The Interplay of Brand, Brand Origin and Brand User Stereotypes in Forming Value Perceptions

Although the importance of various brand-related stereotypes held by consumers has been acknowledged in the marketing literature, their interplay has not yet been considered. We derive four alternative theoretical models of the relationships between brand, brand origin and brand user stereotypes and their effects on consumers’ value perceptions. The four models are subsequently tested on two nationally representative samples in Germany and Serbia, respectively. The best-performing model in both studies predicts that brand origin and brand user stereotypes act as complementary drivers of the brand stereotype. In turn, value perceptions mediate the positive impact of the brand stereotype on purchase intent. This investigation is the first to: (a) juxtapose and empirically assess alternative theoret- ical perspectives on the interplay of brand-related stereotypes; (b) reveal the combined influence of multiple stereotypes on consumers’ value perceptions and through them, on intentions to buy the brand; (c) determine the relative influence of distinct stereotype content dimensions (i.e. warmth vs. competence) on consumers’ perceptions of value; and (d) highlight the extent to which the interplay among different stereotypes in influencing consumer outcomes is stable across different market settings. Implications of the findings are considered and future research directions identified.


Introduction
Three parallel streams of literature have emerged showing that consumer evaluations of and behaviour towards brands are influenced by stereotypical perceptions of brands (Aaker, Vohs and Mogilner, 2010;Kervyn, Fiske and Malone, 2012), brands' country of origin (Halkias and Diamantopoulos, 2020;Magnusson, Westjohn and Sirianni, 2019) and users of brands (Antonetti and Maklan, 2016;Ziano and Pandelaere, 2018), The authors thank the Austrian Science Fund (FWF; Project No. I 3727-G27) and the Slovenian Research Agency (Grant N5-0084) for supporting this research. respectively. Brand stereotypes reflect consumers' shared beliefs about brands as intentional agents (Kervyn, Fiske and Malone, 2012) and are based on the notion that individuals can relate to brands similarly as they relate to people and, therefore, perceive brands similarly as they perceive humans (Fournier, 1998). Country stereotypes refer to oversimplified beliefs about traits, intentions and behaviours that are characteristic of people living in a country, which are formed through direct or indirect experience (Chattalas and Takada, 2008;Maheswaran, 1994;Samiee, 1994). They not only reflect cognitions of single individuals, but are shared representations of the stereotyped entity within a society (Stangor and Schaller, 1996), 2 P. Gidaković et al. expanded to all members of the stereotyped entity when the respective categorization is applied (Bodenhausen, Kang and Peery, 2012). Brand user stereotypes capture the shared beliefs about individuals or social groups perceived to be users of a brand (Antonetti and Maklan, 2016).
Each of these stereotypes has been linked -in separate studies -to managerially relevant outcomes. For example, stereotypical perceptions of brands have been linked to consumers' willingness to buy (e.g. Aaker, Vohs and Mogilner, 2010), firm reputation (e.g. Shea and Hawn, 2019), brand loyalty and intention to switch (e.g. Davvetas and Halkias, 2019); stereotypical perceptions of brands' origin have been found to influence brand affect (e.g. Halkias and Diamantopoulos, 2020) and customer-based brand equity (e.g. Magnusson, Westjohn and Sirianni, 2019); and stereotypical perceptions of brand users have been related to social imitation (e.g. Antonetti and Maklan, 2016) and negative word-of-mouth (e.g. Hickman and Ward, 2013). Just one recent study has made a first attempt to explore the link between stereotypical perceptions of brands and their origin, suggesting stereotype content transfer from brands' country of origin to brand stereotypes (Diamantopoulos et al., 2021). A summary of extant research is offered in the Appendix (online Supporting Information).
While these research streams have provided valuable insights into how each of the three brandrelated stereotypes affects relevant outcomes, they have been silent (with the exception of Diamantopoulos et al., 2021) on whether and how these stereotypes relate to one another and whether they jointly shape consumers' brand evaluations and subsequent behaviour. This is surprising since marketers often depict brands, their origin and their users simultaneously in brand communications in an attempt to purposefully stimulate consumers' stereotypical perceptions via utilizing stimuli that activate favourable associations in consumers' minds. Consider the advertisements in Figure 1. Each ad features the brand, its origin and its user(s). For example, in the Volvo ad, marketers' reliance on the brand's favourable country of origin stereotype is evident, as this information is placed eye-catchingly in the middle of the print ad. Marketers may directly co-create stereotypical beliefs regarding the warmth and competence of their brands (e.g. through emphasizing sustainability and high product quality), as well as capitalizing on favourable country stereotypes associated with the brand origin. Moreover, they can influence brand buyer stereotypes by consistently depicting a certain type of group of users/ambassadors of their brands in their brand communications. Yet, do consumers rely solely on some cues (e.g. the brand's origin) and disregard others (e.g. the brand user), or do they simultaneously utilize several/all cues when evaluating the brand?
From a managerial perspective, simultaneously communicating cues related to the brand, its origin and its users (as shown in Figure 1) would obviously not be warranted if the influence of each stereotype is solitary. Marketers could simply rely on one type of cue (e.g. brand logo, slogan and tagline) and related stereotypical perceptions to influence brand evaluations and subsequent behavioural responses. However, if multiple brand-related stereotypical perceptions do indeed function in concert, marketers can benefit from knowing how the three stereotypes are connected and how they jointly impact consumers' responses.
From a theoretical perspective, extant literature largely considers each brand-related stereotype in isolation, which either implies that brandrelated stereotypes essentially capture the same phenomenon and can thus be used interchangeably or that they are distinct but completely unrelated constructs. The former assumption can be tested by concurrently measuring the three brandrelated stereotypes and examining their discriminant validity; yet, prior research has failed to do so. The latter assumption, however, may well lead to model misspecification if brand-related stereotypes are, in fact, interdependent predictors of consumer outcomes. An assumption of such interdependence appears to be theoretically warranted because several studies indicate that consumers transfer their beliefs between brands, countries and users (Diamantopoulos, Schlegelmilch and Palihawadana, 2011;Diamantopoulos et al., 2021;Fennis and Pruyn, 2007;Nebenzahl, Jaffe and Usunier, 2003).
Another shortcoming of extant research on brand-related stereotypes lies in overlooking their relation to consumers' perceived value -defined as the overall utility of a brand according to what is received and what is given (Zeithaml, 1988). The concept of brand value is central to marketing theory and practice alike (Kumar and Reinartz, 2016;O'Cass and Ngo, 2011). From a theoretical standpoint, marketing is 'the activity, set of institutions, and processes for creating, communicating, delivering and exchanging offerings that have value for customers, clients, partners and society at large' (American Marketing Association, 2017). From a practical perspective, consumers' perceptions of value are important determinants of satisfaction, loyalty, word-of-mouth and purchase in-tentions (Vieira, 2013); they are also a source of competitive advantage and corporate success (Vogel, Evanschitzky and Ramaseshan, 2008;Watson et al., 2015).
There is initial evidence suggesting that consumers' stereotypical perceptions of brands can positively affect the underlying mechanisms of value perception (Kolbl et al., 2020;Shea and 4 P. Gidaković et al. Hawn, 2019) and that brand origin stereotypes affect consumers' quality assessments (Shea and Hawn, 2019), the latter being a major component of perceived value (Gallarza, Gil-Saura and Holbrook, 2011;Vieira, 2013). However, the role of brand user stereotypes, as well as the combined impact of multiple stereotypes on consumers' value judgments, has not been investigated so far (see Appendix). Consequently, very little is known regarding the value-enhancing effects of different stereotypes and their usefulness as explanatory constructs for consumers' purchase decision-making.
Against this background, this paper aims to investigate the relationships between brand, brand origin and brand user stereotypes and assess their combined ability to predict managerially relevant outcomes. Specifically, drawing on theories from cognitive psychology and consumer culture theory, we propose and empirically test four alternative conceptual models linking the stereotype content of the three brand-related stereotypes to purchase intentions through value perceptions. To consistently capture the content of the three brandrelated stereotypes, we apply the stereotype content model (SCM; Fiske et al., 2002), which is the most prominent model for capturing stereotype content (Fiske, 2018;Halkias and Diamantopoulos, 2020). To enhance generalizability, we conduct our research in two contrasting empirical settings (respectively comprising a developed country and a country in transition) and employ both global and local brands as stimuli.
Our intended contribution is fivefold. First, we contribute to the stereotyping literature by offering the first attempt to conceptually and empirically assess how different types of brand-related stereotypes are interlinked. Given the rich body of research on consumer stereotypes (see Appendix), it is essential to understand how consumers navigate their brand preference through simultaneously stereotyping brands, their origin and their users. Our investigation is, to the best of our knowledge, the first to shed light on this issue.
Second, we reveal the combined effect of multiple stereotypes on consumers' value perceptions and, through them, on purchase intentions. We thus contribute to the still underdeveloped body of research on consumer value (Kumar and Reinartz, 2016) and, specifically, to the nascent research on the link between consumer stereotypes and perceived value (Kolbl et al., 2020). Third, we provide evidence of the stereotype content (i.e. warmth and competence) transfer across the three distinct stereotypes and assess the relative importance of warmth and competence in such transfer. Furthermore, by delineating the value-building role of each dimension, we contribute to the body of literature on the diagnosticity of warmth and competence in predicting consumer outcomes (Güntürkün, Haumann and Mikolon, 2020;Kolbl, Arslanagic-Kalajdzic and Diamantopoulos, 2019).
Fourth, we contribute to the international marketing literature by establishing the extent to which the interplay among different stereotypes in influencing consumer outcomes is stable across market settings differing sharply in their degree of economic development.
Finally, from a managerial perspective, we provide implications for practitioners aiming to leverage the content of different brand-related stereotypes in order to favourably influence consumers' perceptions of value and subsequently boost brand purchase intentions. Our findings suggest that managers should emphasize both the warmth and the competence of their brands, as both appear to be equally important drivers of value perceptions. Moreover, brand communications portraying the brand's origin and its users as both warm and competent can directly enhance consumers' shared beliefs about the brand; however, brand user warmth and competence appear to be more effective drivers of positive brand stereotypes.

Alternative conceptualizations of the relationships between brand-related stereotypes
In conceptualizing the relationships between brand, brand origin and brand user stereotypes and their mutual influence on brand evaluation, we contrast four theoretical perspectives ( Table 1). The first two perspectives (cue utilization theory and consumer learning by analogy (CLA)) have their theoretical origin in cognitive psychology and view stereotypes as energy-saving devices (Macrae, Milne and Bodenhausen, 1994) and schemas (Sherman, 1996), governed by the principles of categorization (Macrae and Bodenhausen, 2000;Macrae, Milne and Bodenhausen, 1994). This perspective seems to dominate current literature  (Luedicke, Thompson and Giesler, 2010;Thompson and Haytko, 1997) on brand-related stereotypes (see Appendix). The other two approaches (model of the structure and movement of the cultural meaning, and consumer lay theories) originate in consumer culture theory and support the notion that stereotypes can be viewed as cultural meaning (Luedicke, Thompson and Giesler, 2010;McCracken, 1986;Thompson and Haytko, 1997), governed by the principles of social and cultural influences (Haslam et al., 2002). While the four models differ in their assumptions about the relations between brand, brand origin and brand user stereotypes, they share the idea that across different stereotyped objects, the same dimensions of stereotype content are relevant. Indeed, research on the perception of stereotyped targets has revealed very similar dimensions of stereotype content in different contexts (Abele et al., 2008). We therefore assume that they are also relevant for the perception of brands, brand origin and brand users, and that the mutual influence of the different stereotypes is channelled through these fundamental dimensions.
The most prominent theoretical framework capturing the content of stereotypes is the SCM (Fiske et al., 2002). The SCM posits that people organize their perceptions along two core dimensions of social cognition -warmth and competence. Warmth captures beliefs about the target's intentions towards the perceiver or their social group (e.g. kind, friendly and good-natured). Competence represents beliefs about the target's ability to enact her/his intentions (e.g. capable, efficient and intelligent). The SCM has been successfully applied in prior research on social groups (Cuddy, Fiske and Glick, 2007;Fiske, 2018;Shiu, Hassan and Parry, 2015), brands (Bernritter, Verlegh and Smit, 2016;Ivens et al., 2015;Kervyn, Fiske and Malone, 2012), countries of origin (Diamantopoulos et al., 2017;Halkias and Diamantopoulos, 2020;Magnusson, Westjohn and Sirianni, 2019) and brand users (Antonetti and Maklan, 2016;Hickman and Ward, 2013;Ziano and Pandelaere, 2018). Brand-related stereotypes, as conceptualized along the two fundamental dimensions of warmth and competence, are distinct from other wellestablished constructs such as brand personality, brand attitude, country image and brand user image (for a detailed discussion, see Kervyn, Fiske andMalone, 2012 andKolbl et al., 2020). A main distinguishing feature is that stereotypes capture consumers' socially shared beliefs (common consensus in society), while related constructs capture a consumer's individual beliefs. In addition, research addresses related constructs as either antecedents or outcomes of brand-related stereotypes in empirical research models. For example, Ivens et al. (2015) treat brand personality as an antecedent of brand stereotypes, while Motsi and Park (2020) model country image as an outcome of country stereotypes.

Model 1 -lens model
Model 1 assumes that brand-related stereotypes are independent and unrelated (see Figure 2). Warmth and competence beliefs about each target 6 P. Gidaković et al. (brand, brand origin and brand user) are thus considered as unique and independent pieces of information that consumers potentially consider when forming value judgments. Formally, this model implies no covariation between the three stereotypes, but allows for covariation of the warmth and competence dimensions within each stereotype, thus accounting for potential halo effects (see Judd et al., 2005). Since this model compares the importance of several exogenous variables, it is useful in answering questions such as whether and to what extent consumers use stereotypical information in their brand evaluations.
We derive Model 1 from the 'lens model' of cognitive psychology, which proposes that humans 'must accumulate and combine cues' (Brunswik, 1955, p. 208) to make valid judgments. The underlying notion is that a person sees the world through several different lenses, or a set of imperfect cues, and these cues inform decision-making. Thus, human judgment is modelled as a linear function of a set of cues (Karelaia and Hogarth, 2008), with the warmth and competence dimensions of each target (brand, brand origin and brand user) functioning as 'independent variables' (Orquin, 2014, p. 271). Cues that strongly suggest one categorization (e.g. BMW is competent) over alternative categorizations (e.g. BMW drivers are cold) are said to be diagnostic. In this vein, multiple cues can predict brand evaluations; however, some cues will be more diagnostic, and consequently more influential, than less informative or ambiguous cues (Birnbaum, 1972;Nisbett, Zukier and Lemley, 1981). Diagnostic cues serve to make particular brand-related stereotypes 'more salient and more compelling to a perceiver, whereas less diagnostic information is less suggestive' (Skowronski and Carlston, 1987, p. 690). This means that some highly diagnostic cues (i.e. a specific brandrelated stereotype content) may be sufficient to form one's value judgments about a certain brand, independent of the other evidence available (i.e. other stereotypes and their content). For example, if Huawei users are stereotyped as competent, this might suggest that the brand will be perceived as being of value, regardless of the perceived (in)competence of the brand's origin (Reeder and Coovert, 1986).
The theoretical rationale of Model 1 is consistent with cue utilization theory (Purohit and Srivastava, 2001), which suggests that the extent to which a cue is used in assessing brands varies with its perceived diagnosticity (Richardson, Dick and Jain, 1994) -that is it depends on the presence/valence of other cues in the decision-making environment. Consumers rely on different cues associated with the brand, its origin and/or users. Through stereotyping, they make cognitive assessments that influence their brand-related responses (Cuddy, Fiske and Glick, 2007). According to Model 1, the brand, brand origin and brand user stereotypes are energy-saving devices (Macrae, Milne and Bodenhausen, 1994) and independent predictors of perceived value (Purohit and Srivastava, 2001). Consumers' stereotypical assessments of the brand/brand origin/brand user (and their underlying dimensions of warmth and competence) can thus facilitate informational transfer for assessing potential benefits and sacrifices (i.e. perceived value) to reduce uncertainty (Kolbl et al., 2020). Consequently, consumers make inferences about the brand's characteristics based on the stereotypes' diagnosticity for the judgment of value.

Model 2 -personification model
In Model 2, brand origin warmth and competence are mapped onto the brand user warmth and competence which, in turn, channel these stereotypes to brand warmth and competence ( Figure 3). The model also includes a direct relation from brand user warmth and competence to value perceptions, accounting for the fact that social groups, such as brand users, can have a direct influence on an individual's value perceptions (Akerlof and Kranton, 2000).
We derive Model 2 from the CLA model, which asserts that 'consumers learn through a process of internal knowledge transfer from familiar to novel domains' (Gregan-Paxton and John, 1997, p. 226). A fundamental assumption of learning by analogy is that existing knowledge structures (e.g. stereotypical schemas) serve to facilitate the achievement of specific learning objectives. When consumers evaluate brands, they rely on stereotypes about brands' warmth and competence (Kolbl et al., 2020;Shea and Hawn, 2019). Brand warmth and competence thus become learning targets and consumers need to identify analogues (i.e. bases) from which brand stereotype content can be inferred (Gregan-Paxton and John, 1997). Warmth and competence of brands' origin provide such a base because consumers often hold pre-existing  stereotypes about brands' country of origin (Chattalas and Takada, 2008;Maheswaran, 1994;Samiee, 1994).
CLA further assumes that once a base and a target have been identified, consumers construct a mapping by identifying elements that are common to both the base and the target (Gregan-Paxton and John, 1997). According to research on country-of-origin personification (Nebenzahl, Jaffe and Usunier, 2003), people (i.e. the nation and brand users) represent a common element of a country and brands originating from it. This stream of research shows that consumers personify country-of-origin stereotypes through 'people who buy products manufactured in a particular country' (de Moura Engracia Giraldi, Ikeda and Campomar, 2011, p. 101).
Accordingly, warmth and competence of brand users can function as mappings that transfer consumers' stereotypes about the brand origin's warmth and competence to stereotypes about a brand's warmth and competence. For example, Michael Jordan personified 'Americana' and transferred this stereotype to brands like Nike and the NBA (Rosenthal, 2020). In this context, the stereotypical properties of a country become associated with a brand via a person (e.g. an endorser) or a reference group (brand users) (Childers and Rao, 1992;Muniz and O'Guinn, 2001), who co-create and reinforce the brand stereotype (Bellezza and Keinan, 2014).

Model 3 -irradiation model
Model 3 (Figure 4) proposes that the warmth and competence dimensions of the brand origin stereotype are mapped onto the warmth and competence dimensions of the brand stereotype which, in turn, 8 P. Gidaković et al. channel this stereotype content to the brand user stereotype.
We derive Model 3 from the model of the structure and movement of the cultural meaning of consumer goods, according to which 'cultural meaning is located in three places: the culturally constituted world [e.g. countries], the consumer good [e.g. brands] and the individual consumer [e.g. brand user], and moves in a trajectory at two points of transfer: world to good and good to individual' (McCracken, 1986, p. 71; examples added). Cultural meaning thus begins in the physical and social world constituted by the categories and principles of the prevailing culture, moves to consumer goods and, finally, to the consumer. For instance, Apple users may be seen as competent because Apple channels stereotypes about American competence to its buyers.
Stereotypes can be considered 'as meaningmaking devices' (Castano, Bonacossa and Gries, 2016, p. 3) that help 'identify and construct mental representations of expected relationships between people, places, objects and ideas' (Heine, Proulx and Vohs, 2006, p. 90). In this sense, brandrelated stereotypes carry cultural meanings (Allen, Fournier and Miller, 2008;Torelli, Rodas and Stoner, 2017), which may transfer to the overall brand assessment, including the brand's value (Holt and Thompson, 2004;McCracken, 2009). The cultural meaning of brands connotes a consensual understanding -a public agreement -on what the brand represents as a symbol of a social group (Elliott, 1994;MacInnis, Torelli and Park, 2019).
Several studies in consumer research support the suggested movement of meaning (for a review, see Batra, 2019). In an international marketing context, a recent study by Diamantopoulos et al. (2021) found that there is stereotype content transfer between country warmth (competence) and brand warmth (competence). Such transfer of meaning is consistent with the process of irradiation (e.g. Diamantopoulos, Schlegelmilch and Palihawadana, 2011), through which consumers transfer their beliefs of brands' origin to brands. Animosity/affinity towards a country also influences consumers' perceptions of brands stereotypically associated with the prejudiced/liked nation (Oberecker and Diamantopoulos, 2011;Russell and Russell, 2010). Furthermore, the brand trait transference literature suggests that consumers transfer their impressions of brands to brand users (Antonetti and Maklan, 2016;Fennis and Pruyn, 2007). Therefore, a typical user of a particular brand can be reflective of that brand (Helgeson and Supphellen, 2004) and plays an important role in consumers' information processing and brand attitude formation (Mazzocco and Brock, 2006). These findings imply that brand stereotypes can potentially mediate the effects of brand origin stereotypes on brand user stereotypes.

Model 4 -brand-mediated model
Model 4 ( Figure 5) specifies the content of brand origin and brand user stereotypes as distinct antecedents of the brand stereotype content. In contrast to Model 1, this model acknowledges that brand origin and brand user stereotypes are combined, adapted and juxtaposed (Thompson and Haytko, 1997), warranting covariation between these two stereotypes.
We derive Model 4 from research on consumers' use of lay theories about brands, brands' origin and brand users (Fournier and Alvarez, 2019;Kardes et al., 2008;Luedicke, Thompson and Giesler, 2010;Thompson and Haytko, 1997). 'Lay theories (also referred to as "implicit theories", "naïve theories" and "folk theories") have been defined both as ontological assumptions (beliefs about what is true in the world) and as narrative representations (frameworks that explain and organize the world)' (Plaks, Levy andDweck, 2009, p. 1069). Stereotypes can be conceptualized as lay theories because they capture both the ontological assumptions about warmth and competence of social targets and the narrative assumptions about how social targets acquire warmth and competence (Fiske et al., 2002;Schneider, 2005). Importantly, previous research shows that consumers routinely rely on implicit theories in brand evaluation processes (Kardes, Posavac and Cronley, 2004;Kardes et al., 2008).
Consumers often hold lay theories related to warmth and competence of both brands' country of origin and social groups that are perceived to use a brand. For instance, Levi's is believed to be representative of its American origin, implying a high-warmth and high-competence stereotype (Aaker, Benet-Martínez and Garolera, 2001). However, Levi's is also believed to be representative of its user group, namely gay men (Kates, 2004), who are stereotyped as warm but less competent (Fiske et al., 2002). Consumer culture theory suggests that consumers contrast and integrate lay theories related to brands' origin and brand users, and use them to construct brand meanings (Luedicke, Thompson and Giesler, 2010;Thompson and Haytko, 1997). In this process, stereotypes about brand users and brand origin are transferred to the respective brand (Holt, 2004;Price and Coulter, 2019). For example, the Harley-Davidson brand is infused with the image of an 'iconic American brand [i.e. its origin], symbolizing the American way' (Schembri, 2009(Schembri, , p. 1305; example added) and the outlaw biker (i.e. brand user) that dominates the stereotype of Harley owners (Schouten and McAlexander, 1995). Similarly, the MG brand encapsulates both its British heritage (i.e. origin) and the vintage car collector (i.e. brand user) subculture (Leigh, Peters and Shelton, 2006). This stream of research implies that both brand origin and brand user stereotypes independently influence the brand stereotype and, through it, brand evaluations.

Empirical studies
We test and compare the previously introduced four conceptual models in two complementary studies: Study 1 in a developed West European country (Germany), using global and local brands as stimuli and Study 2 in an East European country in transition (Serbia), with a new set of global and local brands from the same product/service categories as in Study 1. Replication studies conducted with different populations (e.g. countries) or dif-ferent aspects of research design (e.g. measures or stimuli) are considered to constitute conceptual instead of exact replications (Lynch et al., 2015). We opted to conduct such a conceptual replication in order to reduce the uncertainty of model selection (Preacher and Merkle, 2012), as well as to enhance the generalizability of our findings (Durvasula et al., 1993).
The two selected countries differ substantially in terms of the KOF index of globalization, which is the most comprehensive measure of globalization (Gygli et al., 2019), as well as in terms of their per capita income, which is a good indicator of purchasing power and economic development. Germany is classified as a developed economy, while Serbia is an economy in transition (UN, 2020). Compared to Germany, Serbia represents an economically and socially less advanced European country (e.g. GDP per capita_Germany = $50,800, GDP per capita_Serbia = $15,100; urban population Germany = 77.5%, urban population Serbia = 56.4%; CIA, 2020). Moreover, Serbia is a less globalized country (rank 36) than Germany (rank 7 on the KOF Globalisation Index; Gygli et al., 2019). In terms of country brand strength, Germany is a much stronger country brand (ranked 9 worldwide and ranked 3 in Europe) than Serbia (ranked 103 worldwide and ranked 35 in Europe; Bloom Consulting, 2020). Moreover, consumers in developed countries and countries in transition often ascribe different meanings to branded products and have different levels of brand-related knowledge (Coulter et al., 2005;Strizhakova, Coulter and Price, 2008). Thus, there is evidence of relevant cross-national differences between the selected countries, providing a basis for assessing model stability across national samples (Cadogan, 2010) and ensuring that the results are applicable to both developed and transition countries.

Study 1: participants, design and measures
In an online survey, we recruited 404 paid participants from Germany via a professional marketing research agency employing a nationally representative sample in terms of gender and age (M age = 47.8, SD age = 16.3; 52.2% female). Respondents were informed that they were participating in an academic study on branding and were assured anonymity. Next, respondents were randomly assigned to evaluate one of eight brands from four product categories featuring local/global brands: coffee (Melitta/Lavazza), home appliances (Loewe/Samsung), banking services (Deutsche Bank/Hypo Vereinsbank) and telecommunications (mobilcom Debitel/Vodafone). These brands were selected, following consultation with local marketing experts, due to their differing product category, utilitarian and hedonic properties and country of origin. We included both local and global brands, since brand globalness and localness have been shown to affect brand stereotypes (Davvetas and Halkias, 2019;Kolbl, Arslanagic-Kalajdzic and Diamantopoulos, 2019). In all four models, we also control for the effects of product category involvement and brand familiarity on purchase intentions.
To measure the content of the three brandrelated stereotypes, we adopted four items per dimension from Fiske et al. (2002): warmthfriendly, good-natured, kind and warm; competence -capable, competent, efficient and intelligent. Furthermore, participants provided ratings on established multi-item scales for perceived value (Netemeyer et al., 2004), purchase intentions (Putrevu and Lord, 1994) and product category involvement (Kapferer and Laurent, 1993). We measured brand familiarity with a slider scale used by Kolbl, Arslanagic-Kalajdzic and Diamantopoulos (2019). Scale items, factor loadings and Cronbach's alphas are shown in Table 2. Measurement model fit was satisfactory (χ 2 = 1,055.55, DF = 516, CFI = 0.960, NNFI = 0.954, RMSEA = 0.051, SRMR = 0.045) and all scales proved to be reliable and conform to standards of convergent and discriminant validity (Table 3).
To assure data quality and minimize common method bias (CMB), we included two attention checks with a directed query (e.g. 'Please select number five (5) on this list'; see Abbey and Meloy, 2017). We also relied on different rating formats for measurement scales and obtained responses to the dependent variables before measuring the independent variables (MacKenzie and Podsakoff, 2012). Moreover, we tested for potential CMB with a variation of the marker variable approach developed by Malhotra, Kim and Patil (2006). Out of 538 initially significant zero-order correlations among the measurement items (indicators) of our constructs, 521 (96.8%) remained significant after subtracting the second smallest positive zero-order correlation of 0.011. Thus, CMB did not confound the study results.

Study 1: results
We estimated four structural equation models corresponding to the conceptual specifications in Figures 2-5. The results (Table 4) reveal that the brand-mediated model (Model 4) exhibits the best fit based on a combination of fit indices (Hu and Bentler, 1999). Since the four models are not nested but include the same variables, we also compared them on information-based criteria (i.e. AIC, CAIC and ECVI; Rust, Lee and Valente, 1995); these also support Model 4 as the bestfitting model. Additionally, to minimize the uncertainty related to the comparison and the selection of non-nested models, we employed a two-step approach developed by Merkle, You and Preacher (2016) to formally compare Model 4 with its closest rival, Model 3. In the first step, we rejected the null hypothesis that the two models are indistinguishable (ω 2 = 0.358, p < 0.05), which warrants the second step, namely the model comparison by Vuong's test (Vuong, 1989). Application of the latter rejected the null hypothesis that both models fit equally well in favour of the alternative hypothesis that Model 4 provides a significantly better fit to the data than Model 3 (z = 2.862, p = 0.01).
In Model 4, brand origin warmth and brand user warmth jointly explain 55% of the variance in brand warmth, while brand origin competence and brand user competence together explain 52% of the variance in brand competence. Brand warmth and competence further account for 36% of the variance in perceived value which, together with the two control variables (brand familiarity and      product category involvement), explains 67% of the variance in purchase intentions. In effect size terms, all these explained variances imply strong relationships (Cohen, 1988). Model estimation results are summarized in Figure 6. Model 4 offers three additional insights into the predictive ability of brand-related stereotypes. First, by imposing equality constraints on the paths from the brand origin and user stereotypes on the brand stereotype (e.g. setting brand origin warmth → brand warmth = brand user warmth → brand warmth), we find that the brand user stereotype (warmth -γ = 0.67, p < 0.001; competence -γ = 0.62, p < 0.001) is a significantly stronger predictor of the brand stereotype compared to the brand origin stereotype (warmth γ = 0.11, p < 0.01; competence -γ = 0.18, p < 0.001; warmth -χ 2 DF = 1 = 48.31, p < 0.001; competence -χ 2 DF = 1 = 35.79, p < 0.001). Second, by imposing equality constraints on the paths from brand warmth and competence to perceived value (i.e. brand warmth → perceived value = brand competence → perceived value), we find that brand warmth (β = 0.20, p < 0.001) and brand competence (β = 0.26, p < 0.001) are equally strong predictors of perceived value ( χ 2 DF = 1 = 1.69, p = 0.194). Third, we computed the indirect effects using bootstrapping with 5,000 subsamples and 95% confidence intervals (CIs; see Table 5). We observe significant indirect effects of brand user warmth (β = 0.27, p < 0.01, CI  As an additional analysis (available upon request from the authors), we performed a robustness check to investigate whether the structural relations as proposed by our best-fitting model (i.e. Model 4) remain constant on the subsamples of participants who rated local versus global brands; no significant differences could be established.
Study 1 offers several important insights into the relationships between brand-related stereotypes and their effects on perceived value. First, by measuring the stereotype content (i.e. warmth and competence) of three stereotypes, we establish their discriminant validity and demonstrate that these are distinct constructs. Second, we offer the first empirical investigation of the interrelationships among different brand-related stereotypes by comparing four alternative theoretical explanations. These comparisons resulted in the rejection of Models 1-3 in favour of the brand-mediated model (Model 4). Third, we offer empirical evidence showing that brand-related stereotypes indeed affect consumers' value perceptions (and through them, their behavioural intentions).
Fourth, we establish that the brand stereotype is more influenced by stereotypical perceptions of brand users rather than the brand's origin, and that brand warmth and competence have similar (positive) effects on perceived value. Finally, we demonstrate that brand user warmth, brand user competence and brand origin competence have positive indirect effects on purchase intentions.

Study 2: participants, design and measures
In Study 2, we rely on the same research design and measures as in Study 1; however, we apply them in the context of a country in transition (Serbia) using a different set of brands as stimuli. We again included both global and local brands in the same four product categories as in Study 1, based on the same selection criteria indicated earlier: coffee (Nescafe/Grand), home appliances (Bosch/Alfaplam), banking services (ErsteBank/Komercijalna Banka) and telecommunications (Vip/MTS).
In collaboration with a marketing research agency, we recruited 438 respondents for an online survey (M age = 42.6, SD age = 12.2; 52.7% female), employing again a nationally representative sample in terms of gender and age. We used the same measurement scales as in Study 1 (see Table 2) and applied the same data collection procedures and steps to improve data quality. We again tested for potential CMB with a variation of the marker variable approach (Malhotra, Kim and Patil, 2006). After partialling out the second smallest positive zero-order correlation (0.02), 551 of initially 567 (97.2%) significant zero-order correlation coefficients remained significant. Thus, CMB did not materially affect the observed relationships in Study 2 either. Measurement model fit was again satisfactory (χ 2 = 1,084.583, DF = 516, CFI = 0.964, NNFI = 0.959, RMSEA = 0.050, SRMR = 0.029), and all scales proved to be reliable and valid (see Table 3).

Study 2: results
The estimation of the four structural models (Figures 2-5) and inspection of their respective fit indices (Table 4) indicates that, again, Model 4 exhibits the most satisfactory overall fit (Hu and Bentler, 1999) and also outperforms the other three models in terms of information-based criteria (Rust, Lee and Valente, 1995). We again employed the procedure proposed by Merkle, You and Preacher (2016) to formally compare Model 4 with its closest rival (Model 3). In the first step, we rejected the null hypothesis that the two models are indistinguishable (ω 2 = 0.536, p < 0.001). Using Vuong's test (Vuong, 1989) in the second step, we rejected the null hypothesis that both models fit equally in favour of the alternative hypothesis that Model 4 has a better fit (z = 4.78, p < 0.001).
Looking at the path coefficient estimates, Model 4 replicates the effects observed in Study 1 (Figure 5). Brand origin warmth and brand user warmth together explain 57% of the variance in brand warmth, while brand origin competence and brand user competence jointly predict 55% of the variance in brand competence. The two brand stereotype dimensions account for 48% of the variance in perceived value that, along with the two control variables, explains 61% of the variance in purchase intentions. As in Study 1, the relevant effect sizes imply strong relationships (Cohen, 1988).
We again imposed equality constraints to examine the relative magnitude of the effects of brand origin and brand user stereotypes on the brand stereotype (e.g. brand origin warmth → brand warmth = brand user warmth → brand warmth). The chi-square difference tests again produced significant differences for both warmth ( χ 2 DF = 1 = 92.097, p < 0.001) and competence ( χ 2 DF = 1 = 121.100, p < 0.001) dimensions, indicating that the brand user stereotype (warmth γ = 0.71, p < 0.001; competence -γ = 0.70, p < 0.001) acts as a stronger predictor of brand stereotype compared to the brand origin stereotypes (warmth -γ = 0.06, p < 0.05; competenceγ = 0.08, p < 0.001). Furthermore, we compared the magnitude of the effects of brand warmth (β = 0.36, p < 0.01) and competence (β = 0.38, p < 0.001) on perceived value. As in Study 1, the chi-square difference test did not produce a significant result ( χ 2 DF = 1 = 0.01, p = 0.92), indicating that the magnitude of effects of brand warmth and competence on perceived value is equivalent.
We again computed all indirect effects implied by Model 4 and used bootstrapping with 5,000 subsamples to construct 95% CIs for these indirect effects ( Table 5) on purchase intentions are also significant. The respective CIs again overlap, which indicates that both dimensions of the brand stereotype content function as equally important predictors of consumers' purchase intentions. A robustness check involving a comparison of the structural relations on subsamples of participants who respectively assessed global versus local brands failed to reveal any significant differences. Thus, as in Study 1, Model 4 is stable across both global and local brands.
Study 2 fully replicates the findings of Study 1 and thus confirms the insights regarding the interrelationships of brand-related stereotypes and their effects on perceived value and purchase intentions. Model 4 again proves to be the bestfitting model in a country in transition. Study 2 further supports the findings from Study 1, showing that the brand user stereotype functions as a stronger predictor of the brand stereotype than the brand origin stereotype. In turn, brand warmth and brand competence have a similar influence on consumers' value perceptions. Regarding the indirect effects of the brand-related stereotypes on perceived value and purchase intentions, we find a similar pattern as in Study 1 -that is, the effects of brand user warmth, brand user competence and brand origin competence have positive indirect relations to purchase intentions.

General discussion
While an increasing number of studies have drawn on the SCM (Fiske et al., 2002) and applied a stereotyping perspective to brands, their origins and their users (see Appendix), extant research has failed to investigate how these brand-related stereotypes are interrelated and how they jointly influence managerially relevant outcomes. Our research empirically addresses this gap in both a developed country and a country-in-transition setting by linking different brand-related stereotypes to perceptions of value and, through them, to purchase intentions. Our findings have several important implications for theory and practice, as outlined below.

Theoretical implications
First, our research contributes to extant literature on brand-related stereotypes (e.g. Antonetti and Maklan, 2016;Diamantopoulos et al., 2017Diamantopoulos et al., , 2021Kervyn, Fiske and Malone, 2012). By applying the SCM, this stream of research has provided valuable insights into how the two fundamental dimensions of warmth and competence of either brands, their origin or their users impact relevant attitudinal and behavioural consumer outcomes. However, except for a recent study by Diamantopoulos et al. (2021), it has been silent on how these stereotypes relate to one another. Our work demonstrates that the three stereotypes are distinct yet interdependent constructs. Specifically, the brand origin and brand user stereotypes influence the content of the brand stereotype which, in turn, impacts value perceptions and, ultimately, purchase intentions. This is an important finding as it demonstrates that consumers navigate their brand preference through simultaneously stereotyping brands, their origin and their users. Therefore, investigating stereotypes in isolation of each other, as documented in previous research (see Appendix), can lead to a fragmented and possibly biased view of the role of brand-related stereotypes in a consumer behaviour context.
Second, when conceptualizing brand-related stereotypes, scholars have applied different theoretical lenses, such as energy-saving devices (e.g. Kolbl et al., 2020), meaning (e.g. Xu, Leung and Yan, 2013) and schemas (e.g. Magnusson, Westjohn and Sirianni, 2019). We use a novel approach to theoretically underpin the three brand-related stereotypes, namely through consumer lay theories (Luedicke, Thompson and Giesler, 2010;Thompson and Haytko, 1997), and corroborate earlier findings by showing that consumers rely on implicit theories in brand evaluation processes (e.g. Kardes, Posavac and Cronley, 2004;Kardes et al., 2008). Lay theories may therefore offer a promising approach for studying stereotypes by investigating how consumers apply and integrate narratives about brands, their origin and their users when evaluating brands.
Third, we offer first empirically based insights on how the three brand-related stereotypes relate to consumers' value perceptions. Perceived value is a key concept in (international) marketing (AMA, 2017); however, there is scant research on whether and how brand-related stereotypes drive consumers' value perceptions (for notable exceptions, see Kolbl et al., 2020;Shea and Hawn, 2019). Our findings show that perceived value is an important mediating variable that translates consumers' brand-related stereotypical perceptions into behavioural intentions. By considering perceived value as an outcome variable and concurrently drawing on social psychology as well as consumer culture theory, we contribute to 'the emergence of a truly interdisciplinary approach to the concept of value and to the process of value creation' (Gallarza, Gil-Saura and Holbrook, 2011, p. 188). Moreover, we address a recent call to explore the antecedents and consequences of customer value (Zeithaml et al., 2020).
Fourth, our findings contribute to the ongoing debate on the relative importance of warmth versus competence in predicting consumer outcomes (Güntürkün, Haumann and Mikolon, 2020;Kirmani et al., 2017;Kolbl, Arslanagic-Kalajdzic and Diamantopoulos, 2019). We find that the brand user stereotype has a stronger influence (in terms of both warmth and competence) on the brand stereotype compared to the brand origin stereotype. In other words, the extent to which a particular brand will be perceived as being warm or competent depends more on how the buyers of the brand are perceived to be on these dimensions and less on the origin of the brand. Furthermore, brand warmth and brand competence were found to have equally strong influences on perceived value, together explaining more than a third of the variance in the latter. This impressive predictive ability serves to emphasize the diagnosticity of both brand stereotype dimensions in predicting consumers' value judgments.
Fifth, prior research suggests that consumers from mature and emerging markets differ both in their evaluations of different country origins and their judgments of global/local brands (e.g. Batra et al., 2000), implying differences in the assessment of at least brand origin and brand stereotypes. Moreover, while there is evidence to suggest that brand origin stereotypes can differ depending on the country of study (e.g. Cuddy et al., 2009), whether such variations also apply to brand and brand user/buyer stereotypes has not yet been examined. We contribute to extant knowledge by demonstrating that the interplay among different stereotypes in influencing consumer outcomes is stable across different market settings (i.e. developed markets vs. markets in transition) and different sets of brands (i.e. global vs. local). Thus, the observed effects -relating to both the interrelationships among the three stereotypes as well as their influence on value perceptions and subsequent purchase intentions -are not contingent on the market setting/brand stimuli involved, pointing to stable and generalizable relationships.

Managerial implications
Our results are particularly important for marketers wishing to strategically utilize different brand-related cues to boost consumer perceptions of value. Specifically, developing brand communications aimed at triggering brand-related stereotypes is a promising strategy for enhancing perceived value to customers, regardless of the country setting involved.
Brand communications providing cues relating to the brand (e.g. logo, slogan and tagline), the brand's origin (e.g. 'made in/designed in…', featuring the origin country's typical landscape or symbols) and the brand user (e.g. celebrity endorser, expert and typical consumer) offer important information to consumers, who utilize these cues in their overall brand evaluations. While brand users (e.g. Zlatan Ibrahimović) and brand origin (e.g. Sweden) cues directly influence stereotypical brand associations (e.g. Volvo), the latter impact consumers' purchase intentions through their value perceptions (e.g. 'Volvo is good value for money'). Therefore, co-developing and deliberately depicting different stimuli in brand communications can provide a strategic lever for managers to enhance their brand's value, and consequently, stimulate purchase behaviour.
We also find that consumers put a greater emphasis on cues related to the brand user as opposed to the brand origin when forming their stereotypical judgments about the brand. Managers can thus favourably influence their brands' stereotypes through the careful selection of brand users (e.g. endorsers, influencers or reference/social groups) to be depicted in brand communications. While marketers need to be aware that the positive influence of brand origin on the brand stereotype is less powerful than that of the brand user(s), brand origin can additionally facilitate consumers' shared beliefs about the brand.
Concerning the stereotype content, our findings suggest that both warmth and competence of brand-related stereotypes matter. Therefore, emphasizing the warmth and competence of the brand, its origin, as well as its users can positively influence value perceptions, and through them, purchase intentions. For example, depicting typical users of the brand as friendly and skilled is likely to enhance consumers' stereotypical perceptions of the brand itself. Similarly, associating the brand's country of origin with welcoming and innovative people in brand-related communications is likely to benefit the brand. However, as previously noted, emphasizing the brand's origin may bear fewer fruits compared to emphasizing brand users and may be limited to competence stereotypes about the brand origin. As for the brand stereotype, when brands are perceived as being both warm and competent, consumers show a higher willingness to buy them (Aaker, Vohs and Mogilner, 2010). Such brands belong to the 'golden quadrant' (Aaker, Garbinsky and Vohs, 2012), in which brands -as social objects -are perceived to be simultaneously cooperative, sincere and friendly as well as intelligent, successful and possessing leadership skills. Our findings show that scoring high on both dimensions can provide brands with extra benefits in terms of their perceived value, eventually boosting purchase intentions. This can be achieved, for example, by featuring the brand's advanced technological properties (i.e. competence) as well as its friendly design (i.e. warmth) in marketing communications.

Limitations and future research
Our research is not without limitations and our investigation should be seen as a springboard for future research. While we have focused on the brand, brand origin and brand user stereotypes, consumers may rely on additional stereotypes, such as employee/service provider stereotypes (Kirmani et al., 2017;Wang et al., 2017). Incorporating such additional stereotypes in the theoretical models developed in this paper would offer further insights into consumers' stereotypical perceptions of brands, especially in the context of service brands (Rosenbaum and Walsh, 2012). In addition, the relationship between brand-related stereotypes and behaviour may be mediated by emotions (Cuddy, Fiske and Glick, 2007). For example, Ivens et al. (2015) found that emotions mediate the impact of brand warmth and competence on consumers' attitudinal and behavioural responses towards brands. Consequently, incorporating emotions in the brand-mediated model would extend the insights on the relation between the brand-related stereotypes and managerially relevant outcomes.
Regarding the non-significant indirect effects of brand origin warmth on consumer responses (both perceived value and purchase intentions), future research should investigate the reasons behind them. The manner in which value judgments are formed is theoretically similar to the manner in which quality judgments are formed (e.g. Aaker, 1996;Gallarza, Gil-Saura and Holbrook, 2011), and there is wide consensus on quality being a key input to overall value (e.g. Snoj, Pisnik Korda and Mumel, 2004;Teas and Agarwal, 2000;Zeithaml, 1988). Brand origin competence (indicating capability and efficiency) may therefore be more strongly related to perceived value through brand competence than brand origin warmth. In this vein, future research should shed more light on the role of brand warmth and brand competence in mediating the relationship between brand origin/brand user stereotypes and consumers' value versus quality perceptions, in an effort to disentangle potential differences regarding how brand-related stereotypes relate to these consumer perceptions.
In the present research, we did not take into account that consumers differ in how certain they are about stereotypical inferences (Stangor, 1988;Tormala, 2016). For example, consumers might perceive users of a brand as a homogeneous and meaningful entity, might easily retrieve this stereotype from memory and be certain that the stereotype indeed applies to the users of this brand. However, as regards another brand, consumers might be less certain about what the stereotype is and whether it applies to all users. Research suggests that stereotypes are more likely to shape judgments and behaviour when stereotyped groups are perceived as homogeneous and meaningful entities, and when certainty about the stereotype is high (Florack, Bless and Piontkowski, 2003;Spencer-Rodgers, Hamilton and Sherman, 2007). Hence, it would be insightful to study whether the relations observed in the present research become stronger with an increase in stereotype certainty, homogeneity and entitativity of the stereotyped group.
As we used the same product categories across both our studies, it is not likely that product category idiosyncrasies could have had an impact on the observed results. However, future research may investigate the potential moderating role of some key contextual factors (such as psychological distance, utilitarian vs. hedonic products, etc.) in potentially affecting the homogeneity of the structural relationships.
While research on country image shows that consumers' image of a particular country can shape their perceptions of brands originating from that country (Johansson, Douglas and Nonaka, 1985;Thakor and Katsanis, 1997;Thakor and Lavack, 2003), 'reverse' effects have also been noted 'whereby consumers use their perception of a country's product to imbue their perceptions of the country' (Lee and Lockshin, 2012, p. 508; see also Kleppe, Iversen and Stensaker, 2002;Magnusson et al., 2014). Therefore, exploring the possibility that stereotypical content associated with (a set of) brands with the same origin may transfer over to the stereotype content of that origin might be another potentially promising approach.
Finally, replications with consumer samples from other countries characterized by different economic and/or cultural conditions would be helpful to establish the robustness and generalizability of our findings.