Table of Contents
1 Theoretical Foundation
1.1 The Brand Personality Concept
1.2 Measuring Brand Personality
1.2.1 Evolution of Measurement Tools
1.2.2 The Five Dimensions of Brand Personality
1.2.3 Binary Free Choice Approach
1.3 Defining Culture
1.3.1 Approaches of Cultural Studies
1.3.2 Operationalizing Culture
1.3.3 Cultural Differences between Poland and Germany
2 Empirical Research
2.1.1 Data Specification
2.1.2 Data Manipulation
2.2 First Survey
2.2.1 Descriptive Statistics of the Sample
2.3 Second Survey
2.3.1 Descriptive Statistics of the Sample
3.2 Comparison of Findings
4.2 Future Direction
List of Figures
Figure 1: The Five Dimensions of Culture for Germany and Poland According to Hofstede (itim international 2008, cf. table 4)
Figure 2: Differences in Perception of the Five Dimensions of Brand Personalities between Germans and Polish (First Survey)
Figure 3: Differences in Perception of the Five Dimensions of Brand Personalities between Germany and Poland (Second Survey)
Figure 4: Differences in Perception of the Five Dimensions of the Brand Personalities ofAdidasbetween Germans and Polish
Figure 5: Differences in Perception of the Five Dimensions of the Brand Personalities ofAllianzbetween Germans and Polish
Figure 6: Differences in Perception of the Five Dimensions of the Brand Personalities ofBMWbetween Germans and Polish
Figure 7: Differences in Perception of the Five Dimensions of the Brand Personalities ofCoca Colabetween Germans and Polish
Figure 8: Differences in Perception of the Five Dimensions of the Brand Personalities ofGooglebetween Germans and Polish
Figure 9: Differences in Perception of the Five Dimensions of the Brand Personalities ofL’Orealbetween Germans and Polish
Figure 10: Comparing Total Means of the Five Dimensions for the First and Second Survey
List of Tables
Table 1: Aaker’s Brand Personality Scale (Aaker 1997, p. 354)
Table 2: Approaches in Comparative Cultural Studies (Hofstede 1980, p. 35 and Srnka 2002, p. 9)
Table 3: Comparing Poland and Germany (Eurostat 2008 and CIA Factbook 2008)
Table 4: Analysis of Differences in Voting Behavior in the First versus Second Survey
Table 5: Product Categories (Aaker 1997, p. 349)
Table 6: The Five Dimensions Computed with Weighted Means
Table 7: Received Ratings in the Second Survey
Table 8: Differences of the Five Dimensions in the Second Survey
Anthropomorphization happens whenever human characteristics are imputed to inanimate objects. This aptitude is observable in all cultures because it simplifies the interaction with nonliving objects (Fournier 1998, p.344). In search for ways to encourage lasting costumer relationships the brand personality concept becomes a more prominent anthropomorphization in marketing. Brand personality is: “the set of human characteristics associated with a brand” (Aaker 1997, p. 347).
Aaker’s brand personality scale was chosen for the purpose of this paper to create and analyze an online survey among Polish and German people in order to assess cultural differences in the perception of brand personalities.
Successful brands can constitute a considerable share of a corporation’s equity (Batra 1993; Gregory 2007). This value can be created primarily through brand loyalty (Aa-ker 1996, pp. 105) brand name awareness, perceived quality, brand associations, and other intangible proprietary brand assets (Aaker 1991, p. 28-29, Kiley 2007). In order to tap into this pool of value drivers, researchers suggest active investment in creating an anthropomorphization of their brand (Fournier 1998). The current term explaining this is brand personality, which in turn is a unique “set of human characteristics associated with a brand” (Aaker 1997, p. 347). It is expected to enhance consumer preference and usage (Biel 1993; Sirgy 1982; Freling and Forbes, 2005b). Thus a great amount of attention is given to this concept. More and more brands operate worldwide and thus have a wide range of customers from different cultural backgrounds. This work will be investigating to what extend a common perception of brand personalities by members of different cultures can be assumed or not and whether brand personality can be proposed as a common denominator for marketers in order to appeal to customers from various cultural backgrounds.
The research question that should be answered by this work is: What are the cultural differences in the perception of brand personalities – using the example of Germany and Poland?
Therefore this work consists of two main sections: The first is devoted to the theoretical foundation of the topic and the second one applies these findings empirically by using the example of Germany and Poland, in order to narrow down the scope of this work.
The first chapters of section one are dedicated to set in place a well-laid theoretical foundation of brand personality, to review previous attempts to measure it, and to find a suitable measurement scale especially for the cross-cultural aspect of this study. Afterwards the term ‘culture’ is examined and an adequate meaning as well as definition is dissolved from the bulk of meanings available. These efforts all aim to find a suitable way to nail down and operationalize the concept that is the most applicable for the aim of this study.
The second section consists of an account of the conducted empirical study among German and Polish people. It will particularly usher the used two-step model, the online survey, the applied rating scale and stimulus of brands and describe the samples statistics. The focus of the section is laid on the analytical process and the presentation of findings.
A critical review of the limitations of this study, a comparison of the findings and the implications of these findings are discussed in a third section. In the fourth section this study is summarized leading to an anticipation of future direction as conclusion of the work.
1 Theoretical Foundation
In order to identify how individuals from different cultural backgrounds perceive brands, a well defined, valid, theoretical framework has to be applied. The first challenge is to operationalize the two terms ‘brand personality’ and ‘culture’ in a meaningful way as to create a generalizable construct for empirical inquiry.
1.1 The Brand Personality Concept
Relationship principles have replaced the paradigm of the market as a place of mere anonymous exchanges (Webster 1992). Customers’ loyalty becomes an increasingly important term and has been identified to make up for a considerable part of a successful firm’s equity (Aaker 1996). If today researchers talk about “brand loyalty” (Fournier 1998, p. 343), they imply that customers can enter into a lasting relationship with brands as far as developing emotional bonds and thus exhibit stronger consumer preference and higher usage for respective brands (Biel 1993; Sirgy 1982). Thus ever since marketers started to focus on how firms can build relationships with their customers, more attention was given to relationship theory.
Relationships between customers and their brands are suggested to refer to similar associated concepts as interpersonal relationships (Fournier 1998). This transference of concepts and cognitions from human interaction to inanimate objects by assigning human characteristics to them is called anthropomorphization. This inevitable and thus universal human propensity (Brown 1991) is performed to simplify human interaction with inanimate objects. Anthropomorphization simplifies interaction in so far as people can only interpret their observations along the rich collection of their own experience (Freling 2005). Due to these underlying paradigm changes and extension of relationship theory the brand personality concept became a more and more prominent anthropomorphization in marketing.
The concept was indicated first by Gardener and Levy (1955, p. 39) urging marketers to view advertisements not as separated units trying to please and invite different target groups but rather as an orchestrated effort to build up a brand personality. Its impact on performance was studied in a retail store environment (Martineau 1958, p. 47) and mentioned by Franklin Evans (1959), when comparing the brands and brand users of Ford and Chevrolet. Though he couldn’t find salient differences between the examined models he suggested that brands take on personalities that differentiate them even when physical configuration does not (Evans 1959, p. 340).
Likewise Plummer (1985, p. 28-30) talks about three different ways to describe brands: physical attributes, functional characteristics and characterizational attributes. Whereby he equates the characterizational attributes of a brand as the brands personality. He describes the dynamics behind the formation of brand personality. He indicates that it happens when consumers form a mental picture by observing behavior of distinct incarnations of brands and their mascots. Furthermore social demographics like origin, age, gender etc. can occasionally be attributed to brands (Thakor and Kohli 1996, p. 29). Thus brand personality can best be summarized as: “the set of human characteristics associated with a brand” (Aaker 1997, p. 347).
Certain brand personalities reflect in phrases like: “Let’s see what google has to say about”, “Maybe ebay has a cheaper solution”, “That’s no problem with my Mac” or “Look at that snobbish BMW”. Statements like these, that are drawn from ordinary life, demonstrate how brands are perceived to do things for their customers, help them or possess certain characteristics. Freling and Forbes (2005a, p.152-153) identify three major reasons why people assign personalities to brands: First it seems to familiarize the product, second it seems to provide more reassurance and comfort using the brand, and third it reduces the perceived risk.
The cardinal benefit of a strong brand personality is that it is most likely to be perceived advantageous in terms of the abovementioned three areas and hence will be memorized more easily and bring forth a greater brand loyalty (Freling and Forbes 2005a, 2005b).
Talking about brand personality as an anthropomorphization raises questions about the governing rules of this transference of characteristics to a brand and their effect on customers’ loyalty. Human (personality) traits can easily be accessed by observing an individual’s physical characteristics, body language, behavior, demographic characteristics, language and terminology, expressed attitudes, values, and believes.
Unlike human personalities, brand (personality) traits are inferred in a direct and indirect way (Aaker 1997). Direct inference can be made from observing people who for some reason are associated with the brand. Brand users, employees, CEOs (chief executive officers), prominent company founders or owners, athletes or mascots can serve for this direct inference on brand personality. Indirect inference on brand personality can be made from brand name, symbols, logos, the country of origin, product category associations, product-related attributes, functional attributes, prices, and distribution channels (Aaker 1997). Brand personality makes up for the nonfunctional benefit that users derive from using a certain brand. Brand personality causes symbolic and self-expressive (Levy 1959; Malhotra 1988) benefit. From this point of view it can also be seen as a value-added and thus as a positive influence on the cost-benefit ratio of products. Acknowledging that this symbolic or self-expressive benefit differs from person to person, it can be noted that further theoretic insight into the dynamics and measures of brand personality is crucial.
1.2 Measuring Brand Personality
A variety of different characteristics can be associated to brands. Different human personality traits can make their holder more or less successful in life (cf. Richter et al. 2005, p.5-6) likewise it can be assumed that different brand personalities and brand personalities from different product categories vary concerning their impact on a successful relationship with customers. The challenges are spotting successful brand personalities and brand personality traits, comparing the success of different brand personalities over in various product categories, and especially different cultures. Research in these fields could allow the systematic creation of successful brand personalities across cultures. In order to meet these challenges the concept needs a set of sophisticated analytical tools. Brand personality itself already is a good common denominator to market a brand across cultures (Plummer 1985a). Measurement scales of brand personality should be generalizable and validated across different product categories and cultures for the purpose of this study. Different attempts have been made to find measures of what stands at the core of a successful customer relationship.
1.2.1 Evolution of Measurement Tools
Initially consumer research attempted to explore the correlation of consumers’ characteristics like self-concept, preferences, needs, demographic variables, and purchase of certain products as well brands (e.g. Koponen 1960; Evans 1959). Analytical tools used in this field do not differ much from those used in psychology. For instance the (human) personality inventory ‘Edwards Personal Preference Schedule’ (EPPS) was used. The EPPS was used to measure individuals needs or motives in 15 areas such as: achievement, order, autonomy, dominance, change, endurance and so on. Disappointingly this approach could not explain more than 10 % of the variance in consumer behavior (Plummer 1985b, p 28).
The lack of predictive accuracy for brand choice resulting from these explorations paved the way for an inverse view of the relationship. The focus changed from the customer’s personality to that of the brand. The research question changed from: “What characteristics of a person encourage a purchase decision for a certain brand?” to “What characteristics of a brand encourage consumers’ purchase decisions?” (Plummer 1985b).
Two major groups of measurements were used by the antecedent studies of brand personality. First tools that tended to be created by researchers situational for a specific research situation without prior testing of validity, reliability, and no chance for comparability. Second tools sharing concepts from human personality research that were validated for human personality research but not for brand personality research. Researchers using the first approach of brand personality research created photo-sorts (Plummer 1985b), adjective checklists (Wells 1957) asking subjects to rate applicability of adjectives either using Likert-scales (point rating from applicable to not applicable) or semantic-differential (e.g. good vs. bad, modern vs. traditional etc.) scales, sentence completion, psychodramatic exercises or qualitative interviews (Plummer 1985b) to measure case-specific associations of consumers regarding brands. Often these scales were created for specific research design disregarding other works in that field. A number of studies aimed at accessing symbolic analogies between self image and brand image with similar rating scales, which raises the questions about common method bias (Dolich 1969, Bellenger 1976). Bellenger et al. (1976) tried to access the impact of congruence between subjects’ self image and store image for explaining brand loyalty to retail stores. Wells et al. (1957) wanted to reveal associations with certain automobile brands like Ford, Chevrolet, Buick, Cadillac and Plymouth. An array of 108 familiar adjectives was given to 100 students asking them to rate the applicability of these adjectives using forced choice, no multiple checks and no omissions.
The second sort of tools borrowed from human personality research was often used in combination with the abovementioned array of methods (Evans 1959) or models like the “Big five” of human psychology (Caprara et al. 2001) and were directly applied to brands, for sake of their thoroughly laid conceptual foundation in contrast to the abovementioned first way of assessing brand personality. The big five factors of human personality that were used are: extraversion, agreeableness, conscientiousness, emotional stability (versus neuroticism) and openness to experience/creativity. They are the result of more than seventy years of joined effort of psychologists and social scientists to find universally applicable personality traits (Goldberg 1990).
So far no empirical proof could be given to the applicability of this big five factors of human personality to brand personality. Researchers used to claim that through anth-ropomorphization human personality traits and brand personality traits should be the same.
It is likely to assume that a branded pair of jeans is bought for different reasons than is an electronic gadget (Aaker, 1997) and that different product categories vary in symbolic and utilitarian benefit. Plummer (1985b) argues that brands can be described according to their physical, functional, and characteristic attributes (p. 2728). Why then did antecedent research not bear in mind, that brand personality might be less important for products bought for their functional benefits than for products bought for symbolic or self-expressive reasons according to this renown determination?
This lack of discrimination for products of predominantly utilitarian nature and those of predominantly symbolic nature limits all of the abovementioned studies’ comparability. Furthermore, all of these approaches are lacking in certain aspects of insufficient theoretical conceptualizations, generalizability or the validity for isolating influencing aspects of successful brand personalities. Thus very soon it became obvious that “...consumer behavior researchers must develop their own definitions and design their own instruments...” (Kassarjian 1971, p. 415).
1.2.2 The Five Dimensions of Brand Personality
A comprehensive measurement tool with a good theoretical background is necessary for the aim of this study. Furthermore, sufficient empirical testing and thus generali-zability, cross-cultural validity, and reliability as well as an adequate treatment of the difference in symbolic or utilitarian features of brands exerting to address the above-mentioned shortcomings in sufficient way is crucial to be qualified for a cross-cultural analysis of the perception of brand personality.
Jennifer Aaker’s seminal work “Dimensions of brand personality” (Aaker 1997) was the first, and so far only, attempt to create a comparable framework in marketing as described for human personality structure (Goldberg 1990). Addressing the lack of instruments and definitions the paper gave a prediction concerning the relationship of brand personality and human personality dimensions. Thus the model can help to find answers to the governing rules of the anthropomorphization called brand personality (see section 1.1). Obviously Aaker’s model has drawn a lot of attention there are 731 (Google scholar 2008a) citations recorded for Aaker (1997) by Google scholar since its publication. Aaker starts formally defining brand personality as “the set of human characteristics associated with a brand” (Aaker 1997, p. 347). After this she runs a series of procedures and thus developed a scale that conforms to the challenges mentioned above (1.2.1).
For raising a comprehensive list of possible traits beyond the ad hoc scales created by academics and practitioners and to ensure content validity, Aaker started with a collected list of 309 candidate traits that can be associated to a brand. For compiling this list she borrowed from lists of previous marketing research of practitioners and academics, asked subjects to do a free association by writing down all the traits that came to their minds in the three product categories symbolic, utilitarian, and symbolic and utilitarian products, and used lists from psychology. This step was introduced in order to account for the challenge of enabling a more holistic measure of brand personality.
In order to make the list useful by facilitating the execution of practical research it went through several stages of refinement eliminating redundancies and ambiguities cutting it down to 114 traits that best described a brand.
For providing certainty that the resulting tool would measure all kinds of brands’ personality in the same accurate way showing no blind spots or bias for certain sorts of brands, stimuli selection had to be done circumspective. Thirty-seven brands were chosen from a national study for the U.S., in which brands were rated according to salience and “brand personality”. The brands were picked with a balance of the above-mentioned three product categories in mind. The 37 brands were allocated in four groups observing similarity of the groups according to the individual brands’ features taking one common brand as reference in each of the four groups. This was done to ensure generalizability of the traits that were to be assessed. The most important issue for the aim of this study is to make sure that the measuring tool contains no potential for cultural bias. Spotting perception of brand personality is all about the describing adjectives. Thus it has to be guaranteed that members of different cultural societies understand exactly the same by the given traits. Thus a sample of 1200 prospects was picked U.S wide among non-students according to five demographic concerns: gender, age, household income, ethnicity, and geographic location as to have a representative sample. The quality of the U.S.’s cultural variety for covering possible connotations that members of other cultures may have, will have to be further identified throughout this study. Equal proportions of subjects where picked for every one of the 4 different subgroups of brands. Using a five-point Likert scale subjects were asked to evaluate to what extend the 114 traits described the brands presented. A totality of 631 subjects responded and their responses where analyzed. First a correlation matrix between the 114 traits across the 37 brands was computed by averaging the scores of each personality trait on each brand across the 631 subjects. Out of that another matrix was generated correlating each of the 114 traits with all other 114 traits, in order to identify which traits described more or less the same issues in subjects’ eyes. Applying a principal component analysis, five factors where extracted which fulfilled the criteria of possess high loadings, communalities, and eigenvalues greater than one. The factor analyses was repeated in four subsamples consisting of: males (n=278), females (n=353, younger subjects (n=316), as well as older subjects (n=315) and the same five factors were obtained, which again indicates a certain stability across different groups of people. Furthermore, the originally 114 traits where reduced to 45 by running a factor analysis on every single one of the five factors – the “big five” of brand personality - identifying factors of these “big five” dimensions. Aaker calls them facets. Fifteen facets where extracted and individual traits where associated to them by cluster analysis choosing the one with the highest item-to-total correlation to give the name of the cluster and checking for the next two traits’ that correlated most with the first one. To ensure reliability of the resulting scale a test-retest was performed on a sample of 81 subjects and test-retest correlation was obtained. Three traits with test-retest correlation smaller than 0,6 had to be removed for reliability’s sake from the original construct. The remaining 42 traits of the construct could be approved as seen in table 1.
Aaker’s brand personality scale is not a one-dimensional concept, merely revealing one characteristic of a certain brand personality, but rather is multi-dimensional by placing each brand on a scale with five dimensions. Each brand can score differently on each of the dimensions of characteristics. There might be brands scoring high on one dimension and low on all others or brands scoring high on all dimensions and so on. Aaker identifies five brands that score especially high on one of the five dimensions respectively. She names: Hallmark cards for sincerity, MTV for excitement, Wall Street Journal for competence, Guess Jeans for sophistication and Nike tennis shoes for ruggedness.
illustration not visible in this excerpt
Table 1: Aaker’s Brand Personality Scale (Aaker 1997, p. 354).
Source: Own depiction from Aaker (1997, p. 354).
1.2.3 Binary Free Choice Approach
Practitioners in today’s marketing research industry have developed a binary free choice approach that is much quicker to administer. Subjects are given a list of attributes and are asked to indicate which brands from a given list they would associate with these respective attributes. Despite of a high correlation to Aaker’s dimensions of brand personality there is no question that this method is not explorative in nature, but rather descriptive (Romaniuk 2008). Thereby it can serve as a way for brand managers to visualize a brand’s personality vis-à-vis to its competitors with the aim to undertake corrective measures. Furthermore it is also inferior to Aaker’s brand personality scale when it comes to identifying differences because it only gives a “yes/no” option instead of a rating scale. Thus subjects who feel strongly for the brand cannot be distinguished from those who only see a slightly association. Neutral and negative responses can bring level, and thus bury valuable insights. (Romaniuk 2008) For this study’s research purposes a more precise tool namely Aaker’s approach is used.
1.3 Defining Culture
The aim of this study is to find cultural differences. Culture itself is a very broad concept with several meanings attached to it. Thus the first task will be to sort out a set of different meanings and definitions. Again it is a concept that needs to be defined and operationalized for the purpose of this study. Starting at the semantic origin, culture comes from the latin “colere“ which means to nurture, cultivate or to care for. Hansen (2000) takes up this direction by pointing out that this term is used in agriculture, medicine, and geography. In the broader sense the meaning is: culture is a process that nurtures an intended condition of mind, behavior, and body (Busche 2000). Neither of these meanings is certainly not what a marketing academic would be looking for. Culture can furthermore be viewed as the intended result of educational or molding effort in manner, courtesy, and fine arts, which can become an evaluative term for a classification of the undertaken nurturing effort (Busche 2000). This evaluation is highly subjective which means that it is neither suitable for the scientific approach of this study. The third possible meaning that Hubertus Busche (2000) highlights is culture of an era and/or human group that sets up the circumstances of an individual’s life by institutions that are bearers of values. This meaning is rather descriptive for the aim of this study and gives valuable hints for further exploration. The fourth meaning indicated by Busche (2000) is the mental wealth of groups of people, as it reflects in arts, values, philosophy, and science. Considering that artifacts of a culture can give insights to the mental set up of its creators, this meaning could provide some useful information. Still this approach is more likely to be applied by anthropologists rather than by marketers. From this review of meanings the third one ‘culture of an era and/or human group’ is suitable for this study. Like Busche and Hansen many authors are well aware of the fact that there is not one common scientific definition that can give a clear-cut picture of what culture is and what it is not. This short excursion to different routs and meanings of culture, shows what Hofstede sums up in his words “‘One thing we have to agree on: culture does not exist.’ Culture is a concept that we made up which helps us understand a complex world, but it is not something tangible like a table or a human being. What it is [sic!] depends on [sic!] the way in which we define it.” (Hofstede 2006) His statement confirms what the abundance of results of the investigation for the meaning and definitions of the term ‘culture’ imposes. For the aim of this work, like any other prior research the term has to be defined independently and operationalized in a meaningful way.
As a result of their review of literature Kroeber and Kluckhohn (1967) found 164 definitions. The striking feature of all of these given definitions is, that different disciplines define the concept differently - apparently to make the unclear construct workable for their various specific purpose of research. The aim of Kroeber and Kluckhohn was to give a definition comprising all or at least the majority of the reviewed definitions. The definition which they considered comprising enough is this: ”Culture consists of patterns, explicit and implicit, of and for behavior, acquired and transmitted by symbols, constituting the distinctive achievements of human groups, including their embodiments in artifacts; the essential core of culture consists of traditional (i.e. historically derived and selected) ideas and especially their attached values; culture systems may, on the one hand, be considered as products of action, on the other as conditioning elements of further action.” (Kroeber and Kluckhohn 1963, p. 181)
This definition lists a lot of features of culture worth discussing in a marketing environment. It goes along with the third meaning from Busche (2000) that culture constitutes circumstances of an individual’s life and that it is determined by predominant values of an era and/or human group. Beyond this Kroeber and Kluckhohn’s definition reveals how exactly culture influences individuals.
First culture consists of behavior patterns, which means consistent behavior is shown. Explicit and implicit is a distinction made to comprise written or taught (explicit) behavior patterns as well as such that are copied from childhood until dead by common “ways of thinking, feeling and reacting” (Kluckhohn 1951, p. 86). These behavior patterns and the subjacent thinking and feeling, which take the form of values, are “acquired and transmitted by symbols” (Kroeber and Kluckhohn 1963, p.