The impact of imprisonment by strong product associations on the extension potential of brands

(1)

Master Thesis

Prison Brands

The impact of imprisonment by strong product associations on the extension

potential of brands

Student: Annick Werndly Studentnumber: 5965772

Date of submission: 31-08-2015 (final version) Msc. In Business Administration – Marketing Track Faculty of Economics & Business (FEB)

1st_{Supervisor: Jorge Labadie}

(2)

Statement of Originality

This document is written by Student Annick Werndly 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.

(3)

Abstract

The value of brands is widely recognized by businesses and academics. However, strong, favorable and unique associations can also be a burden to a brand. If brands are strongly associated with the product category, this can limit the brand in extending to other product categories. The drivers of a successful brand extension are extensively discussed in prior research. However little attention has been paid to the role of the transferability of the brand associations in this evaluation process. In this thesis three experiments are performed, exploring the importance of the brand associations in extension evaluation by examining the concept prison brands. Prison brands have strong product associations and are therefore more difficult to extend to another product category. The first experiment is a general experiment to test existing literature and reveals that product class similarity has a direct effect on the extension evaluation. A higher product class similarity results in a higher extension evaluation. In a second experiment, it is shown that the effect of product class similarity on extension evaluation is different for prison brands versus not prison brands. High Fit Line Extensions are evaluated similar for prison brands versus not prison brands. High Fit Category Extensions are evaluated better for not prison brands than for prison brands. Low Fit Category Extensions are evaluated similar for prison brands versus not prison brands. In a third experiment a prison brand is primed with an abstract association. Again these primed prison brands are compared to prison brands. This experiment reveals that priming enhances the extension evaluation. Both High Fit Line Extensions and High Fit Category Extensions are evaluated better for primed prison brands than for prison brands. Low Fit Category Extensions are evaluated similarly for both prison and primed prison brands.

(4)

1. Introduction

1.1 Downside to a strong brand

The value of brands nowadays is widely recognized by businesses and academics. The question is: is there also a downside to a strong brand? Can strong favorable and unique associations also limit the long term success of a brand? When looking at the superbowl 2015, all kinds of brands fight with their advertisements to stand out and be the strongest (Brandchannel, 2015). A strong brand, which is referred to as a brand with high brand equity, ads value to its products or

services, compared to similar products and services, only by displaying the brand. Brand equity is defined as “the differential effect of brand knowledge on consumer response to the marketing of a brand” (Keller, 1993, p. 17). Brands try to generate a high brand equity by trying to evoke strong, favorable and unique associations in the consumers mind (Keller, 1993). However, at the same time, strong, favorable and unique associations can be a burden to a brand. For

example, if brand associations are related to a certain product category, these brand associations can limit the brand in extending to other product categories. Haveman & Labadie (2003) argue that such brands do not allow other brand associations and are locked in their existing

associations. Therefore this type of brands is called: prison brands. Brands can be imprisoned by any kind of associations, such as user imagery, price associations and product associations. However in this research the focus is on brands that are imprisoned by strong product associations. A strong product association can prevent the development of new associations, because consumers do not allow new associations anymore. This process can be explained by the basis of a psychological principle called “associative inhibition” which can be explained as: “If a is already connected with b, then it is difficult to connect it with k; b gets in the way (Kline, 1921). Some brands (a) have such strong product associations (b) that they cannot be connected with another product (k) anymore. Good examples of brands with strong product associations are prototypical brands. Prototypical brands are brands that are typical of a product class, such as Kleenex, Luxaflex, Brinta and band aid. Because these brands are typical of the product class, the product associations are so strong that the brands are considered to be less likely

(7)

extendable. Kleenex for example, tried to extend to the disposable diaper category but this extension failed. The only way to make a successful extension to this product category, was by introducing a new brand called ‘Huggies’. Another example of a prison brand is Brinta. Brinta is so strongly associated with porridge that it could not be connected to anything else. Therefore the parent brand Honig decided to bring a new breakfast drink on the market under a new sub brand, called Wake-up (Franzen & Moriarty, 2009). One big disadvantage of subbranding, as Honig did, is that the subbrand does not profit from the brand image of the parent brand as much as in a brand extension.

1.2 The importance of performing brand extensions

In general, prison brands do not have to be problematic (Haveman & Labadie, 2003). A brand can have strong product associations and benefit from that for a long time. Rindfleisch and Inman (1998) describe that brands that are strongly linked to a product category are more preferred and of higher value to consumers. However, because of the difficult extendibility, prison brands cannot make optimal use of brand extensions. Brand extensions are important for firms for several reasons. Firstly, according to Milberg, Park & Carthy (1997) brand extensions can strengthen existing brand beliefs and attitudes which feeds into a higher brand equity. Secondly brand extensions can decrease the costs of gaining distribution, and enhance the efficiency of the marketing of the brand, especially compared to setting up a new brand (Aaker & Keller, 1990). Thirdly, brand extensions offer companies a strategic opportunity to evaluate and redefine the nature and direction of their business (Tauber, 1981). Furthermore, brand

extensions can also be important for risk spreading. If the brand is solely depending on one product and the environment changes in a way in which the product needs adaptation or has become redundant, not only the product, but also the whole brand is damaged.

1.3 Problem definition

(8)

Brand extensions are a popular way of entering a new product category because the extensions can make use of an established brand name. This makes the extension less costly than

introducing a new brand. However some practical examples of Kleenex and Brinta, as described above, show that it is not easy to perform a successful brand extension. According to Völckner & Sattler (2006), 80% of the brand extensions in FMCG fail. There has been extensive research in the field of examining the formation of attitude towards line and category extensions (Aaker & Keller, 1991; Boush & Loken, 1991; Broniarczyk & Alba, 1994; Park, Milberg & Lawson, 1991; Sood & Keller, 2012). Most studies concern research to the fit and quality perceptions of the brand and the brand extensions (Aaker & Keller, 1991; Boush & Loken, 1991; Broniarczyk & Alba, 1994; Park, Milberg & Lawson, 1991). For prison brands, brand extensions can be difficult because of the strong product associations, which often do not match a new product category. Therefore this thesis explores the role of those strong product associations on the evaluation of different extensions. Hence, the following research question is proposed:

What is the impact of imprisonment by strong product associations on the extension potential of brands?

1.3.2 Subquestions

To have a better understanding of how the strong product associations of prison brands make the extension to another product category difficult, firstly key characteristics of prison brands should be determined. Besides, several types of extensions with different levels of fit are investigated. Next, the difficulties for prison brands in those extension types are researched. Lastly, some studies indicate that it is more easy to transfer an abstract association to a broader set of product classes (Aaker & Keller, 1990; Park et al, 1991). Therefore it would be interesting to understand whether prison brands are able to focus on an existing more abstract association which can function as an umbrella for all the existing associations. This results in the following list of sub questions:

(9)

-What are the key characteristics of prison brands, particularly those with strong product associations?

-What is the role of the similarity in product class of extensions done by prison brands? - What are ways for prison brands to move away from imprisonment by strong product associations?

1.3.3 Delimitations

The goal of this study is to determine how extensions of prison brands are evaluated compared to evaluations of not prison brands. In previous studies already several drivers of extension evaluation are investigated and confirmed. One important driver that is identified, is the quality of the extended brand. In this research there is no focus on the quality of the original brand, but this is as much as possible controlled for. In addition, it is not the mean of this study to

investigate all possible drivers but a clear focus is on the importance of brand associations and in particular strong product associations. The other types of prison brands as discussed by Haveman & Labadie (2003) are therefore not included in this research. Furthermore the research to priming abstract associations takes a more practical approach. No research to priming abstract associations has been done so far. Therefore the purpose of this experiment is exploratory and no generalizable results should be expected.

1.4 Contribution

1.4.1 Theoretical contributions

Many studies examined the drivers for the success of a brand extension (Aaker & Keller, 1991; Boush & Loken, 1991; Sood & Keller, 2012). In general two important drivers for brand extension evaluation are identified: the quality of the brand and the product class similarity between the brand and the extension. Broniarczyk & Alba (1994) and Park, Milberg & Lawson (1991) also investigate the impact of brand-specific associations on extension evaluation. This research will further elaborate on the role of brand-specific associations, by comparing the

(10)

extension potential of two brands: prison brands and not prison brands. In the literature the concept prison brands is explained, however there is no information on how extensions are evaluated for prison brands. Moreover this research investigates a way in which prison brands can move away from the strong product associations and perform a successful brand extension. By priming an abstract association this research builds on Aaker & Keller (1990) who suggest that abstract associations are more easily transferable to a new product category.

1.4.2. Managerial contributions

Managers often use brand extensions to enter a new product category because the extension can profit from the established brand name. Although brand extensions are a cost-efficient way to bring a new product or service on the market, there are also many examples of extensions that did not succeed. For managers it is very important to know how consumers evaluate brand extensions and which drivers play a role. This research examines the importance of brand associations and the transferability of those associations to another product category. Besides this study also looks for ways in which prison brands can move away from the strong product associations. This can provide a practical tool for managers to prevent brand extension failures.

1.5 Structure

This thesis consists of 9 chapters in which the process of extension evaluation for prison and not-prison brands is extensively described and tested. In chapter 2 3 & 4 existing literature is reviewed which gives insight on brand associations, brand extensions and the transferability of brand associations. In chapter 5 the hypotheses for two experiments are described and

explained. In chapter 6 the research methodology of a qualitative analysis and two experiments is presented. In chapter 7 the results of the qualitative analysis provides insight in how brand associations differ between prison and not-prison brands. Besides the results of the quantitative analysis test the importance of the different established drivers for the extension evaluation. In chapter 8 the findings are discussed and linked back to theory. Moreover theoretical and

(11)

managerial implications are discussed. In chapter 9 a summary and conclusion of the research is provided. Also several limitations and opportunities for further research are discussed.

(12)

2. Brand associations

According to Aaker (1991), brand associations are the brand’s assets and liabilities that include anything linked in memory to a brand . Keller (1993) defines brand associations as

informational nodes, linked to the brand node in memory, which form the meaning of the brand for consumers. Each brand has a set of associations that vary in strength, uniqueness and favorability (Keller, 1993). These brand associations are created in the mind of consumers and can vary from product attributes (Heineken and beer) to perceptions of people (Nike and Tiger Woods), places (Toyota and Japan) occasions (Pepsi and a birthday party) and values (Bodyshop and sustainability) (Henderson et al., 1997). Brand associations play an important role in

building a strong brand since they are essential components of brand image, brand equity and brand knowledge (Farquhar & Herr, 1993; Keller, 1993, Schmitt et al., 1993). Brand associations are important for both marketers and consumers. For marketers, brand associations help to differentiate the brand from other brands and create brand equity (Low & Lamb, 2000). A higher brand equity can lead to lower costs, more revenues and larger profits (Keller, 2003). Furthermore brand associations can provide a basis for a brand from which new products and extensions can benefit (Henderson et al., 1997). For consumers, brand associations help to organize and retrieve the information in memory to make fast purchase decisions (Aaker, 1991). To understand how brand equity can be enhanced, firstly the role of brand associations in

building brand equity is discussed in chapter 2.1. Secondly, in chapter 2.2 a more cognitive approach explains in more detail how brand associations are linked and organized in memory by discussing association networks.

2.1 The role of brand associations in building brand equity

Keller (2001) explains the overall evaluation of the brand in terms of a customer-based brand equity model. This model illustrates the process that is concerned with the building of strong brands. Therefore it is interesting to explore the dimensions of this customer-based brand equity model and explore the role of brand associations in this process. The overall goal of this

(13)

model is to identify the building blocks of high customer-based brand equity, which is defined as: “the differential effect of brand knowledge on consumer response to the marketing of the brand” (Keller, 1993, p.8). This can be explained as follows: If consumers respond more positive to the product, price, promotion or distribution of a brand than those of a fictitious or unnamed product, although the element is the same, then the brand has a positive customer-based brand equity (Keller, 1993). The customer-based brand equity model, developed by Keller (2001), shows the building blocks for creating high brand equity. The customer-based brand equity model, as portrayed in figure 1, consists of four levels that have to be developed in a fixed order, from the bottom to the top. The specific levels will be discussed in moe detail in the chapters below.

Fig. 1 Keller (2001) Customer-based brand equity pyramid

2.1.1 Brand Identity

The first and lowest level of building brand equity is concerned with creating brand salience. This means that consumers need to be aware of the brand and the products that the brand sells. Brand awareness refers to the ability of consumers to identify the brand, both by recognizing the brand when displayed but also by the ability to recall the brand when the product category is given (Rossiter & Percy, 1987). Therefore a link between the brand and the product class needs to be present in the mind of the consumer. This link is formed by the retrieval of product-related brand associations, such as product properties and product use (Franzen & Moriarty, 2009 refer to Timmerman, 2007).

(14)

2.1.2 Brand Meaning

The second level is concerned with creating both brand imagery and brand performance. These dimensions relate to how consumers think about a brand in terms of values and meaning and how the performance of a brand’s products is evaluated. Brand performance refers to the way in which a product meets the functional needs of a consumer. Brand imagery refers to what

consumers think of the brand and this is the part where brand associations play an important role (Keller, 1993). Brand image is determined by the type of associations that belong to the brand and the favorability, strength and uniqueness of those associations. These constructs are discussed below in paragraph 2.1.2.1 and 2.1.2.2

2.1.2.1 Types of associations

There are several types of brand associations that can be distinguished by their level of

abstraction (Alba & Hutchinson, 1987; Russo & Johnson, 1980). According to Franzen & Moriarty (2009) brand associations can be classified using a means-end chain. The lowest meaning level consists of the most concrete brand associations, such as name, logo, color & packaging. The highest meaning level is concerned with central value associations, such as striving towards harmony. Keller (2003) classifies association types in three categories: attributes, benefits & attitudes. The category attributes consists of product-related and non-product related attributes of a brand (Keller 2003 & franzen & moriarty, 2009). The category benefits can be described as the personal value consumers attach to the product. The category attitudes is defined as the overall evaluation of a brand. (Keller, 1993).

2.1.2.2 Favorability, strength and uniqueness of associations

The different types of associations as discussed above, can vary according to how favorable, strong and unique they are. Firstly, brand associations need to be favorable. If an association of a brand is not favorable at all, or not favorable in the specific product category, consumers are not likely to evaluate the brand positively. Favorability of associations is created when the attributes and benefits of a brand satisfy the need and wants of the consumer. With this an overall positive

(15)

attitude towards the brand is formed. In this process, it is important that the product offering of a brand is in line with the consumers’ needs. Therefore brands have to keep up with the quality of products of other brands. This is described by Keller (2001) as points of parity. These points of parity consist of all the brand associations that are shared by brands. Points of parity are important in the consumers decision making process, because these provide certain checks for consumers and make consumers consider the product (Keller, 2001). In buying shampoos for example, consumers look for a shampoo with a smell. This smell is a point of parity because almost all shampoos have a smell. However, having a smell is still necessary for a shampoo brand to even be considered in the buying process.

Secondly, the strength of an association is important to enhance brand image. The strength of an association depends on how the information enters the memory of the consumer and how it is stored. Even though an association is really favorable for a brand, if the strength is low, most consumers will not retrieve this favorable association while thinking about the brand. Therefore the association has not much impact.

As described above, strong and favorable associations help to create a better brand image. However it is also important to distinguish from competitors. Therefore a brand should not only have strong and favorable, but also unique associations. Uniqueness of the associations can be explained as the brand associations that are not similar for competitor brands. To distinguish from other brands, brands want to create meanings that differ from other brands. These points in which a brand differs from competitors are described by Keller (2001) as points of difference. The points of difference consist of all the brand associations that are different for one brand compared to the other. Points of difference are also important in the consumers decision making process because they provide rational and emotional reasons to buy a certain product. A good example of this is the comparison between Albert Heijn Pilsener and Bavaria. In the product category beer, consumers look for certain quality (POPs) but also for associations that make the brand stand out from other brands. AH Pilsener and Bavaria are almost the same beers since

(16)

they are both produced by Bavaria. However, Bavaria has more unique associations, such as student beer, young users and holiday and therefore distinguishes from AH Pilsener.

2.1.3 Brand responses

The third level is concerned with creating both brand judgments and brand feelings, which can be explained as the way consumers think and feel about the brand. Brand judgments relate to the rational evaluation of the brand in terms of quality, reliability and superiority. Brand feelings relate to the emotional meaning of the brand for the consumer. Examples are warmth, pleasure, social acceptance and self-respect.

2.1.4 Brand relationships

The fourth and highest level of brand equity is concerned with creating brand resonance. Brand resonance is described as the willingness of consumers to engage in a long-lasting relationship with the brand. At this level, the consumer identifies him or herself with the values of the brand and considers the brand not as a product but as a way of life. Although this level is desirable for brands, it is beyond the scope of this research and will therefore not be explained in too much detail.

2.2 Association Networks

To understand the way in which brand associations are retrieved in the consumers mind, a network of brand associations can be identified (Roedder John et al., 2006; Henderson et al., 1997). In such an association network, there can be shown how brand associations are linked to the brand and to each other. These association networks are displayed by a brand map. An example of a brand map is given in figure 1. The nodes in this brand map represent the brand (McDonald’s), product categories (burger, fries), product-brands (Big Mac), product features (quality), and people and occasions (family, social involvement).

(17)

The consumer retrieves the brand associations related to McDonald’s in a way that can be explained by spreading activation theory (Henderson et al., 1997, refer to Collins and Loftus, 1975). First, the consumer thinks of one of the associations directly connected to McDonald’s, such as meals or family. After that the thinking spreads to one of the nodes connected with the associations earlier retrieved, such as quality, which is connected to meals. The associations that come first to mind, when eliciting a cue, are called core associations (Franzen & Moriarty, 2009). A core association of McDonald’s can be meals, service, value, family or social involvement. These are the associations that in the map are directly linked to the brand.

The amount of activation that can spread from the brand name is limited. Franzen & Moriarty (2009, p. 384) state that the more elements are associated with a brand, the less activation will spread to any particular element. This is considered as the fan effect. A consequence of the fan effect is that if a brand has many associations, it takes longer to retrieve a single brand

association.

Spreading theory plays not only a role in evaluating a brand, but also in the brand choice

process. When consumers want a product from a certain product category, they need to choose a certain brand. The activation then comes from a product category and spreads to the brands that have the strongest association to that product category (Franzen & Moriarty, 2009). In the product category beer, the association moves first toward the brands with the strongest associations (for example, Heineken and Grolsch) and then to the other brands (Pilsener). This whole process of retrieving brands and information, only takes a few seconds, for example 3,4 seconds for beer and only 1,2 seconds for milk (Franzen & Moriarty, 2009 refer to Brouwer and van der Weyde, 1998).

(18)

(19)

3. Brand extensions

3.1 Types of brand extensions

There are two approaches to use an established brand name to enter new markets, a line extension and a category extension (Aaker & Keller, 1990; Franzen & Moriarty, 2009). A line extension is an extension where the brand name is used to introduce a new product in the same product class (Aaker & Keller, 1990). Examples are Coke Zero by Coca-Cola in the coke category and Sensation White Toothpaste by Colgate in the toothpaste category.

A category extension is in the literature sometimes referred to as a brand extension, (Aaker & Keller, 1990; Tauber, 1981; Franzen & Moriarty; 2009) however in this research, a different understanding of a brand extension is used. In correspondence to Völckner & Sattler (2006) the term brand extension is used as a generic term for all the extensions that make use of established brand names. A category extension can be described in more detail by the

description of Franzen & Moriarty (2009, p. 385) as: “the process by which a new product

category is brought to the market under an established brand that until then has not been attached to that category.” An example of a category extension is the introduction of Frozen Pudding Pops by Jello in the ice-cream category.

3.2 The importance of performing brand extensions

Brand extensions are important for firms for several reasons. Firstly, according to Milberg, Park & Carthy (1997) brand extensions can strengthen existing brand beliefs and attitudes which feeds into a higher brand equity. This implies that established strong brand names can be used of which the firm can benefit. The firm moves into a category from a strong position: consumer awareness already exists for the brand and the consumers perceptions of the brand can be transferred to the extension (Tauber, 1981). Secondly brand extensions can decrease the costs of gaining distribution, and enhance the efficiency of the marketing of the brand, especially compared to setting up a new brand (Aaker & Keller, 1990). Thirdly, brand extensions offer companies a strategic opportunity to evaluate and redefine the nature and direction of their

(20)

business (Tauber, 1981). These are internal reasons to perform brand extensions. Furthermore, brand extensions can also be important for risk spreading. If the brand is solely depending on one product and the environment changes in a way in which the product needs adaptation or has become redundant, not only the product, but also the whole brand is damaged.

3.3 Dimensions of brand extension evaluation

Several studies examined the formation of attitude towards line and category extensions (Aaker & Keller, 1991; Boush & Loken, 1991; Broniarczyk & Alba, 1994; Park, Milberg & Lawson, 1991; Sood & Keller, 2012). Sood & Keller (2012) researched the effect of brand name structure on the attitude towards brand extensions, but most studies concern research to the fit and quality perceptions of the brand and the brand extensions (Aaker & Keller, 1991; Boush & Loken, 1991; Broniarczyk & Alba, 1994; Park, Milberg & Lawson, 1991). According to those studies, consumer reactions are formed by a process dominated by two important drivers: the quality perceptions of the original brand and the perceived fit between the original brand and the brand extension categories. These drivers, adapted from the model of Aaker & Keller (1990) agree with the main findings of other studies, although there is a slight difference in the terms used. Broniarczyk & Alba (1994) for example, classify the drivers in terms of brand affect transfer and product category similarity but describe those in the same way as Aaker & Keller (1990) describe perceived quality of the brand and the perceived fit.

The model of Aaker & Keller (1990) is used as a basis for this research because Bottomley & Holden (2001) extensively tested the drivers by analyzing eight comparable

studies. Aaker & Keller (1990) originally included a third driver, which is the perceived difficulty of making the extension. However, the contribution of this driver cannot significantly be proven by several of the studies compared by Bottomley & Holden (2001) and is therefore excluded from this research.

(21)

The overall attitude towards the original brand is an important predictor for the evaluation of a brand extension because consumers are likely to transfer their affect to the brand extension if they like the original brand (Aaker & Keller, 1990; Bottomley & Holden, 2001; Broniarczyk & Alba, 1994). Brand attitude is conceptualized by Aaker & Keller (1991, p. 29), as: “the consumers perception of the overall quality of the brand”. Zeithaml (1988) explains the cognitive structure of perceived quality by several levels of abstraction, using a means-end chain approach. A low level of abstraction consists of product attributes, which are the physical characteristics of products. Perceived quality is on a higher level of abstraction than product attributes and can therefore be seen as a form of overall evaluation or value judgment of the brand (Olshavsky, 1985; Holbrook & Corfman, 1985). If the overall perceived quality is good, consumers are more positive towards the brand and will therefore evaluate the brand extension better and are more willing to try the brand extension. Reversely, if the overall perceived quality is bad, consumers are not positive towards the brand and this will have a negative impact on the evaluation of the brand extension and the willingness to try the extension.

3.3.2 Product class similarity

In addition to the perceived quality of the brand, the product class similarity is extensively described as an important predictor of brand extension evaluation (Aaker & Keller, 1991; Broniarczyk & Alba, 1994; Czellar, 2002; Park, Milberg & Lawson, 1991). Product class similarity can be described as the fit between the product class of the original brand and the product class of the brand extension (Park, Milberg & Lawson, 1991). Previous studies show that the perceived similarity between the brand and the brand extension category enhances the transfer of affect, knowledge and buying intentions. If a brand such as Heinz, which is famous for its ketchup, would extend to the toothpaste category, consumers would think of the toothpaste in terms of ketchup and will be confused. However, if Heinz makes other products with

tomatoes, such as soup and pasta sauce, the extension is more related to the original product category and therefore evaluated better.

(22)

The measure for product class similarity can vary from low to high. There are several ways to evaluate the product class similarity. From a consumers perspective, the product class similarity can be based on product ingredients or the place in the shelves in the supermarket, but also on less obvious comparisons such as calcium content (Broniarczyk & Alba, 1994). From a manufacturer perspective, the product class similarity can be based on the way of processing and the materials used. However, since this research is focussing on consumers perceptions of a brand, product class similarity is only explained by the consumers perspective, in accordance to Aaker & Keller (1990).

Product class similarity is described by Aaker & Keller (1990) in three dimensions, which are complement, substitute and transfer. Complement refers to the extent of which consumers view the product category of the original brand and the brand extension as complementary. If Marlboro would extend to the category lighters, lighters and cigarettes would be complementary and therefore the product class similarity would be higher. Substitute refers to the extent to which consumers view the product category of the original brand and the brand extension as substitutes. If Heinz would extend from bolognaise pasta sauce to basil pasta sauce, the two pasta sauces would be substitutes and therefore the product class similarity would be higher.

Transfer refers to the view of consumers on the ability of the manufacturer of the

original product, to produce the brand extension as well. If Heinz would extend to the toothpaste category, consumers will be likely to doubt the ability of Heinz to deliver good dental care and therefore the product class similarity will be lower.

For line extensions, substitutability is a very important component. A line extension is an extension within the same product category as the original brand and because of these related product categories, consumers are usually more convinced of the manufacturers ability to produce the extension. Besides, the substitutability is high because line extensions are products that are a variant of the original product and therefore often substitutes.

For category extensions, substitutability and transfer are lower than for line extensions. Since category extensions are extensions to another product category, the degree of

(23)

substitutability and transfer relies on how close the product categories are. For category extensions with a high product category fit with the original products, substitutability and transfer will be higher than for category extensions with a low product category fit with the original products. Category extensions can also be complementary because the new product class complements the original product class.

The total product class similarity depends on how the product classes are defined. If the product classes are defined very broad, the product class similarity will be lower for a new product category then when the product classes are defined narrow.

(24)

4 The transferability of brand-specific associations

In correspondence with Broniarczyk & Alba (1994) and Park, Milberg & Lawson (1991) there is an important effect of brand-specific associations on extension evaluation. Their findings show that brand-specific associations can override the effects of product class similarity. Although product class similarity is an important driver of extension evaluation, consumers can have their own theories on why products can belong in the same category, rather than having the same product features (Murphy & Medin, 1985). Park, Milberg & Lawson (1991) found that the evaluation of a brand extension is higher when the brand and the extension category share the same associations. The principle behind this fit is cognitive consistency (Franzen & Moriarty, 2009). If consumers can find a link between the different products of the extensions and the brand, the brand concept is stronger and the extension is more easily accepted. Just one single brand association can provide a connection between the brand and the extension. Both Broniarczyk & Alba (1994) and Park, Milberg & Lawson (1991) demonstrate that these brand-specific associations can overrule the effects of product class similarity.

The impact of brand-specific associations on the extension evaluation is different

depending on the way in which a consumer processes the brand extension. The brand extension can be evaluated by piecemeal processing and category-based processing (Aaker & Keller, 1990). Category-based processing relates to the similarity between the product category of the original brand and the brand extension and piecemeal processing relates to the importance of brand-specific associations. Boush & Loken (1991) demonstrated that when consumers evaluated products that were found ‘the same’ as the original brand products, category-based processing dominated. However, when consumers evaluated products that were found ‘moderately different’ from the original brand products, piecemeal processing dominated. So especially when the extension has a lower product class similarity, more piecemeal processing takes place and brand-specific associations are more important. Therefore brand-specific associations will play a bigger role in the evaluation of category extensions than in the evaluation of line extensions.

(25)

Brand-specific associations can provide a connection between the brand and the extension, however it can also be the other way around. Some brands have difficulties performing an extension despite a high product class similarity. That way brand-specific associations prevent the brand from having a successful extension. A good example is the extension of Pampers into the paper tissue category. This extension failure cannot only be related to product class similarity. Both pampers and tissues are in the non-food category and Pampers has the ability to produce strong and absorbing paper and therefore would probably make high quality tissues. However, it is the association with poop that is limiting the brand into the paper tissues extension. Absorbing poop is a strong favorite association in the pampers category, but is not relevant, or even negative in the extension category. Another example of the impact of brand-specific associations is described by Zeithaml (1988). The association of pulp is compared for both orange juice and apple juice. In orange juice, pulp stands for high quality , whereas in apple juice, pulp stands for low quality. In this case the association pulp is both perceived as positive and negative, depending on the extensions product category.

Park, Milberg & Lawson (1991) researched a practical example of the importance of brand-specific associations by comparing two watch brands, Timex and Rolex. Both brands share many product associations and therefore an extension into another category, based on product class similarity, would result into the same attitude towards the extension. However, Rolex distinguishes from Timex in its luxury and high status. These associations are important for the evaluation of the brand extension as well. An extension of Rolex into another luxury category, such as cologne, neckties and cuff links, is received better than an extension of Timex in that same category. Therefore it is important that brands leverage their strong, unique and favorable associations in such a way that the associations will have a positive influence in the extension category.

(26)

In the previous paragraphs is described how brand associations play an important role in the evaluation of a brand extension. The brand associations play a role in both the perceived quality of the brand and in the perceived fit between the brand and the brand extension. The customer-based brand equity model shows that if brands have strong, unique and favorable associations, brands have higher brand equity and are perceived as strong brands. Besides, brand-specific associations play in a role in the connection between the brand associations and the extension category. This fit is more difficult to achieve for brands that are strongly determined by one or a few strong associations. Because the association networks of these brands are less extensive, much activation will rapidly spread to those few but strong association (Franzen & Moriarty, 2009). Besides, the strength of the brand’s core association has a substantial impact on the development of new associations (Franzen & Moriarty, 2009). Different associations can compete against each other in the mind of consumers. This competition between associations can result in a situation where old associations complicate the development of new ones and consumers do not allow new associations anymore. This process stems from a psychological principle called “associative inhibition” (Franzen & Bouwman, 1999). Muller & Schumann (1894) first described associative inhibition which is quoted by Kline (1921) as: “If a is already connected with b, then it is difficult to connect it with k; b gets in the way”. A practical example of how associative inhibition works comes from the study of Aaker & Keller (1990). In their research, the hypothetical brand extension of Heineken into popcorn is included. Common reactions of consumers were that this popcorn would be unappetizing and taste like beer. Consumers connect Heineken (a) with beer (b), which makes it difficult to connect Heineken with popcorn (k); beer stands in the way. For brands that are strongly determined by only one strong association, associative inhibition can be problematic, since extension possibilities are very limited. Brands that do not allow new associations are called “prison brands” because they are imprisoned by their own associations (Haveman, 2003). In general, prison brands can be imprisoned by every association: a price, a product class, a user moment or a user image. However within this research, the focus is on brands that are strongly associated with the

(27)

product class. Therefore the definition of a prison brand is in this research set as follows: a prison brand is a brand that is locked in a certain product category because the strong product associations prevent the brand from having success in another product category.

An example of a prison brand is Xerox. Xerox is strongly associated with the product class copiers. Although Xerox was very successful in this product category, over time Xerox tried to broaden its product range by entering the computer market (Franzen & Moriarty, 2009). However, customers associated Xerox so strongly with copiers that the introduction of a Xerox personal computer flopped. Xerox was not able to make the switch from copiers to computers, even though for Xerox this seemed a logical step. Not only extensions into totally new product categories can be a difficult step for prison brands. Even if the product class of the extension appears to be quite similar to the product class of the original brand, an extension of a prison brand can be difficult. An example of this is the extension made by Biotex. Biotex is famous for its soaking and prewashing detergent and therefore strongly associated with prewash. Biotex decided to introduce a detergent for the main wash, which in terms of product category, seems quite similar to a detergent for the prewash. However this extension did not work out and the new washing detergent flopped (Franzen & Moriarty, 2009).

4.2 Abstract versus product-related associations

The transferability of brand associations to extensions is not only determined by the strength of the associations but also by the type of associations. Park, Milberg & Lawson (1991) found in their research that prestige brands had higher extendibility to products in another product category, than functional brands. Prestige brands have more abstract or symbolic associations such as luxury, exclusivity and good quality, whereas functional brands have more product-related associations, such as strong, durable, smell and color. Franzen & Moriarty (2009)

provide a classification system to explain the importance of association types on brand meaning. In this classification, brand associations are categorized in several hierarchical levels, ranging from concrete to abstract. This hierarchical classification implies that brands with more abstract

(28)

brand associations have higher brand meaning than brands with more product-related

associations. Aaker & Keller (1991) confirm this idea in their research to extensions for several brands. They found that the brand Vuarnet, a sunglasses brand that was associated with style, had a high ability to be extended to other product classes such as clothing, shoes and wallets. Both Park, Milberg & Lawson (1991) and Aaker & Keller (1991) address that brands with more abstract associations have higher extendibility. Park, Milberg & Lawson relate abstraction in their research to prestige, but practical examples indicate that even functional brands can have abstract brand associations. An example of a functional brand with abstract associations is Bertolli. Bertolli used to sell olive oil as their main product, but because of their association: the Italian way of life, Bertolli is able to sell all products related to Italy.

Milberg (1993) shows that if a brand introduces brand extensions consistent with a broader, more abstract product category, it changes not only the brands’ core business

definition but also enlarges the extendibility to different product categories. A good example of this abstraction process is described by Milberg, Park & Carthy (1997). In their research, the difficulties of Xerox in extending to another product category are addressed. As mentioned earlier, the brand Xerox was strongly associated with copiers and as a result an extension into the personal computers failed. The solution for Xerox was to reposition the brand as “the document company”, which is on a higher level of abstraction than the product-related association of copiers. The document company was more easy to transfer into many other product categories such as paper, ink, document and organization software.

4.3 Possibilities to move away from strong product associations

Brands that are imprisoned by strong product associations have several possibilities to enhance the transferability of their associations. The first possibility is to provide a brief elaboration on the attributes of the brand extension, so that the spill-over of the associations is limited. Aaker & Keller (1990) provided the following brief elaboration on a hypothetical brand extension of Heineken into popcorn: “In regular and Cheddar cheese flavors”. According to Aaker & Keller

(29)

(1990) this led to more favorable extension evaluations. A disadvantage of this method is that it costs much time and money to reach all consumers. Besides, consumers have to process the message actively, which requires ability and motivation (Petty and Cacioppo, 1986).

A second possibility links to the paragraph above and is concerned with creating brand associations that are more abstract. Aaker & Keller (1990) argue that is more easy to transfer an abstract attribute to a broader set of product classes. For prison brands it is not necessary to leave all existing strong brand associations behind. The strong , favorable and unique

associations after all make the brand stand out. Therefore the added abstract association could function as an umbrella for the existing associations. Above the example of how Xerox

repositioned itself is provided. Another example of a brand that created an abstract association as an umbrella, is Osborne. Traditionally, Osborne is a brandy- and sherrybrand but it wanted to produce and sell wines as well. Because most consumers associated Osborne with sherry, the extension into wine was difficult. Therefore, Osbore created the abstract association “Spain” as an umbrella association where both quality sherry and wine could be associated with. To fully embrace this abstract association, Osborne opened Spanish food restaurants and started to sell Spanish food as well. At the same time, Osborne wine was served at these restaurants, which made it possible to gradually introduce wine. This way the sherry association did not limit the extension into wine, but the quality associations and strong brand image were kept.

(30)

5. Hypotheses

To provide insight on the developed main question and related sub questions, this chapter will elaborate on the different hypotheses to be tested in this research. These hypotheses are building upon the literature presented in chapter 2, 3 and 4. At first two hypotheses are presented regarding the difference in extension potential between two brand types

accompanied by different brand-specific associations. Lastly two hypotheses on the effect of priming an existing association are presented.

5.1 Extension potential of prison brands vs not-prison brands

As described in chapter 3, product class similarity is an important predictor of brand extension evaluation (Aaker & Keller, 1990; Broniarczyk & Alba, 1994; Czellar, 2002; Park, Milberg & Lawson, 1991). A higher similarity between the brand and the extension category enhances the transfer of affect, knowledge and buying intentions and results in a better brand extension evaluation. Again, product class similarity can be described as the fit between the product class of the original brand and the product class of the brand extension (Park, Milberg & Lawson, 1991). However brand extensions occur in several forms. As described by Franzen & Moriarty (2009) and Aaker & Keller (1990) we can distinguish between line extensions and category extensions. For line extensions, substitutability is a critical component. Because of the related product categories, consumers are more confident in the product quality of the extension. For a category extension, more new associations are needed than for a line extension, since

associations with the new product category should be created in the mind of consumers (Aaker & Keller, 1990). Therefore it is expected that line extensions generally will be evaluated more positively by consumers than category extensions.

Since category extensions can also differ from each other on the level of fit, we add to the experiment a distinction between category extensions with a high fit with the initial product category versus category extensions with a low fit with the initial product category. For category extensions with a high product category fit with the original products, substitutability will be

(31)

higher than for category extensions with a low product category fit with the original products. Consumers perceive the High Fit Category Extension as a more natural match with the initial brand and therefore are more likely to evaluate the positively.

This results in the following three types of brand extensions: High Fit Line Extensions, the High Fit Category Extensions and the Low Fit Category Extensions. Due to the argumentation above, a direct effect of extension type on extension evaluation is expected. The corresponding hypotheses are defined as follows:

H1a. High Fit Line Extensions (HFLE) are evaluated more positively than High Fit Category Extensions (HFCE).

H1b. High Fit Category Extensions (HFCE) are evaluated more positively than Low Fit Category Extensions (LFCE).

Besides the direct effect of extension type on extension evaluation, also the impact of brand positioning should be investigated. As explained in chapter 4, brands that do not allow new associations are called “prison brands” because they are imprisoned by their own associations (Haveman & Labadie, 2003). Specifically, in this research, the focus is on brands that are strongly associated with the product class. The definition of a prison brand is set as follows: a prison brand is a brand that is locked in a certain product category because the strong product associations prevent the brand from having success in another product category. Broniarczyk & Alba (1994) found in their research that there was an interaction effect between brand-specific associations and the product class similarity on extension evaluation. A brand can successfully extend to another category if the brand-specific associations are sought benefits in the new product category. Prison brands have strong product associations, which might in the mind of consumers be conflicting with the new product category. According to Trout & Ries (1979) and Aaker & Keller (1990) it is critical for a brand to have a unique selling proposition, because this

(32)

convinces consumers to purchase a particular brand. This unique proposition is referred to by Keller (1993) as Points of Difference. For prison brands, that are limited in extending their associations to other product categories, it is difficult to create PODs in a new product category. Furthermore it is hard for prison brands to successfully create Points of Parity in another product category. This is because prison brands in this situation fail to enable positive product associations with the consumer and are therefore not considered positively by the end user (Walker & Olson, 1991). Furthermore Park et al (1991) suggest that the level of abstraction of an association impacts the transferability of an association. A more abstract brand has a higher ability to extend towards product categories unrelated to the initial brand, compared to a more functional brand. Since prison brands are based on more concrete product associations, it is believed that it is more difficult for this type of brand to successfully extend out of their initial product category.

As part of this research we will look at the effect of being a prison brand on the evaluation of the three above mentioned extension types. Since High Fit Line Extensions are extensions done in the same product category, category-based processing will dominate (Boush & Loken, 1991). Because prison brands are strongly associated with the product category, it is expected that HFLE done by prison brands are evaluated higher than HFLE done by not-prison brands. For High Fit Category Extensions this is expected to be the other way around. In a HFCE evaluation piecemeal processing will dominate and therefore the fit between the brand-specific associations and the new product category is more important (Boush & Loken, 1991) Since prison brands have strong product-related associations, it would be more difficult to extend to another product category although the product category is, in terms of fit, quite close to the original product category. A not-prison brand, which has in general more associations that are not product-related is therefore expected to be extended more easily to another product category.

For a Low Fit Category Extension, no difference in evaluation between prison brands and not-prison brands is expected. For both brands, the product category has a low fit with the

(33)

original product category and the associations of both brands are not expected not to match the extension well. Regardless of whether the brand is imprisoned, it is likely that the LFCE is considered too unrelated and therefore consumers might be hesitant in their evaluation. In general, no direct effect of brand positioning on extension evaluation is expected. This is because both prison and not-prison brands are both expected to have higher and lower extension

evaluations, but this is dependent on the type of extension. Summarizing, Extension type and brand positioning are expected to interact such that High Fit Line Extensions (HFLE) are evaluated more positively for prison brands versus not prison brands, while High Fit Category Extensions (HFCE) are evaluated more positively for not prison brands versus prison brands. Furthermore High Fit Category Extensions (HFCE) are expected to be evaluated more positively for not prison brands versus prison brands, while Low Fit Category Extensions (LFCE) are expected to be evaluated equally for not prison brands and prison brands. This results in the following hypotheses:

H2a. The effect of hypothesis 1a will be more pronounced for prison brands than for not prison brands.

H2b. The effect of hypothesis 1b will be more pronounced for not-prison brands than for prison brands.

Hypotheses 1 and 2, as described above, are summarized in a graphical way in a conceptual model below (figure 3).

(34)

Fig. 3 Extension potential of prison vs not prison brands

5.2 Extension potential of primed prison brands vs prison brands

Prison brands have difficulties extending to other product categories because the strong product associations prevent the brand from having success in another product category. Since prison brands do not immediately allow new product associations, it is useful to look at how possible abstract brand associations can be better put forward. Both Park, Milberg & Lawson (1991) and Aaker & Keller (1991) confirm that brands with more abstract associations have higher

extendibility. As described above, Xerox repositioned itself as the document company and succeeded to move away from the copier association. Another practical example of a company who did this is Osborne. With the introduction of the association ‘Spain’ Osborne was able to move away from sherry to wine and other Spanish products.Both companies focused on bringing forward associations that were already present, but were not core.By priming an already existing association, consumers accept the association more easily and the more abstract association helps the brand in extending to other product categories. The examples above come from practice, but are not tested in research yet. Therefore the second part of this research focuses on the extension potential of primed prison brands in comparison to prison brands. In line with hypothesis 1a and 1b, a main effect of the extension type on the extension evaluation is expected, where a higher similarity between the brand and the product class of the

(35)

extension results in a better brand extension evaluation. Again three extension types are

compared The High Fit Line Extension, the High Fit Category Extension and the Low Fit Category Extension. The following hypotheses are proposed:

H3a. High Fit Line Extensions (HFLE) are evaluated more positively than High Fit Category Extensions (HFCE).

H3b. High Fit Category Extensions (HFCE) are evaluated more positively than Low Fit Category Extensions (LFCE).

Again an interaction effect between brand positioning and extension type is expected. Both prison brands and primed prison brands have strong product associations, which are most relevant in category-based processing. Therefore it is expected that for High Fit Line Extensions the extension is evaluated similarly for both brands. For High Fit Category Extensions it is

expected that extensions of primed prison are evaluated better than extensions of prison brands. In a HFCE evaluation piecemeal processing will dominate and therefore the fit between the brand-specific associations and the new product category is more important. Prison brands have strong product-related associations, that make it more difficult to extend to another product category although the product category is, in terms of fit, quite close to the original product category. The primed prison brand has a primed abstract association which can have a better fit with another product category and therefore be extended more easily. For a Low Fit Category Extension, no difference in evaluation between prison brands and primed prison brands is expected. For both brands, the product category has a low fit with the original product category and the associations of both brands are not expected not to match the extension well.

Summarizing, extension type and brand type are expected to interact such that High Fit Line Extensions (HFLE) are evaluated equally for prison brands and primed prison brands, but High Fit Category Extensions (HFCE) are evaluated more positively for primed prison brands versus

(36)

prison brands. Extension type and brand type are expected to interact such that High Fit Category Extensions (HFCE) are evaluated more positively for primed prison brands versus prison brands, while Low Fit Category Extensions (LFCE) are evaluated equally for prison brands and primed prison brands. All of these expectations for the interaction effect are summarized in the last two hypotheses:

H4a. The effect of hypothesis 3a will be more pronounced for prison brands than for not prison brands.

H4b. The effect of hypothesis 3b will ben more pronounced for not-prison brands than for prison brands.

Hypotheses 3 and 4, as described above, are summarized in a graphical way in a conceptual model below (figure 4).

(37)

6. Method

In this study two experiments are performed researching consumers evaluations of a variety of hypothetical brand extensions for three brands that are positioned differently (prison brand, not prison brand and primed prison brand) under three extension types (High Fit Line Extension, High Fit Category Extension and Low Fit Category Extension) in two product categories (deodorant and tea). The evaluation of the brand extensions is examined by means of rating scales. The brands are represented in each product category by two actual brands of which one functions as both a prison and a primed prison brand.

6.1 Qualitative Analysis

Open-ended brand associations were obtained for all four brands. Subjects were asked to write down the first things that come to mind when thinking of the brands. All subjects answered this question for all of the four brands. There was no minimum and a maximum of 9 associations for each brand. Subjects were also asked to rate each of the given associations in terms of strength on a 5-point scale (1=very low, 5=very high).

Afterwards, the open-ended responses to the questions were classified by a coder in three categories: product category associations, abstract associations and other associations.

6.2 Quantitative Analysis

6.2.1 Design

Two 2x3x2 mixed designs are used with brand positioning (Prison, not prison or prison, primed prison) as the between-subjects factor and extension type (HFLE, HFCE and LFCE) and product category (tea, deodorant) as the within-subjects factors. Each respondent participated in two repetitions of the experiment. In the first part, brands within the product category tea are included and in the second part, brands within the category deodorant are included. Each respondent received a subset of three extensions in two product categories and evaluated in total six extensions.

(38)

6.2.2 Stimuli

6.2.2.1 Independent Variables

Brand Positioning. Brand positioning was manipulated as a between-subjects variable. A pre-test was conducted among a focus group to select potential brands. In total, four brands were selected in two product categories. Within each product category one prison brand and one not-prison brand were selected so that the two brands could be compared. The brands were selected on the criteria of being well-known, perceived as high quality and not being extended to broad previously. High quality brands were chosen so that this measure could not interfere with the extension evaluations.

Focus group subjects were asked to come up with prison brands that are associated strongly with the product category. At the same time subjects were asked to come up with not-prison brands in the same category as the not-prison brand, that are not associated strongly with the product category.

The focus group came up with several prison and not-prison brands in the same product category. The following brands were chosen as the most representative: Pickwick and Lipton in the category tea and Rexona and Dove in the category deodorant. Pickwick and Rexona were selected as the prison brands and Lipton and Dove as the not-prison brands. Subsequently, 10 people were asked to list up the first things that come to mind when thinking of the brands Pickwick, Lipton, Rexona and Dove. The results of this pre-test indicated that both Pickwick en Rexona are more associated with their product category than Lipton and Dove. ‘Tea’ is more mentioned as the first association at Pickwick (7 out of 10) than at Lipton (2 out of 10) and ‘deodorant’ is more mentioned as the first association at Rexona (6 out of 10) than at Dove (1 out of 10). Besides, the mentioned associations were used to construct a prime for the chosen prison brands. Out of all the associations mentioned for Pickwick and Rexona, the following abstract associations were selected: Natural for Pickwick and Sportive for Rexona. The primed prison brands were constructed by adding a brief elaboration on these abstract associations, in correspondence with Aaker & Keller (1991). Aaker & Keller (1991) state that providing brief

(39)

elaborations or cues is a way to improve the extension evaluation. For Pickwick this resulted in: “Pickwick is all about natural flavours. With every sip, the feeling of pure nature arises.” For Rexona, this resulted in: “Rexona is perfect for every sport experience. From running to soccer, Rexona makes you feel a true athlete.”

Extension type. Extension type was manipulated as a within-subjects variable. A pre-test was conducted among a focus group to select potential brand extensions. In total, six extensions, two High Fit Line Extensions, two High Fit Category Extensions and two Low Fit Category Extensions were chosen. Within each product category three extensions were selected that were equally likely for both the prison and the not-prison brands. The extensions were selected on a few criteria. Firstly, the brand extensions had to involve products that the brands did not offer yet. Secondly, the brand extensions had to fit with the abstract associations introduced for Rexona and Pickwick. Besides, the brand extensions needed to be equally suitable for both the prison brand and not-prison brand. Focus group subjects came up with several brand extensions of which the following were chosen as most representative:

HFLE HFCE LFCE

Pickwick Asian Sensation Tea Lemon & Mint water Oatmeal cookies Pickwick (primed) Asian Sensation Tea Lemon & Mint water Oatmeal cookies Lipton Asian Sensation Tea Lemon & Mint water Oatmeal cookies Rexona Long lasting deodorant Freshen up body tissues Sports drink Rexona (primed) Long lasting deodorant Freshen up body tissues Sports drink Dove Long lasting deodorant Freshen up body tissues Sports drink Table 1 Brand extensions distinguished in terms of extension type and and brand positioning

6.2.2.2 Dependent Variables

Brand Extension Evaluation: Brand extension evaluation was built up by two measures: the quality of the brand extension and the intentional buying behaviour. In correspondence to Aaker

(40)

& Keller (1991), the perceived overall quality of the brand extension is measured by a 7-point scale (1=inferior, 7=superior) as well as the likelihood of trying the extension, assuming a purchase was planned in the product class (1=not at all likely, 7=very likely).

6.2.3 Manipulation Checks

Product category fit

The fit between the brand and the product category is important to check the brand positioning manipulations. Product category fit is measured for all brands by four questions (7-point likert scale):

-To what extent do you associate Pickwick with tea? -To what extent do you associate Lipton with tea -To what extent do you associate Dove with deodorant -To what extent do you associate Rexona with deodorant?

Fit between the brand and the extension

The fit between the brand and the brand extension can be an important indication for the

success of the brand extension and is also important to check the HFLE, HFCE and LFCE category manipulations. Fit is measured by two questions that are used in Keller & Aaker (1992) (both 7-point likert scale): “There is a good fit between the brand and the extension” and “It is appropriate for the brand to make the extension”.

6.2.4 Control Variables

Quality of the original brand

Since the quality of the original brand is a control variable, one measure (5-point likert scale) is used: “How do you perceive the overall quality of the following brands?”.

The impact of imprisonment by strong product associations on the extension potential of brands

Master Thesis

Prison Brands