Kevin Lane Keller. Handbook of Marketing. Editor: Barton A Weitz & Robin Wensley. Sage Publication. 2002.
Although brand management has been an important activity for some companies for decades, branding has only emerged as a top management priority for a broad cross-section of organizations in the last decade or so. A number of factors have contributed to this trend, but perhaps the most important is the growing realization that one of the most valuable assets that firms have is the intangible asset that is their brands. As will be outlined below, creating strong brands can have a number of bottom-line and other benefits to a firm. As a consequence, branding principles have been applied in virtually every setting where consumer choice of some kind is involved, e.g., with physical goods, services, retail stores, people, organizations, places, or ideas. Driven in part by this intense industry interest, academic researchers and marketing commentators have explored a number of different brand-related topics in recent years, generating literally hundreds of papers, articles, research reports, and books on branding.
The purpose of this chapter is to review and provide some context and interpretation to this explosion of research. The goal is to highlight what has been learned, from an academic perspective, in the study of branding and brand equity, as well as what gaps still exist. Although emphasis is placed on research published since 1990 or so, earlier ‘classics’ or noteworthy studies are highlighted where appropriate. We concentrate on those important issues in building, measuring, and managing brand equity that have received some academic attention. Consequently, we do not cover all brand management issues and challenges, e.g., important managerial branding issues such as brand recovery, revitalization of classic or heritage brands, brand architecture design, and brand stewardship are not addressed in detail. Note, however, that many of these topics have been discussed in the large number of trade books in the area.
We begin by reviewing the fundamentals of branding in terms of some conceptual foundations and brand equity measurement issues. Next, we turn to a consideration of brand intangibles—the means by which a brand transcends a product—and review research concerning brand personality, experiences, relationships, and communities as well as corporate images. Third, we discuss factors affecting the choice and design of brand elements such as brand names and logos, as well as some related legal issues with respect to branding. Fourth, we review a number of issues with respect to brand extensions—the most widely studied area of branding—examining moderating and mediating factors and managerial issues in detail. Fifth, we examine brand strategies and brand alliances and the means by which brands leverage their brand equity, as well as borrow brand equity from others. Finally, the chapter concludes by offering some summary observations and identifying research priorities in the study of branding and brand management.
Before considering how brand equity has been conceptualized and measured, it is useful first to define what a brand is. According to the American Marketing Association, a brand is ‘a name, term, sign, symbol, or combination of them that is designed to identify the goods or services of one seller or group of sellers and to differentiate them from those of competitors.’ Technically speaking then, whenever a marketer creates a new name, logo, symbol, etc. for a new product, he or she has created a brand. It should be recognized that many practicing managers, however, refer to a brand as more than that—defining a brand in terms of having actually created a certain amount of awareness, reputation and prominence in the marketplace. In some sense, a distinction can thus be made between the AMA definition of a ‘small “b” brand’ versus the sometimes industry practice of a ‘big “b” brand’—i.e., a ‘brand’ versus a ‘Brand.’ It is important to recognize this distinction as disagreements about branding principles or guidelines can often revolve around the definition of what is being meant by a ‘brand’ as much as anything. With this caveat in mind, we turn to the topics of conceptualizing branding effects and measuring brand equity.
A number of studies have, one way or the other, explored the various effects of brands on consumer behavior and the effectiveness of marketing programs (see Hoeffler and Keller (2001) for a comprehensive review, as well as Yoo et al. (2000) for some empirical tests). In these studies, ‘brand type’ or some such variable has been included directly in the research design as an independent variable, or indirectly in the design as a moderator variable interacting with one or more other independent variables. Regardless of how incorporated, this research has revealed numerous positive effects and advantages from having created a ‘strong’ brand, where brand strength may reflect macro brand considerations such as market leadership or market share position, as well as more micro brand considerations such as consumer familiarity, knowledge, preferences, or loyalty. Some of the findings from these studies include the following.
Brand name has been shown to be positively associated with consumer product evaluations, perceptions of quality and purchase rates (Brown & Dacin, 1997; Day & Deutscher, 1982; Dodds et al., 1991; Leclerc, et al., 1994; Rao & Monroe, 1989). This tendency may be especially apparent with difficult-to-assess ‘experience’ goods (Wernerfelt, 1988) and as the uniqueness of brand associations increases (Feinberg et al., 1992). In addition, familiarity with a brand has been shown to increase consumer confidence, attitude toward the brand, and purchase intention (Feinberg et al., 1992; Laroche et al., 1996) and mitigate the potential negative impact of a negative trial experience (Smith, 1993). Chaudhri & Holbrook (2001) show that brand trust and brand effect combine to determine purchase loyalty and attitudinal loyalty, and, in turn, that purchase loyalty leads to greater market share and attitudinal loyalty leads to higher relative price for the brand. Although these various factors have contributed to long-term category leadership for some brands, as an important caveat, Golder (2000) found that, based on an unbiased sample of 100 categories, many leading brands have lost their leadership over a 76-year period. A number of factors and changes in the marketing environment helped to contribute to such changes in brands’ fortunes.
Several studies have demonstrated that brand leaders can command larger prices differences (Agrawal, 1996; Park & Srinivasan, 1994; Sethuraman, 1996; Simon, 1979) and are more immune to price increases (Bucklin et al., 1995; Sivakumar & Raj, 1997). In a competitive sense, brand leaders draw a disproportionate amount of share from smaller share competitors (Allenby & Rossi, 1991; Grover & Srinivasan, 1992; Russell & Kamakura, 1994). At the same time, prior research has demonstrated that market leaders are relatively immune to price competition from these small share brands (Bemmaor & Mouchoux, 1991; Blattberg & Wisniewski, 1989; Bucklin et al., 1995; Sivakumar & Raj, 1997). In addition, lower levels of price sensitivity have been found for households that are more loyal (Krishnamurthi & Raj, 1991). Advertising may play a role in the decreases in price sensitivity (Kanetkar et al., 1992). Boulding et al. (1994) claim that unique advertising messages (e.g., product differentiation for high quality products and low price messages for low price leaders) led to a reduction in the susceptibility to future price competition.
A number of communication effects have been attributed to well-known and liked brands (Sawyer, 1981). Brown & Stayman (1992) maintain that ‘halo effects’ related to the positive feelings toward a brand can positively bias the evaluation of advertising of the brand. Humor in ads seems to be more effective for familiar or already favorably evaluated brands than for unfamiliar or less-favorably evaluated brands (Chattopadhyay & Basu, 1990; Stewart & Furse, 1986; Weinburger & Gulas, 1992). Similarly, consumers appear to have a more negative reaction with ad tactics such as comparative ads (Belch, 1981), depending on the nature of the brand involved (see also Kamins & Marks, 1991). Consumers are more likely to have a negative reaction to ad repetition with unknown as opposed to strong brands (Calder & Sternthal, 1980; Campbell & Keller, 2000). Familiar brands appear to withstand competitive ad interference better (Kent & Allen, 1994). Van Osselaer & Alba (2000) showed that when the relationship between brand name and product quality was learned prior to the relationship between product attributes and quality, inhibition of the latter could occur.
In addition, panel diary members who were highly loyal to a brand increased purchases when advertising for the brand increased (Raj, 1982). Other advantages associated with more advertising include increased likelihood of being the focus of attention (Dhar & Simonson, 1992; Simonson et al., 1988) and increased ‘brand interest’ (Machleit et al., 1993). Ahluwalia et al. (2000) demonstrated that consumers who have a high level of commitment to a brand are more likely to counter-argue with negative information (see also Laczniak et al., 2001). This may be the reason why strong brands were shown to be better able to weather a product-harm crisis (Dawar & Pillutla, 2000).
Finally, Montgomery (1975) found that products that were from the top firms in an industry had a much higher chance of being accepted in the channel and gaining shelf space in supermarkets. Also, research suggests that stores are more likely to feature well-known brands if they are trying to convey a high quality image (Lal & Narasimhan, 1996).
In short, across a wide range of marketing activity, there have been demonstrable advantages from creating a strong brand.
As with advertising and other marketing phenomena, a number of different theoretical mechanisms and perspectives have been brought to bear in the study of branding. Although there are a number of industry perspectives that highlight important concepts and relationships with respect to branding and brand management, three main streams of academic research that have formally defined or conceptualized brand equity, based on either consumer psychology, economics, or biology and sociology, are briefly summarized here.
Researchers studying branding effects from a cognitive psychology perspective frequently have adopted associative network memory models to develop theories and hypotheses, in part because of the comprehensiveness and diagnostic value they offer (see Krishnan (1996) and Henderson et al. (1998) for empirical demonstrations, as well as Lassar et al. (1995)). The brand is seen as a node in the memory with a variety of different types of associations, varying in strength, linked to it. Relatedly, prior research has also often adopted a categorization perspective to memory representations of branding (Boush & Loken, 1991). This approach assumes that consumers see brands as categories that, over time, have come to be associated with a number of specific attributes, based in part on the attributes associated with the different products that represent individual members of the brand category (Loken & Roedder John, 1993).
Researchers have also relied on numerous concepts and principles from social psychology and social cognition in developing models of consumer brand-related decisions, e.g., such as affect referral mechanisms, attributional processes, accessibility-diagnosticity considerations, expectancy value formulations, and so on. Researchers have also used models of consumer inference-making fairly extensively. Teas & Grapentine (1996) construct a framework of the role of brand names in consumer purchase decision-making processes from a marketing research perspective that highlight some of these considerations.
Two well-established models of brand equity that rely in various ways on consumer psychology principles in their development are highlighted here (see also Farquhar, 1989). In three books and numerous papers, Aaker (1991, 1996; Aaker & Joachimsthaler, 2000) has approached brand equity largely from a managerial and corporate strategy perspective but with a consumer behavior underpinning. He defines brand equity as a set of four categories of brand assets (or liabilities) linked to a brand’s name or symbol that add to (or subtract from) the value provided by a product or service to a firm and/or to that firm’s customers:
- Brand awareness
- Perceived quality
- Brand associations
- Brand loyalty
In his writing, he has developed a number of distinct and useful concepts related to brand identity, brand architecture, and brand marketing programs, and has addressed a number of managerial branding challenges (Aaker, 1994; Aaker & Joachimsthaler, 1999; Joachimsthaler & Aaker, 1997).
Keller (1993, 1998) has approached brand equity from somewhat more of a consumer behavior perspective. He defines ‘customer-based brand equity’ as the differential effect that brand knowledge has on the consumer or customer response to the marketing of that brand. According to this model, a brand is said to have positive customer-based brand equity when customers react more favorably to a product and the way it is marketed when the brand is identified, as compared to when it is not (e.g., when it is attributed to a fictitiously named or unnamed version of the product). Customer-based brand equity occurs when the consumer has a high level of awareness and familiarity with the brand and holds some strong, favorable, and unique brand associations in memory. Keller views brand building in terms of a series of logical steps: establishing the proper brand identity, creating the appropriate brand meaning, eliciting the right brand responses, and forging appropriate brand relationships with customers (Keller, 2001). Achieving these four steps, according to his model, involves establishing six core brand values—brand salience, brand performance, brand imagery, brand judgments, brand feelings, and brand resonance. He also develops a number of different concepts and considers a number of different managerial applications (Keller 1999a, 1999b, 2000).
Despite their somewhat different foundations, the Aaker & Keller models share much in common with each other, as well as with other psychologically based approaches to brand equity. Most importantly, both acknowledge that brand equity represents the ‘added value’ endowed to a product as a result, in part, of past investments in the marketing for the brand. It should be noted that the Aaker & Keller models rely to some extent on spreading activation processes from an associative network model of memory. Janiszewski & van Osselaer (2000) offer some evidence to suggest that a connectionist model of brand-quality association may provide a more robust explanation of consumer reactions to various branding strategies than a spreading activation account, under certain conditions (see also van Osselaer & Janiszewski, 2001). With this model, consumers are assumed to be adaptive learners who are ‘learning to value,’ as opposed to the spreading activation perspective, which they argue is more relevant for consumers who are ‘learning to recall.’ Note too that Meyers-Levy (1989) showed that a large number of associations were not necessarily advantageous and could produce interference effects and lower memory performance.
Although behavioral models have been perhaps the dominant basis to studying branding effects and brand equity, as noted above, other valuable viewpoints have also emerged. For example, Erdem (1998a, 1998b) takes an information economics perspective on the value (or equity) ascribed to brands by consumers (see also Montgomery & Wernerfelt, 1992; Sappington & Wernerfelt, 1985; Wernerfelt, 1988). Based in part on a premise of the imperfect and asymmetrical information structure of markets, Erdem’s approach centers on the role of credibility as the primary determinant of what she dubs ‘consumer-based brand equity.’ When consumers are uncertain about product attributes, according to Erdem, firms may use brands to inform consumers about product positions and to signal that their product claims are credible. By reducing consumer uncertainty, brands are seen as lowering information costs and the risk perceived by consumers. She provides empirical support for these signaling mechanisms in an umbrella branding application to the oral hygiene market.
Similarly, Rao et al. (1999) have argued that a brand name can credibly convey unobservable quality when it is the case that false claims would result in intolerable economic losses, due to either losses of reputation, sunk investments, or losses of future profits. In a brand alliance application with hypothetical television brands, they showed that consumers’ evaluations of the quality of a product with an important unobservable attribute were enhanced when the brand was allied with a second brand that was perceived to be vulnerable to consumer sanctions.
Sociology- and Biology-based Approaches
Some researchers have studied branding from more of a sociological, anthropological, or biological perspective. For example McCracken (1986, 1993) considered the broader cultural meaning of brands and products (see also Richins, 1994). As outlined in subsequent sections, other researchers have explored topics such as brand communities (Muniz & O’Guinn, 2001; Schouten & McAlexander, 1995; Solomon & Englis, 1992) and brand relationships (Fournier, 1998). (See Ratneshwar et al. (2000) for some recent commentary.)
Other researchers have adopted more of a perceptual or even subconscious approach to branding. For example, as described in more detail below, Schmitt (1999a, 1999b) views branding in a more experiential way in terms of the effects on all five senses. Zaltman (Zaltman & Higie, 1995; Zaltman & Coulter, 1995) use metaphors as a guiding theme and qualitative research techniques to uncover the mental models driving consumer behavior with respect to brands.
All three of these theoretical approaches to branding have their roots in basic disciplines and thus, to some extent, share the corresponding strengths and weaknesses of those disciplines. Like other areas of marketing, however, adopting or at least recognizing the advantages of multiple perspectives can potentially offer deeper and richer understanding of branding and brand equity.
Finally, it should be recognized that the concept of brand equity is not without its critics (see Feldwick, 1996 for some insightful commentary). For example, the principle of ‘double jeopardy’ (DJ) is based on the robust observation that large share brands have more buyers who buy more often and who exhibit unusually high behavioral loyalty (Ehrenberg et al., 1990; Ehrenberg, et al., 1997). In downplaying the importance of brand equity, Ehernberg (1997) interprets this pattern to mean ‘… there are large brands and small ones rather than any evidence of strong brands and weak ones…’
Critics, however, have countered this charge. Dyson et al. (1997) point out that the DJ model describes ‘aspects of buyer behavior in steady markets with readily substitutable brands’ but notes that the task of marketing is often to change the setting or situation to the benefit of the brand. In other words, the role of marketing often may be to create a violation of the DJ assumptions. Similarly, Baldinger & Rubinson (1997) maintain that loyalty cannot be assumed and that ‘in order to become a large brand and stay a large brand, consumers must not only buy it, but like buying it.’ Finally, Fader & Schmittlein (1993) note that part of the explanation for the DJ effect is the existence of an extremely brand-loyal segment for high share brands, and their increased availability at retail locations (smaller stores that carry fewer brands are likely to carry the high share brand)—both indicators of brand equity. (See Chaudhuri (1999) for some additional discussion.)
Resolving the debate depends in part on assumptions and beliefs about market stability and the power of marketing actions to influence consumers. DJ proponents find short-term marketing actions relatively impotent, whereas brand equity proponents believe marketing activities can be disruptive and influence consumer behavior.
Measuring Brand Equity
The manner by which brand equity is conceptualized has obvious implications for how it is measured. Keller & Lehmann (2001) provide a broad, integrative perspective on measuring brand equity (see also Srivastava et al. (1998), Ambler (2000), and Epstein & Westbrook (2001)). They define the ‘Brand Value Chain’ in terms of a series of three steps in the creation of a value of a brand. According to this model, the first step in value creation is when an investment in marketing activity affects the consumer/customer mind set or brand knowledge (e.g., in terms of brand awareness, associations, attitudes, attachment, and activity). The second step is when brand knowledge, in turn, affects market performance (e.g., in terms of price premiums and elasticities, cost savings, market share, profitability, and expansion success). Finally, the third step is when market performance affects shareholder value (e.g., in terms of stock price and market capitalization).
Keller & Lehman identify key measures associated with each stage of this value creation process, as well as a set of ‘filters’ or moderator variables that impact the transfer or flow of value between stages of the model. Although a review of all the possible marketing research methods, techniques, and measures associated with each of the three different stages of brand-value creation is beyond the scope of this chapter, it is useful to highlight some notable recent research advances for each stage.
Consumer/Customer Brand Knowledge
In terms of measuring brand awareness, Hutchinson et al. (1994) developed a general Markov model of brand name recall and explored the implications of three special cases of the model as applied to the soft drinks and beverages categories. Their model analysis addressed a number of managerial issues and showed that 1) market structure played an important role in determining brand name recall, and, as a result, brands in certain situations could therefore be completely ignored; and 2) usage rates, advertising expenditures, market penetration and various product attributes were found to be significant predictors of recall latency. In an entirely different approach, Duke (1995) showed how indirect memory measures of awareness—the Ebbinghaus Savings Test and word fragment completion—could supplement more traditional measures of free recall.
In terms of measuring brand image, the Zaltman Metaphor Elicitation Technique (ZMET) uses qualitative methods to tap into consumers’ visual and other sensory images (Zaltman & Coulter, 1995a, 1995b). Specifically, based on a belief of the importance of nonverbal channels of communication, as part of the research process ZMET attempts to: 1) reveal the ‘mental models’ that drive consumer thinking and behavior; and 2) characterize these models in actionable ways using consumers’ metaphors. In part, ZMET requires study participants to take photographs and/or collect pictures from magazines, books, newspapers or other sources, and use these visual images to indicate what the brand means to them in various ways.
Several researchers have applied conjoint analysis to measure aspects of brand equity. For example, Rangaswamy et al. (1993) used conjoint analysis to explore how brand names interact with physical product features to affect the extendability of brand names to new product categories. Swait et al. (1993) proposed a related approach to measuring brand equity, which designs choice experiments that account for brand name, product attributes, brand image, and differences in consumer socio-demographic characteristics and brand usage. They defined the equalization price—a proxy for brand equity—as the price that equated the utility of a brand to the utilities that could be attributed to a brand in the category where no brand differentiation occurred. They illustrated their approach with an application to the deodorant, athletic shoe, and jeans market, and described its managerial implications.
Using similar techniques, Bello & Holbrook (1995) found comparatively little evidence of price premiums across a number of categories (see also Holbrook, 1992), but suggested that this absence may be due in part to the preponderance of ‘search’ goods as opposed to ‘experience’ goods in their sample. Finally, Mahajan et al. (1994) described a methodology to assess the importance of brand equity in acquisition decisions. Their approach involves defining relative attributes for acquisition such as financial performance, product market characteristics, and marketing strategy-related variables, and having key executives provide ratings of real and hypothetical firms based on that information. They illustrate their approach in the all-suites segment of the hotel industry.
Several researchers have employed ‘residual approaches’ to estimate brand equity. According to these (e.g., Bong et al., 1999; Srinivasan, 1979), brand equity is what remains of consumer preferences and choices after subtracting out objective characteristics of the physical product (although some researchers, e.g., Barwise et al. (1989), have challenged the separability assumption implicit in these approaches). Kamakura & Russell (1993) employed a single-source, scanner panel-based measure of brand equity that modeled consumer choices as a function of two factors: 1) brand value (perceived quality, the value assigned by consumers to the brand, after discounting for current price and recent advertising exposures); and 2) brand intangible value (the component of brand value not directly attributed to the physical product and thus related to brand name associations and ‘perceptual distortions’). In an application to the laundry detergent market, they show that brand equity was closely related to order of entry of brands and their cumulative advertising expenditures.
Park & Srinivasan (1994) also have proposed a residual methodology that estimates the relative sizes of different bases of brand equity by dividing it into two components: 1) the attribute-based component of brand equity, defined as the difference between subjectively perceived attribute values and objectively measured attribute values (e.g., collected from independent testing services such as Consumer Reports or acknowledged experts in the field), and 2) the non-attribute-based component of brand equity, defined as the difference between subjectively perceived attribute values and overall preference. They proposed a survey procedure to collect information to estimate these different perception and preference measures, and illustrate their approach in the toothpaste and mouthwash categories.
Dillon et al. (2001) present a model for breaking down ratings of a brand on an attribute into two components: 1) brand-specific associations (i.e., features, attributes or benefits that consumers link to a brand), and 2) general brand impressions (i.e., overall impressions based on a more holistic view of a brand). They empirically demonstrate their model properties in three product categories: cars, toothpaste, and paper towels. Finally, using established notions of health found in the epidemiology literature, Bhattacharya & Lodish (2000) defined ‘brand health’ in terms of ‘current well-being’ and ‘resistance.’ They provided an empirical application of these constructs using store scanner data, demonstrating that their proposed resistance indicator was able to predict the share loss that was suffered by the existing brands in a category in the event of a new product introduction.
Several researchers have studied how the stock market accounts for and reacts to the brand equity for companies and products. Simon & Sullivan (1993) developed a technique for estimating a firm’s brand equity derived from financial market estimates of brand-related profits. Under the assumption that the market value of the firm’s securities provides an unbiased estimate of the future cash flows that are attributable to all of the firm’s assets, their estimation technique attempted to extract the value of brand equity from the value of the firm’s other assets. They illustrated their approach in part by tracing the brand equity of Coca-Cola and Pepsi over three major events in the soft drink industry from 1982–1986.
Aaker & Jacobson (1994) examined the association between yearly stock return and yearly brand equity changes (as measured by EquiTrend’s perceived quality rating as a proxy for brand equity) for 34 companies during the years of 1989 to 1992. They also compared the accompanying changes in current-term return on investment (ROI). They found that, as expected, stock market return was positively related to changes in ROI, but that there was also a strong positive relationship between brand equity and stock return. They concluded that investors can and do learn about changes in brand equity—although presumably indirectly through learning about a company’s plans and programs.
Using data for firms in the computer industry in the 1990s, Aaker & Jacobson (2001) found that changes in brand attitude were associated contemporaneously with stock return and led accounting financial performance. They also found five factors (new products, product problems, competitor actions, changes in top management, and legal actions) that were associated with significant changes in brand attitudes. Similarly, using Financial World estimates of brand equity, Barth et al. (1998) found that brand equity was positively related to stock return and that this effect was incremental to other accounting variables such as the firm’s net income.
Adopting an event study methodology, Lane & Jacobson (1995) showed that a stock market participant’s response to brand extension announcements, consistent with the trade-offs inherent in brand leveraging, depended interactively and non-monotonically on brand attitude and familiarity. Specifically, with their sample the stock market appeared to respond most favorably to extensions of high esteem, high familiarity brands and to low esteem, low familiarity brands (in the latter case, presumably because there was little to risk and much to gain with extensions). The stock market reaction appeared to be less favorable (and sometimes even negative) for extensions of brands where consumer familiarity was disproportionately high compared to consumer regard, and to extensions of brands where consumer regard was disproportionately high compared to familiarity.
At the heart of branding and brand management is the brand itself, which can be thought of as composed of various brand elements. Brand elements can be defined as those trademarkable devices that serve to identify and differentiate the brand (e.g., brand names, logos, symbols, characters, slogans, jingles, and packages). A number of broad criteria have been identified as to how to choose and design brand elements to build brand equity (Keller, 1998):
- Aesthetic appeal
- Transferability both within and across product categories as well as across geographical and cultural boundaries and market segments
- Adaptability and flexibility over time
- Legal and competitive protectability and defensibility
Although a robust industry exists to help firms design and implement these various elements (see Kohli & LaBahan, 1997 for a descriptive account of the brand name selection process), comparatively little academic attention, even in recent years, has been devoted to this topic. Nevertheless, several research studies and programs have emerged around designing and legally protecting brand elements, as follows.
Sensory or Phonetic Considerations
Research on choice criteria for brand names extends back for years (see Robertson, 1989 for an overview). A number of studies have considered sensory or phonetic aspects of brand names. For example, in a study of computer-generated brand names containing random combinations of syllables, Peterson & Ross (1972) found that consumers were able to extract at least some product meaning out of these essentially arbitrary names when instructed to do so. Specifically, ‘whumies’ and ‘quax’ were found to be remindful of a breakfast cereal, and ‘dehax’ was remindful of a laundry detergent.
Researchers studying phonetic symbolism have considered how the sounds of even individual letters can contain meaning that may be useful in developing a new brand name (see Klink (2000) and Yorkston & Menon (2001) for a review of the conceptual mechanisms involved and some managerial applications). For example, some words begin with phonemic elements called ‘plosives’ (i.e., the letters b, c, d, g, k, p, and t), whereas others use ‘sibilants’ (i.e., sounds like s and a soft c). Because plosives escape from the mouth more quickly than sibilants and are hasher and more direct, they are thought to make names more specific and less abstract, and be more easily recognized and recalled (Vanden Bergh et al., 1984). A survey of the top 200 brands in the Marketing and Media Decision’s lists for the years 1971 to 1985 found a preponderance of brand names using plosives (Vanden Bergh et al., 1987). On the other hand, because sibilants have a softer sound, they tend to conjure up romantic, serene images and are often found with products such as perfumes (e.g., Cie, Chanel, & Cerissa) (Doeden, 1981). Heath et al. (1990) found a relationship between certain characteristics of the letters of brand names and product features: as consonant hardness and vowel pitch increased in hypothetical brand names for toilet paper and household cleansers, consumer perception of the harshness of the product also increased.
Similarly researchers have examined some cultural and linguistic aspects of branding. Leclerc et al. (1994) showed that certain hypothetical products that had brand names acceptable in both English and French (e.g., Vaner, Randal, & Massin) were perceived as more ‘hedonic’ (i.e., providing much pleasure) and better liked when pronounced in French than in English, although these effects did not completely generalize in a follow-up replication study (Thakor & Pacheco, 1997). Schmitt et al. (1994) showed that Chinese speakers were more likely to recall stimuli presented as brand names in visual rather than spoken recall, whereas English speakers were more likely to recall the names in spoken rather than visual recall. They interpreted these findings in terms of the fact that mental representations of verbal information in Chinese are coded primarily in a visual manner, whereas verbal information in English is coded primarily in a phonological manner.
Extending that research, Pan & Schmitt (1996) found that a match between peripheral features of a brand name (i.e., ‘script’ aspects such as the type of font employed, or ‘sounds’ aspects as to how the name is pronounced) and the associations or meaning of the brand resulted in more positive brand attitudes than a mismatch: Chinese native speakers were affected primarily by script matching; English native speakers’ attitudes were primarily affected by sound matching. Pan & Schmitt also interpreted these results in terms of structural differences between logographic systems (e.g., Chinese, where characters stand for concepts and not sounds) and alphabetic systems (e.g., English, where the writing of a word is a close cue of its pronunciation), and their resulting visual and phonological representations in memory.
Finally, Zhang & Schmitt (2001) present a conceptual framework for managing brand name creation in an international, multilingual market, e.g., China. Their empirical results indicated that the choice of a translation should be guided by considerations of contextual factors: which brand name (the English or Chinese name) will be emphasized, and which translation approach (phonetic, which preserves the sound of the original name, versus phonosemantic, which preserves the sound of the original name and creates product category and brand associations) for similar products are considered the standard in the marketplace.
In terms of more direct semantic meaning, applying basic associative memory theory, Keller et al. (1998) showed that a brand name explicitly conveying a product benefit (e.g., PicturePerfect televisions) led to higher recall of an advertised benefit claim consistent in meaning with the brand name (e.g., picture quality), compared with a non-suggestive brand name (e.g., Emporium televisions). On the other hand, a suggestive brand name led to lower recall of a subsequently advertised benefit claim unrelated in product meaning (e.g., superior sound), compared with a non-suggestive brand name.
Sen (1999) explored how a brand name’s semantic suggestiveness interacted with the type of decision task involved in an initial encounter with a brand to influence the brand information encoded and recalled during a subsequent encounter with a proposed extension. He found that when information about an efficient set of new brands was learned through a choice task, brand names that suggested general superiority appeared to benefit subsequent brand extensions more than names that were suggestive of category-specific, attribute-based superiority. After a judgment task, however, the category-specific names appeared to benefit brand extensions more than the general superiority names.
Finally, in an exploratory study of alpha-numeric brand names (i.e., containing one or more numbers in either digit form (e.g., 5) or in written form (e.g., ‘five’), Paiva & Costa (1993) found that alpha-numeric brand names were more favorably evaluated when designating technology-related products. This effect was moderated by a number of factors, including the visual or aural aspects of the name, the actual numbers that were used in the name, and the words or letters, along with the number(s), that comprised the brand name.
Little academic research has explored the consumer behavior effects of logo design or other visual aspects of branding (see Schmitt & Simonson (1997) for some background discussion). Henderson & Cote (1998) conducted a comprehensive empirical analysis of 195 logos that were calibrated on 13 different design characteristics in terms of their ability to achieve different communication objectives. They interpreted their findings as suggesting that: 1) logos with a high recognition goal should be very natural, very harmonious and moderately elaborate; 2) low investment logos intended to create a false sense of knowing and positive affect should be less natural and very harmonious; and 3) high image logos intended to create strong positive affect without regards to recognition should be moderately elaborate and natural but with high harmony. Overall, the results suggest that logos should generally be natural and fairly elaborative, but not overly so.
Janiszewski & Meyvis (2001) suggest that a dual-process model is most applicable in describing consumer responses to repeated exposure to static brand names, logos, and packages. The dual process model assumes a passive processing system by consumers, and posits that their response to a stimulus is a function of sensitization and habituation. They provide experimental evidence to that effect.
Several academics have considered legal issues involved with branding. Cohen (1986, 1991) has argued that trademark strategy involves proper trademark planning, implementation, and control, reviewing issues in each of those areas. Zaichovsky (1995) provides a comprehensive treatment of brand confusion (see also Foxman et al., 1992). Simonson (1994) provides an in-depth discussion of these issues and methods to assess the likelihood of confusion and ‘genericness’ of a trademark. He stresses the importance of recognizing that consumers may vary in their level or degree of confusion, and that as a result it is difficult to identify a precise threshold level above which confusion ‘occurs.’ He also notes how survey research methods to assess confusion must accurately reflect the consumers’ state of mind when engaged in marketplace activities.
Simonson has provided behavioral perspectives to consider a number of issues related to appropriation, dilution, etc. (e.g., Simonson, 1995). Harvey et al. (1998) considered the legal and strategic implications of the look-a-like ‘trade dress’ practice of major food chains that adopt the visual cues (e.g., shape, size, color and the like) of national brands in branding their own, private labels (see also Kapferer, 1995; Loken et al., 1986). Oakenfull & Gelb (1996) have described how to avoid ‘genercide’—i.e., when consumers employ the brand name as the product category label—through research and advertising. Finally, Sullivan (2001) considered the optimal number of registered trademarks to protect a brand.
An important and relatively unique aspect of branding research is the focus on brand intangibles -aspects of the brand image that do not involve physical, tangible, or concrete attributes or benefits. For a review of some seminal work in this area, see Levy (1999). Brand intangibles are often a means by which marketers differentiate their brands with consumers (Park et al., 1986) and transcend physical products (Kotler, 2000).
Brand personality has been defined as the human characteristics or traits that can be attributed to a brand. Aaker (1997) examined 114 possible personality traits and 37 well-known brands in various product categories to create a brand personality scale composed of five factors:
- Sincerity (e.g., down-to-earth, honest, whole-some, and cheerful)
- Excitement (e.g., daring, spirited, imaginative, and up-to-date)
- Competence (e.g., reliable, intelligent, and successful)
- Sophistication (e.g., upper class and charming)
- Ruggedness (e.g., outdoorsy and tough)
In a cross-cultural study exploring the generalizability of this scale outside the United States, Aaker, Benet-Martinez & Berrocal (2001) found that three of the five factors applied in Japan and Spain, but that a ‘peacefulness’ dimension replaced ‘ruggedness’ in both countries, and a ‘passion’ dimension emerged in Spain instead of ‘competency.’
Aaker (1999) also explored how different brand personality dimensions affected different types of people in different types of consumption settings. She found that self-congruity (i.e., brands which were chosen whose personality matched those of subjects) was enhanced for low versus high self-monitoring individuals, whereas situation congruity (i.e., brands which were chosen whose personality matched those of the situation involved) was enhanced for high versus low self-monitoring individuals. She interpreted these results in terms of a ‘malleable self,’ which is composed of self-conceptions that are chronically accessible or made accessible by a social situation.
Relatedly, Graef (1996) found that increased self-monitoring was associated with a greater effect of image congruence on consumers’ evaluations of publicly consumed brands, but not privately consumed brands. Moreover, consumers’ evaluations of publicly consumed brands were also more affected by the congruence between brand image and ideal self-image, than actual self-image. In the case of privately consumed brands, however, these effects were equal. Graef (1997) also found that when considering a specific consumption situation, the congruence between brand image and a dynamic measure of consumers’ situational ideal self-image was more strongly correlated with consumers’ brand evaluations than was the congruence between brand image and static measures of either their actual or ideal self-image.
Arguing that marketers and brand managers have largely ignored sensory, affective, and creative experiences, Schmitt (1997) outlines the SOOP (‘superficial—out of profundity’) model of the branding of customer experiences. He distinguishes between three type of experiential brands—‘sense,’ ‘feel,’ and ‘think’ brands—based on the primary type of appeal that they present and the type of experience they target. Schmitt argues that common to all types of brands is the idea of providing value to consumers by enhancing relations between the brand and the consumer through rewarding, sensory stimulation, emotional binds, or creative rewards. He also defines five types of sensory experiences (which he calls ‘SEMs,’ Strategic Experiential Modules): ‘Sense’ experiences involving sensory perception, ‘Feel’ experiences involving affect and emotions, ‘Think’ experiences, which are creative and cognitive, ‘Act’ experiences, involving the physical and possibly incorporating individual actions and lifestyles, and ‘Relate’ experiences that result from connecting with a reference group or culture. (See also Schmitt (1999a, 1999b).)
Fournier (1998) extended the metaphor of interpersonal relationships into the brand domain to conceptualize the relationships that consumers form with brands (see also Fournier & Yao (1997) and Fournier et al. (1998)). Fournier views brand relationship quality as multi-faceted and consisting of six dimensions beyond loyalty/commitment along which consumer-brand relationships vary: 1) self-concept connection; 2) commitment or nostalgic attachment; 3) behavioral interdependence; 4) love/passion; 5) intimacy; and 6) brand partner quality. Based on lengthy, in-depth consumer interviews, Fournier defined 15 possible consumer-brand relationship forms: 1) arranged marriages, 2) casual friends/buddies; 3) marriages of convenience; 4) committed partnerships; 5) best friendships; 6) compartmentalized friendships; 7) kinships; 8) rebounds/avoidance-driven relationships; 9) childhood friendships; 10) courtships; 11) dependencies; 12) flings; 13) enmities; 14) secret affairs; and 15) enslavements. Additionally, Fournier (2000) developed the Brand Relationship Quality (BRQ) scale to empirically capture these theoretical notions.
Aaker, Fournier & Brasel (2001) conducted a two-month longitudinal investigation of the development and evolution of relationships between consumers and brands. They found that two factors—experience of a transgression and personality of the brand—had a significant influence on developmental form and dynamics. Specifically, brands associated with Sincerity traits (e.g., sincere, wholesome, sentimental) relative to those with Exciting traits (e.g., exciting, young, trendy) demonstrated increasing levels of relationship strength over time, but those results held only when relationship development proceeded without the experience of a transgression. In cases where a transgression occurred, relationship strength dramatically suffered for sincere brands, whereas some aspects of relationship strength eventually rebounded for exciting brands.
Muniz and O’Guinn (2000) have defined ‘brand communities’ as a specialized, non-geographically bound community, based on a structured set of social relationships among users of a brand. They note that, like other communities, a brand community is marked by 1) a shared consciousness, 2) rituals and traditions, and 3) a sense of moral responsibility. They demonstrated these characteristics in both face-to-face and computer-mediated environments for the Apple Macintosh, Ford Bronco, and Saab brands.
Somewhat relatedly, Schouten & McAlexander (1995) have defined a ‘subculture of consumption’ as a distinctive subgroup of society that self-selects on a basis of a shared commitment to a particular product class, brand, or consumption activity. They note that characteristics of a subculture of consumption include an identifiable, hierarchical social structure; a unique ethos, or set of shared beliefs and values; and unique jargons, rituals, and modes of symbolic expression. As an illustration, they presented results of a three-year ethnographic field study with Harley-Davidson motorcycle owners. They later expanded their investigation to include the Jeep brand and explore various relationships that consumers could hold with the product/possession, brand, firm, and/or other customers as a measure of loyalty (McAlexander & Schouten, 2001).
Much research has considered corporate image in terms of its conceptualization, antecedents, and consequences (e.g., see reviews by Barich & Kotler, 1991; Biehal & Shenin, 1998; Dowling, 1994; Schumann et al., 1991). A corporate image can be thought of as the associations that consumers have in their memory to the company or corporation making the product or providing the service as a whole. Corporate image is a particularly relevant concern when the corporate or company brand plays a prominent role in the branding strategy adopted. In establishing a corporate image, a corporate brand may evoke associations wholly different from an individual brand, which is only identified with a certain product or limited set of products (Brown, 1998).
For example, a corporate brand name may be more likely to evoke associations of common products and their shared attributes or benefits; people and relationships; and programs and values. Brown & Dacin (1997) distinguish between corporate associations related to corporate ability (i.e., expertise in producing and delivering product and/or service offering) and corporate social responsibility (i.e., character of the company with regard to important societal issues). In terms of the latter consideration, much research has explored the implications of cause-related marketing (e.g., Drumwright, 1996; Sen & Bhattacharya, 2001; Varadarajan & Menon, 1988) and green marketing (e.g., Journal of Advertising, 1995; Menon & Menon, 1997) strategies.
Keller & Aaker (1992, 1998) defined corporate credibility as the extent to which consumers believe that a company is willing and able to deliver products and services that satisfy customer needs and wants. Based on past consumer behavior research (e.g., Sternthal & Craig, 1984), they identified three possible dimensions to corporate credibility:
- Corporate expertise—the extent to which a company is seen as being able to competently make and sell their products or conduct their services
- Corporate trustworthiness—the extent to which the company is seen as motivated to be honest, dependable, and sensitive to consumer needs
- Corporate likability—the extent to which the company is seen as likable, prestigious, interesting, etc.
Experimentally, Keller & Aaker (1992) showed that successfully introduced brand extensions can lead to enhanced perceptions of corporate credibility and improved evaluations of even more dissimilar brand extensions. Keller & Aaker (1998) showed that different types of corporate marketing activity, by impacting different dimensions of corporate credibility, as well as perceptions of extension fit and attribute beliefs, differentially affected consumer evaluations of a corporate brand extension. Specifically, corporate marketing activity related to product innovation produced more favorable evaluations for a corporate brand extension than corporate marketing activity related to either the environment or, especially, the community. Their findings also revealed that corporate marketing activity influenced evaluations even in the presence of advertising for the extension (see also Brown & Dacin, 1997).
The area of branding research that has seen by far the most attention in recent years has been on how firms should leverage brand equity, especially in terms of brand extensions. To develop managerial guidelines, research has addressed a number of factors affecting extension success.
Modeling Consumer Evaluations of Brand Extensions
Much research has adopted a categorization perspective in modeling consumer evaluations of brand extension (e.g., Boush & Loken, 1991; Hartman et al., 1990; Loken & Roedder John, 1993). According to this perspective, if a brand were to introduce an extension that was seen as closely related or similar in ‘fit’ to the brand category, then consumers could easily transfer their existing attitude about the parent brand to the extension. On the other hand, if consumers were not as sure about extension similarity, then they might be expected to evaluate the extension in a more detailed, piece-meal fashion. In this case, the favorability of any specific associations that would be inferred about the extension would be the primary determinant of extension evaluations.
Fit has thus been identified as a key moderator as to how consumers evaluate brand extensions (Boush et al., 1987). Consistent with these notions, Aaker & Keller (1990) found that a perception of fit between the original and extension product categories, as well as a perception of high quality for the parent brand, led to more favorable extension evaluations. A number of subsequent studies have explored the generalizibility of these findings to markets outside the US. Based on a comprehensive analysis of 131 brand extensions from seven such replication studies around the world, Bottomly & Holden (2000) concluded that this basic model clearly generalized, although cross-cultural differences influenced the relative importance attached to the model components. In attempting to understand why some brands have been able to extend into ‘perceptually distinct’ domains, however, Klink & Smith (2001) show that effects of fit can disappear when attribute information is added to extension stimuli and are applicable only for later product adopters, and that perceived fit increases with greater exposure to an extension.
On the other hand, applying Mandler’s congruity theory, Meyers-Levy et al. (1994) showed that suggested products associated with moderately incongruent brand names could be preferred over ones that were associated with either congruent or extremely incongruent brand names (see also Zinkhan & Martin, 1987). They interpreted this finding in terms of the ability of moderately incongruent brand extensions to elicit more processing from consumers that could be satisfactorily resolved (assuming consumers could identify a meaningful relationship between the brand name and the product).
Bases of Extension Fit
Prior research has examined a number of different factors concerning fit perceptions. In general, this research has reinforced the importance of taking a broad and contextual view of fit. Adopting a demand-side and supply-side perspective of consumer perceptions, Aaker & Keller (1990) showed that perceived fit between a parent brand and extension product could be related to the economic notions of perceived substitutability and complementarity in product use (from a demand-side perspective), as well as the perceived ability of the firm to have the skills and assets necessary to make the extension product (from a supply-side perspective).
Subsequent research has shown that virtually any association about a parent brand held in memory by consumers may serve as a potential basis of fit. Park et al. (1991) distinguished between fit based on ‘product-feature-similarity’ and ‘brand-concept-consistency’ (i.e., how well the brand concept accommodates the extension product). They also distinguished between function-oriented brands, whose dominant associations relate to product performance (e.g., Timex watches), and prestige-oriented brands, whose dominant associations relate to consumers’ expression of self-concepts or images (e.g., Rolex watches). Experimentally, Park et al. showed that the Rolex brand could more easily extend into categories such as grandfather clocks, bracelets, and rings than the Timex brand; but, on the other hand, that Timex could more easily extend into categories such as stopwatches, batteries, and calculators than Rolex. They interpreted these results to suggest that, in the former case, there was high brand-concept-consistency for Rolex that overcame a lack of product-feature-similarity; in the latter case there was enough product-feature-similarity to favor a function-oriented brand such as Timex.
Broniarczyk & Alba (1994) showed that a brand that may not even be as favorably evaluated as a competing brand in its category—depending on the particular parent brand associations involved—may be more successfully extended into certain categories. For example, in their study, although Close-Up toothpaste was not as well liked as Crest toothpaste, a proposed Close-Up breath mint extension was evaluated more favorably than one from Crest. Alternatively, a proposed Crest toothbrush extension was evaluated more favorably than one from Close-Up.
Broniarczyk & Alba (1994) also showed that a perceived lack of fit between the parent brand’s product category and the proposed extension category could be overcome if key parent-brand associations were deemed valuable in the extension category. For example, Froot Loops cereal—which has strong brand associations to ‘sweet,’ ‘flavor,’ and ‘kids’—was better able to extend to dissimilar product categories such as lollipops and popsicles than to even similar product categories such as waffles and hot cereal because of the relevance of their brand associations in the dissimilar extension category. The reverse was true for Cheerios cereal, however, which had a ‘healthy grain’ association that was only relevant in similar extension product categories.
Thus, extension fit is more than just the number of common and distinctive brand associations between the parent brand and the extension product category (e.g., see Bijmolt et al., 1998). Along these lines, Bridges et al. (2000) refer to ‘category coherence’ of extensions. Coherent categories are those categories whose members ‘hang together’ and ‘make sense.’ According to this view, to understand the rationale for a grouping of products in a brand line, a consumer needs ‘explanatory links’ that tie the products together and summarize their relationship. For example, the physically dissimilar toy, bath care, and car seat products in the Fisher-Price product line can be united by the link, ‘products for children.’
Similarly, Schmitt & Dube (1992) proposed that brand extensions should be viewed as conceptual combinations. A conceptual combination (e.g., ‘apartment dog’) consists of a modifying concept or ‘modifier’ (e.g., apartment) and a modified concept or ‘header’ (e.g., dog). Thus, according to this view, a proposed brand extension such as McDonald’s Theme Park would be interpreted as the original brand or company name (e.g., McDonald’s) acting on the ‘head concept’ of the extension category (e.g., theme parks) as a ‘modifier.’ Bristol (1996) adopted a similar view and suggested that consumers often seem to construct on-line ‘conjunctive inferences’ about extensions without exclusively using knowledge about the retrieved brand, product class, product-type category or subcategory, or exemplar.
Finally, researchers have explored other, more specific aspects of fit. Boush (1997) provides experimental data as to the context sensitivity of fit judgments. Similarity judgments between pairs of product categories were found to be asymmetrical, and brand name associations could reverse the direction of asymmetry. For example, more subjects agreed with the statement ‘Time magazine is like Time books’ as compared to the statement that ‘Time books are like Time magazine,’ but without the brand names, the preferences were reversed. Smith & Andrews (1995) surveyed industrial goods marketers and found that the relationship between fit and new product evaluations was not direct, but was mediated by customers’ sense of certainty that a firm could provide a proposed new product. Finally, Martin & Stewart (2001) explore the multidimensional nature of product similarity ratings and the moderating effects of goal congruency.
Extension Feedback Effects
Much research has considered the reciprocal effects on the parent brand from having introduced an extension. Keller & Aaker (1992) and Romeo (1991) found that unsuccessful extensions in dissimilar product categories did not affect evaluations of the parent brand. The Keller & Aaker study also showed that unsuccessful extensions did not necessarily prevent a company from ‘retrenching’ and later introducing a more similar extension. Sullivan (1990) found that the ‘sudden acceleration’ problems associated with the Audi 5000 automobile had greater spillover to the Audi 4000 model than to the Audi Quattro model, which she interpreted as a result of the fact that the latter was branded and marketed differently.
Loken & Roedder John (1993) found that dilution effects were evident, but only under certain conditions (e.g., when extensions were perceived to be moderately typical and when extension typicality was not made salient to consumers), and for certain types of beliefs (e.g., gentleness beliefs rather than quality beliefs). In their study, however, dilution effects were largely unrelated to the similarity of the brand extension. Roedder John et al., (1998) found that dilution effects were less likely to be present with ‘flagship’ products and occurred with line extensions, but were not always evident for more dissimilar category extensions.
Gürhan-Canli & Durairaj (1998) extended the results of these studies by considering the moderating effect of consumer motivation and extension typicality. In high motivation conditions, they found that incongruent extensions were scrutinized in detail and led to the modification of family brand evaluations, regardless of the typicality of the extensions. In low motivation conditions, however, brand evaluations were more extreme in the context of high (versus low) typicality. As the less typical extension was considered an exception, its impact was reduced. In a similar vein, Ahluwalia & Gürhan-Canli (2000) adopted an accessibility-diagnosticity perspective to explain the effects of brand extensions on the family brand name and observed comparable experimental results. Finally, Milberg et al. (1997) found that negative feedback effects were present when 1) extensions were perceived as belonging to a product category dissimilar from those associated with the family brand, and 2) extension attribute information was inconsistent with image beliefs associated with the family brand.
In terms of individual differences, Lane & Jacobson (1997) found some evidence of a negative reciprocal impact from brand extensions, especially for high need-for-cognition subjects, but did not explore extension similarity differences. Kirmani et al. (1999) found dilution effects with owners of prestige-image automobiles when low-priced extensions were introduced, but not with owners of non-prestige automobiles or with non-owners of either type of automobile. Using national household scanner data, Swaminathan et al. (2001) found positive reciprocal effects of extension trial on parent brand choice, particularly among non-loyal users and prior non-users of the parent brand, and consequently on market share. Category similarity, however, appeared to moderate the existence and magnitude of these effects. Evidence was also found for potential negative effects of unsuccessful extensions among prior users of the parent brand, but not among prior non-users. Additionally, experience with the parent brand was found to have significant impact on extension trial, but not on extension repeat.
Morrin (1999) examined the impact of brand extensions on the strength of parent brand associations in memory. Two computer-based studies revealed that exposing consumers to brand extension information strengthened rather than weakened parent brand associations in memory, particularly for parent brands that were dominant in their original product category. Higher fit also resulted in greater facilitation, but only for non-dominant parent brands. Moreover, improvements in parent brand memory due to the advertised introduction of an extension was not as great as when the same level of advertising directly promoted the parent brand.
In reviewing this literature, Keller & Sood (2001a) put forth a conceptual model that posits that changes in consumer evaluations of a parent brand as a result of the introduction of a brand extension, is a function of three factors: 1) strength and clarity of extension evidence; 2) diagnosticity or relevance of extension evidence; and 3) evaluative consistency of extension evidence with parent brand associations. One implication of their conceptual model is that existing parent brand knowledge structures, in general, will be fairly resistant to change. In order for parent brand associations to change, consumers must be confronted with compelling evidence that is seen as relevant to existing parent brand associations, and they must be convinced that the evidence is inconsistent enough to warrant a change (Roedder John et al., 1997).
Thus, an unsuccessful brand extension can potentially damage the parent brand only when there is a high degree of similarity or ‘fit’ involved, e.g., in the case of a failed line extension in the same category. When the brand extension is farther removed, consumers can ‘compartmentalize’ the brand’s products and disregard its performance in what is seen as an unrelated product category. Keller & Sood provided experimental evidence consistent with all of these conjectures.
Number and Types of Extensions
Prior research (Farquhar 1989; Keller 1998) has distinguished between line extensions (i.e., when the parent brand is used to brand a new product that targets a new market segment within a product category currently served by the parent brand) and category extensions (i.e., when the parent brand is used to enter a different product category from that currently served by the parent brand). Several studies have examined the market performance of real brands to determine the success characteristics of line extensions (see Quelch & Kenney (1994) for some useful managerial insights and guidelines, and Hardie et al. (1994) for some commentary).
Using data on 75 line extensions of 34 cigarette brands over a 20-year period, Reddy et al. (1994) found that:
- Line extensions of strong brands were more successful than line extensions of weak brands
- Line extensions of symbolic brands enjoyed greater market success than those of less symbolic brands
- Line extensions that received strong advertising and promotional support were more successful than line extensions that received meager support
- Line extensions entering earlier into a product subcategory were more successful than line extensions entering later, but only if they were line extensions from strong brands
- Firm size and marketing competencies also played a part in a line extension’s success
- Earlier line extensions helped in the market expansion of the parent brand
- Incremental sales generated by line extensions could more than compensate for the loss in sales due to cannibalization.
Based on data collected from 166 product managers in packaged goods firms, Andrews & Low (1998) found that meaningful product line extensions had been launched in companies that: 1) had longer planning and reward horizons; 2) encouraged risk taking; 3) rotated brand assignments regularly; 4) had a product-focused management structure; 5) required comparatively more evidence to justify new stock keeping units and 6) utilized smaller new product development teams.
In an empirical study of the determinants in product line decisions in the personal computer industry from 1981–1992, Putsis and Bayus (2001) found that firms in the industry expanded their product lines when there were low industry barriers (e.g., few market-wide introductions, low industry concentration) or perceived market opportunities (e.g., due to high market share, recent market entry). High market share firms were found to expand aggressively their product lines, as did firms with relatively high prices or short existing product lines. They also found important substantive differences between the factors affecting the direction of a product line change (i.e., expansion or contraction of its current line) and the magnitude of any line change (i.e., how many products were introduced or withdrawn).
In the context of fast moving packaged goods, Cohen et al. (1997) developed a decision support system to evaluate the financial prospects of potential new line extensions. The model incorporated historical knowledge about the productivity of the firm’s new product development process, as well as R&D resource factors that affect productivity, to provide shipment forecasts at various stages, allow for a product line perspective, and facilitate organizational learning.
For market reasons or competitive considerations, firms often consider introducing lower priced versions of their established brand name products. Additionally, some marketers have attempted to migrate their brand up-market through brand extensions. Many strategic recommendations have been offered as to the topic of ‘vertical extensions’ (e.g., Aaker, 1994). For example, Farquhar et al. (1992) describe how to use ‘sub-branding’ strategies as a means to distinguish lower-priced entries, and ‘super-branding’ strategies to signal a noticeable, although presumably not dramatic, quality improvement (Ultra Dry Pampers or Extra Strength Tylenol). Comparatively little empirical academic research, however, has been conducted on this topic.
In an empirical study of the US mountain bicycle industry, Randall et al. (1998) found that brand price premium was significantly positively correlated with the quality of the lowest-quality model in the product line for the lower quality segments of the market; and that for the upper quality segments of the market, brand price premium was also significantly positively correlated with the quality of the highest quality model in the product line. They concluded that these results suggest that managers only wishing to maximize the equity of their brands would offer only high-quality products and avoid offering low-quality products, although overall profit maximization could dictate a different strategy.
Kirmani et al. (1999) examined the ‘ownership effect’—by which owners have more favorable responses than non-owners to brand extensions—in the context of brand line stretches. They found that the ownership effect occurred for upward and downward stretches of non-prestige brands (e.g., Acura) and for upward stretches of prestige brands (e.g., Calvin Klein and BMW). For downward stretches of prestige brands, however, the ownership effect did not occur, because of the owners’ desire to maintain brand exclusivity. In this situation, a sub-branding strategy protected owners’ parent brand attitudes from dilution.
Multiple Brand Extensions
Keller & Aaker (1992) showed that when consumers did not already have strongly held attitudes, the successful introduction of a brand extension improved evaluations of a parent brand that was originally perceived only to be of average quality. By changing the image and meaning of the brand, the successfully introduced extension also could make subsequent brand extensions, that otherwise may not have seemed appropriate to consumers, make more sense and be seen as a better fit. Experimentally, they showed that by taking ‘little steps,’ i.e., by introducing a series of closely related but increasingly distant extensions, it was possible for a brand ultimately to enter product categories that would have been much more difficult, or perhaps even impossible, to have entered directly. Similarly, Dawar & Anderson (1994) showed that undertaking extension introductions in a particular order could allow distant extensions to be perceived as coherent, and that following a consistent direction in extension strategy allowed for greater coherence and purchase likelihood for a target extension (see also Jap, 1993).
Boush & Loken (1991) found that far extensions from a ‘broad’ brand were evaluated more favorably than from a ‘narrow’ brand. similarly, Dacin & Smith (1994) showed that if the perceived quality levels of different members of a brand portfolio were more uniform, then consumers tended to make higher, more confident evaluations of a proposed new extension. They also showed that a firm that had demonstrated little variance in quality across a diverse set of product categories was better able to overcome perceptions of lack of extension fit (i.e., as consumers would think, ‘whatever they do, they tend to do well’).
Dawar (1996) showed that for brands with a single product association, brand knowledge and context interacted to influence evaluations of fit for extensions to products that were only weakly associated with the brand. Specifically, retrieval inhibition effects appeared to reduce the activation of ‘less relevant’ product associations, lowering the perceived fit of extensions closer to such products, especially for individuals more knowledgeable about the brand. For brands strongly associated with more than one product, only context—by selectively eliciting retrieval of brand associations—influenced the evaluations of the fit of the extensions.
In an empirical study of 95 brands in 11 nondurable consumer goods categories, Sullivan (1992) found that, in terms of stages of the product category life cycle, early-entering brand extensions did not perform as well, on average, as either early-entering new-name products or late-entering brand extensions. DeGraba & Sullivan (1995) provided an economic analysis to help interpret this observation. They posited that the major source of uncertainty in introducing a new product is the inability to know if it will be received by customers in a way that will allow it to be a commercial success. They further argued that this source of uncertainty could be mitigated by spending more time on the development process. Under such assumptions, they showed that the large spillover effects triggered by introducing a poorly received brand extension caused introducers of brand extensions to spend more time on the development process than did introducers of new-name products. Finally, Sullivan (1990) found that when Jaguar launched its first new model in 17 years, the older models of Jaguar experienced an increase in demand as a result of advertising used to promote the new 1988 model.
Characteristics of Consumers
Perceptions of fit and extension evaluations may depend on how much consumers know about the product categories involved. Muthukrishnan & Weitz (1990) showed that knowledgeable, ‘expert’ consumers were more likely to use technical or manufacturing commonalties to judge fit, considering similarity in terms of technology, design and fabrication, and the materials and components used in the manufacturing process. Less knowledgeable ‘novice’ consumers, on the other hand, were more likely to use superficial, perceptual considerations such as common package, shape, color, size, usage, etc. Relatedly, Broniarczyk & Alba (1994) also showed that perceptions of fit on the basis of brand-specific associations were contingent on consumers having the necessary knowledge about the parent brand. Without such knowledge, consumers tended again to rely on more superficial considerations, such as their level of awareness for the brand or overall regard for the parent brand, in forming extension evaluations. Finally, Zhang & Sood (2001) showed that 11–12-year-old consumers, relative to adults, evaluated brand extensions by relying less on category similarity between the parent brand and the extension category and more on the inherent characteristics of the name itself used to brand the extension (see also Achenreiner & Roedder John, 2001; Nguyen & Roedder John, 2001).
In a cross-cultural study, Han & Schmitt (1997) found that for US consumers, perceived fit was more important than company size in extension evaluations, but that for Hong Kong consumers, company size mattered for low fit extensions. They suggest that the value of collectivism may explain the relative higher importance of corporate identity as a quality cue for East Asian consumers.
Characteristics of the Parent Brand
Perceptions of fit and extension evaluations may also depend on aspects of the parent brand involved. Prior research has shown that one important benefit of building a strong brand is that it can be extended more easily into more diverse categories (e.g., Rangaswamy et al., 1993). High quality brands are often seen as more credible, expert, and trustworthy at what they do (Keller & Aaker 1992). As a result, even though consumers may still believe a relatively distant extension does not really fit with the brand, they may be more willing to give a high quality brand the ‘benefit of the doubt.’ When a brand is seen as more average in quality, however, such favorable source attributions may be less forthcoming, and consumers may be more likely to question the ability or motives of the company involved.
As a caveat to the previous conclusion, prior research (e.g., Farquhar & Herr, 1992) has shown that if a brand is seen as representing or exemplifying a category too much, it may be difficult for consumers to think of the brand in any other way. Farquhar et al. (1992) define a master brand as an established brand so dominant in customers’ minds that it ‘owns’ a particular association—the mention of a product attribute or category, a usage situation, or a customer benefit instantly brings a master brand to mind. Recognizing that the exceptionally strong associations of a master brand often make it difficult to extend it directly to other product categories, they have proposed various branding strategies to extend master brands indirectly by leveraging alternative master brand associations that come from different parts of their brand hierarchies (e.g., via sub-branding, super-branding, brand-bundling, and brand-bridging).
The limits to extension boundaries potentially faced by master brands may be exacerbated by the fact that, in many cases, brands that are market leaders have strong concrete product attribute associations (Farquhar et al. 1992). In general, concrete attribute associations may not transfer as broadly to extension categories as more abstract attribute associations. For example, Aaker & Keller (1990) showed that consumers dismissed a hypothetical Heineken popcorn extension as potentially tasting bad or like beer; a hypothetical Vidal Sassoon perfume extension as having an undesirably strong shampoo scent; and a hypothetical Crest chewing gum extension as tasting like toothpaste or, more generally, tasting unappealing. In each case, consumers inferred a concrete attribute association with an extension that was technically feasible even though common sense might have suggested that a manufacturer logically would not be expected to introduce a product with such an attribute.
Because of their intangible nature, on the other hand, more abstract associations may be seen as more relevant across a wide set of categories (Rangaswamy et al., 1993). For example, the Aaker & Keller study also showed that the Vuarnet brand had a remarkable ability to be exported to a disparate set of product categories, e.g., sportswear, watches, wallets, and even skis. In these cases, complementarity may have led to an inference that the extension would have the ‘stylish’ attribute associated with the Vuarnet name, and such an association was valued in the different extension contexts.
Two caveats should be noted, however, concerning the relative extendibility of concrete and abstract associations. First, concrete attributes can be transferred to some product categories (Herr et al., 1996). Farquhar et al. (1992) have argued that if the parent brand has a concrete attribute association that is highly valued in the extension category because it creates a distinctive taste, ingredient, or component, an extension on that basis can often be successful. Second, abstract associations may not always transfer easily. Bridges et al. (2000) examined the relative transferability of product-related brand information when it was either represented as an abstract brand association (e.g., ‘durable’) or when it was represented as a concrete brand association. Surprisingly, the two types of brand images extended equally well into a dissimilar product category, in part because subjects did not seem to believe that the abstract benefit would have the same meaning in the extension category (e.g., durability does not necessarily ‘transfer’ because durability for a watch is not the same as durability for a handbag).
Finally, Joiner & Loken (1998), in a demonstration of the inclusion effect in a brand extension setting, showed that consumers often generalized possession of an attribute from a specific category (e.g., Sony televisions) to a more general category (e.g., all Sony products) more readily than they generalized to another specific category (e.g., Sony bicycles). This inclusion effect was attenuated when the specific extension category increased its typicality to the general category (e.g., Sony cameras vs. Sony bicycles).
Characteristics of the Extension Category
Prior research has shown that some seemingly appropriate extensions may be dismissed because of the nature of the extension product involved. If the product is seen as comparatively easy-to-make, such that brand differences are hard to come by, then a high quality brand may be seen as incongruous or, alternatively, consumers may feel that the brand extension will attempt to command an unreasonable price premium and be too expensive. For example, Aaker & Keller (1990) showed that hypothetical extensions such as Heineken popcorn, Vidal Sassoon perfume, Crest shaving cream, and Häagen-Dazs cottage cheese received relatively poor marks from experimental subjects, in part because all brands in the extension category were seen as being about the same in quality, suggesting that the proposed brand extension was unlikely to be superior to existing products. When the extension category is seen as difficult to make, on the other hand, such that brands potentially vary a great deal in quality, there is a greater opportunity for a brand extension to differentiate itself, although consumers may also be less sure as to what exactly will be the quality level of the extension (Kardes & Allen, 1990).
Other researchers have looked at the choice context faced by the extension in the new category. For example, McCarthy et al. (2001) found that brand extension advantages over suggestive new brand names were confined to situations of limited information processing and better fit. When consumers processed information more deeply, however, new brands could perform as well as or better than brand extensions. The researchers also found differences in their experimental study between consumer attitudes and their choices, suggesting an important caveat to generalizations from lab studies (see also Hem & Iverson, 2001).
Characteristics of the Extension Marketing Program
Prior research suggests that introductory marketing programs for extensions can be more effective than if the new product was launched with a new name (Kerin et al., 1996; Smith, 1992; Smith & Park, 1992). Moreover, a number of studies have shown that the information provided about brand extensions through its supporting marketing programs, by triggering selective retrieval from memory, may ‘frame’ the consumer decision process and affect extension evaluations (e.g., see Boush, 1993).
For example, Aaker & Keller (1990) found that cueing or reminding consumers about the quality of a parent brand did not improve evaluations for poorly-rated extensions. Because the brands they studied were well-known and well-liked, such reminders may have been unnecessary. Elaborating briefly on specific extension attributes about which consumers were uncertain or concerned, however, by clarifying the nature of an important attribute of the new product, appeared to be effective in inhibiting negative inferences and in reducing the salience of consumers’ concerns about the firm’s credibility in the extension context, leading to more favorable evaluations.
Bridges et al. (2000) found that providing information could improve perceptions of fit in two cases when consumers perceived low fit between a parent brand and an extension. When the parent brand and the extension shared physical attributes but the parent brand image was non-product-related and based on abstract user characteristics, consumers tended to overlook an obvious explanatory link between the parent brand and extension. Information that raised the salience of the physical relationship relative to distracting non-product-related associations—a ‘relational’ communication strategy—improved extension evaluations. When the parent brand and the extension only shared non-product associations and the parent brand image was product-related, consumers often made negative inferences on the basis of existing associations. In this case, providing information that established an explanatory link on an entirely new, ‘reassuring’ association—an ‘elaborational’ communication strategy—improved extension evaluations. They concluded that the most effective communication strategy for extensions appeared to be one which recognized the type of information that was already salient for the brand in the minds of consumers when they first considered the proposed extension, and highlighted additional information that would otherwise be overlooked or misinterpreted (see also Chakravarti et al., 1990).
Lane (2000) found that repetition of an ad that evoked primarily benefit brand associations could overcome negative perceptions of a highly incongruent brand extension. Moreover, for moderately incongruent brand extensions, even ads that evoked peripheral brand associations (e.g., brand packaging or character) could improve negative extension perceptions with sufficient repetition. Process measures revealed that, as a result of multiple exposures to ads for incongruent extensions, positive extension thoughts and thoughts of extension consistency eventually overcame the initial negative thoughts of inconsistency. In a somewhat similar vein, Barone et al. (2000) experimentally demonstrated that positive mood primarily enhanced evaluations of extensions viewed as moderately similar (as opposed to very similar or dissimilar) to a favorably evaluated core brand. These effects of mood were found to be mediated by perceptions of the similarity between the core brand and the extension, as well as the perceived competency of the marketer in producing the extension.
Finally, research has explored several other aspects of extension marketing programs. Keller & Sood (2001a) found that ‘branding effects,’ in terms of inferences based on parent brand knowledge, operated both in the absence and presence of product experience with an extension, although they were less pronounced or, in the case of an unambiguous negative experience, even disappeared. In considering the effects of retailer displays, Buchanan et al. (1999) found that evaluations of a ‘high equity’ brand could be diminished by a unfamiliar competitive brand when: 1) a mixed display structure led consumers to believe that the competitive brand was diagnostic for judging the high equity brand; 2) the precedence given to one brand over another in the display made expectations about brand differences or similarities accessible; and 3) the unfamiliar competitive brand disconfirmed these expectations.
Branding Strategies and Alliances
Brand strategies concern how brand elements such as the brand name are employed across the products of a firm. LaForet & Saunders (1994) conducted a content analysis of the branding strategies adopted by 20 key brands sold by each of 20 of the biggest suppliers of grocery products to Tesco and Sainsbury, Britain’s two leading grocery chains, and found that a variety of different branding approaches had been adopted. Essentially, these different approaches involved using common and/or distinct brands in various ways across the products sold by the firms (see also LaForet & Saunders, 1999).
One frequently employed branding strategy, a sub-brand strategy—whereby an existing name is combined with a new name to brand new products—has received some research attention. Prior research has shown that a sub-branding strategy can enhance extension evaluations, especially if the extension is more removed from the product category and less similar in fit (Keller & Sood, 2001a; Milberg et al., 1997; Shenin, 1998). Moreover, it has also been shown that a sub-brand can also protect the parent brand from any unwanted negative feedback (Janiszewski & van Osselaer, 2000; Kirmani et al., 1999; Milberg et al., 1997), but only if the sub-brand consists of a meaningful individual brand that precedes the family brand (Keller & Sood, 2001b). Thus, it appears that the relative prominence of the brand elements determines which element(s) becomes the primary one(s) and which element(s) becomes the secondary one(s) (see also Park et al., 1996).
Wänke et al., (1998) showed how a sub-branding strategy could help to set consumer expectations. Experimentally, a new compact car manufactured by a sports car company that had existing sports cars branded Winston Silverhawk, Winston Silverpride, and Winston Silverstar, received a more sports-car typical evaluation when its name reflected the continuation (Winston Silverray) rather than the discontinuation of previous models (Winston Miranda). The contrast effect created by the name discontinuation strategy was more pronounced for non-experts than experts.
In an econometric study of ‘twin automobiles’ (i.e., made in the same plant with essentially the same physical attributes but different names, such as Ford Motor Company’s Ford Thunderbird and Mercury Cougar), Sullivan (1998) found that twins were not perceived as perfect substitutes—their relative prices differed. Most importantly, parent brand quality reputation affected the relative prices of the twin pairs. Specifically, the average quality of the parent line increased the demand for used cars sold under the parent brand name.
Finally, Bergen et al. (1996) studied branded variants—the various models that manufacturers offer different retailers (see also Shugan, 1989). They argued that as branded variants increased, consumers’ cost of shopping for a branded product across stores increased, leading to less such behavior. Because reduced shopping implies reduced competition, they reasoned that retailers presumably should be more inclined to carry the branded product and provide greater retail service support. An empirical examination with data from three retailers across 14 product categories supported these notions.
Brand alliances, where two brands are combined in some way as part of a product or some other aspect of the marketing program, come in all forms (Rao, 1997; Rao et al., 1999; Shocker et al., 1994). Prior research has explored the effects of co-branding, ingredient branding strategies, and advertising alliances.
Park et al., (1996) compared co-brands to the notion of ‘conceptual combinations’ in psychology. Experimentally, they explored the different ways that Godiva (associated with expensive, high-calorie boxed chocolates) and Slim-Fast (associated with inexpensive, low-calorie diet food) could introduce a chocolate cake mix separately or together through a co-brand. The findings indicated that a co-branded brand extension that combined two brands with complementary attribute levels appeared to have a better attribute profile in the minds of consumers than either a direct extension of the dominant brand, or an extension that consisted of two highly favorable, but not complementary, brands.
Consistent with the conceptual combination literature, they also found that consumers’ impressions of the co-branded concept were driven more by the header brand (e.g., Slim-Fast chocolate cake mix by Godiva was seen as lower calorie than if the product was called Godiva chocolate cake mix by Slim-Fast; the reverse was true for associations of richness and luxury). They also found that consumers’ impressions of Slim-Fast after exposure to the co-branded concept were more likely to change when it was the header brand than when it was the modifier brand. Their findings show how carefully selected brands can be combined to overcome potential problems of negatively correlated attributes (e.g., rich taste and low calories).
Simonin & Ruth (1998) found that consumers’ attitudes toward a brand alliance could influence subsequent impressions of each partner’s brands (i.e., such that spillover effects existed), but these effects also depended on other factors such as product ‘fit’ or compatibility, and brand ‘fit’ or image congruity. Brands less familiar than their partners contributed less to an alliance but experienced stronger spillover effects than their more familiar partners. Similarly Voss & Tanshuhaj (1999) found that consumer evaluations of an unknown brand from another country were more positive when a well-known domestic brand was used in an alliance.
Finally, Levin & Levin (2000) explored the effects of dual branding, which they defined as a marketing strategy in which two brands (usually restaurants) share the same facilities while providing consumers with the opportunity to use either one or both brands. They found that when two brands were linked through a dual-branding arrangement and both brands were described by the same set of attributes, then the effect of dual branding was to reduce or eliminate the contrast effects. When two brands were linked through a dual branding arrangement and the target brand was less well-specified than the context brand, then the effect of dual branding was to increase assimilation effects.
A special case of co-branding is ingredient branding, which involves creating brand equity for materials, components, parts, etc. that are necessarily contained within other branded products (McCarthy & Norris, 1999; Norris, 1992). Carpenter et al. (1994) found that the inclusion of a branded attribute (e.g., ‘Alpine Class’ fill for a down jacket) significantly impacted consumer choices even when consumers were explicitly told that the attribute was not relevant to their choice; subjects evidently inferred certain quality characteristics as a result of the branded ingredient. Brown & Carpenter (2000) offer a reasons-based account for the trivial attributes effect and show that the effect will depend on the choice context involved. Finally, Broniarszk & Gershoff (2001) show that the effect of trivial differentiation is more pronounced with strong brands.
Desai & Keller (2001) conducted a laboratory experiment to consider how ingredient branding affected consumer acceptance of an initial line extension, as well as the ability of the brand to introduce future category extensions. Two particular types of line extensions, defined as brand expansions, were studied: 1) slot filler expansions, where the level of one existing product attribute changed (e.g., a new type of scent in Tide detergent); and 2) new attribute expansions, where an entirely new attribute or characteristic was added to the product (e.g., cough relief liquid added to Life Savers candy). Two types of ingredient branding strategies were examined by branding the target attribute ingredient for the brand expansion with either a new name as a self-branded ingredient (e.g., Tide with its own EverFresh scented bath soap) or an established, well-respected name as a co-branded ingredient (e.g., Tide with Irish Spring scented bath soap). The results indicated that with slot filler expansions, although a co-branded ingredient facilitated initial expansion acceptance, a self-branded ingredient led to more favorable subsequent extension evaluations (see also Janiszewski & van Osselaer, 2000). With more dissimilar new attribute expansions, however, a co-branded ingredient led to more favorable evaluations of both the initial expansion and the subsequent extension.
Finally, Venkatesh & Mahajan (1997) derived an analytical model based on bundling and reservation price notions to help formulate optimal pricing and partner selection decisions for branded components. In an experimental application in the context of a university computer store selling ‘486 class’ laptop computers, they showed that at the bundle level, an all-brand Compaq PC with Intel 486 commanded a clear price premium over other alternatives. The relative brand ‘strength’ of the Intel brand, however, was shown to be stronger in some sense than that of the Compaq brand.
Samu et al., (1999) showed that the effectiveness of advertising alliances for new product introductions depended on the interactive effects of three factors: the degree of complementarity between the featured products, the type of differentiation strategy (common versus unique advertised attributes with respect to the product category), and the type of ad processing (top-down or bottom-up) that an ad evoked (e.g., explicitness of ad headline).
Conclusions and Future Prospects
Branding and brand management has clearly become an important management priority in the past decade or so. The academic research that was reviewed has covered a number of different topics and incorporated a number of different studies that have collectively advanced our understanding of brands. Before considering future research opportunities, it is worth taking stock of the progress that has been made and the kinds of generalization that might be suggested by the research reviewed.
Prior research has convincingly demonstrated the power of brands. Branding effects are pervasive, and the effects of virtually any marketing activity seem to be conditioned or qualified by the nature of the brands involved. In particular, consumer response to a product and its prices, advertising, promotions, and other aspects of the marketing program have been shown to depend on the specific brands in question. In understanding how these effects are manifested, essentially all the theoretical approaches one way or another interpret these branding effects in terms of consumer knowledge about the brand and how that knowledge affects consumer behavior. The particular dimensions or aspects of brand knowledge that drive these differences vary, however, by theoretical account and by the particular problem being investigated. Similarly, the exact mechanism involved also varies according to the brand setting under study.
Because of the primary explanatory role that brand knowledge plays as an antecedent, however, branding effects are highly dependent on the context involved. Highlighting brand-related information can activate certain brand associations and not others in a manner to produce different outcomes. This differential accessibility may be a result of the cues in the marketing environment from the marketing program, or through other means. Branding effects can thus be surprisingly complex. Moreover, there is inherent complexity with brands themselves as brand names, logos, symbols, slogans, etc. all have multiple dimensions, which each can produce differential effects on consumer behavior.
As a result, brand management challenges can be especially thorny. There are a number of input variables that come into play and a number of outcome variables that may be of interest. Only by understanding the totality of the possible antecedents and consequences of brand marketing activity and the possible mechanisms involved can proper analysis be conducted and decisions executed.
Although much progress has been made, especially in the last decade or so, a number of important research priorities exist that suggests that branding will be a fertile area for research for years to come (see Shocker et al. (1994) for an historical overview of branding progress and research agenda, as well as Berthon et al. (1999)). The review of the different areas highlights a number of specific research directions in those various research programs. The summary observations above help to suggest a broad research agenda for branding and brand management research. Six potential lines of such research are highlighted here.
Deeper, More Integrated Branding Research
In a general sense, progress in our understanding of branding and brand management will depend critically on three key areas: 1) the development of new conceptual models of brand equity; 2) the calibration of new brand equity measures and metrics; and 3) increased attention to the effects of integrated marketing programs and activities. The reality is that marketing has become increasingly complex as more and more ways to communicate with consumers, distribute products, etc. are being adopted by leading-edge firms. As a consequence, to provide greater managerial insight and guidance, there also need to be more sophisticated conceptual approaches to branding, as well as correspondingly more varied and detailed measures of branding effects. Specifically, research imperatives in these three areas and their rationale are as follows:
- Develop richer, more comprehensive, and actionable models of brand equity. There is still a need to develop more fully articulated models of brand equity that provide greater expository detail and which are useful in aiding managers across a broad range of decision settings. At the same time, there also need to be more parsimonious representations of these models to facilitate their applicability and use. Because of their central role, models that provide a more complete articulation of brand knowledge are especially needed. Along those lines, more work needs to be done to better conceptualize ‘brand intangibles’ in terms of structure and processes. In all these pursuits, much opportunity exists for multi-disciplinary work that relies on complementary theoretical approaches and perspectives.
- Design better ‘brand metrics’and more insightful measures of brand equity. As in virtually any area of marketing, there is a need to develop more insightful, diagnostic measures of branding phenomena. As is often the case, this will clearly involve multiple methods and measures. As suggested above, different decision makers have different interests and therefore will require different types of information. To practicing managers, it is especially important to develop better measures that are able to directly relate marketing activity to actual brand performance. To senior management, it is especially important to develop highly reliable brand valuation techniques and means of assessing returns on brand investments. Of critical importance in that regard is achieving a better understanding of the value and contribution of brands in various joint branding contexts and arrangements, be it with another brand, a person, an event, or whatever.
- Understand the effects of the increasingly diverse array of marketing activities and how to optimally integrate these various activities to maximize their collective contribution. As new, alternative marketing approaches supplement more traditional strategies and tactics -especially in terms of communication and distribution—it will be important to fully understand their effects on consumers and how they should best be designed and implemented. At the same time, there is a critical need to understand how these different marketing activities should best be assembled as part of the marketing program. The advantages and disadvantages of different marketing activities and their potential interactions must be understood so that they can be properly ‘mixed and matched’ and integrated so that the ‘whole can be greater than the sum of the parts.’
Broader Branding Research
It will be especially important for branding research to take a broad perspective and account for all the different forces involved. In particular, it will be important to take a strong internal branding point of view and consider the role of employees and other individuals inside the company in building and managing brand equity. At the same time, it has become harder and harder for companies to ‘go it alone’ with their brands in the marketplace. As a result, it will also be important to take a strong external point of view and consider how brands can ‘borrow’ equity in various ways. These realizations suggest the fourth and fifth lines of research.
- Consider organizational, ‘internal’ branding issues. A relatively neglected area of branding is prescriptive analysis of how different types of firms should best be organized for brand management. Additionally, there needs to be more insight into how to align brand management within the organization and those efforts directed to existing or prospective customers outside the organization.
- Obtain greater understanding of how meaning ‘transfers’ to and from brands. As brands are often linked with other entities—people, places, companies, brands, events, etc.—it is important to understand how the knowledge about these other entities impacts brand knowledge. In what ways do the images of country-of-origin or country-of-brand, celebrity spokespeople, retail store, etc. change or supplement the image of a brand? At the same time, it is important to understand how the meaning of a brand transfers to other brands, products, etc.
More Relevant Branding Research
Finally, as branding is applied in more and more different settings, brand theory and best practice guidelines need to be refined to reflect the unique realities of those settings, as suggested by the following.
- Develop more refined models for specific application areas of branding. Finally, greater attention must be applied to understanding similarity and differences in branding for different types of application areas. Obviously, one priority area of study is ‘virtual’ branding and how brands should be built on the Internet. In a general sense, how broadly applicable are the guidelines that emerge from academic research? Which principles are valid and which ones need to be modified or supplemented in some way?
One challenge in pursuing research in this area, as well as the other five areas, will be to achieve the necessary rigor to satisfy the highest academic standards, while also achieving the necessary relevance to satisfy the most demanding industry practitioners.