Co-branding : the curious case of brand alliances with low brand concept consistency

Author: Siemon Feenstra (_11111119) Supervisor: dhr. drs. R.E.W. Pruppers

University of Amsterdam: Business School MSc. Business Administration: Marketing Track 27th January 2017

Master thesis

Co-branding: The curious case 


of brand alliances with 


Statement of originality

This document is written by student Siemon Feenstra who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.


Table of Contents

1. Introduction 1

1.1. Co-branding 1

1.2. The paradigm of brand fit 2

1.3. Contributions 4

1.4. Research outline 5

2. Why co-branding? 6

2.1. Brand equity 6

2.2. Associative network 6

2.3. Leveraging brand equity 7

2.4. Co-branding 9

3. Evaluations and the role of fit 13

3.1. Evaluations of co-branding 13

3.2. Categorization theory 14

3.3. Perceptual fit 15

4. Bases of brand fit 18

4.1. Functional and symbolic associations 18

4.2. Product Feature Similarity 19

4.3. Brand Concept Consistency 19

4.4. Integration 20

5. Hypotheses development 22

5.1. Introduction 22

5.2. Brand concept consistency 23

5.3. Host brand concept contingency 23

5.4. Mediation of brand fit 24

— Study I —

6. Methodology (Study I) 26

6.1. Research design 26

6.3. Stimuli development 27

6.4. Method 35

7. Results (Study I) 39

7.1. Manipulation checks 39

7.2. Hypotheses testing 44

8. Discussion (Study I) 52

— Study II —

9. Theoretical framework (Study II) 54

9.1. Introduction 54

9.2. Problem statement - Study II 54

9.3. Contribution 55

9.4. Theoretical framework 55

10. Methodology (Study II) 58

10.1. Research design 58

10.2. Stimuli development 58

11. Results (Study II) 63

11.1. Manipulation checks 63

11.2. Results - Study II 69

11.3. Additional analysis 74

12. Discussion (Study II) 75

13. General discussion & conclusion 78

13.1 Summary of Study I and II 78

13.2 Conclusions 78

13.3. Theoretical implications 80

13.3. Managerial implications 82

13.4. Limitations and future research 83

References 85

Appendices 1

A. First appendix - Additional analysis 1

B. Second appendix - Questionnaires Study I 8

C. Third appendix - Questionnaires Study II 17

Abstract

Over the years a popular strategy for introducing new consumer products is by two brands pursuing a co-branding strategy. There seems to be a consensus among scholars that one of the key success factors in predicting co-branding success is the customers’ perceived brand fit between the two brands forming the brand alliance. The paradigm that a high perceived brand fit will lead to suc-cessful brand alliances is widely accepted, however in practice we still see many failures of high-brand fit product launches, while on the other hand low-high-brand fit products can be very successful. This study aims to get a clearer understanding on how these mis-matches on paper can still be favor-ably evaluated in practice. The central premise of this study is based around the categorization theo-ry and how situational variables can provide context to how consumers judge the fit between two brands. Specifically in this research the role division of the two brands is used as a situational variable to influence the categorization process that consumers use. The results indicate that the context of which brand concept is perceived as the host brand changes the categorization process and subse-quently influences the perceived brand fit and alliance evaluation. It was found that the positive effect of brand concept consistency, in other words the increase in brand fit due to brand concept consis-tency changing from low to high, is more articulated for a symbolic host brand than a functional host brand. Based on the results we argue that the effect of high brand concept consistency is still very important and certainly exists for brand alliances, but we think that how important this is, is strongly contingent on the hosts’ brand concept. We even argue that this interaction between the host brand concept and brand concept consistency is so extreme that it’s not only a matter whether this effect is more articulated for one of these two conditions, but at certain conditions there is hardly any differ-ence at all.

1. Introduction

1.1. Co-branding

In the continuing pursuit of growth, companies are constantly looking for new growth oppor-tunities. One of the widely used strategies for growth are brand extensions, which is a new product with a well-known brand name but then in a new product category. Next to brand extensions, a pop-ular strategy for introducing new consumer products is co-branding, which is the paring of two or more brands to form a separate and unique product examples are e.g. a McFlurry with M&M’s, and Lenovo laptops with integrated Dolby Audio speakers (Park, Jun, & Shocker, 1996). A co-branding strategy is used by marketers to try to transfer the positive associations of the two partner brands to a newly formed co-branded product and strengthen the parent brand and extend customer value per-ceptions to a new product (Park et al., 1996; Washburn, Till, & Priluck, 2000). According to signaling theory, the combination of two brands provides consumers with greater assurance about product quality, which then results in higher product evaluations (Rao, Qu, & Ruekert, 1999). However ac-cording to Thompson and Stratton (2012) the academic research on co-branding’s impact on con-sumer decision making is still in its infancy. Therefore it is certainly not a closed case on why certain co-branded products succeed while others might fail.

1.1.2. Brand fit & categorization

There seems however to be a consensus among academia that one of the key success factors in predicting co-branding success is the customers’ perceived fit between both the product and the brands (Tauber, 1988; Simonin & Ruth, 1998; Baumgarth, 2003; Park et al., 1991; Desai & Keller, 2003; Dwivedi, Merrilees, & Sweeney, 2009; Völckner & Sattler, 2006; Helmig et al., 2008; Gam-moh, Voss, & Chakraborty, 2006). One of the reasons why brand fit is seen as one of the most impor-tant aspects for the success of a co-branded product is that the transfer of associations, such as per-ceived quality and other benefits of the two brands will increase when customers understand the rela-tionship between the brands and that this combination makes sense (Aaker & Keller, 1990; Park et al., 1996).

In order for consumers to perceive a brand fit they need to categorize two brands as belong-ing to the same cognitive category, which needs a relationship among the characteristics of the two brands (Simonin & Ruth, 1998). This relationship can determine on the level of abstraction in the brand associations, whereas brand associations can vary broadly from concrete physical product-re-lated attributes to abstract non-product reproduct-re-lated attributes that are more symbolic in nature, such as

e.g. ‘luxury’ or ‘status’ (Park, Jaworski, & MacInnis, 1986). For instance in the example of the watch brands Casio and Rolex, at the concrete level of the physical product features both brands share many attributes, however on a more abstract level of the brand images they are highly inconsistent, whereas only Rolex has been associated with the concepts of luxury and high status. With this insight in mind, consumers may perceive a weak relationship between Rolex and Casio based on their diver-gent brand images, which results in a low brand fit. While at the same time a consumer may perceive a high fit between Rolex and another luxury brand such as Jaguar or Louis Vuitton. Although on the concrete abstraction level of product features Rolex and Jaguar do not share many physical product features, the two brand images of luxury and status can provide context for why the two companies cooperate together (Barsalou, 1982; Medin, Goldstone, & Gentner 1993). This context can then in-fluence the perceived fit since consumers can categorize the abstract brand images of the two brands into the same cognitive category and then understand the alliance (Barsalou, 1985; Ratneshwar, Pechmann, & Shocker, 1996).


1.2. The paradigm of brand fit

The paradigm that a high perceived brand fit will lead to successful brand alliances is widely accepted, and has been researched by many scholars such as e.g. Simonin and Ruth (1998), Baum-garth (2003), and Desai and Keller (2003), however in practice we still see many failures of high-brand fit product launches, while on the other hand low-high-brand fit products can be very successful. For instance the perceived fit between the brand concepts of Acer and Ferrari could be marked as tenu-ous, however a laptop launched by Acer and Ferrari has been marked as one of the most successful alliances in the tech-industry (Lindström, 2010). This example is certainly not a rare isolated case as there are numerous examples of other partnerships with two highly divergent brand concepts that are allegedly successful in the marketplace, such as a line of kitchen appliances made by e.g. Philips and premium Italian design firm Alessi, and smartphones by LG and Prada. So even though almost all studies regarding brand fit have indicated that brand fit is a key success factor for a favorable al-liance evaluation these alal-liances seem to indicate that co-branded products can be evaluated favor-ably despite their low brand concept consistency. In regard to this insight, an interesting question has emerged: Could it be that under certain conditions brand fit is of less importance, and could this be due to how consumers categorize two brands differently based on contextual cues?

1.2.1. Problem definition & statement

As a summary it can be concluded that how consumers categorize two brands is influenced by context (Medin et al., 1993) and it has been researched how this categorization can be influenced by

the brand concept of the parent brand (Park et al., 1991), but in the setting of co-branding there is not a single brand, but both the brand concepts of the two brands are involved and can provide con-text to the consumer. Although brand concept consistency has proven to be a success factor for favor-able alliances evaluations (Simonin & Ruth, 1998; Baumgarth, 2003; Desai & Keller, 2003), alliances with low brand concept consistency can however succeed in the marketplace. This research therefore aims to get a clearer understanding on how these mis-matches on paper can still be evaluated favor-ably in practice. The central premise of this research is based around categorization theory and how situational variables, such as the brand concept of the parent brand can provide context for the con-sumer. However with co-branding there are two brand concepts involved, so especially when there are two highly divergent brand concepts involved, how do consumers categorize this complex context of the two brand concepts?

One specific form of co-branding that has received attention from scholars is ingredient branding (Desai & Keller, 2002; Uggla, 2004), in which the key attributes of one brand are incorpo-rated into a co-branded products as an ingredient, e.g. the M&M’s in a McFlurry. With ingredient branding the role division between the two brands in the alliance is clearly defined as a host and a partner ingredient brand, whereas the host brand is the primary brand associated with the secondary partner brand (Uggla, 2004). Since the categorization process of the consumer is dependent on con-text, this research will use this specified role division as a situational variable to try to influence how consumers categorize two brands based on the brand concept of the most dominant brand; the host brand. By doing so, this research aims to understand how this categorization process may be influ-enced by the context of which brand is perceived as the most dominant brand; the host brand. And to see whether this context of the role division may change the categorization process, which ulti-mately may influence the perceived brand fit and alliance evaluation. To answer this underlying ques-tion the following research quesques-tion has been formulated:

How can the detrimental effect of low brand concept consistency be avoided by the de-liberate choice and positioning of the partner brand?

1.2.2. Subquestions

In order to answer the research question several sub-questions were made as a guidance to address the relevant concepts and its theoretical foundation.

- What is co-branding and what are its benefits?

- What are the bases of brand fit and why is it important?

- What is brand concept consistency and what is its effect on the categorization process?

- How is brand fit perceived and processed by consumers and what is its effect with the customers’ evaluation of a co-branded product?

1.3. Contributions

This research builds onto two streams of existing literature, evidently the literature of brand extensions and co-branding in particular, and secondly at a more fundamental level of the catego-rization process and specifically how context can influence this process. First of all this paper will shed light on the relationship between three major constructs in both the brand extension and the co-branding literature, which are brand concept consistency, brand fit and alliance evaluation. Is it granted that high brand concept consistency always leads to a higher brand fit and subsequently to a higher alliance evaluation, or are there specific situations in which brand fit is of less importance? One of those situations is the brand concept, and this paper in an attempt to create a consensus of conflicting arguments made by Park et al. (1991) and Lansing & Olsen (2010) regarding a differential effect of brand concept consistency among brand concepts. Park et al. (1991) claimed that brand ex-tensions that belong to the same concept as the parent brand (high brand concept consistency) are evaluated more favorable than extensions that hold another brand concept (low brand concept con-sistency), and this argument was found to be valid for both functional as symbolic brands. However in a co-branding setting this argument had been rejected recently by a study of Lanseng and Olsen (2010), who claimed that the argument for a higher brand fit due to a high brand concept consistency was only valid for functional brands and not for symbolic brands.

Most studies in a co-branding setting distinguish between high and low brand concept consis-tency (e.g. Simonin & Ruth, 1998), or between functional-functional, symbolic-symbolic, and mixed-alliances (e.g. Lanseng & Olsen, 2012), but an important contribution of this study is that this paper is one of the first articles that distinguishes a brand alliance with low brand concept consistency by the role division of the two brands to see whether the categorization process is influenced by the brand concept of the host brand. This is important as Swaminathan, Gürhan-Canli, Kubat, & Hayran (2015) pointed out that researchers in previous studies (e.g. Simonin & Ruth, 1998; Park et al., 1996) have overlooked how the context of situational factors may have influenced how consumers catego-rize and interpret the fit between brand alliances. By testing whether the role division of the two brands in the brand alliance can influence the categorization process of consumers, this research builds on the knowledge of those situational factors and how it influences the categorization process.

1.3.2. Managerial contributions

The findings of this study provides implications for managers and marketers when a co-branding strategy is pursued. This research provides insight in how consumers judge the fit between two brands in a co-branding setting at different levels of brand concept consistency, and how context may influence this brand fit leading to more favorable alliance evaluations. First of all is this impor-tant for managers finding the right partner to form an alliance with, by providing insights in making the right trade-offs in partner selection. This is not only beneficial in partner selection, but by influ-encing the categorization process of consumers with framing and positioning the role division, the fit perception and alliance evaluation can be influenced. Meaning that even with low brand concept consistency brand alliances can be successful in certain situations, which may lead to complete new expanding opportunities for managers of both functional and symbolic brands.


1.4. Research outline

The following three chapters will provide the theoretical background with the most relevant findings of previous studies to create the foundation for this research. In Chapter 2 the concept and benefits of a co-branding strategy will be described with an overview of the main motives and bene-fits for firms to pursue a co-branding strategy, and the definitions and scope of this research. Chapter 3 gives an overview of how consumers evaluate co-branded products based on categorization theory and the associative network, and it will be explained what the role of perceptual fit entails in combi-nation with a brief understanding of different types of fit. One of those types of fit is brand fit, which will be more thoroughly described in Chapter 4. The combination and integration of these central concepts will lead to the hypothesis forming in Chapter 5. The research design incorporates two studies, which will be analyzed and discussed separately. Study I will start in Chapter 6 which describes the research design and the methodology being used. The results from Study I will be pre-sented in Chapter 7 followed by the discussion of these results in Chapter 8. The introduction, methodology, results and discussion of Study II are written in Chapters 9 till 12. Chapter 13 provides a general discussion and conclusion of the results from both studies and answers the research ques-tion in combinaques-tion with both the theoretical and managerial implicaques-tions, limitaques-tions and possible directions for future research.


2. Why co-branding?

2.1. Brand equity

A brand can be defined as “a name, term, sign, symbol, or design, or combination of them which is intended to identify the goods and services of one seller or group of sellers and to differenti-ate them from those of competitors” (Kotler, 1991, p. 442). The brand name of a product is a cue for consumers and represents those images that are formed due to previous experiences with the brand (Swait, Erdem, Louviere, & Dubelaar, 1993). A brand name can add incremental utility or value to a product, which is called brand equity. In one of the most widely cited articles in the field of branding, Keller (1993) describes the concept of consumer-based brand equity, which is “the differential effect of brand knowledge on consumer response to the marketing of the brand” (p. 1). This brand knowl-edge influences what a consumer thinks of a brand and determines what comes to mind. According to Keller (2003) all different kinds of information, such as awareness, attributes, benefits, images, thoughts, feelings, attitudes, and experiences may become part of consumer memory and affect con-sumer response to marketing activities. This is mainly important because by the use of marketing programs, marketers try to create a differentiated consumer response that gives a personal meaning towards the brand (Peter & Olson, 2001; Keller, 2003). This will ultimately lead to consumers holding familiar, and favorable, strong and unique brand associations in their memory, which creates positive consumer-based brand equity (Keller, 1993). Although brand equity can also refer to a variety of fi-nancial measurements, this research will take a customer’s perspective on brand equity, and so

throughout this paper the term brand equity refers to the consumer-based brand equity as previously defined by Keller (1993).

2.2. Associative network

As mentioned in the previous paragraph a product’s brand name represents images that have been formed based on information and experiences, and acts as a cue for consumers that influences the overall perception of the product (Swait et al., 1993). Those perceptions, preferences and choices are linked to a brand and form brand associations in the mind of the consumer (Aaker, 1990). Brand associations can vary broadly, from physical product attributes to perceptions regarding people, places, and occasions (Keller, 2003). Brand associations are crucial elements for brand equity, brand image and brand knowledge (Keller, 1993). To be able to influence these brand associations it is im-portant to look at the underlying foundation of what these associations are and how they are config-ured.

One of the most widely accepted conceptualizations of how these associations are stored in the consumer’s minds is that of a network of information (Anderson & Bower, 1973). This model views memory or knowledge as a network consisting of sets of nodes and links. Nodes are the infor-mation elements connected by links that vary in strength. Any node linked to the network can acti-vate already stored information or create new information. The stronger the link between nodes, the higher the probability that most of these nodes will be activated instead of just receiving information from a single node (Collins & Loftus, 1975; Raaijmakers & Shiffrin, 1981; Bourne, Dominowski, Lof-tus, & Healy, 1986). Links between two nodes implicate an association in the mind of the consumer (Krishnan, 1996).

Consistently with the associative network memory model, Aaker (1990) and Keller (1993) de-scribe how a brand can be seen as a network where a node is the brand itself and the other nodes are the associations. As an illustration, in the memory of the consumer the node Nike (brand) can be linked with the node of athletic shoes (product) and a node of durability (attribute) (Krishnan, 1996). Aaker (1991, p. 148) defines these brand associations as “everything linked in memory to a brand” and they represent what the brand means for a consumer. The set of associations will build up during time and although many of these associations may not be based on product attributes, Krishnan (1996) and Park and Srinivasan (1994) argue that it is however valuable to understand these to get a complete picture of how the brand is perceived by consumers. As an example, Park and Srinivasan (1994) describe how the masculine-related associations of Marlboro have nothing to do with actual product attributes, but may play a significant role in purchase preferences by consumers. As the asso-ciative network becomes richer when the number of associations increases, it is also becoming more complex (Krishnan, 1996). This increase creates more possible paths and thus makes it easier to ac-cess a particular brand node (Park & Srinivasan, 1994), but might also lower the brand memory be-cause of these associations interfering with it (Meyers-Levy, 1989).

2.3. Leveraging brand equity

In a continuous pursue for growth, companies are facing markets cluttered with competing brands, higher costs and flattening demand. For new products it is therefore difficult to establish a unique differentiated position, so the risks of launching new brands are high, with failure rates rang-ing from 80 to 90 percent (Leuthesser, Kohli, & Suri, 2002). In the mean time due to the same com-petitive environments, the pressure for firms to be more efficient in their marketing expenses increes (Keller, 1993). As a rincreesult there is a continuing princreessure for companiincrees to leverage their brand as-sets (Desai & Keller, 2002), which is argued by Tauber (1988) and Keller (1993) as the most valuable

asset for a company. Keller (2003) describes this as the process of brand equity leverage, and he de-scribes how the importance of the brand-leveraging process is increasing, which is the linking of a brand to another entity, such as a person, place, thing, or brand (Figure 1). This way marketers try to increase their brand equity by borrowing equity from others, which gives opportunities to marketers to build and leverage brand knowledge that otherwise could be difficult to achieve through marketing programs directly (Keller, 2003). The focus in this study is on the top part of the brand equity lever-aging with the leverlever-aging of other brands (accentuated in Figure 1).

The leveraging of brand equity can especially be of major importance when companies want to en-ter into new categories. Instead of creating new brands for new categories with all the associated high risks and costs, already existing brands can leverage their brand equity to enter these new cat-egories. Tauber (1988) quite accurately describes this as: “Brands have become the barrier to entry, but they are also the means to entry’’ (p. 27). One of the classical strategies for companies to intro-duce new products in a competitive environment are brand extensions, which uses the parent brand to introduce products in different categories out-side the parent brand category (Tauber, 1988). This strategy offers major advantages, one of them is that this brand extension capitalizes the brand equity of the existing brand, and therefore reduces the costs of entry and the risk of failure when the con-sumer already attaches attributes and benefits towards the brand that are desired in the new category (Tauber, 1988). A successful example of a brand extension strategy is e.g. Unilever’s Bertolli, which extended from Italian olive oil as its core product towards all kinds of Italian-related food products such as pasta sauces, spaghetti and now even butter spread and mayonnaise.

An alternative strategy to leverage brand equity for launching new products is the use of a co-branding strategy (in Figure 1 phrased as Alliances) (Keller, 1993). Co-co-branding involves the paring of two or more brands to form a separate and unique product (Park et al., 1996). Brand extension and co-branding strategies are both an attempt to strengthen the parent brand and extend customer value perceptions to a new product, therefore co-branding is sometimes referred to as a special case of brand extensions (Park et al., 1996). However by comparing the co-branding literature there are some interesting similarities and differences between the far more sophisticated research on brand

Figure 1 - Secondary sources of brand equi-ty (Keller, 2003)

extensions (Helmig et al., 2008), therefore co-branding will be discussed more thoroughly in the next paragraphs.

2.4. Co-branding

There is no universally accepted definition of co-branding and several terms are being used interchangeably such as brand alliances, composite branding, ingredient branding, multi-branding, joint or dual branding. In the broad sense, co-branding can be described as any deliberate pairing of two brands in a marketing context such as advertisements, products, product placements and distrib-ution outlets (Grossman, 1997). A more narrow definition is written by Park et al. (1996) who define co-branding as the pairing of two or more brands to form a separate and unique product. When a co-branding strategy is used to launch a new product, it usually signals to the customer a commit-ment of the brands for a long term relationship (Leuthesser et al., 2002). This notion of the ‘long-term’ time element was added in a more recent definition by Helmig et al. (2008, p. 360): “Co-brand-ing represents a long-term brand alliance strategy in which one product is branded and identified si-multaneously by two brands”. This research will adapt the more narrow definition, because despite the lack of universal agreement on the definition, Leuthesser et al. (2002) point out that it appears that there is a general agreement that co-branding involves the creation of a single product using two brands (Levin, Davis, & Levin, 1996; Shocker, 1995). This criterion is most often used to distinguish co-branding from other types of brand alliances and offers an alternative for brand extensions (Leuthesser et al., 2002), therefore this paper will use the definition of Helmig et al. (2008).

2.4.2. Co-Branding versus other Brand Alliances

By distinguishing co-branding as a form of a brand alliance strategy, a clear boundary be-tween co-branding and several other terms and forms used in previous research is drawn. Although other brand alliance strategies, such as joint sales promotions, advertisement alliances, dual branding, and product bundling, are sharing similar underlying motives, there are clear differences with co-branding. Joint sales promotion is an effort of two or more brands to pool their promotional re-sources to capitalize on joint opportunities for sales growth and profits (Varadarajan, 1985). A newly introduced brand can benefit by the association with a well-established brand through the enhance-ment of its image, which reduces the perceived risk for the consumer to try this new brand

(Varadarajan, 1985). A further step to capitalize on a joint opportunity for sales growth is product bundling, which is another form of a brand alliance where two or more products are packaged in a single offering (Guiltinan, 1987). In a bundle one brand can influence the overall evaluation of the other brands bundled in the same package, but bundling is based around the theory of additivity

as-sumption and anchoring (Tversky & Kahneman, 1974; Yadav, 1994) and lacks similarities with co-branding such as the transfer of associations. In the context of another brand alliance, dual brand-ing, Levin and Levin (2000) researched the transfer of affect from one brand to another. They found that when two brands share attributes but differ in the level of these attributes, consumers tend to compare and contrast the attribute levels of both brands. The effect of dual branding is to reduce or eliminate contrast effects by assimilation (Levin & Levin, 2000). Despite the similarity of the transfer of affect in this matter it is not as relevant in a co-branding setting. This is because with dual brand-ing the choice for the consumer is between two separate offerbrand-ings, while the consumer cannot choose between the two brands in a co-branded context since it is a single product.

In conclusion and as a clear distinction between co-branding and the other brand alliance strategies it can be summarized that although these brand alliance strategies share similarities with co-branding in regards of the main underlying motives such as the interdependent image improve-ments by partnering with another brand, and signaling benefits such as greater assurance about product quality and higher product evaluations (Helmig et al., 2008; Rao et al., 1999; Wernerfelt, 1988; Erdem & Swait, 1998) there is a clear fundamental difference: Co-branding is the only strategy in which a single product is branded and identified simultaneously by two brands in contrast to the other strategies where two separate products and brands are being combined in one package, loca-tion or promoloca-tion.

2.4.3. Benefits of Co-Branding

Co-branding has been around for quite some time and the academical research regarding the benefits of co-branding started in the mid-1990’s (e.g. Norris, 1992; Rao & Ruekert, 1997; Rao, 1997; Rao et al., 1999; Hillyer & Tikoo 1995). Despite its longer existence the usage of co-branding strategies has drastically increased in the last decade (Voss & Gammoh, 2004; Helmig et al., 2008). Previous research has indicated the benefits of co-branding for the participating brands forming the brand alliance. Park et al. (1996) and Washburn et al. (2000) both describe that a positive perceived attribute of one of the partnering brands can transfer to the co-branded product and influence con-sumer choice and preference. The combination of two brands also provides concon-sumers with greater assurance about the product quality compared with a mono-branded product, such as a brand exten-sion, and that will lead to higher product evaluations and premium prices (Rao, Qu, & Ruekert, 1999; McCarthy & Norris, 1999). Helmig et al. (2008) argue that adding a second brand can con-tribute a perception of additional value to the co-branded product that the primary brand cannot achieve on its own. Academic research on co-branding’s impact on consumer decision making is however still in its infancy, and the research on co-branding focuses mainly on the ability of two

brands to improve perceptions of co-branded products (Thompson & Stratton, 2012). Based on the previous mentioned literature it can be summarized that co-branding offers an attractive alternative to brand extensions, because co-branded products can acquire the salient attributes of both parent brands (Leuthesser et al., 2002).

2.4.4. Ingredient branding

One specific form of co-branding is ingredient branding, in which the key attributes of one brand are incorporated into a co-branded products as an ingredient (Desai & Keller, 2002). With the use of a branded ingredient an existing attribute of the host can be changed and will be reinforced by the presence of the partner brand which enhances the overall perception of the co-branded product (Desai & Keller, 2002; Uggla, 2004). There are two forms of ingredient partner brands, at first there are ingredient partner brands that cannot be bought separately and which are always linked to a host brand (Riezebos, Kist, & Kootstra, 2003). For instance Intel acts as a branded ingredient for many computer brands, such as e.g. Dell, HP, Acer, and the camera lenses of Carl Zeiss act as a branded ingredient in e.g. Sony cameras and Nokia smartphones, but both products are hardly sold separately. Yet in another form, individual brands can also be positioned as ingredient partner brands, for in-stance M&M’s can be bought in nearly every supermarket, but the brand M&M’s is also used as an ingredient partner brand in the McFlurry M&M’s, or Oreo that acts as an ingredient partner brand in a variant of Milka chocolate. These brands are positioned as ingredient brands, but can also be bought outside a co-branding context.

With ingredient branding the role division between the two brands in the alliance is clearly defined as a host and a partner brand. The alliance between a host and a partner brand can be based on an asymmetrical or a symmetrical collaboration. When one brand is generally more dominant than the other, Uggla (2004) speaks of an asymmetrical brand alliance, in contrast a symmetrical col-laboration refers to a more balanced relationship between the host and partner brand. The host brand is the primary brand associated with the secondary partner brand (Uggla, 2004). An important characteristic of the host brand is that it is the category driver, so this means that the associations of the partner brand are secondary to the identity and immediate territory of the host brand (Hillyer & Tikoo, 1995; Uggla, 2004). In short, Uggla (2004) summarizes that the host brand defines the prod-uct category and the customer base, whereas the partner brand modifies the host brand’s prodprod-uct category attributes. For example, Philips launched a range of kitchen appliances that were developed together with Italian design firm Alessi. In this case, the kitchen appliances of the host brand Philips are enhanced towards a more premium segment by the transfer of premium design associations due to the presence of the partner brand Alessi. The symbolic associations from the partner brand are

borrowed and are then used to reinforce the host brand. Both Alessi and Philips contributed to the alliance with its own associations and together they created a better attribute profile than Philips as a host brand or Alessi as a partner brand could have achieved by a direct brand extension.

A partner brand can reinforce the value proposition of a leader brand in functional and or symbolic ways. It can bring functional brand associations to a leader brand, such as the refined taste of Oreo cookies in a Milka chocolate bar. Or the partner brand can reinforce symbolic associations for a leader brand (Uggla, 2010), such as the Italian design associations of Alessi towards the kitchen appliances of Philips. In both the examples of Oreo as a branded ingredient in Milka chocolate, and the symbolic design associations of Alessi in Philips’s kitchen appliances, the strong partner brand is a way for the host brand to differentiate its own position and create a competitive advantage over other brands with a similar positioning (Uggla, 2004). In general an ingredient branding strategy is a strat-egy that can be used to differentiate the extension from the competition by signaling a strong message to consumers that the co-branded product offers the combined benefits of two quality brands in one (Desai & Keller, 2002).

3. Evaluations and the role of fit

3.1. Evaluations of co-branding

In the previous chapter it was discussed that the research on co-branding is mainly focused on the ability of two brands to improve perceptions of co-branded products (Thompson & Stratton, 2012), but another stream of research is focused on the exploring of the determents for favorable co-branding evaluations. One of the pioneers in the co-co-branding literature, Simonin and Ruth (1998) showed that consumers’ attitudes towards the co-branded product can positively influence their atti-tudes towards both the partner’s brand, and the other way around from the partner brands towards the co-branded product. Washburn et al. (2004) found a direct link between the positive effects of brand equity and co-branded products, where the perceived brand equity of the co-branded product will be improved by a high brand equity of the partner brands. Brand familiarity also plays a role with the effect of customers’ prior attitude decreasing when a less familiar partner brand is constitut-ed (Simonin & Ruth, 1998), and more familiar brands will contribute more to the evaluation of the branded product (Baumgarth, 2004). Voss and Tansuhaj (1999) however find evidence that co-branded products can increase evaluations of a previously unknown brand if the unknown brand partners with a well-known brand.

Especially the findings of Simonin and Ruth (1998) and Washburn et al. (2004) confirmed the underlying motives for firms to pursue a co-branding strategy as a brand equity leverage strategy. This is as mentioned due to the transfer of associations and images from the partner brands to the co-branded product. As discussed in the previous chapter firms are trying to influence the brand as-sociations stored in the memory of the consumer by marketing programs (Keller, 1993). This will ul-timately lead to consumers holding familiar and favorable, strong and unique brand associations in their memory, which creates positive consumer-based brand equity (Keller, 1993). However, the asso-ciations that consumers store in their memory can also be shared between the brand and the product category (Krishnan, 1996). For instance in the given example of the previous chapter, the node dura-bility can lead to Nike in particular, but just as easy can it be connected to the category of athletic shoes in general. If the brand shares associations with the category then it can become a member of that category (Krishnan, 1996). This last notion is part of the process of categorization and is an im-portant factor in the evaluation of an extension. In order to get a better understanding under what circumstances these associations and images can transfer, a closer look at the fundamental categoriza-tion theory is needed and will be provided in the next paragraphs.

3.2. Categorization theory

Several theories, such as semantic generalization (Mazanec & Schweiger, 1981), schema theo-ry (Thompson, 1988), cognitive consistency theotheo-ry (Pimentel & Reynolds, 2004), and categorization theory (Aaker & Keller, 1990) have been used to explain the transfer of associations and predict the customer’s evaluations of brand extensions. Categorization theory has in time emerged as the richest foundation for most of the extension research (Thompson & Stratton, 2012). Based on the catego-rization theory, Sujan (1985) suggests that a consumer would evaluate a new product by two possible strategies. The first being a piecemeal-based process, whereby the product is evaluated on an at-tribute-by-attribute basis, where each piece of information is assessed. The overall value of the prod-uct is determined by the integration of the (most important) attributes. The other approach Sujan (1985) describes is that of a category-based process. To simplify a complicated environment and to make understanding more efficient, customers tend to categorize by brand association or image given their knowledge and previous experience on perceived similarity and resemblances (Sheng & Pan, 2009). In the categorization-based approach, affects, attitudes, or judgments are cued by the catego-rization process rather than a rational evaluation of the (most important) attributes of the product (Cohen & Basu, 1987). The fundamental premise of categorization theory is that consumers’ percep-tions of the parent brand will impact the perceppercep-tions of the brand extensions (Thompson & Strutton, 2012). In this case the parent brand resembles a category and the brand extension belongs to that category cued by the same brand name (Boush & Loken, 1991). When consumers are exposed to this new product with the same brand name, the category of the parent brand gets activated. Once the category has been activated the cognition and affective judgments that are already associated with the parent brand will be used to evaluate the brand extension (Cohen & Basu, 1987). In co-branding the situation is similar, where the two partner brands both resemble a category and consumers may use these brand names as a cue to transfer their associations and images as a judgment to the new co-branded product (Thompson & Strutton, 2012).

The transfer of associations and images is however not an automatic process which is activat-ed as soon as a company leverages their brand equity with another entity. This transfer may happen specifically if consumers perceive a similarity or ‘fit’ between the brand and the other entity. The ar-rows on the brand equity leverage model of Keller (2003) (Figure 1) between the brand and the dif-ferent entities represent this ‘fit’. For instance in the case a brand wants to partner with a celebrity to leverage positive associations from this celebrity, e.g. Michael Jordan and Nike, scholars found that the product must ‘fit’ the image of the celebrity for a brand to leverage associations from an endorser, which is called the match-up theory (e.g. Baker & Churchill 1977; Joseph 1982; Kahle & Homer

1985). Another example of leveraging associations is that of a brand sponsoring an event, e.g. Adidas sponsoring the Olympics. Scholars found that in similarity with the match-up theory, there should be a so-called sponsor-brand fit (e.g. Gwinner, 1997; Gwinner & Eaton, 1999), which is the ‘fit’ between a brand and a sponsored event which is needed to transfer event-related images to the sponsor brand. A brand-cause fit (e.g., Drumwright 1996; Strahilevitz & Myers, 1998) describes the fit between a brand and the cause and this fit influences the responses from consumers towards the campaign. In the context of brand extensions (Aaker & Keller, 1990) the transfer of associations and images from the brand towards the new product is similar and this transfer may happen specifically if consumers perceive a similarity or ‘fit’ between the new product and/or the existing product classes. The fit then acts as a catalyst for the transfer of associations from the existing category towards the new product (Cohen, 1982; Boush & Loken, 1991; Aaker & Keller, 1990). When the fit is considered as a poor fit there might be a negative spillover effect on how the co-branded product is evaluated (Keller, 1993).

3.3. Perceptual fit

Thompson and Strutton (2012) point out that in the brand extension and co-branding con-text various definitions have been used to define this ‘perceptual fit’ (Tauber, 1988), such as ‘extension typicality’ (Boush & Loken, 1991), ‘fit’ (Keller & Aaker, 1992), ‘logical consistency’ (Thompson, 1988), and ‘category-to-category relatedness’ (Farquhar, Herr, & Fazio, 1990). But they all have in common that in a general sense, perceptual fit reflects consumers’ perceptions to the degree to which the new product is consistent, or is a logical outgrowth, with their parent brand(s) (Thompson & Strutton, 2012). Many scholars acknowledge that the perceived ‘fit’ is a key succes factor for the eval-uation of both brand extensions and co-branded products (Tauber, 1988; Dwividi et al., 2009; Gammoh et al., 2006; Aaker & Keller, 1990; Broniarczyk & Alba, 1994; Park et al., 1991). Despite its many apparent similarities the fit bases of a brand extension and co-branding are diverged in impor-tant aspects (Uggla, 2004). Therefore it is conceptually imporimpor-tant to distinguish brand extension fit from the fit in a co-branding context (Simonin & Ruth, 1998).

With a brand extension the original and the extension category classes have to be perceived closely related in order for consumers to perceive a fit (Aaker & Keller, 1990). Aaker and Keller de-fine this as product category fit, or widely used by other scholars as product fit. Park et al., (1991) also argue that consumers do not just evaluate brand extensions on product category similarity, but that the brand concept also plays a crucial role. The evaluation of a brand extension is depending on how a brand concept can accommodate the extension as seen as consistent in the consumers’ mind (Park et al., 1991). The authors define this as brand concept consistency, also referred to as brand fit.

3.3.1. Conceptualization differences in fit

In the context of co-branding the terms product fit and brand fit are also being used, but de-spite the use of the same terms, the conceptualization and operationalization of these definitions are however different. Figure 2 illustrates the conceptualization and operationalization of product- and brand fit in a branding setting. Product fit refers to the fit between the partner brands and the co-branded product on a concrete level of the product related similarities (Park et al., 1996). In order to

maximize the transfer of relevant associations from the individual brands to the co-branded product, the two brands should have a high degree of product fit (Aaker & Keller, 1990; Park et al., 1991). With a co-branded product both of the partners’ brand images are involved and both the brands’ associations play a role in the evaluation of the co-branded product (Varadarajan, 1986; Simonin & Ruth, 1998; Broniarczyk & Alba, 1994). In regard to these two brands, Simonin & Ruth (1998) find that a co-branded product will be evaluated more positively if there is an overall perception of fit between the two brands because it helps con-sumers to more easily combine the brand attitudes. The term brand fit therefore describes the perceived fit between the two brands (brand fit will be discussed more thoroughly in Chapter 4). 


Not only the conceptualizations of the terms are different between a brand extension and a co-branded product, how the consumer has to manage and deal with all these types of fit is also dif-ferent. In the setting of co-branding the role and perception of fit is certainly becoming more com-plex than with a brand extension. Consumers not only have to deal with an additional source of product fit, the fit between both the brands and the co-branded product separately, but also the brand fit between the two brands forming the brand alliance. The evaluation of the fit between the two brands is completely unique to the context of co-branding, therefore it is not surprising that this concept has received attention of many scholars (e.g. Park et al., 1996; Simonin & Ruth, 1998; Baumgarth, 2004; Besharat, 2010). Simonin and Ruth’s (1998) study showed how the evaluation of a new co-branded product is dependent on this fit between both brands and they find that it is the most important success factor. This finding is later confirmed by Baumgarth (2010), who concluded that the co-branded product was particularly influenced by the brand fit. The fit between brands can be perceived in many ways. According to Keller (2003) the potential brand fit can be as broad as any association held in the customers’ mind about the parent brand. Murphy and Medin’s (1985) even

B

1

B

2

P

1

Illustration of product fit and brand fit in co-branding 
 P1: co-branded product


claim that consumers will form their own theories of brand fit while evaluating new products. Since there is no clear conceptualization of brand fit, researchers believe it is essential to determine the causes of fit between two brands in a brand alliance (Helmig et al., 2008). In order to get a better understanding of brand fit, a more thorough explanation of the theory of brand fit will be provided in the next chapter.

4. Bases of brand fit

Park et al. (1996) argued that one of the most important aspects of a co-branded product is that customers understand the relationship between the characteristics of the two brands and that the combination makes sense. As mentioned in the previous chapter the overall perception of fit between the two brands is important for a favorable evaluation (Simonin & Ruth, 1998). Despite this existence of consensus of brand fit being a key success factor, there is not a single conceptualization of this construct among researchers. For instance Simonin and Ruth (1998) use the terms product fit and brand fit in their article. However they conceptualize the product fit as the compatibility between the two product categories of the two brands, which is as earlier conceptualized in this article a form of brand fit. Their conceptualization is in line with the product feature similarity as defined by Park et al. (1991). Simonin and Ruth (1998) use the operationalization of brand concept consistency to de-fine the brand fit.

An important stream of the literature on determining the perceived brand fit is focussing on the level of abstraction in the brand associations. In the previous chapter it was discussed that many researchers take the view that brand fit is about the physical attribute similarity (Aaker & Keller, 1990; Boush & Loken, 1991; Park et al., 1996). In other words the extent to which consumers per-ceive the compatibility between the product categories of both the brands at the functional product level (Simonin & Ruth, 1998). Although consumers might perceive a high fit based on brand specific physical attributes, at a more abstract level these brand associations may be inconsistent (Keller, 1993; Park et al., 1991; Park et al., 1996). Other scholars therefore conceptualize brand fit also beyond the physical attributes and are also considering the brand concept fit (Park et al., 1991; Broniarczyk & Alba, 1994; Dwivedi et al., 2009). Park et al. (1991) define these two dimensions of fit based on their abstraction level as product feature similarity (concrete) and brand concept consistency (abstract).

4.1. Functional and symbolic associations

The theoretical basis for the level of abstraction lies in the associative network (discussed in Chapter 2.2). Brand associations can vary broadly from functional to symbolic consumer benefits (Park et al., 1986). The functional benefits usually relate to the product-related attributes and are of-ten related to fairly basic motivations such as solving a current problem, prevent a poof-tential problem, resolve conflict (Park et al., 1986; Keller, 1993). Symbolic benefits usually comprehend non-product-related attributes and relate to underlying needs for social approval or personal expression or

self-en-hancement (Park et al., 1986; Keller, 1993). An example of these symbolic associations could be the associations of prestige and exclusivity linked to a brand. Most brands offer a mixture of symbolic and functional benefits, and it therefore is possible for brands to develop a brand image with both concepts. However Park et al. (1986) state that in practice this is very difficult to manage and there-fore they argue that brands should select one brand concept from the beginning and maintain this concept over time consistently. This last notion is important because although brands can have both concepts in their brand image, in general it means that brands have a focus and are either symbolic or functional in the minds of the consumer. Bhat and Reddy (1998) summarized this as that func-tional brands are satisfying the immediate and practical needs, and the symbolic brands satisfy sym-bolic needs (e.g. personal expression or self-enhancement) and their practical use is only incidental. An example that is used by several scholars is the comparison between brands in the category of watches, e.g. Seiko and Rolex (Park et al., 1991) and Casio and Movado (Bhat & Reddy, 1998). For instance the brand Casio would be seen as a functional brand because it is primarily used to tell the correct time. Rolex however would be considered a symbolic brand because its primary use lies in the symbolic aspects such as its appeal of ‘luxury’ and ‘status’, and the ability to tell the time correctly is only an incidental reason for its use.


4.2. Product Feature Similarity

In order for consumers to categorize two brands as belonging to the same category, there needs to be a relationship among the characteristics of the two brands (Simonin & Ruth, 1998). Many scholars, (e.g. Aaker & Keller, 1990; Park et al., 1996) take the view of brand fit being about the physical attribute similarity. On the concrete level of product feature similarity the fit bases of brand extensions and co-branding are very similar. In a co-branding setting this dimension can also be used to determine the perceived fit between the two brands. In this case the product feature simi-larity will not be about the compatibility between the new product and the parent brand’s product category but rather about the compatibility between the product categories of both the brands in the alliance. For instance in the example of the McFlurry M&M’s, the concrete physical attribute of chocolate and crunchy nuts by M&M’s is perceived as fitting with the creamy ice-cream of the

McFlurry. Or the audio speakers by Dolby that fit nicely with the laptop product category of Lenovo.

4.3. Brand Concept Consistency

Although consumers might perceive a high fit based on brand specific physical attributes, at a more abstract level these brand associations may be inconsistent (Keller, 1993; Park et al., 1991; Park

et al., 1996). For instance in the earlier example of the watch brands Casio and Rolex, it is illustrated how the brand names of Casio and Rolex both belong to the watch product category and share many physical attribute associations. However they are inconsistent on a more abstract level where only Rolex has been associated with the concepts of luxury and high status. This distinction is impor-tant to recognize and therefore other scholars conceptualize fit also beyond the physical attributes and are also considering the brand concept fit (Park et al., 1991; Broniarczyk & Alba, 1994; Dwivedi et al., 2009). In the example of Rolex consumers might not perceive a high fit with Casio, which is in fact a very similar brand regarding the product feature similarity, but consumers may more likely per-ceive a high fit with another luxury brand such as Jaguar. Although on the concrete abstraction level of product features these two brands do not share many physical attribute associations, but as already described the two brand images of the brands can provide context for why the two companies coop-erate together (Barsalou, 1982; Medin, Goldstone, & Gentner 1993). This partner context can then influence the perceived fit since consumers can categorize the abstract brand images of the two brands into the same cognitive category (Barsalou, 1985; Ratneshwar et al., 1996). Therefore it can be argued that when consumers see a relationship between the images of the two brands it will lead to a higher overall fit perception compared to when there is no perceived relationship (mis-fit) at the abstract level of the brand images.

4.4. Integration

Product category fit and brand concept consistency are often discussed as two independent types of brand fit. But Park et al. (1991) have however argued that the degree of perceived fit is a function of both product feature similarity and brand concept consistency perceptions. Therefore they found that consumers react more positively when there is a fit on both dimensions rather than a fit perception on a single one. In additional findings of their influential paper, Park et al. (1991) found an interaction effect between the product feature similarity and brand concept consistency, although this is in a classical brand extension context. For their examples of a functional brand concept and a symbolic brand concept they found that for both examples the brand concept consistency yielded higher evaluations with a high product feature similarity than a low product feature similarity. And the other way around the same effect can be witnessed where the product feature similarity yielded higher evaluations with a high brand concept consistency than with a low brand concept consistency. Park et al. (1991) also found that although the brand concept consistency had a strong effect on both functional and symbolic brand concepts, the brand concept consistency appeared to have a relatively greater effect on the symbolic brand than the functional brand. It is argued that for symbolic brands

the product feature similarity is of less importance for the consumer, this might be because with sym-bolic brands tangible attributes are not present in the evaluation of the consumer (Lanseng & Olsen, 2010). These attributes are rather recoded into cultural meaning and social symbols (McCracken, 1986; Solomon, 1983). As already described with categorization theory, the importance of two brands to be perceived as belonging to the same category is important for the transfer of favorable associations and judgments from the individual partner brands towards the extension product (Cohen & Basu, 1987). If there is both a product feature similarity and high brand concept consistency be-tween the both brands then consumers will perceive a high (brand) fit (Jap, 1993).

5. Hypotheses development

5.1. Introduction

The previous chapters described the phenomenon and laid the theoretical foundation that is required to further investigate the effects of brand concept consistency on brand fit and subsequently the evaluation of the co-branded product. In Chapter 2 the concept and benefits of a co-branding strategy have been described, which is mainly based on the leveraging of associations from a second brand towards the co-branded product in an attempt to yield more favorable evaluations (Keller, 2003). This is done by the transfer of relevant associations from the individual brands towards the co-branded product (Keller, 2003). The transfer of associations and images is however not an automatic process which is activated as soon as a company leverages their brand equity with another entity. This transfer may happen specifically if consumers perceive a similarity or fit between the brands and the other entity (Aaker & Keller, 1990; Park et al., 1991). Chapter 3 explained that this perceptual fit is a key success factor for a favorable evaluation, which is the perceived fit between the product categories of the brands and the product separately (product fit), and the perceptual fit between the two brands that are forming the alliance (brand fit) (Varadarajan, 1986; Simonin & Ruth, 1998; Broniarczyk & Alba, 1994). The latter has been described in Chapter 4, whereas the brand fit basis can vary on ab-straction level, with product feature similarity (concrete) and brand concept consistency (abstract) as the two dimensions of brand fit (Park et al., 1991). The degree of perceived brand fit is a function of both product feature similarity and brand concept consistency perceptions (Park et al., 1991). There-fore they found that consumers react more positively when there is a fit on both dimensions rather than a fit perception on a single one. Out of the direct scope of their research, Park et al. (1991) in-terestingly found that although for a brand expansion the brand concept consistency had a strong effect on both functional and symbolic brand concepts, the brand concept consistency appeared to have a relatively greater effect on the symbolic brand than the functional brand. In a co-branding setting not only is the perceived fit more complex, also the conceptualization of brand fit is complete-ly different from for brand extensions.

Building on these insights an interesting research question has emerged, which is whether in a co-branding context, (a) high versus low brand concept consistency will lead to a higher perceived brand fit, and (b) if so whether the strength of this effect is contingent on the host brand concept, and (c) will this increase in perceived brand fit subsequently lead to a more favorable evaluation?

5.2. Brand concept consistency

Park et al. (1991) found in their influential research about brand extensions that consumers reacted more favorable to brand extensions when the extension belonged to the same brand concept as the parent brand (high brand concept consistency) than when the product belonged to a different brand concept (low brand concept consistency). This argument was proven to hold ground for both functional as symbolic brands (Park et al., 1991). In a setting of co-branding consumers are also more likely to easily perceive fit based on the same brand concept, than when the two brand concepts are different, therefore the first hypothesis is:


H1 Perceived brand fit will be higher when brand concept consistency is high than when brand concept consistency is

low.


5.3. Host brand concept contingency

Park et al. (1991) argued that the attributes of brands can vary from concrete levels (e.g. prod-uct performance, solving problems etc.) to more abstract levels (e.g. luxury, high status etc.). A func-tional brand concept is primarily understood in terms of their product-related attributes or aspects related to product performance. A symbolic brand concept is argued to be primarily understood in terms of images linked to non-product related attributes and consumers’ expression of who they are or want to be (Keller, 1998; Park et al., 1986).

Two brands that might not be perceived as two similar brands regarding their product feature similarity, may still be seen as members of the same cognitive category when these two brands share associations on a higher abstraction level, such as luxury and status (Barsalou, 1985; Lanseng & Olsen, 2012). Park et al. (1991) describe this as a superordinate concept category that links the sym-bolic brands in the consumer’s mind, whereas functional brands are primarily stored under their product category. Context influences the fit bases of consumers’ judgments (Medin et al., 1993), and different host brand concepts may therefore result in different contexts on how consumers make the judgment of the fit between two brands.

Since a functional brand is primarily understood in product related attributes, it is likely that in an alliance with a functional host the categorization process of the consumer is driven by finding a clear link with the product category of the partner brand. For both a functional and a symbolic part-ner this link is relatively easily made and only requires matching the two product categories on com-plementary attributes (Lanseng & Olsen, 2012). We argue that therefore in the context of fit bases,

product feature similarity is more salient than brand concept consistency. Thus the incongruity be-tween a functional host and a symbolic partner of the associations on a higher abstraction level would then have a smaller impact on the overall perceived brand fit as those abstract associations are less salient in the categorization process.

In contrast, a symbolic brand is primarily understood in non-product related attributes and consumers’ expression. The relationship between the product related attributes and the brands’ bene-fits may even be obscured (Lanseng & Olsen, 2012). Therefore we argue that non-product related attributes (abstract) are more salient in the categorization process of a symbolic host brand compared to the categorization process for a functional host brand. The categorization process for a symbolic host brand is thus likely to be more intuitively driven by finding a link with the superordinate concept category of the partner brand. There is an incongruity between the symbolic host and the functional partner on the associations of the higher abstraction level and this incongruity will then have a large influence on the overall perceived brand fit. The product categories of both brands could still be per-ceived as complementing each other, but we argue that consumers are likely to perceive this alliance as a misfit, since the non-product related abstract associations are made salient in the categorization process. Therefore the following hypothesis has been developed:

H2 The effect of brand concept consistency on brand fit (hypothesis 1) is more pronounced when the host brand is

sym-bolic than when the host brand is functional.

5.4. Mediation of brand fit

In virtually all brand extension studies there were simple direct effects found of perceptual fit on the extension evaluation (e.g. Aaker & Keller, 1990; Boush & Loken, 1991; Thompson and Strat-ton 2012). Simonin & Ruth (1998) found that the perceived fit between the two brands had a signifi-cant positive effect on alliance evaluation, and research by Hadjicharalambous’s (2001) confirmed this direct effect in their study. These findings are consistent with the brand extension literature as a higher perceived fit leads to a higher evaluation. So as a conclusion the last hypothesis is stated as fol-lows:

Figure 3 illustrates the conceptual model with the underlying relationships between the vari-ables and the hypotheses. 


Brand Concept Consistency Brand Fit + H1 + H2 Alliance 
 Evaluation H3 Host Brand 
 Concept

Figure 3 - Conceptual model

+

— Study I —

6. Methodology (Study I)

6.1. Research design

The objective of this research is to obtain a more thorough understanding of the complexity of brand fit in a co-branding setting, and whether brand concept consistency and its contingency of the role division of the two brands in the alliance influences the evaluation of a co-branded product. In order to be able to examine the effects of brand concept (in)consistency on alliance evaluation and its contingency of the host brand concept, an experimental design was developed. This study exam-ined consumers’ reactions to two different co-branded products using the same brand set consisting of four brand names. Different scenarios were developed in which the same brand set was used to manipulate the role division between the host and the partner brand concept. These co-branded products were manipulated under four different conditions that varied on the host brand concept and partner brand concept. Four brands from two different categories and with two different brand con-cepts were used, which resulted in a total of eight different brand alliances. Therefor the experimen-tal design consisted of a 2 (host brand concept: functional or symbolic) x 2 (brand concept consisten-cy: low or high) x 2 (product replicate: product A, product B) between subjects factorial experiment.

The research was conducted in two phases. At first the stimuli were developed by both a qual-itative and a quantqual-itative pretest. The qualqual-itative pretest was to find appropriate sets of co-branded products that facilitated the interchangeability of the host and the partner brand using the same set of brands. This was needed to exclude the possibility that by using different brands the manipulations would be driven by the change in brands rather than the change in role division. Also a first selection of brands was made based on presumptions from the qualitative pretest. Based on the results of the

Product A Product B

Functional

Partner Symbolic
Partner Functional Partner Symbolic
Partner

Brand 3,

Category B Category BBrand 4, Category ABrand 1, Category ABrand 2,

Functional Host


Brand 1, Category A (High BCC)F + F
 (Low BCC)F + S


Functional Host


Brand 3, Category B (High BCC)F + F
 (Low BCC)F + S
 Symbolic Host


Brand 2, Category A (Low BCC)S + F
 (High BCC)S + S


Symbolic Host