Testing the concept of Brand Personality in Spontaneous Associations - A critique on the insights provided by brand personality associations

i Testing the concept of Brand Personality in Spontaneous Associations Bárbara Inês Dias Almeida A critique on the insights provided by brand personality associations Dissertation presented as partial requirement for obtaining the Master’s degree in Statistics and Information Management ii NOVA Information Management School Instituto Superior de Estatística e Gestão de Informação Universidade Nova de Lisboa TESTING THE CONCEPT OF BRAND PERSONALITY IN SPONTANEOUS ASSOCIATIONS A critique on the insights provided by brand personality associations by Bárbara Almeida Dissertation presented as partial requirement for obtaining the Master’s degree in Information Management, with a specialization in Marketing Research & CRM Advisor / Co Advisor: Diana Orghian, PhD November 2021 iii DEDICATION I dedicate this project to my family, specially my parents and sister, who are always here to support me and always believe in me. I also want to thank my friends Ana Guerreiro, Luísa Pereira and Cristiana Correia, and partner, Cláudio Nunes, who have helped me throughouth my studies and shared their own experiences with me. Lastly, I want to thank my advisor, who gave me the freedom to explore this topic, while providing useful and leading insights in my research. iv ABSTRACT The concept of Brand Personality has been explored considerably over the years. However, previous research seems to lack integration and, most importantly, practical applicability in Marketing. Our research explores the evolution of the concept, further extending the critique on the methodology of previous studies, and the integration and redundancy of brand personality in the studies of brand associations. In a free association study performed with fifteen brands from different categories, we successfully confirm that consumers rarely make personality associations spontaneously, while they tend to create a greater number of personality associations when conditioned to do so through projective methods. KEYWORDS Brands, Brand Associations, Brand Category, Brand Identity, Brand Image, Brand Personality, Brand Positioning, Free Association, Human Brand Associations, Marketing Strategy, Personality, Projective Methodology, Projective Techniques, Symbolic, Utilitarian. v INDEX Introduction ................................................................................................................................... 1 Literature review ........................................................................................................................... 3 1.1. Brand personality – evolution of the term ....................................................................... 3 1.2. Projective Techniques ...................................................................................................... 5 1.3. The integration of the concept of Brand Personality in Brand studies and literature ..... 7 1.3.1. Brand Associations .................................................................................................... 7 1.3.2. Human Brand Associations (HBAs) ........................................................................... 9 1.3.3. Symbolic and Functional Positioning of Brands ........................................................ 9 Methodology ............................................................................................................................... 11 1.4. Procedure ....................................................................................................................... 11 1.5. Participants ..................................................................................................................... 14 1.6. Turning free association results into numeric values ..................................................... 14 Results and discussion ................................................................................................................. 15 1.7. Condition Main Effect ..................................................................................................... 15 1.8. Effects of Symbolic and Functional Positioning of Brands ............................................. 16 1.9. Covariate Effects: Loyalty, Liking, Experiences ............................................................... 18 Conclusions .................................................................................................................................. 20 Limitations and recommendations for future works .................................................................. 22 REFERENCES ................................................................................................................................ 23 Appendix...................................................................................................................................... 27 vi LIST OF FIGURES Figure 1 - Comparing the factors of Brand identity (on the left) with Brand Personality Drivers (on the right). .............................................................................................................................................. 8 Figure 2 - Spontaneous Condition, free association. The word "personality" is not referred. ............. 11 Figure 3 - Priming Condition, free association using projective measures. The word "personality" is referred. ........................................................................................................................................ 12 Figure 4 - Measuring brand attitudes (loyalty, liking, good experiences). ............................................ 13 vii LIST OF TABLES Table 1 - Comparing human personality and brand personality ............................................................. 4 Table 2 - Classification of words resulting from respondents' free word association task................... 14 Table 3 - Descriptive statistics for the 15 dependent variables (personality means) ........................... 15 Table 4 - Brand Positioning.................................................................................................................... 16 Table 5 - Descriptive statistics for brand positioning as identified by respondents ............................. 17 1 INTRODUCTION As objects of value for people, brands have proven to simplify the load in choices between alternatives, building relationships and trust with the consumer through repeated consumption (Bagozzi et al, 2020). Besides, brands and products serve as a representation of the self, our status, ideologies, lifestyle, and how we want to be perceived in general. How customers perceive brands, and specially the concept of brand personality, has an effect on their motivation to buy and engage with those brands, mainly when we talk about the belonging needs, esteem needs, or the self-actualization needs from Maslow’s Hierarchy of Needs (Maslow, 1943). There have been several studies stating that we learn, define, and remind ourselves of who we are by our possessions, supporting that the driving force that motivates consumers to purchase goods and services is the need to express their own self, their own personality, through the use of a brand (Sirgy, 1986; Belk, 1988). Personality traits can be associated with a brand in a direct way by the people that represent the brand, such as the brand’s product endorsers, the company’s employees or CEO, and the brand’s user imagery. Therefore, it is expected that the personality traits of the people associated with the brand are transferred directly to the brand (McCracken, 1989). Many researchers have explored the concept of how personality is related to consumption and to how consumers respond to brands, many times referring to terms such as brand awareness, brand associations, brand image, brand attitudes, brand attributes, or brand personality. The latter is formally defined by Aaker’s pioneer study in 1997 as “the set of human characteristics associated with a brand”, which are expected to differentiate the brand from the competition, creating competitive advantage. Aaker (1997) studied the dimensions of brand personality, suggesting that they operate in different ways than human personality dimensions, and developed a Brand Personality Framework (BPF) with 5 dimensions: Sincerity, Excitement, Competence, Sophistication and Ruggedness. Sincerity refers to brands that are down-to-earth, honest, wholesome and cheerful; excitement refers to brands as being daring, spirited, imaginative and up to date; competence refers to being reliable, intelligent and successful; sophistication comprises characteristics like being charming and upper-class; finally, ruggedness refers to a rugged and outdoorsy brand. Over the years, Aaker’s Brand Personality model has been used as a basis for brand personality studies, with some adaptations done by, for example, considering only the most relevant dimensions, using a reduced scale validated by other studies, using expert judgement to select some items from the scale, or by adding or adjusting items in order to create a better fit to the object of scrutiny, whether it is a specific store, geographical area, and others (Aaker et al., 2001; Ekinci & Hosany, 2006; Murphy et al.,2007) – nonetheless, most studies prove the generalizability of Aaker’s framework. 2 When thinking about the business applicability of the term, researchers of Brand Personality tend to describe the concept as important to help brand strategists gain insights on how consumers perceive brands, how they might use brands to express themselves, and ultimately, to whom should the brand’s strategy and communication effort be directed for. But what do most studies have in common? In all of them, projective techniques were applied, in which respondents were given a very specific stimulus to identify personality traits, whether it was a direct, open-ended question (e.g. Aaker, 1997: “[…] write down the personality traits that first come to mind when thinking about […]” a brand; Ekinci & Hosany, 2006: “[…] think of the last tourism destination that you have been to as if it were a person.”) mainly basing their methodology on Aaker’s BPF, or were requested to classify on a scale how much they identified a brand with a number of given personality traits (e.g. Murphy et al., 2007: subjects were asked to indicate on a 1 to 5 scale how strongly they associated with the destination 5 brand dimensions and 15 corresponding brand facets). Aaker’s BPF does have a wide applicability, however, in the study of Oklevik, Supphellen, and Maehle (2020), research reveals that in a qualitative study of free associations, there is practically nonexistence of personality traits ascribed to brands, describing the major problems that arise from brand personality research over the years. Motivated by the research of Oklevik, et al. (2020), we question as to whether or not consumers ascribe humanlike traits to brands when they process information about the brand in a spontaneous way. If not, are these scales really helpful for brand strategists, or should there be a less redundant and easier way to gain insights on how customers perceive brands, whether we are talking about their human-like traits, or not. In this study, we explore the evolution of brand personality studies and frameworks, the arguments of brand personality as a metaphor (Avis et al, 2012), and how, ultimately, the use of projective methods in most research regarding brand personality may lead to an inflated level of perceived human traits in most associations people do regarding brands. 9 depiction of the grille of a car, coupled with a depiction of the shape and angle of the headlights, is enough to impact perceptions of the brand's aggressiveness. Another example is the “Butterfinger Story” (Aaker, 1996), where the perception of the brand Butterfinger (which was perceived as being outdated) was immediately changed when Nestlé started using the irreverent character Bart Simpson as the new face of the brand, strengthening this candy bar brand’s relationship with its target market. The same change could occur when a brand adjusts the look and feel of their stores, the attire of their store workers, among other things. That said, Oklevik’s et al. (2020) suggestion of analysing Human Brand Associations (HBAs) would be much more applicable, and diminish the redundancy of any analysis made to brand identity factors. 1.3.2. Human Brand Associations (HBAs) Olkevik et al. (2020) suggest a new approach, based on Associative Network Theory (ANT) (Anderson, 1983; Collins & Loftus, 1975) – this theory explains that brand associations vary according to the frequency with which any specific association is activated in internal and/or external thought processes. Brand Personality associations, or humanlike characteristics associated to brands, may play a role when personal figures represent the brand. The more often these personal figures are exposed to consumers, the more their human characteristics become a part of their thought processes about the brand. Oklevik et al. (2020) suggest that there are three main types of HBA: brand representatives, brand characters and stereotyped brand users. Despite needing further research regarding methodology, this approach leaves the assumption that consumers see brands as humans, rather assuming that consumers associate brands with certain human figures, which in turn are associated with specific human characteristics. There’s the example of Nespresso, being associated with George Clooney – maybe some consumers, in a free-word association, will associate the brand directly with the actor, while others may come up with personality traits that describe the actor, as a way of describing the brand (brand character). Another example is having a friend that is long time user of some specific brand – after a while, we would start associating that brand with that friend, thus associating their personality traits with the brand (stereotypical brand user). 1.3.3. Symbolic and Functional Positioning of Brands Since the concept of Brand Personality also fits into the definition of brands as a passive object with utilitarian and symbolic meaning (Bagozzi et al., 2020), it is important to explore the Symbolic and 10 Functional positioning of brands and how does this positioning contribute to the perception of Brand Personality. It is widely accepted that consumers go to extensive efforts to seek information not only depending on the type of involvement of the purchase decision, but also for purchases that may have an impact on the expression of their lifestyle and self-expression. This results in two main dimensions that are considered in the purchasing process, those being the purchase-decision involvement and the think-feel purchase type. These dimensions form one of the most popular product-classification schemes in advertising and consumer research: the FCB grid. The purchase-decision involvement is more related to how actively consumers seek information about the product and how they engage in brand or product comparisons (Houston and Rothschild, 1977). For this study, we will focus mainly on the think-feel dimension of the FCB grid, in which the perception of brand personality has a higher influence, since one of the measurements used in this dimension is how much the decision to buy some type of product expresses one’s personality and is related to consumers rational or emotional decisions. Previous studies show that consumers are more likely to make rational decisions based on the evaluations of utilitarian benefits offered by different brands when purchasing think products, and more likely to make emotional decisions based on evaluations of hedonic-affective (or symbolic) attributes of various brands when purchasing feel products (Cheong & Cheong, 2020). Products’ utilitarian benefits can include some form of functional or practical values – products bought for “utilitarian needs”, meaning the consumer cares more about functional performance – whereas symbolic attributes may include more sensory or experiential values – products bought for “symbolic needs”, when the purchase is driven by the consumer’s gratification, social acceptance, or sensory stimulation (Ratchford, 1987; Dhar & Wertenbroch, 2000). A consumer’s buying process tends to include both utilitarian and symbolic considerations, because although the degrees of both forms of considerations may vary according to the product type, there’s always some degree of symbolic consideration. One of Aaker’s criteria for what brands to include in her study was to include brands of the following categories: symbolic, utilitarian and both symbolic and utilitarian (as identified by Ratchford, 1987), and for the sake of a contrast evaluation between this and previous studies, we also used this criterion. The most likely scenario that we hypothesize is if a brand is perceived as a symbolic, or symbolic and utilitarian brand, it is more likely to be characterized with personality terms: H2: If the brand is perceived as symbolic, or symbolic and utilitarian, it is more likely to be characterized with personality terms. 11 METHODOLOGY 1.4. PROCEDURE We conducted an experimental study with two randomly assigned conditions, for 15 brands. The following instruction appeared on the beginning of the questionnaire, independently of the assigned condition: “Next, you’ll have questions regarding 15 brands. You don’t need to use a lot of time thinking about the following questions, just try to answer with the first thing that comes to your mind when you think about each of these brands.” Regarding each condition, one of them was our Spontaneous Condition – a free three-word association asking about 3 things that first come to mind when seeing the brand’s logo and name (in Figure 2 there’s an example of the Spontaneous Condition for the brand Nike). That was our control condition. The other, our Priming Condition, was a free three-word association in which respondents were shown the term “personality” when asking about considerations or opinions about the brands (in Figure 3 there’s an example of the priming condition for the brand Nike). Figure 2 - Spontaneous Condition, free association. The word "personality" is not referred. 12 There was a recommendation in the beginning of the questionnaire for respondents to change the displayed language to Portuguese if they were Portuguese, to make word association easier and more reliable in both conditions. The objective of this questionnaire was to have the two conditions being as similar as possible, practically only differing on the use of the word “personality”, as observed in figures 2 and 3. This way we could understand if consumers do ascribe humanlike traits to brands when they process information about the brand in a spontaneous way, by associating brands with a similar number of personality traits independently of the condition, or if these traits are only referred given the specific question about the brand’s personality. Additionally, we tried to understand how consumers attitudes towards the brand might influence their behaviour/characterization. Using a 9-point Likert scale, we asked respondents to indicate how often they bought from the brand (which we called loyalty), if they liked the brand or had good experiences with the brand (question in figure 4) – those questions didn’t change for each condition. Figure 3 - Priming Condition, free association using projective measures. The word "personality" is referred. 13 If respondents stated they didn’t know the brand in the first question (in figures 2 and 3), they wouldn’t be presented with the second question, regarding word association, nor the questions in Figure 4. We also asked if the respondents worked on some of these brands, trying to validate if working in a company could increase the chances of someone seeing the brand as intrinsically having a personality, since they would have a closer relationship with the companies values and customer message – this question didn’t get more than 2 positive responses, so it was ignored in our analysis. Finally, we asked respondents to characterize each of the 15 brands as Symbolic, Utilitarian, or Utilitarian and Symbolic, further asking them to rate their understanding of the terms “Symbolic” and “Utilitarian”. The fifteen brands used on this questionnaire were chosen on a basis of popularity, especially for the Portuguese consumer, since it is the country where the study takes place. The objective was to study three brands of each of the three categories included in Aaker’s study (1997): symbolic (e.g., jeans, cosmetics, and fragrance), utilitarian (e.g., computers, electronics, and appliances), and both symbolic and utilitarian (e.g., automobiles, beverages, and athletic shoes). We chose 15 brands from a list of 95 brands, compiled from 3 different sources: Forbe’s annual list of the world’s most valuable brands for 2020 (Forbes, 2020), a study done by Havas Group, ranking the 50 most Meaningful Brands for Portugal (Havas Group, 2019), and the Brand Finance ranking of the most valued Portuguese Brands in 2019 (Brand Finance, 2019). We merged the three rankings, categorizing each brand as Symbolic, Utilitarian or both Symbolic & Utilitarian, and proceeded to choose 5 brands of each category, resulting in the following list of brands to be presented to respondents: Figure 4 - Measuring brand attitudes (loyalty, liking, good experiences). 14 • Symbolic – Disney, Louis Vuitton, Marlboro, Nike and Superbock; • Symbolic & Utilitarian: Apple, Coca-Cola, Mercedes-Benz, Starbucks and Zara; • Utilitarian; Asus, Continente, Gillette, Ikea and Olá. Our subjects were randomly assigned to either one of both conditions and, within each condition, the order of the questions regarding each brand (for all 15 brands) was also randomized. You can find the model of the questionnaire for both conditions in the Appendix. 1.5. PARTICIPANTS The questionnaire was published on social media (Facebook, Instagram and LinkedIn), and passed on individually to various communities and groups, for one month period. The sample used in this study was originally 168 respondents, which ended up being reduced to 119, excluding very incomplete answers related to respondent fatigue – only answers that filled more than 50% of the 45 words (3 words per brand) required to complete the free word association for both conditions were considered. The sample had nearly 48% female respondents, the average age of respondents was 28.62 years old (SD = 8.544), and more than 50% of the sample had a bachelor’s degree or higher. 1.6. TURNING FREE ASSOCIATION RESULTS INTO NUMERIC VALUES As a result of the sample’s free word association for both conditions, there were about 4159 terms to analyse (an average of 40 terms per person). All 4159 words were labelled as either a personality term (PERS), or a non-personality term (NO_PERS), which were then turned into 1 and 0 (for PERS and NO_PERS, respectively), in order to calculate means, and other statistical estimates. The requirements to consider each word as a personality term were the following: the term had to be an adjective, applicable to people or applicable to brands considering Aaker’s (1997) scale items. All terms were classified independently of the condition or brand they were applying to, in order to minimize errors and possible bias – please bear in mind that, in a traditional investigation, there would be a specific team or even blind coders to perform this task and reduce classification error and bias, but for the context of this dissertation there was only one person performing the coding of all personality terms. Examples of the classification of the words can be seen in Table 2. Table 2 - Classification of words resulting from respondents' free word association task Label Example of terms provided by respondents PERS (1) Amigável, Cool, inovador, inovadora, moderna, arrogante, comunicativa, jovem, original, sneeky, observadora, exigente, innovative, atraente, atrativa, snob, trendy, gananciosa, capitalista. NO_PERS (0) Diferenciada, qualidade, cara, pouco acessível, publicidade criativa, simples de conceito, durabilidade, rapidez, pouca originalidade, premium, funcional, eficácia, design, desonestidade, inovação 15 RESULTS AND DISCUSSION 1.7. CONDITION MAIN EFFECT When trying to understand if using projective techniques, by directly asking people about a brand’s personality, would cause respondents to identify more personality traits than when the term was not mentioned, the analysis showed that the average of identified personality terms was higher for the priming condition, that showed the word “personality”, than for the condition that didn’t mention the word “personality” (spontaneous condition, see Table 3), considering PERS=1 and NO_PERS = 0. Our dependent variable is the mean of personality terms identified. Table 3 - Descriptive statistics for the 15 dependent variables (personality means) Brands Spontaneous condition (M, SD) Priming condition (M, SD) Apple M = 0.12, SD = 0.21 M = 0.48, SD = 0.34 Asus M = 0.09, SD = 0.19 M = 0.37, SD = 0.34 Continente M = 0.08, SD = 0.34 M = 0.34, SD = 0.39 Disney M = 0.15, SD = 0.26 M = 0.49, SD = 0.41 Gillette M = 0.12, SD = 0.23 M = 0.42, SD = 0.40 Ikea M = 0.08, SD = 0.16 M = 0.36, SD = 0.37 LV M = 0.19, SD = 0.25 M = 0.38, SD = 0.36 Marlboro M = 0.21, SD = 0.24 M = 0.42, SD = 0.40 Mercedes M = 0.14, SD = 0.20 M = 0.42, SD = 0.35 Nike M = 0.19, SD = 0.25 M = 0.60, SD = 0.32 Olá M = 0.13, SD = 0.23 M = 0.52, SD = 0.36 Starbucks M = 0.16, SD = 0.23 M = 0.57, SD = 0.37 Superbock M = 0.10, SD = 0.21 M = 0.48, SD = 0.43 Zara M = 0.22, SD = 0.30 M = 0.47, SD = 0.38 Coca Cola M = 0.18, SD = 0.28 M = 0.50, SD = 0.38 We performed a one-way between subject ANOVA to compare the effect of projective measures on the attribution of personality terms for the Spontaneous Condition, and Priming Condition. There was a significant effect of projective measures on the attribution of personality terms at p < .05 for both conditions [Fmean_pers(1,117) = 33.415, p < .001], hence confirming H1. From this, we can postulate that, when using projective techniques by directly asking people about a brand’s personality, people will identify more personality traits than when the term is not mentioned. 16 These results confirm Avis and Ferguson theory (2014), showing that the projective methodologies used in most Brand Personality studies are able to “create” personality identification stimuli. Also, we validate and further extend on Oklevik’s research (2020), being able to assert that consumers are less likely to identify personality traits in brands in a spontaneous way. 1.8. EFFECTS OF SYMBOLIC AND FUNCTIONAL POSITIONING OF BRANDS Regarding the Symbolic and Functional positioning of brands, we had the original classification from which we selected the five brands as shown in Table 4 – A, and also the classification according to our respondents as show in Table 4 – B. Table 4 - Brand Positioning Symbolic Symbolic & Utilitarian Utilitarian A. Original Classification used to choose the 15 brands Disney Apple Asus Luis Vuitton Coca-cola Continente Marlboro Mercedes-Benz Gillette NIKE Starbucks Ikea Superbock Zara Olá B. Classification according to survey respondents (based on frequency tables) Disney Apple Asus Luis Vuitton Coca-cola Continente Marlboro Mercedes-Benz Gillette Starbucks NIKE Ikea Olá Superbock Zara When grouping the brands for these classifications, either attributed originally (Table 4-A) or by respondents (Table 4-B), we ran the one-way between subject ANOVA again, to compare the effect of projective measures on the attribution of personality terms for the Spontaneous Condition, and Priming Condition. For each category of the original classification, there was a significant effect of projective measures on the attribution of personality terms at p < .05 for both conditions [FSymbolic_brands(1,107) = 36.713, p < .001; FUtilitarian_brands(1,108) = 42.265, p < .001; FSymb.&Util._brands(1,108) = 39.879, p < .001]. For the respondents’ classification there was also a significant effect of projective measures on the attribution of personality terms at p < .05 for both conditions [FCustomerSymbolic_brands(1,105) = 32.153, p < .001; FCustomerUtilitarian_brands(1,108) = 37.263, p < .001; FCustomerSymb.&Util._brands(1,109) = 46.689, p < .001.] Additionally, according to the descriptive statistics for the respondents’ classification, Symbolic and Symbolic & Utilitarian brands had a higher mean of personality term identification, while Utilitarian brands had a lower mean comparatively to other brand positionings in both scenarios (see Table 5). 17 Table 5 - Descriptive statistics for brand positioning as identified by respondents Brands Spontaneous condition (M, SD) Priming condition (M, SD) Customer_Symbolic M = 0.18, SD = 0.19 M = 0.46, SD = 0.31 Customer_Symbolic&Utilitarian M = 0.18, SD = 0.19 M = 0.48, SD = 0.27 Customer_Utilitarian M = 0.12, SD = 0.14 M = 0.40, SD = 0.31 Hence confirming our expectations that, for Symbolic or Symbolic & Utilitarian brands, the probability of being characterized with more personality terms is higher than for Utilitarian brands. To measure the statistical significance of the affirmation above, we started by performing a pairedsamples t-test to compare the attribution of personality terms in the top 3 symbolic brands (the 3 brands with higher classifications as symbolic by respondents – Disney, Louis Vuitton, and Marlboro) and the top 3 utilitarian brands (the 3 brands with higher classifications as utilitarian by respondents – Asus, Continente, Gillette). There was a significant difference in the scores for the attributed personality terms for the top 3 symbolic brands (M = 0.30, SD = 0.30) and for the top 3 utilitarian brands (M = 0.23, SD = 0.29); t(96) = 2.43, p = 0.017. We then measured the same effect, this time for all the brands that respondents characterized as either symbolic and symbolic & utilitarian (a total of 8 brands: Louis Vuitton, Disney, Marlboro, Starbucks, Mercedes, Apple, Coca-Cola, and Nike), or as utilitarian (a total of seven brands: Continente, Asus, Gillette, Ikea, Olá, Zara and Superbock). This time, the paired-samples t-test also resulted in a significant difference in the scores for the attributed personality terms for all symbolic and symbolic & utilitarian brands (M = 0.32, SD = 0.27) and for all utilitarian brands (M = 0.25, SD = 0.28); t(107) = 3.60, p < 0.001. Hence, there’s confirmation of H2 (if the brand is perceived as symbolic, or symbolic and utilitarian, it is more likely to be characterized with personality terms), and our analysis supports previous studies which suggest that personality is mostly associated with symbolic brands (Cheong & Cheong, 2020). However, despite proving the significance of our hypothesis, and turning our attention to a more in-depth analysis of the descriptive measures from table 3, we want to try and explain why we had higher significance when analysing all brands, instead of the top 3, since the latter would be expected to have more significant values. Primarily, we can see that the priming condition exacerbates the dispersion of data for some brands. For example, Louis Vuitton, one of the top 3 symbolic brands (which is classified as Symbolic by the majority of respondents), has one of the lowest attributions of personality terms, and Olá 18 (which is classified as Utilitarian by the majority of respondents), has one of the highest attributions of personality terms. This dispersion can probably be explained through respondents’ familiarity with the brand (knowing the brand and already having used its products, or knowing the brand but never having used its products). Firstly, for Louis Vuitton, our analysis showed that the majority of respondents have never used the brand (more than 80% stated that they knew the brand, but had never used it); additionally, given its status as a luxury brand, it is not a brand that the average consumer sees advertised in common marketing channels, there’s not too many stores and there are not a lot of people using the brand. Hence, it is a brand with very few associations regarding HBAs in consumers’ minds. Then we have the example of Olá, which can be considered as a staple brand in Portugal, that is very easily activated in consumers’ minds (98% of respondents know and have used the brand). Hence, when asked to describe the personality of the brand, despite being utilitarian, consumers will find it easier to associate with any type of HBAs. Thus, these exceptions can also help prove Okleviks’ et al. (2020) HBA approach, that human associate brands with certain human figures, which in turn are associated with specific human characteristics, since this association is easier for brands from which consumers can easily identify human association cues (such as Olá, which is highly advertised and sold in many places). 1.9. COVARIATE EFFECTS: LOYALTY, LIKING, EXPERIENCES Finally, the respondents were also questioned about their loyalty and experiences with each brand (figure 4), as a way of trying to understand if these variables could act as covariates in our model – if people bought the brand more often, would they identify more personality traits for that brand? The same question applies to liking the brand or having good past experiences with the brand. A One-way ANCOVA was conducted to determine a statistically significant difference between the Priming Condition or Spontaneous Condition on the number of personality terms identified, controlling for the following variables: • Loyalty (Likert-scale of 1 to 9: “How often do you buy from [brand name]?”) • Liking (Likert-scale of 1 to 9: “I like the brand [brand name].”) • Experiences (Likert-scale of 1 to 9: “I have good past experiences using the brand [brand name].”) After controlling for the three variables above, there was a confirmed significant difference in the mean of attribution of personality terms for each condition [F(1,112) = 36.441; p < .001]. 25 MacInnis, D. J., & Folkes V. S. (2017). Humanizing brands: when brands seem to be like me, part of me, and in a relationship with me. Journal of Consumer Psychology, 27(3), 355–374 Malone, C., and Fiske, S. T. (2013). The human brand: How we relate to people, products and companies. New York, NY: Jossey-Bass. Martineau, P. (1958). The personality of a retail store. Harvard Business Review, 36(1), 47-55 Mayer, J. D. (2007). Asserting the definition of personality. The Online Newsletter for Personality Science, 1, 1–4 McCracken, G. (1989). Who Is the Celebrity Endorser? Cultural Foundations of the Endorsement Process. Journal of Consumer Research, 16(3), 310-321. McCrae, R. R., & Costa, P. T., Jr. (2008). The five-factor theory of personality. In O. P. John, R. W. Robins, & L. A. Pervin (Eds.), Handbook of personality: Theory and research (159–181). The Guilford Press. Murphy, L., Moscardo, G. & Benckendorff, P. (2007). Using brand personality to differentiate regional tourism destinations. Journal of Travel Research, 46(1), 5-14 Norman, W. T. (1963). Toward an Adequate Taxonomy of Personality Attribute: Replicated Factor Structure in Peer Nomination Personality Ratings. Journal of Abnormal and Social Psychology, 66(6), 574-583. Oklevik, O., Supphellen, M. & Maehle, N. (2020). Time to retire the concept of brand personality? Extending the critique and introducing a new framework. Journal of Consumer Behaviour, 19(3), 211-218 Park, J. K., & John, D. R. (2010). Got to Get You into My Life: Do Brand Personalities Rub Off on Consumers? Journal of Consumer Research, 37(4), 655–669. Pervin, L. A., Cervone, D., & John, O. P. (2005). Personality: Theory and Research (9th ed.). Hoboken, NJ: John Wiley & Sons Plummer, J. T. (2000). How Personality Makes a Difference. Journal of Advertising Research, 40(6), 79–83. Ratchford, Brian (1987). New Insights About the FCB Grid. Journal of Advertising Research, 27(4), 2426. Rook, D. W. (2006). Let’s pretend: projective methods reconsidered. In Belk, R. W. (Ed.), Handbook of qualitative research methods in marketing (143-155). Edward Elgar Publishing Limited. Sirgy M. J. (1986). Self-congruity: toward a theory of personality and cybernetics. Praeger Publishers/Greenwood Publishing Group. Swaminathan, V., Stilley, K. M., & Ahluwalia, R. (2009). When Brand Personality Matters: The Moderating Role of Attachment Styles. Journal of Consumer Research, 35(6), 985–1002. 26 Tupes, E. C. & Christal, R. E. (1958). Stability of Personality Trait Rating Factors Obtained Under Diverse Conditions. USAF WADS Technical Report No. 58-61. Lackland Air Force Base, TX: U.S. Air Force. von Wallpach, S., Voyer, B., Kastanakis, M. & Mühlbacher, H. (2017). Co-creating stakeholder and brand identities: introduction to the special issue. Journal of Business Research, 70, 395-398. Voyer, B., Kastanakis, M. & Rhode, A. (2017). Co-creating stakeholder and brand identities: a crosscultural consumer perspective. Journal of Business Research, 70, 399-410 Yoon, C., Gutchess, A.H., Feinberg, F. & Polk, T.A. (2006). A functional magnetic resonance imaging study of neural dissociations between brand and person judgments. Journal of Consumer Research, 33(1), 31-40. 27 APPENDIX Questionnaire – Spontaneous Condition (example for Gillette) and Priming Condition (example for Superbock) Spontaneous Condition Priming Condition 28 29 30 Page | i