Anthropomorphized vs objectified brands: which brand version is more loved?

Purpose – The purpose of this paper is to deal with the role of the human metaphor (anthropomorphism) and consumers’ liking for the humanized version of the brand as antecedents of three key components of brand love: self-brand integration, positive emotional connection and feelings of anticipated separation distress. Design/methodology/approach – A sample of 399 consumers provided information about a brand from a stated list of 16 brands of clothing. Findings – Both anthropomorphism and consumers’ liking for the humanized brand have positive effects on specific components of brand love. The results confirm that brand anthropomorphism is only desirable when the humanized version of the brand is attractive for consumers. Research limitations/implications – A potential shortcoming is the qualitative technique employed to observe anthropomorphic thought. Collecting ratings of anthropomorphic and non-anthropomorphic traits could be viewed as a method more easily applied in market research surveys. Practical implications – Managers have to control how consumers imagine the brand as a human entity because it affects brand love. For example, by tracking consumers’ opinions and traits of those people associated with the brand and brand user stereotypes can condition consumers’ imagination of the humanized brand. Originality/value – Compared to the limited number of studies about the relationship between anthropomorphism and brand love, this study focuses on the effects of anthropomorphism as a process, and not as a personal trait, on brand love. It also relies on consumers’ imagination instead of brand personification strategies to stimulate anthropomorphism.


Introduction
Since Fournier (1998) identified love as a distinguishable facet of all strong brand relationships, brand love has been considered an intrinsically interesting construct in research on consumer-brand relationships (Batra et al., 2012;Huber et al., 2015;Karjaluoto et al., 2016;Palusuk et al., 2019). The question of how to make consumers develop love relationships with a brand is still a highly pertinent topic of research. Most of the prior literature on the antecedents of this relationship has mainly focused on the following domains: the characteristics of products and brands, consumers' characteristics and identities, and consumers' experiences with brands.
At first sight, product quality (Batra et al., 2012), the hedonic and utilitarian nature of a brand (Huber et al., 2015;Karjaluoto et al., 2016), its symbolic nature or brand personality (Roy et al., 2016), brand prestige and uniqueness (Bairrada et al., 2018) have been found to be antecedents of brand love. Other studies have paid more attention to consumers' characteristics, such as personality traits (Rauschnabel et al., 2013) and age (Rivits-Arkonsuo and Leppiman, 2015). There is also empirical evidence of the positive effects that brand experiences (Bıçakcıoğlu et al., 2018), satisfaction (Carroll and Ahuvia, 2006) and trust (Karjaluoto et al., 2016) exert on brand love. Finally, brands may help to define individuals' identities and connect them with other people. In this sense, brand identification, self-brand connection, self-esteem, self-image, self-congruity, sense of brand community, consumers' self-congruence with Facebook "Likes," and online brand engagement have also been identified as precursors (see Bıçakcıoğlu et al., 2018;Huber et al., 2015;Karjaluoto et al., 2016;Roy et al., 2016;Vernuccio et al., 2015;Wallace et al., 2014Wallace et al., , 2017. To move beyond the current state of the art, an interesting issue is whether the view of brands as social entities would result in a better explanation of brand love. The consumer-brand relationship literature has accepted the idea that consumers may relate to brands as if they were persons (Fournier, 1998) and that for a consumer-brand relationship to exist, a brand has to turn into an appropriate, valuable, human-like relationship partner. A key mechanism that enables a brand to assume the role of an active and personalized agent in a relationship is anthropomorphism (MacInnis and Folkes, 2017). The brand love literature is no exception, and the works of Rauschnabel and Ahuvia (2014) and Hegner et al. (2017) proposed that it could foster love. However, these two studies focused on individuals' anthropomorphism tendency, which is a construct that is theoretically and practically distinct from the process of anthropomorphism (Letheren et al., 2016). This process is a cognitive bias of viewing human characteristics in something non-human, while anthropomorphism tendency is the bias of an individual to anthropomorphize something, rather than the process of doing so. Anthropomorphism tendency is a personality trait that may differ between individuals because it is mostly determined by factors such as age, thinking style or personal connections with animals (see Letheren et al., 2016). Therefore, as marketers may exert more influence on the anthropomorphism process (e.g. conferring human characteristics on a brand) than on the anthropomorphism tendency, the primary aim of this study is to analyze whether the brand anthropomorphism process actually exerts stronger effects on brand love than a non-human perspective of a brand.
More specifically, this study moves beyond the current literature because differs from past studies in the following aspects: (1) It uses as stimuli neutral brands in terms of affection and a broader range of brands in a less anthropomorphic product category (e.g. clothes).
(2) It focuses on different dimensions of the brand love prototype proposed by Batra et al. (2012).
(3) Additionally, because consumers' reactions to anthropomorphized brands may be more complex than is typically assumed, the current research takes a further step by also considering how consumers' liking for the anthropomorphized version of a brand affect brand love.

Theoretical framework and research hypotheses
2.1 Brand love as a consumer-brand relationship concept From a consumer-brand relationship perspective, brand love is not a specific and single emotion but an emotionally intense brand relationship (Reimann et al., 2012) that entails a plurality of aspects (Vernuccio et al., 2015). This results in prototype-based definitions of the EJMBE concept, depending on how consumers experience it (Albert et al., 2008;Batra et al., 2012). Among these definitions, the one developed by Batra et al. (2012) characterized brand love as a subjectively experienced combination of cognitions (self-brand integration), feelings (positive emotional connection, anticipated separation distress, attitude valence and attitude strength) and behaviors (passion-driven behaviors and long-term relationships), which consumers synthesize into a "mental prototype." Because brand love is composed of a complex constellation of elements, all of them are potential ways of building it because, as suggested by Rauschnabel and Ahuvia (2014, p. 376), "any research on multidimensional construct is an investigation of its dimensions." Furthermore, as these elements are neither synonymous nor interchangeable (Batra et al., 2012), a specific cause (e.g. brand anthropomorphism) does not have to affect all the different components of the concept. Therefore, to make the analysis more tractable, this study focuses on three specific dimensions because most previous studies have reached a strong consensus in considering them central aspects of brand love: (1) Self-brand integration is a cognitive merging of the brand and the individual's self so that brand meanings are experienced as the individual's own (Reimann and Aron, 2009). Past research has agreed that loved brands are an integrated part of the consumer's self (see Albert et al., 2008;Batra et al., 2012;Vernuccio et al., 2015) because through this cognitive merging, brands provide social value to the individual. As such, the individual's goals of self-expansion and identity creation (Ahuvia et al., 2009) are accomplished.
(2) Among the variety of elements that characterize brand love, positive emotional connection is perhaps the most important because previous studies (see Batra et al., 2012;Langner et al., 2016) have demonstrated that consumers are emotionally connected to their loved brands. It is defined as the emotional bond and positive feelings that consumers experience when thinking about or using a brand .
(3) Loved brands are characterized as irreplaceable and unique. Consequently, brand love is also associated with feelings of separation distress when consumers anticipate or experience being distanced from loved brands because they cease to be available to them (Batra et al., 2012;Langner et al., 2016).

The brand anthropomorphism process
Research efforts in anthropomorphism have increased recently[1] within the marketing field, especially in the branding literature. Epley et al. (2007) define it as an automatic process of attributing human-like characteristics, features, motivations, intentions or emotions to nonhuman agents (e.g. animals, natural and supernatural phenomena, material states, or abstract concepts). It is not only attributing life (animism) to something non-living but also perceiving a non-human object as being "like us" as a result of having human-like features, personality characteristics and a human-like mind (Epley et al., 2008;MacInnis and Folkes, 2017). The anthropomorphism process in the case of brands goes beyond brands' observable actions and behaviors (e.g. Tesla sells electric-powered cars and clean-energy-generation products) to make inferences about their unobserved personalities, intentions and motivations (e.g. Tesla is concerned with the environment and the zero-emission future that most people want). Research on the formation of person impressions may explain this elaboration process to the extent that observed behaviors are translated into traits, which form the basis for the evaluative concept of a person (Srull and Wyer, 1989). Therefore, anthropomophized brands have to do with the perception of brands as if they were actual human beings in terms of their physical appearance, their personality, intentions and emotions.
The interest in anthropomorphism resides in the fact that viewing non-human objects in human-like terms entails important differences in consumer attitudes and behaviors. The information that comes to mind in the elaboration process is relevant in how people react to and treat such objects (Chandler and Schwarz, 2010). Therefore, individuals make social judgments if they use the self or the concept of "human" as a frame of reference (Fiske, 1982), while they make non-social judgments when the human frame is absent.
Based on this idea, it is reasonable to think that consumers' reactions to anthropomorphized vs objectified brands may be different. In what follows, this paper analyses these reactions in terms of three specific dimensions of brand love.

The effects of brand anthropomorphism
Previous research on anthropomorphism has found that it triggers cognitive, affective and motivational responses in individuals. Specifically, empirical evidence show that: (1) It has positive effects on consumers' brand and product evaluations (Aggarwal and McGill, 2012;Delbaere et al., 2011;Hart et al., 2013). For example, when consumers are unfamiliar with a technology, anthropomorphizing the experiences renders them more likely to have a positive brand attitude (van Esch et al., 2019).
(4) It triggers perceptions of an entity as mindful, thoughtful and intentional (Waytz et al., 2010), with which people can potentially build social relationships.
These results grounded in the idea that when people anthropomorphize objects, they automatically activate human schemas and apply knowledge about the social world to the inanimate world, creating some degree of perceived similarity of the objects to humans (MacInnis and Folkes, 2017). This implies that the mental processes involved in thinking about human beings govern cognition regarding objects (Epley et al., 2007), prompting people to treat these entities as moral agents worthy of respect and empathy (Chandler and Schwarz, 2010). Several authors anticipate that brand anthropomorphism stands in relation to brand love because anthropomorphizing non-human entities makes them be perceived as "social entities" (Chandler and Schwarz, 2010;Chen et al., 2017). Thereby, it qualifies them as legitimate and reciprocal relationship partners (Waytz et al., 2010).
As far as brand anthropomorphism is the process of perceiving a brand as "like us" (i.e. human), it can foster self-brand integration through conferring social meanings and particular human-like traits to the brand (Ahuvia et al., 2009). This helps consumers to signal aspects of their own identities and to communicate to others the types of people they are or would like to be (Escalas and Bettman, 2005), in comparison to the situation where the anthropomorphism process does not occur and a brand is evaluated based on its attributes. Therefore, as a cognitive and perceptual strategy that endows a brand with a social meaning, this process facilitates the integration of the brand into the self. Thus, we hypothesized the following:

H1. Consumers exhibit higher levels of self-brand integration for anthropomorphized brands (vs objectified brands).
It may also create a stronger emotional brand connection. Fiske's (1982) schema-triggered affect theory states that some schemas or categories are associated with affective tags. When an object matched against a category, the stored affective or evaluative component associated with that category is cued. Generally, the human category is more positively EJMBE evaluated and viewed as superior compared to non-human entities (i.e. objects) (Rauschnabel and Ahuvia, 2014). Thus, positive emotions like affect and kindness (Schoefer and Ennew, 2005) are available immediately upon this categorization, as they are inherent to human beings. Therefore, according to schema-triggered affect theory, when a brand is placed into the human category positive emotions emerge because they are associated with that category. More formally put: H2. Consumers exhibit higher levels of positive emotional connections for anthropomorphized brands (vs objectified brands).
This study also predicted higher levels of anticipated separation distress when a brand is anthropomorphized. Judgments of non-social (e.g. products) and social entities (e.g. people) are dominated by different considerations (Chandler and Schwarz, 2010). While the interpretation and evaluation of social entities depend on inferred, abstract, psychological and social information, non-social entities are primarily evaluated with more-pragmatic reasons, such as their functions and concrete attributes. Based on this reasoning, the anthropomorphism process makes product attributes less relevant, with consumers paying more attention to those aspects (e.g. feelings and type of relationship) that characterize the interpersonal or social realm. Therefore, it moves consumers' attention away from pragmatic considerations and makes relational and psychological features more salient.
In the social realm, people are reluctant to replace close others (Chandler and Schwarz, 2010) and have a sense of personal loss and distress when these close others are absent (Thomson et al., 2005). Therefore, it is hypothesized that consumers are less willing to replace anthropomorphized brands because they may feel more reluctance and distress about being separated from them compared to objectified brands. Specifically: H3. Consumers manifest higher levels of anticipated separation distress for anthropomorphized brands (vs objectified brands).
Despite the general finding that consumers' assessments of products with anthropomorphic features are positive, empirical evidence has identified factors that condition the effect of anthropomorphism. These include the congruity between the features associated with products (human-like vs non-human-like) and the schemas under which products are presented (human or non-human) or the affective tags associated with human schemas McGill, 2007, 2012). This is an important consideration because accepting that the effect of anthropomorphism is not universally positive implies that the inferred human-like features associated with a brand, whether positive or negative, could benefit or harm the brand McGill, 2007, 2012). Therefore, it may trigger stronger or weaker consumers' reactions to the anthropomorphized brands.

Effects of consumers' liking for the anthropomorphized brand on brand love
In the context of interpersonal relationships, individuals' judgments of others affect the courses and natures of their relationships. In the same line, the quality, traits or characteristics of anthropomorphized brands also affect the brand relationships (Lau and Lee, 1999). In this sense, Fournier and Alvarez (2012) proposed the "brands as intentional agents frame" to suggest that the ratings of a particular brand as a relational partner determine the brand relationships experienced by consumers.
Accordingly, several studies have found that when people anthropomorphize a brand, they may view it as having both positive and negative human-like features (e.g. name, gender and physical characteristics), personality traits (e.g. warm, cold-hearted and intelligent) and intentions (e.g. trustworthy and unfair) (Fournier and Alvarez, 2012;Sweeney and Brandon, 2006). Based on these perceptions, an anthropomorphized brand is judged as a more-or less-likeable person.
Liking for the human version of a brand, as an evaluative global assessment of the type of person the brand has become (Anselmsson et al., 2008), denotes a certain fondness that people have toward it because it is thought to be a well-liked, pleasant and agreeable person. It is an attraction to the human version of the brand such that the consumer would desire to "be around" it because liking has long been believed to be a powerful human motivator for relationship development and maintenance. For example, people attempt to achieve physical closeness and move toward those they like and move away from those they dislike (Lott and Lott, 1972). Consequently, liking the anthropomorphized brand version might condition consumers' reactions because positive/negative evaluations produce immediate approach/avoidance behavior, respectively (Chen and Bargh, 1999).
Building on the above, those who like the human version of a brand will manifest higher levels of self-brand integration, positive emotional connection and anticipated separation distress. However, those who dislike it will manifest a pattern of contrast to facilitate getting away from the human version of the brand. It is proposed that: H4a. Consumers manifest higher levels of self-brand integration the higher their liking for the anthropomorphized version of a brand.
H4b. Consumers manifest higher levels of positive emotional connections the higher their liking for the anthropomorphized version of a brand.
H4c. Consumers manifest higher levels of anticipated separation distress the higher their liking for the anthropomorphized version of a brand.
Puzakova et al. have recently demonstrated that, in specific circumstances, the anthropomorphized versions of brands have a more detrimental impact on consumers' impressions, judgments and evaluations than the non-anthropomorphized ones. This occurs when humanized brands undergo negative publicity or product wrongdoing (Puzakova et al., 2013a), they require sensitive personal information from consumers (Puzakova et al., 2013b), they increase prices or are associated with situations of distributive injustice (Kwak et al., 2013. Taken together, these studies assert that negative brand information is perceived as more relevant when the brand is anthropomorphized because anthropomorphic representations trigger perceptions of the brand as a living agent with its own intentions, motivations and emotions (Aggarwal and McGill, 2007). This intentional nature leads to the perception that anthropomorphized brands are more responsible for their actions than the non-anthropomorphized ones, resulting in less-favorable consumers' reactions.
Thus, it seems that the human versions of brands that consumers dislike generate lower levels of the three key components of brand love than non-anthropomorphized brands do: H5a. Consumers manifest lower levels of self-brand integration when they dislike the anthropomorphized version of a brand compared to the objectified version.
H5b. Consumers manifest lower levels of positive emotional connection when they dislike the anthropomorphized version of a brand compared to the objectified version.
H5c. Consumers manifest lower levels of anticipated separation distress when they dislike the anthropomorphized version of a brand compared to the objectified version.

Study design and data collection
An experimental study with a one-factor between-subjects design was designed. The factor manipulated was the anthropomorphism process: individuals anthropomorphize the brand EJMBE (anthropomorphism condition); or they objectify it (no anthropomorphism condition). A web-based experiment was used to test the hypotheses. Participants were contacted through an online panel and were randomly assigned to one of the experimental conditions. A sample of 399 individuals participated (50.9 percent females, 56.1 percent workers, average age of 40.66 years).

Procedure
To establish a focal brand, participants were instructed to indicate the fashion brands they had bought in the past from a default set of 16 brands (this set was composed of the most-purchased fashion brands in Spain). Among the subset of brands bought by each individual, a specific brand was randomly and automatically selected by the system and assigned to the individual to later conduct the anthropomorphism task. By doing so, this procedure guaranteed that individuals' answers were not focused on their favorite brands.
In the anthropomorphism condition, participants had the instruction to imagine the brand coming to life as a person and to describe the sort of person the brand would be (personality, physical appearance, opinions, emotions and so on). In the condition where the individuals objectified the brand, the instruction was to describe the brand as a product (features, benefits, design, quality and so on).

Measures
Multi-item scales from the literature were used in a Likert format, scaled from 1 ( ¼ totally disagree) to 7 ( ¼ totally agree).
The scales used to measure the three dependent variables are the ones proposed by Batra et al. (2012). Specifically, self-brand integration was a second-order construct of 12 items reflected by current self-identity, desired self-identity, life meaning and frequent brand thoughts. Positive emotional connection was also conceived as a multidimensional construct that comprised intuitive fit (three items), emotional attachment (three items) and positive affect (three items). Finally, anticipated separation distress was assessed by four items.
The scale of liking for the human version of the brand consist on three items (Aggarwal and McGill, 2012) that were adapted to the purpose of this study.
As a manipulation check for anthropomorphism, and based on Waytz et al. (2010), the participants rated on a one-item seven-point semantic differential scale the degree with which they attributed human-like characteristics, behaviors and capacities to the assigned brand.
Brand love tendency was included as a covariate. It was assessed by asking the participants to name all the brands they loved in an open question (Rauschnabel and Ahuvia, 2014). An independent coder coded the number of mentioned brands. This number was used as a single indicator of the respondents' general brand love tendency. Finally, a few demographic measures were included. Table AI shows the wording of all items included in each measure scale.

Validation of scales and manipulation checks
Because the measurement of self-brand integration and positive emotional connection as second-order constructs was too complex for the purpose of this study, parcels as indicators of the sub-dimensions of these concepts were used (Little et al., 2002;Malär et al., 2011;Rauschnabel et al., 2013). By doing so, the multiple facets comprising both self-brand integration and positive emotional connection were preserved. Specifically, the average of each sub-dimension was estimated, and these values were used as indicators of the two higher-level constructs. Table I provides summary statistics, reliabilities and fit indices that prove that all the scales were statistically reliable. Discriminant validity was also confirmed, as in all cases, the value of 1 did not fall within the confidence intervals of the correlations between the factors and because each latent variable shared more variance with its respective indicators (i.e. the average variance extracted) than with its other correlated latent variables (i.e. the squared correlations between the latent variables) (Fornell and Larcker, 1981).
The suitability of the anthropomorphism manipulation was also verified. One-way ANOVA conducted across the two conditions of the experimental factor showed that significant differences existed across the two conditions (M anthro ¼ 4.88, M no anthro ¼ 2.70; F(1, 397) ¼ 131.790, p o0.00), confirming the appropriateness of the manipulation. To ensure that there were no differences in engagement with the task of describing the brand across both conditions, the word counts were compared. No differences existed between them (M anthro ¼ 41.92, M no anthro ¼ 44.89; F(1, 397) ¼ 3.67, p ¼ 0.056, p W0.05).

Effects of brand anthropomorphism
A MANCOVA test followed by univariate comparisons was used to test H1-H3 because the dependent variables were correlated. Brand anthropomorphism was used as a fixed factor, and the general tendency to love brands was included as a covariate.
The results indicated the existence of homogeneity of covariance (Box's M was not significant) and that the two levels of the fixed factor yielded significant differences across all dependent variables (F(3, 394) ¼ 3.021, p o0.05 (Wilks' λ ¼ 0.978)). The Levene's test results supported the assumption of equal variance for the three dependent variables, and the effect of this relation was small at η 2 ¼ 0.022 (Cohen, 1988). The univariate tests revealed that significant differences existed only for (self-brand integration and positive emotional connection) at p o0.05 between the conditions of the fixed factor (see Table II). Therefore, these results gave support to H1 and H2. However, we could not find support for H3.

Construct
No. of items M (SD) Cronbach's α SCR AVE

EJMBE
The levels of anticipated separation distress were similar when individuals anthropomorphized or objectified the brand.

Effects of liking for the anthropomorphized version of the brand
To test H4, another MANCOVA was conducted with only those respondents assigned to the anthropomorphized condition (n ¼ 300). The sample was first split at the median on the liking score (4.25 on a seven-point scale). This resulted in a low liking group of 156 individuals and a high liking group of 144 (M low fit group ¼ 2.82 M high fit group ¼ 5.44 F(1, 298) ¼ 580.665, po0.00). To test H4, the median-split liking factor was used as a fixed factor.
The assumption of homogeneity of covariance was not satisfied (Box's M was significant at p o0.05). The results indicated that the two levels of the liking variable yielded significant differences across the three dependent variables (F(3, 295) ¼ 64.701, p o0.00 (Pillai trace ¼ 0.397), η 2 ¼ 0.397).
As depicted in Table III, the ANOVA test showed significance differences between the two media groups in terms of all dependent variables. Because the data violated the homogeneity assumption in the case of two of them (self-brand integration F(1, 298) ¼ 6.323, p o 0.05; anticipated emotional distress F(1, 298) ¼ 8.873, p o 0.05), Welch and Brown-Forsythe tests were used to further examine the significance of the results. The results were still significant for both dependent variables. These results gave empirical support to H4a-H4c. Thus, consumers manifest higher levels of self-brand integration, positive emotional connection and anticipated separation distress when they have a higher liking for the anthropomorphized version of the brand.
To further examine the effects of brand anthropomorphism on the three dimensions of brand love, a one-way ANOVA with post hoc tests was conducted to analyze the differences between the three conditions (see Figure 1): the non-anthropomorphized brand, and low or high liking for the anthropomorphized version of the brand.  With self-brand integration as a dependent variable, Games-Howell post hoc testing was used because of the inequality of variances. The ANOVA test showed significant differences between the three media groups (F(2, 396) ¼ 82.595, p o0.00). The data violated the homogeneity assumption, but the Welch and Brown-Forsythe tests indicated that the result was still significant (see Table IV ). The same pattern of results was found for both positive emotional connection and anticipated separation distress. Taken together, these results provided empirical support for H5a-H5c.

Theoretical implications
Past research dealing with the antecedents of brand love has considered the brand as an object that provides hedonic and utilitarian value and that satisfies self and social expression needs. The central shortcoming of this approach is that it neglects the view that the brand can surpass an object nature and be regarded as a human entity. By taking this view, the quality of the consumer-brand relationship could be improved as suggested by the anthropomorphism literature. Despite its importance, few studies have focused on the relationship between brand anthropomorphism and brand love (see Hegner et al., 2017;Rauschnabel and Ahuvia, 2014). The current research differs from these studies in three ways. First, they focused on brands for which consumers had strong preferences, while this study analyzed more-neutral brands in terms of affection. Second, they analyzed anthropomorphic tendency, which is an individual trait that varies among individuals, but it differs from the anthropomorphism process (Letheren et al, 2016). While the latter is a cognitive bias of assigning human traits to an object, the former is the tendency to do that. Third, in contrast to the current research, they did not analyze the superior benefits of the humanization of a brand for brand love compared to the view of a brand as a non-social entity.
In addition, compared to past studies that have focused on the passionate dimension of brands (see Albert and Merunka, 2013;Huber et al., 2015), this research emphasized other specific dimensions of brand love such as self-brand integration, positive emotional connection and feelings of anticipated separation distress.
Furthermore, the empirical evidence was obtained from a broader range of brands in a product category (clothing) less complex compared to other products. For example, consumer electronics and technological products have a very interactive nature because they are highly responsive to their users, making them somewhat anthropomorphic (Aggarwal and McGill, 2007), which may favor brand love.
In keeping with prior work on anthropomorphism, this study found converging evidence that if the human version of a brand is not positively evaluated, the process of brand anthropomorphism may not necessarily produce the kinds of consumer reactions expected. Brands that are imagined as human but lack attractiveness are evaluated less positively on the three dimensions of brand love compared to those brands that are imagined as products.
Taken together, these results suggest that, consumers' reactions toward a brand depend not only on the schemas with which consumers imagine the brand but also on the affective evaluation associated with the humanized brand. In general, the social or human schemas of the brand exert a stronger positive effect on consumers, surpassing the effect of non-human schemas only when consumers like the human version of the brand.
Finally, the current research focused on imaginative anthropomorphism, in which imagination plays a central role. The concept of imagination is widely used in consumer research, but Huang and Mitchell (2014) highlighted that there is a scarcity of literature regarding its role in brand personification and brand relationships.

Managerial implications
The issues addressed also have managerial implications to build and maintain brand love relationships with consumers. First, brand managers have traditionally stimulated brand anthropomorphism by specific brand personification strategies using visual stimuli and specific product designs. The current research suggests that another way to encourage the humanization of a brand is by prompting consumers' imaginations. The empirical findings show that inducing consumers to imagine a brand as a person can be strategically important because they integrate more of the brand into the self and develop stronger emotional connections compared to a situation in which a brand is imagined as a product. One way to encourage consumers' imaginations could be with brand storytelling, which is viewed as a powerful communication tool to bring brands to life and provide them with personalities (Delgado-Ballester and Fernandez-Sabiote, 2016). Embedding the brand in a real or fictional story in which it adopts specific archetypal figures in the story (e.g. the hero, the antihero or the mother of goodness) gives the brand a voice to express its story to consumers, helping them to humanize it. Another way is using a brand personification strategy to activate the anthropomorphism process. This strategy consists of attributing certain human characteristics to the brand (Epley et al., 2007) in such a way that the brand is presented in marketing campaigns in an anthropomorphized form (Delbaere et al., 2011).
Second, brands differ in how naturally they are humanized because the interactive nature of certain products makes some brands more anthropomorphic than others (Batra et al., 2012). Therefore, for brands in less-anthropomorphic product categories, it would be interesting to favor brand interactions with consumers, for example, through social media. Channels such as Twitter, Facebook or Instagram allow direct brand interactions with consumers, making brand behavior a crucial attribute to infer human brand traits.
Third, consumers' judgments of humanized brands lead to even stronger consumerbrand connections if these judgments are positive. Consumers show higher feelings of heartbreak and distress when anticipating being distanced from a humanized brand. These impressions of a brand as a person can be based on the many observed behaviors of the people associated with the brand (e.g. celebrity endorsers, the company's employees, the CEO and the actual people who buy and consume it). As a general recommendation, managers might track consumers' opinions about the behavior, performance and traits of those who act on behalf of a brand. This is especially relevant when consumers have to interact with brand employees (e.g. branded retailers, hotels, airlines and hairdressers).
Managers are also advised to analyze stereotypes about brand users because they may serve as reference points to form impressions about a brand's personality and, consequently, to facilitate the integration of the brand into consumers' self and to strengthen their emotional links with it. In case of the existence of negative stereotypes about brand users, it would be better not to depict them on advertisements but to emphasize the functional characteristics of the brand as a product. Meanwhile, managers should create desired stereotype perceptions as part of their strategic brand positioning.

Limitations and further research
This study had some limitations. It adopted a static approach in addressing the effect of anthropomorphism by not considering the idea that consumer-brand relationships inevitably change over time. For example, Huber et al. (2015) demonstrated that, over time, utilitarian aspects become more substantial because a relationship subtly evolves into a more rational partnership. This being so, the effect of the human brand schemas on brand love (compared to product schemas) may decrease as the relationship matures. This question awaits further research.
Finally, this study focused on three dimensions of brand love. However, these three aspects might also exist in various other relationships that are not necessarily love relationships.
Beyond these limitations and the associated research avenues, several open questions suggest promising routes. First, future studies could take brand stereotypes into account. Ivens et al. (2015) and Kim and Kramer (2015) demonstrated that brand roles (warmth vs competence, servant vs partner) shape consumers' feelings and behaviors toward a brand. It would be interesting to analyze whether humanizing a brand is a more effective mechanism to transmit these brand roles and to influence subsequent consumer behavior via behavioral priming. In other words, do different brand roles operate in a similar manner for humanized and objectified brands?
Second, the product category (clothing) used was not complex. As Hart et al. (2013) demonstrated that the more complex the product, the more inclined consumers are to anthropomorphize it, complexity may also condition the relationships analyzed.
Third, brand anthropomorphism was encouraged by a task based on participants' imagination to think about a brand in human terms. As recent studies (see Huang and Mitchell, 2014) have recognized that not all individuals have the same imagery ability, this ability may moderate the relationships analyzed. Furthermore, as far as some brands have stronger and clearer brand personalities than others, brand personification may also exert a moderating effect. Connected to this idea, further research may explore whether brand humanization would be more beneficial for functional products than for symbolic products. Huang and Mitchell (2014) suggest that symbolic products do not require as much imagination as utilitarian products to personify a brand. This being so, it seems that for utilitarian products, brand anthropomorphism would be more beneficial than for symbolic products.
Another worthwhile research avenue would be to examine other moderators such as individual traits and consumer goals. Recently, Ghuman et al. (2015) suggested the existence of cross-cultural differences in people's tendencies to anthropomorphize, with consumers in collectivist societies more likely to anthropomorphize brands because they have more-accessible human knowledge. In addition, Puzakova and Aggarwal (2018) found that anthropomorphism has a detrimental effect on evaluations of a distinctive brand when consumers' distinctiveness goals are heightened.
Finally, imagination is a form of mental representation that expresses unreality and involves creativity (Huang and Mitchell, 2014); as past research has suggested that interpretations of imagination differ as a function of age, it would be desirable to learn the extent to which the results obtained could be replicated across different age groups.