Categorization Research and Brand Extensions

Laurette DubT, UniversitT de MontrTal
Bernd H. Schmitt, Columbia University
Sheri Bridges, Wake Forest University
[ to cite ]:
Laurette DubT, Bernd H. Schmitt, and Sheri Bridges (1992) ,"Categorization Research and Brand Extensions", in NA - Advances in Consumer Research Volume 19, eds. John F. Sherry, Jr. and Brian Sternthal, Provo, UT : Association for Consumer Research, Pages: 255-259.

Advances in Consumer Research Volume 19, 1992      Pages 255-259


Laurette DubT, UniversitT de MontrTal

Bernd H. Schmitt, Columbia University

Sheri Bridges, Wake Forest University


At a time of escalating marketing costs and an increasing number of new product failures, marketers have turned their attention to brand extensions to reduce the expenses and risks associated with new product introductions. To explain how associations and evaluations of the original brand name can transfer to a brand extension, researchers have found categorization models to be a useful framework. According to the categorization approach, consumers evaluate a brand extension on the basis of the fit between the original brand and the product category to which the brand has been extended. This assumes that the existing brand and the extension product category exist in consumers' minds as relatively stable concepts.

How are brands and product categories cognitively organized? How do consumers mentally represent brands and product classes? Based on Rosch and Mervis's (1975) classic work in cognitive psychology, it has been suggested that consumers form relatively stable, hierarchical representations of product classes (e.g., cars), product types (e.g., sports cars) and brands (e.g., Corvette) in long-term memory. These categories are cognitively represented as collections of features (e.g., product or brand characteristics) which are organized around a central tendency, the prototype. It is further assumed that category exemplars vary in terms of their typicality or similarity to the category prototype and that similarity or "fit" with the category is based on a feature matching process.

Such a feature-based conceptualization, however, cannot account for the elaborative and interactive processes by which consumers perceive and evaluate brand extensions--processes that include the formation of cognitive structures related to the brand, to the existing products in the brand line, the extension product category, and the existing brands in the extension product category.

Recent research in psychology and consumer behavior suggests that the categorization process involved in the perception and the evaluation of brand extensions may be more flexible and context dependent than previously thought (Aaker and Keller 1990; Barsalou 1985; Boush and Loken 1991; Park, Milberg and Lawson 1991). For example, individuals may not use one organizing principle, such as similarity to a prototype, to assess typicality. Instead they may derive a category on the basis of functional goals (Barsalou 1985). Therefore brand extensions, too, may show a large degree of flexibility. Moreover, typicality itself seems to depend on the context in which an object is judged (Roth and Shoben 1983). As a consequence, the relationship between the original brand and the extension may also be mediated by contextual factors. In addition, the influence and affect transfer which occurs in brand extensions may not be unidirectional. That is, perceptions of the core brand may be altered by the way an extension is perceived and categorized.

The session summarized here focused on these recent developments in categorization research on brand extensions providing evidence for the dynamic nature of the link between the original brand and the extension. Farquhar and Herr demonstrated the separate effects of the strength of the category-to-brand association (the dominance) and strength of the brand-to-category association (the typicality) in delineating the limits of using brand equity for brand extensions. MacInnis, Nakamoto and Mani investigated contextual factors that mediate the salience of knowledge domains to be used in consumers' judgments about product categories, and more specifically, in brand extensions. Manipulating the salience of common and distinctive features of brand extensions and existing branded products, Boush reported asymmetry effects as they occur in similarity judgments involved in consumers' perceptions of brand extensions. Based on work by Murphy and Medin demonstrating the role of theories in conceptual coherence ( Murphy and Medin 1985), Bridges proposed a process model of brand extension evaluations, suggesting that the existence of explanatory links between the extension and the brand determines both the type of evaluation process that consumers engage in and the valence of their evaluation. Building on Bridges's model, Milberg and Park showed that the existence of explanatory links among existing products and extensions (product-level linkages) and the degree of consistency between brand extensions and the brand image (image-level linkages) affect the beliefs and attitudes toward the original brand and its current offerings. As additional evidence of the reciprocal impact of brand extensions on the original brand, Loken and Roedder-John showed that the degree of typicality of the individual product offerings to the brand category influences the dilution of the overall brand image.

In light of the evidence for contextual variability and sensitivity to context in consumers' perception of brand extension, the present session introduced a constructivist approach to categorization. Barsalou, who recently suggested that mental frames rather than feature lists may be the most appropriate representations of human cognition (Barsalou 1991), presented, in the introductory paper of the session, the basic tenets of frame theory. According to Barsalou, to capture all the information in natural knowledge, a mental representation must include mental structures for multivalued attributes, conceptual relations, and frame recursion. Such mental structures form the core of frame-like representations but are nonexistent in feature list representations. Viewing brand extensions as conceptual combinations, Schmitt and DubT followed up on Barsalou's recent work and demonstrated that a constructivist approach may be required to account for the emergence of highly contextualized and unique features of brand extensions.


Aaker, David and Kevin Lane Keller (1990), "Consumer Evaluations of Brand Extensions," Journal of Marketing, 54, 27-41.

Barsalou, Lawrence W. (1991), "Components of Conceptual Representations: From Feature Lists to Complex Frames," in I.Van Mechelen, J. Hampton, R. Michalski, and P. Theuns (Eds.), Categories and Concepts: Theoretical Views and Inductive Data Analysis. San Diego, CA: Academic Press.

Barsalou, Lawrence W. (1985), "Ideals, Central Tendency, and Frequency of Instantiation as Determinants of Graded Structure in Categories," Journal of Experimental Psychology: Learning, Memory, and Cognition, 11 (4), 629-648.

Boush, David and Barbara Loken (1991), "A Process-Tracing Study of Brand Extension Evaluations," Journal of Marketing Research, 28, 18-28.

Murphy, Gregory and Douglas Medin (1985), "The Role of Theories in Conceptual Coherence," Psychological Review, July, 289-316.

Park C. Whan, Sandra Milberg, and Robert Lawson (1991), "Evaluation of Brand Extensions: The Role of Product Feature Similarity and Brand Concept Consistency," Journal of Consumer Research,18, 185-193.

Rosch, Eleanor and Carolyn Mervis (1975), "Family Resemblances: Studies in the Internal Structure of Categories," Cognitive Psychology, 7, 573-605.

Roth, Emilie M. and Edward J. Shoben (1983), "The Effect of Context on the Structure of Categories," Cognitive Psychology, 15, 346-378.




Lawrence Barsalou, University of Chicago

According to a common view in categorization research, categories possess the property of graded structure. This paper discusses the various determinants of graded structure for different types of categories (e.g., common taxonomic categories and goal-derived categories) and reviews recent psychological evidence for the instability of graded structure. Moreover, studies are cited which suggest that categories may not be represented as independent feature lists but as attribute-value sets. For example, when individuals think about a particular car, abstract attributes will adopt specific values, such as "Liz" for "driver," "gasoline" for "fuel", "four-cylinder" for engine, and so on. In addition, the cognitive representation of categories may include relations and constraints (e.g., "driver" buys fuel," "fuel" flows to "engine"). In light of the evidence for contextual variability and against feature lists, frames rather than independent feature lists may provide the fundamental representation of knowledge in human cognition. Frames contain attribute-value sets, and the relations of a frame connect attributes into a pattern of systematicity. Therefore, from a categorization perspective, frame theory may offer the most comprehensive theoretical model for the study of brand extensions.



Bernd Schmitt, Columbia University

Laurette DubT, UniversitT de MontrTal

Following recent developments in cognitive psychology and linguistic research, we propose to view brand extensions (e.g. McDonald's Theme Park) as conceptual combinations. In conceptual combinations (e.g. "apartment dog") the former noun ("apartment") is seen as a modifier that contextualizes the attributes of the latter noun ("dog") called the "header." In this process, individuals make extensive use of their world knowledge and attributes may be evoked that are not typical of either constituent concept (e.g., the attribute "neurotic" for "apartment dog"). Similarly, in brand extensions the original brand or company name (e.g. McDonald's) may act on the "head concept" of the extension category (e.g. theme parks) as a "modifier." Using a variety of fictitious extensions, we show that brand extensions exhibit the same degree of flexibility and sensitivity to context as conceptual combinations. Moreover, preliminary evidence suggests that the contextual and relational structure of brand extensions may be best explained by dynamic frame-like structures rather than relatively static feature list representations.



Peter H. Farquhar, Carnegie Mellon University

Peter M. Herr, University of Colorado, Boulder

This paper explores the effects of dominance and typicality on extending a brand across product categories. We begin by defining "dominance" as the strength of the category-to-brand association, and "typicality" as the strength of the brand-to-category association for a consumer. We compare these definitions with others used in categorization research and then examine several operational measures of dominance and typicality.

This paper considers previous research by Herr, Farquhar, and Fazio (1990) that shows how learning associations for a new brand extension depends upon the brand's dominance in the parent category and on the relatedness of the parent and target categories. Since dominance and typicality are correlated for many brands, the present research measures their separate effects on brand extendability. These results on typicality are compared with other findings recently reported in the consumer research literature.

This paper then separates the effects of dominance and typicality in determining the boundaries of a brand franchise. When the same brand has been extended to a variety of target categories, we show how the brand's dominance in the parent category is unaffected, but typicality is diminished. We examine the vulnerability of such a brand against competitive strategies for product and brand positioning.



Deborah J. MacInnis, University of Arizona

Kent Nakamoto, University of Arizona

Guyathri Mani, University of Arizona

Considerable research in the brand extension domain has indicated that similarity between an original and extension product category is a critical predictor of consumers' evaluations of a brand extension (Aaker and Keller 1990; Boush et al 1987; MacInnis and Nakamoto 1991). However, many questions arise about how such similarity judgments are made. First, what are the knowledge domains consumers use to judge comparability or similarity across products? Products may not only share physical attributes, but also benefits, usage occasions, usage locations, users, etc.

Second, are certain knowledge domains consistently more salient (i.e. more strongly tied to the product category than others)? For example, across comparison contexts, are the physical attributes which two product categories share more likely to come to mind than similarities in other domains? Attribute similarities may in fact be more salient due to their physical observability (Paivio 1971) and their role in defining category membership.

Third, what factors moderate the salience or activation of knowledge domains? Consumers' knowledge about individual product categories, context effects created by the specific product categories being compared, the task (similarity, choice, evaluation), and other information inherent in the task context may influence the extent to which certain bases are salient (Chakravari and Lynch 1983). Understanding the moderators of salience is important from a marketing context since marketers are often in a position of communicating attribute similarities or differences between two product categories that may not have been immediately obvious (salient ) to consumers.

Fourth, to what extent will the activation of similarities between product categories in a certain knowledge domain also affect perceived differences among product categories. Will, for example, comparisons of the similarities between two product categories on a goal derived category also cue consumers to differences between product categories on goal-derived categories? The salience of specific bases may not only be a function of knowledge structure and the task (as indicated above) but also of a spreading activation effect achieved through elaboration (Collins and Loftus 1975).

Finally, to what extent will articulated similarities and differences between two product categories in specific knowledge domains map onto global perceptions of similarity. Will articulated dimensions only partially reflect overall similarity judgments?

We report an exploratory study designed to address these issues. We examine the nature of consumers' knowledge structure for a product category, and its influence on the comparison of disparate products. We also examine the factors which affect the salience of various domains of knowledge in comparing products. We investigate the role that articulated similarities have on articulating differences among product categories. These results have both theoretical and methodological implications for research which studies product category comparisons.



Sheri Bridges, Wake Forest University

This research proposes a model of the process by which consumers judge brand extensions, then uses the model to identify variables that affect the evaluative outcomes of the judgment process. Underlying the model is the idea that the perceived fit between the extension and brand expectations determines both the type of evaluation process consumers will engage in and the valence of their evaluations.

Based on work in the categorization literature by Murphy and Medin (1985), the key notion of an explanatory link is developed to explain how consumers make sense of the relationship between the extension and the brand. By forming explanatory links at the product or image level, consumers are able to understand how products in the brand line fit together and, consequently, are able to maintain a cohesive, unified brand schema.

Two experiments support hypotheses concerning variables that affect consumers' ability to find explanatory links and, thereby, affect perceived fit. These variables are type of brand schema, type of relationship between the brand and the extension, and type of extension information provided to consumers. Type of brand schema depends on the dominant type of associations consumers have for the brand: concrete/product-related, abstract/product-related, or image-related. The relationship between the brand and the extension is characterized as being of two types: shared physical product attributes (e.g. a watch and a kitchen timer) or no product-level relationship (e.g. a watch and a handbag). The type of information provided to consumers about the extension may be designed to raise the salience of the explanatory link, or to raise the salience plus relevance (i.e. importance) of the explanatory link.

Specifically, when the brand and the extension share physical product attributes, evaluations are higher if the brand's schema is image-related that if it is product-related. The effects of this interaction between schema type and brand-to-extension relationship are moderated by information provided about the explanatory link between the extension and the brand.

Predictions regarding the process model itself also are supported in an experiment in which thought protocols are elicited from participants. Subjects engage in holistic, category-based processing when the extension clearly fits or clearly does not fit with brand expectations. Perceived fit also affects the incidence of nullifying and supporting argumentation in the evaluation process and, ultimately, the valence of the evaluations.


Murphy, Gregory and Douglas Medin (1985), "The Role of Theories in Conceptual Coherence," Psychological Review, July, 289-316.



Sandra J. Milberg, Georgetown University

C. Whan Park, University of Pittsburgh

This paper examines situations in which brand extensions have either a positive or negative impact on the beliefs and attitudes toward the originating brand and its existing products. In particular, this research explores the role played by perceived category coherence (when the grouping of objects makes sense to the perceiver) in understanding when extensions dilute brand associations, fortify and enhance existing beliefs associated with the originating brand, or systematically modify and expand brand associations to adapt to changing markets.

According to ideas introduced by Murphy and Medin (1985) and applied to brand extension research by Bridges (1990), perceptions of category coherence are based on consumers' ability to recognize explanatory links among existing products and extensions (product-level linkages) and perceptions of how consistent products are with consumers' understanding of the brand image (image-image linkages).

When an extension is inconsistent with the brand image it can weaken existing brand beliefs, causing confusion and negative affect. When there are weak product-level linkages, associations with specific products and their related qualities can be diluted. This can also create confusion as to the definition of the business the firm is pursuing. On the other hand, when consumers perceive sensible product-level linkages and consistency between brand extensions and the brand image, extensions should serve to fortify and enhance existing brand beliefs. Conditions under which existing brand beliefs are systematically modified by the introduction of brand extensions, producing positive outcomes, are also explored.


Bridges, Sheri, 1990, "A Schema Unification Model of Brand Extensions," unpublished dissertation, Stanford University.

Murphy, Gregory and Douglas Medin, 1985, "The Role of Theories in Conceptual Coherence," Psychological Review, July 298-316.



David M. Boush, University of Oregon

The recently-voiced perspective that brands can be viewed as categories (Boush and Loken 1991) stresses both the importance of graded structure (a typicality gradient from the most to the least representative products that share a given brand name) and the flexible, constructive nature of brands as "ad hoc" categories (Barsalou 1983, 1985). Further, the transfer of attitude from a brand's existing products to a new product does not occur in an "all or nothing" fashion. Rather, it is a linear function of the degree to which a product is perceived to represent the brand. The notion of brand attitude transfer based on a typicality gradient suggests that the key to understanding brand extension effects is to understand the structure of brands as categories and the way marketers can influence that structure.

According to Tversky's (1977) contrast model, consumers consider both matching and distinctive features when making similarity judgments (e.g. judgments about the similarity between a potential brand extension and existing branded products.) Perhaps the most significant implication of the contrast model is that it predicts asymmetries in similarity judgments (i.e. the variant is more similar to the prototype than vice versa). Family branded products are frequently (if not usually) characterized by extension from relatively representative (prototypical) products to less representative variants. For example, Hallmark has extended from greeting cards to other products involved with "social expression" such as gift wrap and paper party goods. Asymmetry in similarity judgments implies both that some products might be considerably better than others if you want to enter a new category, and that the flow of attitude from existing products to extensions might be stronger or weaker than reciprocal flow. The contrast model has been demonstrated for consumer products (Johnson 1986) but not in a brand extension context. Data will be reported concerning: (1) the existence of predicted asymmetries in similarity judgments; (2) effects of brand names (including secondary brand names) on similarity judgments; and (3) effects of brand names on asymmetries in product similarity judgments.


Barsalou, Lawrence W. (1983), "Ad Hoc Categories," Memory and Cognition, 11(3) 211-227.

Barsalou, Lawrence W. (1985), "Ideals, Central Tendency, and Frequency of Instantiation as Determinants of Graded Structure in Categories," Journal of Experimental Psychology; Learning, Memory, and Cognition, 11(4) 629-648.



Barbara Loken, University of Minnesota

Deborah Roedder John, University of Minnesota

Researchers are increasingly relying on categorization theories to attempt to understand the manner in which the original product category structure impacts judgments about brand extensions (Boush and Loken 1991; MacInnis and Nakamoto 1990; Park et al 1989). The present research identifies situations in which brand extensions may be more or less likely to "dilute" brand equity associated with the brand's current product offerings. We define brand equity associated with the overall brand image in terms of the attribute associations that consumers have learned to associate with a brand name and that underlie favorable impressions of the brand. Brand equity dilution is, therefore, a decrease in the strength of these beliefs about the originating brand resulting from brand extensions.

However, brand equity dilution may also be seen as having effects on belief about the brand's current individual product offerings. It seems likely that products closely associated with the originating brand name and considered to be more typical exemplars of the originating brand's product offerings would be most affected by situations that dilute overall brand image. Atypical instances are more likely to be disassociated from the category and less likely, therefore, to be influenced by overall category judgments. Thus, we anticipate that the greater the typicality of individual products marketed under the originating brand name, the greater will be the dilution of beliefs about those individual products.

In sum, our research identifies situations that influence dilution of the overall brand image and examines the parallels between these effects and the effects on dilution associated with the brand's current individual product offerings. These latter effects are examined as a function of the degree of typicality of the individual product offering to the brand category.