21-06-2013, 12:52 PM
Separating Brand from Category Personality
Separating Brand.pdf (Size: 998.71 KB / Downloads: 131)
Abstract
Consumers often describe brands by using adjectival descriptors of personality traits, and marketers often create or reinforce these perceptions by their brand positioning. Successfully positioning a brand's personality within a product category requires measurement models that are able to disentangle a brand's unique personality traits from those traits that are common to all brands in the product category. This paper proposes a factor model that separates the two by using category-level and brand-level random effects. It illustrates the model on a data-set about brand personalities in three categories (Jeans, Magazines and Cars), and investigates the marketing implications of the results obtained with the model, through analysis of the parameter estimates. The analyses provide support for our conjecture that entire product categories (or sub-categories), not simply brands within them, are perceived to possess personality characteristics, which can be leveraged for marketing strategy.
Introduction
Over forty years of research in marketing (Levy 1959; Martineau 1958) has shown that the perceptions and associations consumers have about brands go beyond their functional attributes and benefits, and include non-functional, symbolic qualities, often referred to as “brand image.” Among these aspects of brand image are perceptions and associations about the brand's “personality,” the “set of human-like characteristics associated with a brand” (Aaker 1997, p. 347). For instance, among soft drinks, Pepsi is often perceived by consumers as more “young,” Coke as more “real and honest,” Dr. Pepper as more “non-conformist and fun” (Aaker 1997, p. 348).
Not surprisingly, marketers attempt to differentiate and build preference for their brands not only on the basis of how consumers perceive them functionally but also on the basis of these brand personality perceptions (Aaker 1997; Keller 1993). It is believed that consumers prefer those brands which, in addition to satisfying their functional needs and wants, also symbolize those personality aspects that they find most congruent with their own actual or desired (“aspired to”) personality associations (Belk 1988; Dolich 1969). The perceived personality of a brand can be shaped by marketers via “transferring cultural meaning” into it in various ways, such as by associating the brand in communications with an endorser or place that already possesses the personality or meaning considered strategically desirable for that brand (McCracken 1986).
Measuring Brand Personality
The appropriate measurement of existing brand personality imagery has been studied for over twenty years (Plummer 1984-85). Researchers have quite naturally sought to develop a valid and reliable measurement (survey) instrument of brand personality that is generalizable enough to be usable across various product categories and consumer segments, drawing on the extensive literature on human personality (Digman 1990; McCrae and Costa 1987), but going beyond it where necessary (Batra, Lehmann and Singh 1993). The measurement instrument used most often recently is the one developed by Aaker (1997). In her extensive development of this instrument, she sought to develop scales “generalizable across product categories” (Aaker 1997, p. 348), by having 631 respondents rate each of 37 brands on 114 personality traits - with these brands being carefully selected to represent a 4
broad array of product/service categories, a few brands per category. She factor analyzed the between-brand variance after averaging the scores of each brand on each personality trait across multiple respondents. In other words, the data matrix she factor-analyzed was based on pooled data from 37 brands across multiple product categories. Using this aggregated category/brand matrix, she found five factors, labeled Sincerity (sample item: honest), Excitement (daring), Competence (reliable), Sophistication (upper-class), and Ruggedness (tough); her scale is described in more detail below.
Two Explanations of Category Influence
It is widely acknowledged that “most of the research papers on brand personality are now based on Aaker's scale” (Azoulay and Kapferer 2003, p. 144), though her scale is not without its critics. It has been criticized on conceptual grounds, with some critics questioning whether the aspects being measured truly represent “personality” (Azoulay and Kapferer 2003; Caprara, Barbaranelli and Guido 2001). Empirically, some others have complained that it does not replicate well in other countries and consumer samples, especially when it is used to gauge within-category brand personality differences (e.g., Austin, Siguaw and Mattila 2003). Importantly for present purposes, it has also been pointed out that some brand personality scale items (those of Aaker 1997, but also others) appear, depending on the category, to pick up functional product category characteristics rather than brand personality ones.
Intended Contribution of this Study
There is thus a clear need to be able to separately measure “category personality” (CP) and “brand personality” (BP) and separate these as distinct dimensions from brand personality evaluations provided by individuals. To our knowledge, no such methodology exists, and no prior quantitative empirical study of CP has been published, despite the qualitative evidence of its importance described earlier (e.g., Levy 1986; Batra and Homer 2004). This paper will develop such a method, demonstrate an illustrative application as hinted on above, and discuss insights into the “category personality” of the three different categories that emerge from using our method. These insights are illustrative, and we make no claim at this stage that our substantive findings are necessarily generalizable. We will test whether the separation via our method of category from brand personality helps in predicting the preference scores of the brands in our illustrative data.