact consumer outcomes: (1) it helps overcoming the "fear of unknown", (2) it makes consumers more tolerant toward a brand's weaknesses, and (3) it contributes towards reducing the intellectual effort when making choices (Diamantopoulos, Schlegelmilch, & Palihawadana, 2011). Furthermore, prior similar studies have revealed that brand familiarity positively influences purchase intentions (Davvetas & Halkias, 2019;Kolbl et al., 2019).

Research design
We conducted our first study in Slovenia, which belongs to the developed group of countries. With a GDP per capita close to $36,500 (World Bank, 2017), Slovenia ranks among the top 25 countries in terms of the KOF index of globalization, which consists of economic, political and social dimensions of globalization (ETH, 2019). With regards to the Economic Complexity Index (ECI), which measures the relative knowledge intensity of an economy, it is ranked among the 15 most complex economies (OEC, 2019). Consumers in Slovenia are thus exposed to a plethora of brands, which target them with different globalness and localness cues evoking a variety of stereotypical perceptions (Davvetas & Halkias, 2019). Variation in latter is particularly relevant for our investigation, since we are explicitly interested in how the content of brand stereotypes impacts different types of perceived value (see H3 -H5).
The questionnaire was initially developed in English, but was later translated into the official language of Slovenia. We opted for the collaborative approach suggested by Douglas and Craig (2007), rather than simple back-translation (Brislin, 1970), whereby particular attention was devoted to ensuring the equivalence in meaning by involving bilingual local researchers. Specifically, we followed Harkness's (2003) five-stage procedure, involving (a) the independent translation of the questionnaire by two local researchers (native speakers), (b) the review of the translated questionnaires by the entire research team, (c) the adjudication of any differences and preparation of a single questionnaire version, (d) the pretesting of the latter on 20 consumers in Slovenia, and (e) the documentation of the translation procedures as described by steps (a) -(c) above.
Following pretesting, the final questionnaire was disseminated online, using a snowball sampling approach (Fricker, 2008). Participants were invited to answer the survey by research collaborators located in Slovenia, who distributed the link online via different channels, such as universities' learning platforms. Respondents (N = 203, 148 female, M age = 28.65, SD = 10.97) participated in a between-subjects survey design and were randomly exposed to one of fourteen brands from different product categories. We first sought to include product categories which were previously used in studies related to brand stereotypes (e.g. soft drinks, beer, clothing, snacks, sweetsplease see Appendix A in the manuscript). Within these categories, we then identified global brands available for sale in both countries, as well as local brands offered in each of the two countries of investigation. Regarding global brands, we narrowed our choice down to seven brands: Coca-Cola, Ebay, Heineken, IKEA, KITKAT, Pringles, North Face. For the selection of local brands, we engaged local experts to select matching (categorywise) local brands for each country, given that they are most familiar with the local marketplace situation and thus represent a credible source of information regarding the selected brands' competitiveness, familiarity and popularity. We further used a pre-test to examine how the selected brands score on PBG and PBL. The assessment was in line with expectations (higher scores on PBG for global brands and higher scores on PBL for local brands).
Note also that, consistent with relevant literature , different combinations of warmth and competence are captured by our stimuli brands. As shown in Appendix B, while some brands score consistently high or low on both stereotype dimensions, other brands exhibit ambivalent (mixed) stereotypes (i.e. they score high on one dimension but low on the other).
Most model constructs were measured through validated multi-item scales drawn from prior research (PBG: Steenkamp et al., 2003 Table 1 for full details on construct measurement). Brand familiarity was measured with the item "How familiar would you say you are with Brand X" (anchored at 1 = not at all familiar to 7 = very familiar), while brand ownership was measured with a binary (yes/no) variable relating to the question "Have you personally bought Brand X in the past 12 months?". We further used gender as an additional control variable, due to the unbalanced sample of male and female respondents.

Analysis and results
Covariance-based structural equation modeling (SEM) in LISREL 8.8 was used to test our conceptual framework, following a two-step approach (Anderson & Gerbing, 1988) whereby the measurement model was examined first and followed by the assessment of the structural model used to test the hypothesized relationships. Confirmatory factor analysis (CFA) resulted in a good overall fit (χ 2 = 742.53, df = 406, RMSEA = 0.06; NNFI = 0.96; CFI = 0.97). Standardized item loadings  Table 1). Moreover, average variance extracted (AVE) ranged from 0.61 to 0.76 and all AVEs exceeded the corresponding squared inter-construct correlations (i.e. shared variances), thus establishing discriminant validity (Fornell & Larcker, 1981) (see Table 2). Common method bias (CMB) was assessed by applying a variation of the marker variable test proposed by Malhotra, Kim, and Patil (2006). We used the second lowest positive correlation (=0.034) between the observed variables (i.e. indicators) as a proxy for CMB and adjusted the zero-order correlations between all items by subtracting this value from the original estimates. Ninety percent of the adjusted correlations remained significant (total number of significant correlations before vs. after the adjustment: 428 vs. 386) thus revealing no major threat due to CMB. We further assessed CMB by including a common latent factor (i.e. method factor) to the CFA model with equal loadings on all indicators and compared the resulting model fit (χ 2 = 726; df = 405; RMSEA = 0.06; NNFI = 0.97; CFI = 0.97) with that of the CFA model without the method factor (see above). In light of the minor differences in fit, CMB does not seem to pose a problem in our study.
We next proceeded to the estimation of the structural model. The latter also revealed a good overall fit (χ 2 = 899.45; df = 497; RMSEA = 0.06; NNFI = 0.96; CFI = 0.96). The relevant parameter estimates are summarized in Table 3.
As predicted by H1a and H1b, PBG is positively and significantly related to both brand stereotype competence (β = 0.52, p < 0.001) as well as brand stereotype warmth (β = 0.37, p < 0.001). To test for differences in the impact of PBG on competence versus warmth, we compared the original model specification to a model with an equality constraint, whereby the paths from PBG to competence and PBG to warmth were set to be equal. A chi-square difference test produced a non-significant result (Δχ 2 = 2.43, df = 1, n.s.), indicating that the effects of PBG on brand stereotype competence and warmth do not differ. Therefore, contrary as hypothesized, PBG does not have a stronger impact on brand stereotype competence than on brand stereotype warmth; thus, H1c is not supported.
When it comes to PBL, the results are also consistent with H2a and H2b, since PBL is positively and significantly related to brand stereotype competence (β = 0.26, p < 0.001) as well as to brand stereotype warmth (β = 0.42, p < 0.001). Following the above-mentioned procedure of equality constraints, we again compared the original model specification to a model with equal paths from PBL to competence and PBL to warmth. A chi-square difference test failed to produce a significant result (Δχ 2 = 1.44, df = 1, n.s.), indicating that the effects of PBL on both dimensions of brand stereotype content are of a similar magnitude. Thus, H2c is not supported either.
It will be recalled that H1 and H2 were formulated under the premise that PBG and PBL impact both dimensions of the brand stereotype content (reflecting Kolbl et al.'s (2019) "inclusive" approach). To provide a formal test of this specification against the "exclusive" approach suggested by Davvetas and Halkias (2019), we compared our hypothesized model against an alternative model in which the paths from PBG to warmth and PBL to competence were set to zero. In light of the nested nature of the two models, we were able to compare their fit (by means of a Δχ 2 test) as well as the proportion of variance explained in the two brand stereotype dimensions (see Table 4).
Our results on the relationship between PBG/PBL and perceived value largely corroborate the findings of previous research in a retailing context (Swoboda et al., 2012). More specifically, PBG is positively and significantly related to functional value (β = 0.22, p < 0.01), emotional value (β = 0.28, p < 0.001), but not to social value (β = 0.15, n.s.). Similarly, PBL is positively and significantly related to functional value (β = 0.20, p < 0.01), emotional value (β = 0.36, p < 0.001) but also to social value (β = 0.43, p < 0.001). Importantly, and in line with H3, brand stereotypes have a positive and significant effect on perceived value over and above the effects of PBG/PBL. Specifically, brand warmth is positively and significantly related to functional value (β = 0.26, p < 0.01), emotional value (β = 0.19, p < 0.05) and social value (β = 0.16, p < 0.05). Competence, on the other hand is positively related to functional value (β = 0.17, p < 0.05), but not to emotional or social value. Our model explains a relatively high proportion of variance in all three value dimensions (functional value: R 2 = 0.30; emotional value: R 2 = 0.29; social value R 2 = 0.30), indicating large effect sizes (Cohen, 1988) Turning to the relative effects of brand stereotypes on perceived Notes: Squared root AVEs are shown on diagonals in bold, Correlations are shown below the diagonal. *Correlation is significant at the 0.05 level (two-tailed); **Correlation is significant at the 0.01 level (two-tailed).
To assess the relative impact of brand stereotype warmth on the dimensions of perceived value (see H5), we used a similar approach and compared our original model specification with several constrained models. First, we first set the paths from brand stereotype warmth to functional-and emotional value to be equal, which resulted in a nonsignificant chi-square difference test (Δχ 2 = 1.37, df = 1, n.s.). We then compared the paths from brand stereotype warmth to functional and social value; this also resulted in a non-significant chi-square difference test (Δχ 2 = 1.64, df = 1, n.s.). Finally, we compared the paths from brand stereotype warmth to emotional and social value, which again produced a non-significant chi-square difference test (Δχ 2 = 0.25, df = 1, n.s.). Taken together, these results show that the effects of brand stereotype warmth on functional, emotional, and social value are of a similar magnitude (within sampling error). Thus, H5 is not supported.
Our results further show that perceived functional and emotional value are positively and significantly related to purchase intentions (functional value: β = 0.14, p < 0.05; emotional value: β = 0.63, p < 0.001), while social value is not. When constraining the paths from functional value to purchase intentions and from emotional value to purchase intentions to be equal, a chi-square difference test failed to produce a significant result (Δχ 2 = 0.28, df = 1, n.s.); thus, the effects of functional and emotional value on purchase intentions are of a similar magnitude.
Finally, the relative importance of the brand stereotype dimensions was also examined through their total effects on consumer responses. While for brand stereotype competence, non-significant total effects on both purchase intentions (total effect = 0.030) and brand ownership (total effect = 0.015) are observed, the overall effect of brand warmth on purchase intentions is positive and significant (total effect = 0.139, p < 0.01) and the same applies to brand ownership (total Notes: *** -p < 0.001; ** -p < 0.01; *p < 0.05 (one-tailed)

Research design
Previous research suggests that consumers in developed/mature markets may differ in their perceptions of and responses towards global and local brands from consumers in developing/emerging markets (Batra, Ramaswamy, Alden, Steenkamp, & Ramachander, 2000;Schuiling & Kapferer, 2004;Steenkamp & de Jong, 2010). Consequently, the role of PBG and PBL in shaping brand stereotype content as well as perceptions of value may also vary across developed versus developing country settings. Given that the vast majority of extant studies on brand stereotypes (see Appendix A) has been conducted in developed countries, we sought to replicate and test the stability of our research model in a less developed country than Slovenia, namely Bosnia and Herzegovina. In doing so, we followed the approach of prior research, which stresses the importance to assess the stability of models involving global and local brands in developed and developing market contexts (e.g. Kolbl et al., 2019;Sichtmann & Diamantopoulos, 2013). Bosnia and Herzegovina has GDP per capita close to $13,000 (World Bank, 2017), and is ranked between 60th and 70th in terms of the KOF index of globalization (ETH, 2019) and between 40th and 50th in terms of the ECI (OEC, 2019).
Following the same approach of stimuli selection as in Study 1, we ended up with fourteen brands (evenly split between global and local brands). Again, selection of local brands within a given product category was done with the help of local research experts (see Appendix B for a mapping of the chosen brands on the warmth and competence dimensions). We also aimed for consistency in construct measurement, questionnaire translation and distribution, and assessment of CMB by following the same procedures as in Study 1. Respondents (N = 192, 120 female, M age = 24.93, SD = 6.98) were invited to participate in the study with the help of locally-based research collaborators who distributed the survey link online. As in Study 1, we also controlled for gender, due to the unbalanced sample of male and female respondents.

Analysis and results
In order to ensure that respondents in both our studies interpreted the construct measures in the same way, we conducted measurement invariance tests using multi-group CFA (Steenkamp & Baumgartner, 1998). We started by assessing configural invariance focusing on the (cross-national) dimensionality of the focal constructs. Next, we assessed metric invariance by testing whether the factor loadings for each scale item across countries are the same (within sampling error). We then further tested for error variance invariance, as well as factor variance/covariance invariance, by respectively constraining error variances and factor variances/covariances to be equal across the two countries.
We first estimated a multi-group CFA model without any constraints on the parameters across the two countries. The results showed that the model fits the data well (Global χ 2 = 1588.29; df = 812; RMSEA = 0.07; NNFI = 0.97; CFI = 0.97) thus establishing configural invariance. We next tested for metric invariance by constraining all the factor loadings to be equal between the two countries. This also resulted in a good fit (Global χ 2 = 1614.50; df = 835; RMSEA = 0.07; NNFI = 0.97; CFI = 0.97). Given that the second model is hierarchically nested in the baseline model (Murray, Gao, Kotabe, & Zhou, 2007), we conducted a chi-square difference test to assess which model fits the data better. The result of the test was not significant (Δχ 2 = 26.21; Δdf = 23, n.s.) thus supporting full metric invariance. When testing for error variance invariance, model fit was acceptable (Global χ 2 = 1833.547; df = 866; RMSEA = 0.08; NNFI = 0.97; CFI = 0.97), however, there was a significant deterioration in fit when compared to the metric invariance model (Δχ 2 = 218.97; Δdf = 31, p < 0.001). A similar result was obtained when testing for factor variance and factor covariance invariance; the model fit was acceptable (Global χ 2 = 1691.58; df = 871; RMSEA = 0.07; NNFI = 0.97; CFI = 0.97), however, there was a significant deterioration in model fit compared to the metric invariance model (Δχ 2 = 77.53; Δdf = 36, p < 0.001). Overall, we can conclude that full metric invariance was established, meaning that the same pattern of factor loading exists across the two countries and that these factor loadings are identical for each scale item across both countries (Murray et al., 2007).
To test the measurement and structural models in Study 2, we employed the same analytical procedures as in Study 1. A confirmatory factor analysis (CFA) resulted in a good fit (χ 2 = 845.76, df = 406, RMSEA = 0.08; NNFI = 0.97; CFI = 0.98; see Table 1), with standardized item loadings across constructs ranging from 0.73 to 0.97 and composite reliabilities from 0.88 to 0.98 (see Table 1). AVE ranged from 0.65 to 0.90 and all AVEs were larger than the corresponding squared inter-construct correlations (see Table 2). Therefore, discriminant validity was established (Fornell & Larcker, 1981).
As in Study 1, we conducted a CMB analysis using Malhotra, Kim, and Patil's (2006) procedure. After identifying the second lowest correlation (=0.007) and adjusting the zero-order correlations between items, 99% of the correlations remained significant (total number of significant correlations before vs. after the adjustment: 444 vs. 443). Adding a method factor and re-estimating the CFA model only resulted in a minor improvement in fit (χ 2 = 829.85, df = 405, RMSEA = 0.07, NNFI = 0.97; CFI = 0.98). Thus, CMB does not raise any material concerns in Study 2 either.
We then proceeded to the estimation of the structural model, which fitted the data well (χ 2 = 1012.81, df = 497, RMSEA = 0.07, NNFI = 0.97; CFI = 0.97); Table 3 shows the relevant parameter estimates. PBG is positively and significantly related to brand stereotype competence (β = 0.32, p < 0.001), while it is not significant in terms of brand warmth (β = 0.08, n.s.). Hence, H1a is supported, while H1b is not. H1c is also supported (Δχ 2 = 5.82, Δdf = 1, p < 0.05) indicating a stronger impact of PBG on competence than warmth. Regarding PBL, the latter is positively and significantly related to brand stereotype competence (β = 0.49, p < 0.001) as well as to brand stereotype warmth (β = 0.59, p < 0.001) thus offering support for both H2a and H2b. However, when setting the paths from PBL to brand competence and from PBL to brand warmth to be equal, a chi-square difference test failed to produce a significant result (Δχ 2 = 0.47, Δdf = 1, n.s.), indicating that the effects of PBL on brand competence and warmth are of a similar magnitude. Thus, H2c is not supported.
We used the same approach as in Study 1 to compare the "inclusive" and "exclusive" approaches regarding the links between PBG/PBL and the two brand stereotype dimensions (Table 4). Consistent with Study 1, the results of the comparison of the two models indicates a significantly better fit for our hypothesized model (reflecting the "inclusive" approach) as compared to the alternative model (reflecting the "exclusive" approach) (Δχ 2 = 32.65, Δdf = 2, p < 0.05). Again, the proportion of the variance explained by the hypothesized model is higher than that explained by the alternative model for both the warmth (R 2 = 0.34 vs. 0.31) and the competence dimensions (R 2 = 0.28 vs. 0.05).
Focusing on the relationship between brand stereotypes and perceived valuewhile controlling for the influence of PBG/PBLbrand competence is positively and significantly related to functional value (β = 0.17, p < 0.05) as well as to social value (β = 0.24, p < 0.01). In contrast, brand warmth is positively and significantly related to functional value (β = 0.15, p < 0.05) and emotional value (β = 0.19, p < 0.01). Based on these findings, we can conclude that H3 is partially supported. Consistent with Study 1, our model in Study 2 explains relatively high proportions of variance in all value dimensions indicating large effect sizes even after controlling for the impact of PBG and PBL (functional value: R 2 = 0.45; emotional value: R 2 = 0.39; social value: R 2 = 0.47).
To assess the relative impact of brand competence on social value versus functional value, we again used the method of equality constraints. A chi-square difference test failed to produce a significant result (Δχ 2 = 1.75, Δdf = 1, n.s.), indicating a similar magnitude of the relevant effects. Thus, H4 is not supported. When setting the paths from brand warmth to functional value and from brand warmth to emotional value to be equal, a chi-square difference test failed to produce a significant result (Δχ 2 = 1.90, Δdf = 1, n.s.), indicating that the effects of brand warmth are of a similar magnitude on functional and emotional value. Thus, H5 is not supported either.
With regards to the perceived value outcomes, the results mirror exactly the findings in Study 1. Perceived functional and emotional value are positively and significantly related to purchase intentions (functional value; β = 0.29, p < 0.001; emotional value; β = 0.45, p < 0.001), while social value is not. When assessing the strength of the impact of functional and emotional value on purchase intentions, the chi-square difference test turns out to be non-significant (Δχ 2 = 0.63, Δdf = 1, n.s.), thus indicating the similarity of their effects on purchase intentions.
As in Study 1, we evaluated the total effects of brand stereotype warmth and competence on consumer responses. Results are comparable. For brand stereotype competence, the total effect on both, purchase intentions (total effect = 0.068), and brand ownership (total effect = 0.030) is low and non-significant, while brand warmth's effect on purchase intentions (total effect = 0.130, p < 0.01) as well as brand ownership (total effect = 0.057, p < 0.05) is positive and significant 5 . Thus, our results once again demonstrate that brand stereotype warmth is more predictive of consumer responses than brand stereotype competence.

General discussion
While an increasing number of studies have drawn on the SCM (Fiske et al., 2002) and BIAF  and applied a stereotyping perspective to brands (see Appendix A), extant research has failed to investigate how brand stereotype content impacts consumers' perceptions of value. Our investigation empirically addresses this gap in both a developed and a developing country setting by linking consumer judgements of brand competence and warmth to perceptions of functional, emotional and social value and, through them, to important behavioral outcomes (namely, purchase intentions and brand ownership). Importantly, our research model examines the aforementioned linkages by taking consumers' perceptions of brand globalness and localness into account, both of which have been previously shown to impact brand stereotype content as well as consumers' value perceptions (Davvetas & Halkias, 2019;Kolbl et al., 2019;Swoboda et al., 2012). Our findings have several important implications for both theory and practice as discussed below.

Theoretical implications
The first theoretical implication of our research to branding literature lies in establishing a positive link between brand stereotype content and consumer perceived value. In contrast to a large body of prior research on perceived value (more than 196 empirical studies until 2013; Vieira, 2013), our research is the first to reveal how consumers' assessments of brand stereotype content, as reflected in warmth and competence, influences their perceptions of value. Consistent with signaling theory, according to which brand cues/signals reduce the information asymmetry between buyers and sellers and enable the assessment of value by consumers (Spence, 2002), we show that brand competence and warmth have information transfer potential and enhance consumers' perceptions of brand value. Importantly, the brand stereotype effect holds over and above the effect of PBG/PBL on perceived value, implying that warmth and competence have distinct diagnostic capabilities. Hence, while PBG/PBL help translate information about the properties of brands both for the formation of brand stereotypes (Davvetas & Halkias, 2019;Kolbl et al., 2019) and for consumers' brand value perceptions (Swoboda et al., 2012), activated competence and warmth stereotypes serve as additional value enhancers in both developed and developing country settings.
A second theoretical implication relates to the relative diagnosticity of the brand stereotype dimensions in impacting different dimensions of perceived value. Brand competence enhances functional value, but not emotional value; this applies to both a developed and a developing market setting. This implies that brands that are stereotyped for their ability (competence) communicate utilitarian benefits and capacity for functional performance to consumers, but fail to convey benefits relating to the pleasure or joy from owning and/or using the brands.
In contrast, brand warmth consistently positively impacts both perceived functional and perceived emotional value. This implies that brands judged as warm and hence well-intended not only increase (decrease) consumer perceptions of emotional benefits (costs) associated with the focal brand but also communicate functional benefits (e.g. better quality) related to the latter. This pattern of findings applies in both a developed and a developing market setting and highlights the importance of warmth as a stereotype dimension. This importance is further underscored by considering the overall effects of warmth versus competence on the outcome variables, namely purchase intentions and brand ownership. Whereas the total effect of competence on both these outcome variables is not significant, the corresponding total effects of warmth are positive and significant. This discrepancy in total effects can be attributed to the fact that, in addition, to influencing purchase intentions (and, ultimately, brand ownership) through enhancing functional value, there is a second pathway through which warmth exerts its influence on outcomesnamely, through increasing emotional value. This second pathway is not applicable to competence, as competence has no impact on emotional vale (see above). Overall, the picture painted is consistent with previous findings on the crucial role of the brand warmth dimension in impacting such outcomes as consumer-brand identification (Kolbl et al., 2019), brand affect (Davvetas & Halkias, 2019), and purchase intentions .
warmth in a developed country setting and to brand competence in a developing country setting, none of these effects transfer over to purchase intentions and brand ownership. The reason for this is that social value is not significantly related to the outcome variables in either setting. Whether this rather surprising finding reflects sample specificities or is indicative of social value's limited predictive ability in general, is something that only future research can answer. A third theoretical implication of our research relates to PBG's and PBL's roles as antecedents of brand stereotypes content. We extend prior knowledge by formally testing and comparing two contrasting views in the literature in empirical settings involving both global and local brands in a developed and a developing market setting. In both settings and in line with prior research, PBG is positively and consistently related to brand stereotype competence (Davvetas & Halkias, 2019;Kolbl et al., 2019). This implies that brands, which score high on PBG, are sending a signal of ability to consumers and are therefore being stereotyped as competent. This finding is aligned with previous stereotyping research which finds that quality, reliability, and consistency speak about a brand's competence . Similarly, in both country settings and, again, consistent with prior literature (Davvetas & Halkias, 2019;Kolbl et al., 2019), PBL encourages stereotypical perceptions of brand warmth through associations of uniqueness and originality (Özsomer, 2012).
Perhaps the most interesting aspect of what drives brand stereotype content is the uncovering of cross-links between PBG and PBL on the one hand and the brand stereotype content dimensions on the other. One of these cross-links was observed in both country settings and relates to the transfer of localness associations to stereotypical perceptions of brand competence: brands scoring high on PBL are not only seen as being warm (see earlier) but also as more competent. This corroborates Kolbl et al.'s (2019, p. 619) finding that "global brands emphasizing localness cuessuch as associations of uniqueness, originality and national pride (Dimofte, Johansson, & Ronkainen, 2008;Özsomer, 2012) are judged to be both warm and competent, irrespective of the market setting involved" and also extends it to brands in general. 6 The second cross-link was observed in the developed country only and related to the transfer of globalness associations to stereotypical perceptions of brand warmth: brands scoring high on PBG are not only judged as being competent (see earlier) but also as being warm. That this link was found to apply in a developed country setting stands in direct contrast to previous research showing that "while in a developed country context, PBG only enhances stereotypical assessments of brand competence, in a developing country context PBG also impacts brand warmth" (Kolbl et al., 2019, p. 619). Again, whether this discrepancy is due to differences in the country pairs and/or specific brands investigated by Kolbl et al. (2019) and those examined in our own studies or whether this is a rather idiosyncratic (and thus inherently unstable link) is an issue open to future research. 7 Finally, it is worth noting that, in both countries the relative importance of perceived functional value and perceived emotional value as predictors of purchase intentions (and, through them, on brand ownership) is the same. This suggests that consumers in both developed and developing countries are similarly affected by these value dimensions when formulating their behavioral intentions (with clear implications for practitionerssee next section).

Managerial implications
This research provides several managerial implications for brand managers in developing, as well as in developed markets. The positive link between brand stereotype content and consumer perceived value (over and above the effects of PBG and PBL) suggests that developing brand communications aimed at triggering brand stereotypes is a promising strategy for enhancing perceived value to customers, regardless of the country setting involved. In this context, managers should recognize that brand competence and warmth are not equally diagnostic in influencing different dimensions of perceived value. Brands that are stereotyped as competent are perceived as valuable in the functional sense, while brands that are stereotyped as warm enhance both functional and emotional value perceptions.
Brand managers that aim to establish stronger functional value of their brands could choose to focus either on brand warmth or on brand competence, or to emphasize both in their communications and market presence. Emphasizing brand stereotype competence, implies that advertising materials, or product packaging should highlight the brand's ability, through depicting it as capable, efficient and competent . Ways of emphasizing the competence aspect include the use of competence-oriented slogans (i.e. Audi: "Advancement Through Technology") and the communication of long-term market presence (i.e. Heineken: "Established in 1873"). If a company aims at enhancing the perception of the emotional value of its brand, our findings recommend focusing mostly on emphasizing the brand warmth stereotype, meaning that the brand communications should be portraying brand with good intentions (i.e. being nice, warm and kind). This recommendation holds for both developed and developing countries. Importantly, emphasizing warmth also enhances functional value thus supporting communication strategies aimed at highlighting the brand's competence. Thus, whereas following a strategy exclusively focused on highlighting the brand's competence will only impact consumers' assessments of functional value, a strategy exclusively focusing on warmth will bring benefits both in terms of enhanced functional as well as emotional value. In short, from a managerial point of view, brand warmth seems to be the more important stereotype content dimension 8 . Having said that, to the extent that it is feasible, managers would be well advised to highlight both stereotype content dimensions in their communications in order to maximize their beneficial impact on consumers' perceptions of value.
In light of the role of warmth and competence in enhancing consumers' value perceptions, how can managers influence these stereotype dimensions? An answer to this question is provided by our findings on the role of PBL as a driver of stereotype content. Regardless of the country setting, PBL is relevant for both, brand competence and warmth. Brands, which decide to target their consumers with the localness cues, are therefore very likely to stimulate consumers' associations of brand competence and brand warmth. Managers who aim to develop strategies that fit to the local markets should emphasize the brand's local embeddedness, its tailor-making to consumers' wants and needs and use communication that appreciates and builds on local symbols. Therefore, suggestion for managers who seek to portray their brands as high on both warmth and competence and, thus, achieve the status of "popular brands" , is to strengthen their localness associations.
Leveraging PBG, on the other hand, is likely to be trickier particularly in a developing country setting where it has no impact on brand warmth. While stimulating PBG (e.g. through adopting global consumer culture positioning strategiessee Alden, Steenkamp, & Batra, 1999) will boost competence judgements in both developed and developing country settings, its beneficial impact on warmth is only realizable in the former setting. This is somewhat problematic becauseas earlier discussedit is warmth that it is the more important dimension of brand stereotype content in terms of its influence on value perceptions and, ultimately, on behavioral outcomes. Thus, it is the antecedents of brand warmth that are of particular interest to practitioners and, according to our findings, leveraging PBG is unlikely to enhance warmth judgements among consumers in a developing country. Finally, from a practitioner's perspective, a rather surprising finding is that whereasas expectedboth functional and emotional value results in increased purchase intentions and higher brand ownership, social value does not seem to impact these outcomes. Having said that, it might well be the case that social value is more predictive of other, perhaps "softer" outcomes, such as consumer-brand identification, various attitudes towards the brand, and/or social identity (e.g. Stokburger-Sauer et al., 2012). There is recent anecdotal evidence that social influence is not a sufficient condition for the behavioral activation itself. Namely, there is a publicized case of an Instagram influencer who failed to sell even a minimum quantity of t-shirts to her audience of 2 million followers (Evans, 2019). This by no means implies that the social value perceptions are not relevant for consumer behavioral responses; it is rather interesting that social value perceptions do not transfer to purchase intentions directly. Thus, more research is needed to better understand the transfer mechanism of social value perceptions on behavioral outcomes, and the conditions that encourage/discourage such transfer.

Limitations and future research
This study is, to the best of our knowledge, the first to investigate how brand stereotype content influences different dimensions of perceived brand value. Thus, further studies are needed to validate our findings in both developed and developing country contexts and reconcile results that are at odds with previous research (e.g. on the role of PBG in influencing brand warmth). Our study focused exclusively on perceived value as the core construct of interest. While undoubtedly important, value is not the only criterion that affects consumers' purchase decisions. Thus, an important issue for future research is the link between brand stereotype content and perceived risk, the latter being a very important construct in explaining consumers' product choice, negative word of mouth, regret, lower repurchase intentions and higher brand switching intentions (Mitchell, 1999;Tsiros & Mittal, 2000;Zeelenberg & Pieters, 2007). Therefore, future studies should also account for consumers' risk associated with making a wrong decision. Such risk is likely to be particularly high when brands are judged as "cold" and "incompetent" and it would be instructive to see how risk assessments are influenced by brand stereotype content (and whether such influence is stronger/weaker than on value assessments).
Our research revisited the impact of PBG and PBL on brand stereotype content (Davvetas & Halkias, 2019;Kolbl et al., 2019) and also provided insights on possible cross-links. However, the way in which PBG and PBL influence the brand stereotype content may be (at least partly) culturally driven, due to differences in cultural values (Steenkamp & de Jong, 2010). Taking cultural differences into account would improve our understanding of the functioning of PBG and PBL as antecedents of warmth and competence perceptions and also generate more fine-grained recommendations on how to leverage these antecedents in different settings.
Furthermore, attention should be drawn to the role of perceived social value, which turned out to be non-diagnostic of purchase intentions. We therefore encourage future researchers to identify the path through which social value is linked to consumer behavioral responses. Finally, a limitation of the current study is its focus on main effects only and lack of consideration of potential conditioning factors. Future research should therefore seek to refine the insights provided by the main effects by identifying relevant moderator influences in the link between brand stereotype content and perceived value.  (2019) is not included in the review, because it focuses on politicians as brands, which is not in central attention of our study.

Appendix B. Mapping of stimuli brands on the warmth and competence dimensions
Brand stereotypes for Study 1 (Slovenia) Brand stereotypes for Study 2 (Bosnia and Herzegovina)