1
The influence of ownership on the effect of image fit on vertical line
extension evaluation and product fit on product category extension
evaluation for a prestigious brand.
Jasper Buddingh’
10271805
22-07-2018, second draft
MSc. in Business Administration – Marketing Track
University of Amsterdam
2
Statement of originality
This document is written by Jasper Buddingh’ 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
Table of contents
Abstract………...5 1.Introduction………..6 1.1 Background………...…6 1.2 Theoretical introduction………6 1.3 Research goal………8 1.4 Contributions………9 1.5 Thesis structure………...10 2. Theoretical framework………...…………...112.1 Brand extensions: Product category extension or line extension...……..………...11
2.2 Customer brand equity………...………..…13
2.3 Fit………14
2.4 Image fit and product category extension…...………16
2.5 Image fit and vertical line extension………..……….17
2.6 Ownership and vertical line extension……….…...18
2.7 Ownership and product category extension………21
2.8 Conceptual framework………24
3. Methodology……….25
3.1.1 Pre-test 1: Brand selection………26
4 3.1.3 Pre-test 3: Extensions………...28 3.2 Experimental design………31 3.3 Measures………..34 4. Results………...35 4.1 Response analysis………....35 4.2 Data preparation……….….36 4.4 Pre-analysis……….36 4.5 Hypotheses testing………..37 5. Discussion……….41 5.1 Summary……….42 5.2 Discussion………...42 5.3 Implications………46
5.4 Limitations and future research………..47
6. References………49
5
Abstract
This thesis has the goal to asses how ownership moderates the effect of image fit on vertical line extension evaluations and product fit on product category extension evaluation for a prestigious brand. With this research a clear gap in the literature is addressed, because up till now ownership hadn’t been investigated within the context of a product category extension. Two different two-by-two experimental designs were used to answer the research question. Results were gathered from 120 respondents.
The results show that ownership moderates the effect of image fit on vertical line extension evaluation and the effect of product fit on product category extension evaluation. In the context of a vertical line extension, when image fit is high owners have higher evaluations than owners. However when image fit is low owners have lower evaluations than non-owners. In this thesis the theory of self-brand relations is proposed as an explanation.
Within the context of a product category extension results are different. With high product fit owners have higher extension evaluations, in the case of low fit also, however the results are less pronounced. Motivated reasoning theory can provide an explanation for this effect.
For managers the implications of this research are also useful, it demonstrates that brand owners react more favourably than non-owners in most situations, however under certain circumstances the results can be reversed. Managers can use this information to prevent damage to the relations with their current customers. Future research should examine in further detail what exactly the underlying mechanisms are that cause these effects.
6
1. Introduction
1.1 BackgroundLast summer the world saw the launch of a new museum, one that was very different from existing ones. No paintings, sculptures or other artwork is to be seen here. Instead, a green tube of Heinz tomato ketchup, a perfume made by motorcycle company Harley Davidson and microwave dishes made by toothpaste manufacturer Colgate are displayed. What do these three products have in common, one might ask? All three, with everything else that’s
exhibited in ‘Sweden’s Museum of Failure’, are examples of product introductions that were brought to the market by existing companies and that were complete disasters (Venrooij van, 2017). An interesting question that pops up is: Why were these products unsuccessful?
1.2 Theoretical introduction
These failed products are examples of brand extensions, a topic that has interested numerous scholars doing research within this field through the years. A brand extension can be defined as the introduction of a new product under the name of an existing brand. The use of an existing brand name can have great benefits. Consumers have associations with a brand. When a new product is coupled with an existing brand, these associations can be transferred (Aaker and Keller 1990). This is helpful because otherwise the company would have to build up strong, unique and favourable associations from scratch, which can be a difficult task. Furthermore consumers are already familiar with the brand, therefore there is less need for expensive marketing campaigns to create awareness for the new product (Tauber, 1981). Everything taken together, the use of an existing brand name is a lot cheaper, easier and more efficient. Therefore about 80 percent of companies use an extension when they introduce a new product (Martinez and Pina, 2010).
7
extensions, on the other line extensions (Riley, 2015). With a product category extension the new product is launched in a different product category than that of the parent brand (Aaker and Keller 1990). There are two types of line extensions. With a horizontal line extension the product category remains the same, however there are slight modifications to the original product like adding new flavours or features (Riley, 2015). A vertical line extension refers to the situation where the product category again stays the same, however the new product is dissimilar from the original one in terms of quality and price. A step-up line extension has higher price and quality, whereas a step-down line extension has lower price and quality (Riley, 2015).
Scholars have identified different factors that contribute to the success of an extension. The quality and the image of the parent brand play an important role (Aaker and Keller, 1990; Dwivedi, 2009). However the most important determinant of an extension’s success appears to be fit. Fit is described in different ways, however the essence of the term describes the similarity between the extension and the parent brand. In numerous articles it is proven that this similarity has a significant positive impact on the extension evaluations (Aaker and Keller, 1990; Dwivedi, 2009; Park et al. 1991). There are two forms of fit: product fit and image fit. Product fit refers to the similarity between the product category of the extension and that of the parent brand. Do both categories share some features or functionality? Image fit explains the way in which the extension and the parent brand share a same concept. Do they have similar non-product related associations, for example the user and usage imagery? Both are important determinants of the success of an extension (Grime et al., 2002).
Another factor that influences how people evaluate an extension is ownership.
Kirmani et al. (1999) found that in most cases owners evaluate vertical line extensions higher than non-owners. The one exception to this rule is when a prestigious brand extends
8
incongruent with the image of the parent brand. Since a prestigious brand is characterised by high prices and high quality an extension which scores lower on these points has low image fit. However, how owners respond to a product category extension has not been investigated up till now. Thus it is a clear gap in the existing literature. A gap that this thesis aims to address.
1.3 Research goal
Therefore the research question of this thesis can be defined as follows:
How does ownership moderate the effect of image fit on the evaluation of a vertical line extension and the effect of product fit on the evaluation of a product category extension for a prestigious brand?
The author of this thesis proposes that for a vertical line extension in a high image fit situation owners have higher evaluations than non-owners. However, in the case of a low image fit situation owners evaluate the extension lower than non-owners. As an explanation for these effects the following theory is put forward. Since owners are likely to have stronger self-brand relations, they like the brand better than non-owners. In case of a high fit image extension, these feelings are transported from the brand towards the extension, resulting in higher
evaluations for owners compared to non-owners. However, with a low fit image extension the extension is incongruent with the parent brand. Since owners have stronger self-brand
relations and are attached to the image of the brand, they are likely to fear that their self-brand relations are threatened, resulting in lower extension evaluation than non-owners who don’t have the fear of damaged self-brand relations.
For a product category extension the results are likely to be different, since the image often remains the same, so in that case the self-brand relations are not threatened. However
9
the product fit does change. In a high product fit situation owners are likely to have higher evaluations. Owners have higher self-brand relations, causing more affection for the brand, which is likely to be transported to a high product fit extension, resulting in higher
evaluations. In the low product fit situation owners are also likely to have higher evaluations than non-owners, however, the difference is less pronounced. The author of this thesis proposes motivated reasoning theory as an explanation. Low product fit extensions are
generally judged negatively. However, owners will be less negative than non-owners, because they want to keep their existing opinions of the brand intact, so they reason in a biased
manner to form a more positive picture of the extension. Therefore they will have higher evaluations than non-owners. However, motivated reasoning can’t protect the owner entirely from the negativity of a low product fit extension. Thus, the difference between owners and non-owners is less pronounced in a low product fit situation compared to a high product fit situation.
1.4 Contributions
This thesis will make a contribution to the scientific community by extending the current literature, and it will also have managerial contributions. It adds to the existing literature by focusing on the role of consumer characteristics, because up till now there has not been much focus on this topic in the field of brand extensions (Childs, 2017). Furthermore, it will clarify how ownership works within the context of a product category extension.
For managers this thesis will also be valuable. To a company existing customers are of vital importance. It is often a lot more profitable to retarget owners than to attract new
customers (Reichheld, 2003). New customers are often not aware of the existence of your product, therefore marketing spending to create awareness is needed. Furthermore, even when potential customers are familiar with your brand, it will take even more marketing spending
10
and persuasion to make these people buy your brand. However, owners are already familiar with your brand and will have a favourable attitude. So, chances are they can be easily persuaded to buy more products of your brand. For this reasons it is profitable to focus on owners, and not to damage your valuable relations with them, since these are one of the most precious assets of the firm. Therefore, also within the context of brand extensions, it is important to know how owners will react. This thesis investigates when owners react
favourably towards an extension and when not, and how they differ from non-owners in this respect, thus providing an important insight for managers.
1.5 Thesis structure
This thesis consists of different parts. After the introduction, chapter 2 will consist of a
theoretical framework covering the most important topics within the field of brand extensions. The most important constructs that are addressed in this thesis will be explained. Furthermore the hypotheses that will be tested are given and the conceptual framework is provided.
Chapter 3 addresses the methodology that is used. Which brand was selected, which pre-tests were done, how did the research design look, which measures were taken and which sample was drawn? The results will be given in chapter 4, with a thorough analysis of the data. In chapter 5 the results will be discussed, and a conclusion will be given. After that the limitations of this study and the options for future research will be addressed.
11
2 Theoretical framework
2.1 Brand extensions: Product category extension or line extension
As mentioned in the introduction there are different forms of brand extensions. In this paragraph the author will explain which extensions exist, and what their most important benefits and downsides are.
The two main sorts are a product category extension and a line extension. A product category extension refers to the situation when a company uses the brand name to introduce a new product in a different category than the one in which the company currently operates (Kim and Lavack, 1996). These extensions are also referred to as horizontal extensions. However, in this thesis the term product category extension will be used to avoid confusion between a horizontal extension and a horizontal line extension which will be discussed below. Furthermore, sometimes the term brand extension is used to refer only to a product category extension and not a line extension. In this thesis brand extension is used as a broad term that refers to both types of extensions.
An example of a product category extension is the iPod introduced by Apple, a
company which before that was only active in the computer industry. The main benefit of this type of extension is that a company can extend into a totally different market, which implies a strong possibility for growth. As a downside to this type of extension, it is often more difficult to export existing brand associations to a product in a completely new product category than to one in which the company is already active. (Kerin et al., 1996).
The second type of extension is called a line extension. There are two different kinds of line extensions: the horizontal and the vertical. A horizontal line extension consists of the introduction of an existing product with some additional or different features (Riley et al., 2015). Examples include the launch of diet Coke, where the product remains in essence the same, only the amount of sugars is adjusted. The horizontal line extension is useful to satisfy
12
variety seeking needs of the consumers, furthermore these types of extension can be used to adjust to consumer trends (Hardie et al., 1994). For example the aforementioned introduction of diet Coke satisfies the need of modern consumers to live healthier and consume less sugars.
With a vertical line extension the product remains the same, only the price and quality differ in relation to the original product. This can happen in two ways (Riley et al., 2015). When a step-up line extension is introduced, the price and the quality rise. When American Express introduced the American Express Platinum card that offered special benefits for frequent users it was a typical example of a step-up extension. Reasons to extend upwards are to reach the high end of the market and achieve higher margins, to gain a better, more
luxurious image, or to revitalize the brand (Heath et al., 2013).
A step-down line extension refers to the situation where the new product has lower price and quality in comparison to the original version. An example is the Emporio Armani clothing line, which is less pricey and exclusive in comparison to the original Giorgio
Armani. The benefit of such a step-down extension is that you can target new segments of the market. People who are less affluent can also purchase the step-down extension, therefore your target audience becomes much broader and it is possible to sell more products (Bohlen, Carlotti and Mihas, 2010).
However, there are also risks involved in the introduction of a vertical line extension. For a step-down the prestige and image of the original brand line could be tarnished, because the products are now available to a bigger part of the population instead of just a selected few. In this way, the step-down extension can damage the parent brand, and scare off the most loyal and promising customers. A step-up can also hurt the parent brand. When products are introduced with higher price and quality, the consumer can get confused about the main image of the brand, which ultimately could be harmful (Riley et al., 2015).
13
2.2 Customer brand equity
In this paragraph the author will demonstrate why so many companies choose to use a brand extension and why an extension is so useful for a company when introducing a new product to the market.
When a company decides to introduce a new product it has several options to choose from. The first option is to create a completely new brand under which the new product will be sold. The second possibility is to use a brand extension. Third, companies can use a
combination of the two aforementioned options, using the existing brand name in combination with a new one.
According to Martinez and Pina (2010) 80 percent of companies resort to a brand extension when introducing a new product or service. There are several reasons as to why a firm chooses to do this. First of all, by using a brand extension a company can transport existing brand associations to a new product (Aaker and Keller 1990). This is a great benefit because now a company doesn’t have to build an image from scratch, which often can be difficult, but can build upon the existing brand and it’s associations. The second benefit is that an existing brand name already has built up awareness among consumers; as a consequence the money that has to be spent on marketing and promotion is significantly less (Tauber, 1981).
An important concept that explains the aforementioned transportation of brand associations is consumer based brand equity (cbbe). According to Keller (1993) cbbe is: “the differential effect of brand knowledge on consumer response to the marketing of the brand” (p.2). Thus, the knowledge a customer has about a brand plays an important role in his or her response towards the marketing of this brand.
Keller (1993) describes brand knowledge using an associative network model. The brand represents a node in the memory of the consumer which is linked to other nodes
14
forming an elaborate network of associations. Brand knowledge consists of two important aspects. Firstly, brand awareness refers to the creation of the brand node in the mind of the consumer and how easy this node can be activated. Secondly, the brand image refers to the associations that are linked to this node, and how strong, favourable and unique they are. As mentioned before, these associations that form the brand knowledge can be transported to other products using a brand extension, making brand knowledge a valuable asset for the firm.
2.3 Fit
One important aspect that determines the success of a brand extension is the fit between the original product and that of the new one. Fit can be described as the number of shared associations between the extension and the parent brand. When the number of shared associations is higher, the fit will be higher also, resulting in a greater transfer of affection from the parent brand towards the extension. So fit is an important aspect in determining the positive evaluation and ultimately the success of an extension (Aaker and Keller 1990). The concept of categorization theory, drawn from the research field of cognitive psychology, is used to explain this phenomenon. Brands are stored as cognitive categories in the memory of the consumer. When an extension fits with the cognitive category of the parent brand in the mind of the consumer, a transfer of associations is possible, resulting in more positive evaluations of the extension (Czellar, 2002).
Although fit has been one of the most important topics within the research done towards brand extensions, there is not a single, clear definition as to what this concept exactly entails. There are 4 different dimensions of fit that are used across the field of brand extension research. These are similarity, typicality, relatedness and brand concept consistency.
Similarity refers to the features, attributes and/or benefits that the parent brand and the
15
describe how similar the extension category is compared to the parent brand (Farguhar et al.,1990). However, these dimensions are intertwined and often used interchangeably. Usually, for these types of fit, when scholars try to determine if there is high or low fit, consumers are asked questions about the similarity in product features and product category extension between the extension and the original brand (Grime et al., 2002). The difference between the aforementioned three dimensions and brand concept consistency is more clear. The concept consistency refers to similarity between the image of the parent brand and that of the extension (Park et al., 1991), others describe it as prototype fit (Mao & Chrishnan, 2006), but generally the term image fit is used (Salinas and Perez 2008; Lau and Pau 2007; Bhat and Reddy 2001).
Since the first three definitions of fit (similarity, typicality and relatedness) are so similar, in this thesis there will be only a distinction drawn between product fit and image fit as proposed by Bhat and Reddy (2001). The following example drawn from the work of Czellar (2002) illustrates the difference between the two clearly. Imagine if Marlboro launches a new product: a ball-pen. The product fit refers to the shared features between a cigarette and a ball-pen (long and small shape). The image-fit refers to the similarity of the image of Marlboro (rough, western) and that of a ball-pen (functional).
There can, obviously, only be product fit when an product category extension is used. With a vertical line extension a product is introduced in the same product category as that of the original product, only with alterations in quality and price. Since there are no clear differences between the original and the extension on the product level, it is not possible to talk about product fit for a vertical line extension. However, image fit plays a role for both product category extensions and vertical line extension. In the following paragraph image fit for a product category extension is explained. After that the author delves deeper into the topic of image fit for a vertical line extension.
16
2.4 Image fit and product category extension
Image fit can be drawn across different lines. Park et al. (1986) make a distinction between three different images based on the way products are used: functional, symbolic or
experiential. When a brand has a functional image the main reason to use it is to fulfil a specific task or solve a problem. For example, toothpaste is used for cleaning your teeth, and therefore a typical example of a functional brand. Symbolic brands on the other hand are used to express one’s personality, and have the goal of enhancing somebody’s self-esteem or to convey group membership. The most common example of these type of brands are clothing. Experiential brands refer to products that provide sensory pleasure, variety or cognitive stimulation. Food is a typical product that belongs to this category.
According to Park et al. (1991) when a brand with one of the three images extends into a product category with the same image, there is image fit. In their research they used two different types of watch brands to assess image fit and to investigate if this had an impact on the evaluations of the extension. First they surveyed respondents to asses which image the brands had. Then they repeated the same action for the hypothetical extensions which were used in their research. Afterwards respondents judged a functional and symbolic extension, for both a functional brand and an symbolic brand. The results showed that evaluations were higher when there was congruence between the image of the extension and the parent brand, thus proving the importance of image fit
Other scholars have proposed alternative forms of image fit. According to Grohman (2009) some brands have a masculine or feminine image. Harley Davidson and Axe are for example associated with manly behaviour, whereas a brand like Dove exhibits more of a feminine image. Image fit in this situation refers to the way in which the gender image of the parent brand and the gender image of the extension are in coherence. Grohman (2009) manipulated gender image fit by extending a masculine brand (Gillete) and feminine brand
17
(Ivory) with a feminine and masculine extension: low-fat food and pocket knives. The evaluations where significantly higher when there was high gender image fit.
Lau and Pau (2007) demonstrated that brand personality is also an important aspect of the brand image, and that personality fit and image fit are related. Brand personality refers to the set of human characteristics associated with the brand (Aaker 1997). Aaker (1997) developed a scale for measuring the brand personality, which includes 5 different categories: sincerity, excitement, competence, sophistication and ruggedness. These 5 dimensions are congruent with the 5 basic human personalities, drawn from psychology literature.
Diamantopoulos and Grime (2005) used this scale to measure brand personality fit and reported that it has significant results on perceptions of the parent brand’s image.
2.5 Image fit and vertical line extension
There has been less research done towards vertical line extensions compared to product category extensions. Especially the role of fit has received less attention. However, some research has been done that takes into account the role of image fit in the context of a vertical line extension. Desay and Hoyer (1993) argue that consumers will accept a vertical extension as long as it is congruent with the parent brand in terms of price, quality and status
positioning. However, according to Riley et al (2015) the different price levels can cause perceptions of brand concept inconsistency, or in other words: low image fit. For a luxurious brand a vertical line extension could be especially damaging, since a downward extension for these brands is more incongruent than for other brands, as the image of luxurious brands is based on high price and exclusivity. This is also the case for prestigious brands, although to a lesser degree, since these type of brands are also high-end and a vertical line extension can damage the prestige associated with the brand (Riley et al. 2015). Be that as it may, there is also research that proposes different effects for a vertical line extension. According to Liu
18
(2009) people transfer the quality associations of the original product to the vertical line extension, without damaging the view of the original brand and without lower evaluations for the vertical line extension. Furthermore, other research shows that when image fit is high (only a small price difference) evaluations are lower than when image fit is low (big price difference). The reason for this effect is that with a big price difference the new extension is clearly positioned as being different from the original products of the brand, therefore it creates a boundary between this new product and the original products. People perceive the new extension as separated from the parent brand to a certain degree. So the original image of the brand is less affected and more easily maintained with a more incongruent extension, than one with higher fit. Therefore people judge the low fit extension higher than the high fit one (Musante, 2007). It can thus be concluded that the effect of image fit is not clear-cut in the case of a vertical line extension.
2.6 Ownership and vertical line extension
While fit is an important factor that influences the evaluations of an extension, it is not the only one. Numerous scholars have explored other factors that have an effect on the success of a brand extension. These topics include advertisement strategies (Morrin, 1999), the initial evaluation of the parent brand (Dwivedi et al., 2009) and sub-branding strategies (Milberg et al., 1997). However, consumer characteristics have received less attention (Childs, 2017). Yet one can argue that consumer characteristics are getting increasingly important to research. In this day and age, where segmentation and targeting has become more focused and the
importance of microtargeting is steadily increasing, it is valuable to discern how different people react in distinctive ways towards an extension. This is especially the case for owners of the brand. It is much easier to increase your profits by retaining customers you already have and by targeting this group with additional products, then to attract completely new
19
consumers (Reichheld, 2003). Since existing customers are so important, it could be harmful for your brand to scare off this customer group or damage the relationship with them.
Therefore it is important to investigate how ownership influences the evaluations of brand extensions.
Kirmani et al. (1999) conducted research towards ownership in the context of a vertical line extension. They measured the responses of owners and non-owners with two different brand types: prestigious and functional. They conducted three different studies. In the first study two car brands, BMW and Acura, were compared. In the second and third study a comparison between Calvin Klein and Gap was made. The results showed that owners evaluated both a step-up and step-down extension for all four brands higher as compared to non-owners, the only exception being when BMW and Calvin Klein extended downwards. According to Kirmani et al. (1999) the reason that owners had higher evaluations is because they have stronger relations with the brand. In the case of the downward extension of BMW and Calvin Klein, owners valued the downward extension lower than non-owners because this type of extension is incongruent with the brand. BMW and Calvin Klein are prestigious brands that have built an image that communicates high prices, high status and exclusiveness. So a cheaper product is more inconsistent for this type of brand than for functional ones which are mainly bought for their functionality. Owners, who are more strongly tied to the existing image of the brand than non-owners, will experience this change of image more negatively than non-owners.
The concept of self-brand relations attempts to explain this process further. According to Royo-Vela and Voss (2015) when consumers own a product, they build relations with the brand. The brand becomes a part of their life, an expression of their inner being, or in other words: an extended part of their self. It becomes something they identify themselves with, and furthermore the brand helps them to build up a sense of self-esteem. Since owners are much
20
more involved with the brand than non-owners, they have more affection towards the brand. As a result they have higher evaluations of the brand’s products; including extensions with image fit. However, because the brand is so closely connected to the consumer on a personal level, change of the image of the brand is seen as a threat to that relation, and ultimately to the self-esteem of the consumer. Therefore owners, who tend to have high self-brand relations, respond more negatively than non-owners to changes drawn across this line. When the image of the extension is not congruent with the parent brand on which the self-brand relations are based, owners tend to respond more negatively than non-owners. So owners are more likely than non-owners to have lower evaluations of an extension with low image fit. This thesis will try to gather further proof for the described ownership effect found by Kirmani et al. by using another (prestigious) brand from a different product category. So, the first hypothesis of this thesis is as follows:
H1: The effect of image fit on vertical line extension evaluation is moderated by ownership, such that high image fit leads to higher evaluations for owners compared to non-owners, and low image fit leads to lower evaluations for owners compared to non-owners.
2.7 Ownership and product category extension
Up till now there has been no research done towards ownership in the context of a product category extension. It is plausible that different effects will be at play in this situation, which will be explained in this paragraph. The price and quality level of a vertical line extension always differs from the original product, this often results in low image fit. This will be especially the case when a luxurious or a prestigious brand introduces a cheaper version of their products, since the high status associated with the brand will be affected. However the
21
image of a product category extension is often similar to that of the original brand, with the only change being that the new product is introduced in a different product category than the one the brand was originally operating in. For example, Louis Vutton started to sell
sunglasses, which were very different products in comparison to the bags and other products they already sold. Yet the image of the sunglasses and the other products was comparable, because all were luxurious and exclusive. When the image remains the same for a product category extension, it is likely that the brand relations are not threatened, since the self-brand relations are based on the image of the self-brand. So it is likely that for a product category extension where the image remains the same, owners will respond different than with a vertical line extension where the self-brand relations are threatened.
For a high fit image and high fit product category extension evaluations will
presumably be higher for owners than for non-owners. Since owners have stronger self-brand relations, they value the brand more than non-owners. In the case of a high image and product fit extension this affection is likely to be transported from the brand towards the extension, since high fit heightens the transfer from brand associations towards the extension. Even though the extension belongs to a different product category, because of the similarities in image and product features and functionality, owners are likely to have higher evaluations of the product category extension than non-owners.
In the case of a low fit product category extension owner evaluations are also likely to be higher than those of non-owners, however the difference is less pronounced. As mentioned earlier, since the image remains the same the self-brand relations are not threatened for
owners. Presumably another process takes place, which will be described below.
Evaluations of low fit product category extensions are typically negative, for the reason that consumers don’t believe the company has the capabilities to successfully make such a different product. Yet for owners evaluations may not be as negative as for
non-22
owners. Motivated reasoning theory provides the explanation for these results. Research has shown that consumers who strongly identify themselves with an
organization tend be less sensitive towards negative information concerning that organization than consumers who don’t have strong identifications (Ahluwalia et al., 2000; Einwiller et al., 2006). Since owners tend to be more involved and to identify themselves more strongly with the brand than non-owners (Kirmani 1999), it is likely that they also will have a tendency to be less sensitive towards negative information surrounding the brand. The mechanism that underlies this phenomenon can be explained by the ‘negativity effect’, which has been studied by numerous scholars in the field of psychology.
The results from different articles drawn from the impression formation literature demonstrate that people perceive negative information as more diagnostic or informative than positive (e.g., Fiske 1980; Klein 1996; Skowronski and Carlston 1989). According to this theory people simply see negative information as more useful in comparison to positive. However this ‘negativity effect’ is absent when people strongly identify with an organization and they are confronted with negative information in relation to that organization.
Furthermore, consumers with strong identification come up with significantly more counterarguments in response to negative information, so that they will be able to support more positive views. They reason in a biased manner, they don’t take information at face value but gather counterarguments to arrive at a desired conclusion, an example of defensive information processing. The reason that consumers with a high level of brand identification act this way, is that they want to protect their existing beliefs, and don’t want to alter them. Through the use of biased reasoning they can hold on to their existing attitudes (Einwiller et al. 2006).
However Einwiller et al. (2006) found that even though consumers with strong organization identification are less sensitive towards negativity surrounding their favourite
23
organization, their evaluations still drop after exposure to negative information, so they are not immune. Nevertheless, their evaluation of the organization still remains higher than that of consumers who don’t identify themselves with the organization, although the difference between the groups becomes smaller.
So when a product category extension with low fit is introduced, evaluations will be negative. Yet because of motivated reasoning theory, owners will evaluate the low fit product category extension higher than non-owners, since they are less susceptible towards the
negativity surrounding a low fit extension. Presumably they will act this way because with biased reasoning, they can protect their existing attitudes towards the brand and don’t have to alter them. Since they strongly identify with the brand, they are motivated to keep their positive feelings towards the brand intact. So the second hypothesis for this thesis is as follows:
H2: The effect of product fit on product category extension evaluation is moderated by ownership, such that high product fit leads to higher evaluations for owners compared to non-owners; low product fit also leads to higher evaluations for owners compared to non-owners, however the difference is less pronounced at low product fit compared to high product fit.
Motivated reasoning is not likely to be present within the context of the earlier described vertical line extension. Since the vertical line extension is so different from the original brand in terms of the image, owners are not likely to engage in motivated reasoning. Owners form self-brand relations around the image of the brand, however when the image changes so much, there is nothing for the owners left to defend. Why would an owner use biased
24
results from Einwiller et al. (2006), who found that with extreme negative information biased reasoning ceased to exist for people who strongly identified with the organization or brand. So in the case of a vertical line extension motivated reasoning is not applicable, and only the self-brand relation theory, as described in earlier paragraphs, is likely to be relevant.
2.8 Conceptual framework and hypotheses
Figure 1 provides an overview of the different independent variables, moderators and dependent variables used in this research. As can be seen there are two different models, for the first one image is the independent variable and vertical extension evaluation the dependent variable, for the second model product fit is the independent variable and product category extension evaluation the dependent. In both cases ownership is the moderator.
Figure 1: conceptual model
Image fit Vertical extension
evaluation Ownership
Product fit Product category
extension evaluation
25 The hypotheses used for this thesis are the following:
H1: The effect of image fit on vertical line extension evaluation is moderated by ownership, such that high image fit leads to higher evaluations for owners compared to non-owners, and low image fit leads to lower evaluations for owners compared to non-owners.
H2: The effect of product fit on product category extension evaluation is moderated by ownership, such that high product fit leads to higher evaluations for owners compared to non-owners; low product fit also leads to higher evaluations for owners compared to non-owners, however the difference is less pronounced at low product fit compared to high product fit.
3 Methodology
The aim of this research is to examine the effect of image fit on the evaluation of a vertical line extension, the effect of product fit on the evaluation of a product category extension, and to evaluate the moderating role of ownership on these effects. There will be two separate 2-2 models, one will be used to examine ownership and image fit within the context of a vertical line extension, the other will be used to asses ownership and product fit in the context of a product category extension. For both independent variables, namely fit and ownership, there will be two levels. In the case of fit there will be a high and low fit scenario, for ownership there will be a group of owners and non-owners. The methodology section consists of three parts. In the first part the different pre-tests will be discussed. In the second part the
experimental design will be described. Lastly, the different measures that were used in this study are explained.
26
3.1.1 Pre-test 1: Brand selection
The choice for the brand that will be used in this thesis was dependent on a number of factors. Firstly, since the role of ownership is investigated in this thesis, the brand should be real and not be a fictional one, for the reason that owners of a fictional brand don’t exist. It is possible to let people imagine that they possess a hypothetical brand, however with this approach it is doubtful if the results are reliable, since chances are that people will react differently than when a real brand is used.
Secondly, in order to answer the research question the chosen brand has to be a prestigious one. Based on the research performed by Kirmani et al. it is clear that in order to create an extension that has an strong misfit on the level of the image, which is necessary to test the proposed hypothesis, a prestigious brand has to be used. Reason being that a step-down extension is incongruent with the prestige of a prestigious brand, which is likely to threaten the self-brand relations of owners, resulting in lower evaluations for them compared to non-owners in a low image fit situation.
Thirdly, it should ideally be a brand with owners who can be found and approached. Since there are time and monetary constraints tied to this research project, the brand of choice should be one that a lot of people possess, making it is possible to find owners.
Based on the aforementioned criteria the brand Apple was picked as a suitable brand. Apple is a real brand that has a prestigious image making it a suitable brand for this study. Furthermore, a lot of people own Apple products, making owners of the brand easy to find. To asses if Apple really can be classified as a prestigious brand a pre-test was taken by 30 people between the ages of 18 and 25. Perceived prestige was measured using a two item 7-point scale taken from Kirmani et al (1999). The two items were: Apple is prestigious, Apple has high status (1=strongly disagree, 7=strongly agree). Results show that the mean for Apple regarding this measurement was 6.03 (0,67). Thus, Apple scores high on prestige and
27
therefore can be regarded as a prestigious brand, and is suited for use in this study. However, it has to be noted that Apple is a brand that already has numerous extensions. Research demonstrates that consumers react differently when a brand with
numerous extensions extends even further, then when a brand without extensions introduces a new product. In most cases the consumer reacts more positively towards an extension in the case of an already extended brand than towards an un-extended brand. The reason being that they are already familiar with the brand having extensions, so a new extension doesn’t seem odd. Thus, for brands that have never extended before, results could be different. Further research should asses if indeed outcomes would be different using these type of brands.
The reason that the author of this study still chooses to use Apple as a brand is because for brands that have never extended before, there is also a possibility that consumers won’t like their extensions at all, even the high fit ones. These brands are often so tied to their product category, that extensions are not well received. Especially for owners this could be the case since they have stronger relations with the existing brand. For example, for a brand like Rolex people could judge extensions as odd, since Rolex has such strong associations with the product category of a watch. So in order to test the hypothesis of this study, the author deems Apple to be appropriate. Furthermore, as already mentioned, practical reasons also led to the decision to choose Apple, since Apple is one of the few brands that is
prestigious and still has a lot of easy to find and approachable owners.
3.1.2 Pre-test 2: Apple quality and extensions
The second pre-test consisted of an interview with 17 respondents. The goal of the second pre-test was firstly to assess what the respondents perceived to be the most important qualities of the brand, so that the quality of the vertical line extension could be manipulated
28
of the brand Apple, and explain why they thought these were the most important. Secondly, respondents were asked their opinions on possible extensions that could be used for the product category extension, furthermore they were asked to explain why they thought these extensions would be suitable. The most mentioned qualities and extensions are summarized in table 1.
Table 1 : Apple quality/extensions
Apple quality Apple extensions
Fashionable design Smart tv
User-friendly Gaming console
Smart/hi-tech Innovative home appliances
Innovative Self-driving car
3.1.3 Pre-test 3: extensions
The purpose of this pre-test was to find appropriate extensions to use for the experiment design. Both for the product category extension and the vertical line extension a high fit and a low fit extension needed to be found. Based on the propositions of the respondents of pre-test 2 and further contemplation by the author of this thesis, 4 extensions were chosen for the product category extension to asses in pre-test 2. Two of them were thought to represent a high fit extension because they shared a similar product category with Apple, and two a low extension because the product category differed strongly from the ones of Apple. All 4 shared a high-tech, innovative image, therefore the image fit was thought to be the same.
Furthermore, the information gathered about the quality during pre-test 2 was used to manipulate the quality and price for 4 different vertical extensions of the iPhone 8. In this case two were thought to be a high fit extension and two a low fit. Prices were based on current market information, so that the prices of the high fit represented a high-end phone and that of
29
the low fit a low-end phone. Moreover, text was included to describe that the low fit
extensions scored lower and the high fit extensions similar on important aspects of the iPhone, like user-friendliness. 30 people were asked to judge the eight extensions based on
measurement scales (table 2) developed by Aaker and Keller (1990) and Bhat and Reddy (2001). The goal was to test if indeed they were seen as high fit and low fit, and which of the four extensions were most appropriate to use in the research design. For the product category extension product fit was measured. Also image fit was taken into account, this needed to be high for the extensions that would be ultimately chosen, to make sure that the measured effects were caused by product fit and not image fit. For the vertical line extension only image fit was measured.
Table 2: Pre-test measures
Product fit (Aaker and Keller, 1990)
(1=strongly disagree, 7=strongly agree)
1 (complement): The extension product could be used together with the original products of the brand to satisfy the same need.
2 (substitute): The extension product could function as a substitute for the original products of the brand.
3 (transfer): The manufacturer of the parent brand is capable to successfully make the extension product.
Image fit (Bhat and Reddy 2001)
(1:strongly disagree, 7=strongly agree)
1: The extension fits with my idea and image of the brand.
30
In table 3 the results are summarized. For the product category extension the results show that indeed the two high fit extensions (smart tv and gaming console) score high on product and image fit. Also the low fit extensions gave the expected results, scoring low on product but high on image fit. The only exception is that the self-driving car scores below average on image fit. For the vertical line extension the outcomes too were in line with expectations. The first and second scored high on image fit, while the other two scored lower.
Eventually, for the high fit product category extension the smart tv was chosen, since a gaming console is a product for which evaluations are very diverse. Gaming consoles
typically are devices which some people are very fond of, while others are not at all interested in using them. This could problematically influence the homogeneity of variance. For the low fit the smart robotic vacuum cleaner was chosen since the image fit for the vacuum cleaner was more similar to the smart tv than the self-driving car. Also the difference between high and low fit was approximately the same for the vertical and the product category extension, making an comparison possible between the two.
For the high fit vertical extension the second option was chosen, since the first one is very similar to the existing iPhone 8 Plus and for this thesis hypothetical brand extensions are used. Using an already existing extension or something similar to it could give different results since consumers are already familiar with it, which can alter their opinions. In the case of the smart tv, Apple already has a product that is referred to as a ‘tv’, however this product resembles a dream-box which can be used to stream music and movies, but is dissimilar to a smart tv since it doesn’t have a screen, it doesn’t include audio equipment nor can it
broadcast a large selection of television channels. For the low fit extension option 4 was chosen because it scored comparable with the low fit product category extension making a comparison more appropriate.
31 Table 3: means pre-test
Extensions Product fit Image fit
Smart tv 5,87 5,93
Gaming console 5,96 5,67
Smart robotic vacuum cleaner: cleans the floor without
human input
1,94 4,78
Self-driving car 1,58 3,67
Vertical line extension 1: iPhone pro
- 6,07
Vertical line extension 2: iPhone light
- 5,76
Vertical line extension 3: iPhone basic
- 2,53
Vertical line extension 4: iPhone simple
- 1,85
3.2 Experimental design
This study consists of two independent 2-2 between-subjects experimental designs. The goal of this thesis is to assess how ownership moderates the effect of image fit on vertical line extension evaluation and how ownership moderates the effect of product fit on product
category extension evaluation. The independent variables in the designs are ownership, for the product category extension model product fit and for the vertical line extension model image fit. The dependent variable are product category extension evaluation and vertical line extension evaluation.
32
Since the author of this study was interested in how a vertical line context would differ from a product category extension a separation of the models seemed appropriate. Another reason why two separate models were chosen, is because a different independent variable was used for each model, namely image fit and product fit, which made combining the two types of extension in one model difficult. However, the downside is that since the two types of extension were examined in a different model, making a comparison between the two is also more difficult.
Since this study consists of two independent 2-2 experimental designs, 4 different conditions are needed for each design (see table 4 and 5). The product category extension design needs two different products of which one scores high on product fit and one low. As mentioned in the pre-test smart tv and smart robotic vacuum cleaner were selected. For the vertical line extension a high image fit (iPhone light) and low image fit product (iPhone basic) were selected, also based on the results of the pre-test. For both the category extension and the vertical extension the same respondents are used. This has the benefit that differences
between the models are more likely caused by the different conditions, than behavioural differences between the participants. However, a disadvantage is that participating in both experiments may alter the results. When someone has to evaluate both extensions it is possible that results will differ from the situation where only one extension is evaluated. Although the order in which respondents received the two extensions was randomized, making biased results less plausible, there still is a chance that results would be different in a situation where for both models different respondents were used.
For this thesis an online experiment/questionnaire was used. Typically experiments are divided in lab experiments and field experiments. Lab experiments have the advantage of high internal validity, because there is extensive control over the variables and the research
33
reason that the study is being taken in a real-life setting. An online experiment/questionnaire doesn’t possess these advantages in the same degree as the other two formats. However, it is cheaper and more efficient, so taking the monetary and time limits for this study into account, this was the best possible option. Furthermore, often in the extension literature a questionnaire is used, making it a justified approach.
Respondents were partly self-selected and partly approached in the library of a University. So the sample used for this thesis consists of students from the Netherlands. A homogenous sample was chosen because it has the benefit of raising internal validity, since there is less chance that other factors influence the outcome of the study. However the
downside is that it lowers the external validity, it is not sure whether the same results hold for the general population or whether these results are typical for a student population only.
Table 4: Product category extension
Owner – high fit (Smart tv) Non owner – high fit (Smart tv) Owner – low fit (Vacuum cleaner) Non owner – low fit (Vacuum cleaner)
Table 5: Vertical line extension
Owner – high fit (iPhone light) Non owner – high fit (iPhone light) Owner – low fit (iPhone simple) Non owner – low fit (iPhone simple)
3.3 Measures
The survey used in this study used a number of different measures. The survey started off with a short introduction, describing the goal of the study, mentioning the time needed to complete the questionnaire, and thanking the respondent for his or her participation. After that two different blocks of questions followed, one in which respondents gave their opinion on the vertical line extension, in the other the opinion about the product category extension was
34
asked. Both of the blocks were accompanied by a description explaining to the respondent what they were doing. The respondents were randomly assigned the high or low fit scenario in both cases. Also the order in which the respondents received the vertical and product category blocks were randomized. The measures that were used are described in table 6. For both the vertical line extension and the product category extension the same measure was used to asses extension evaluation: since both constructs are similar the same measure could be used. This measure has been used before, in different research, to asses both vertical line extensions and product category extensions.
After this a block of different questions followed. Respondents were asked if they owned any Apple products to determine ownership, the other independent variable in the research. Also, some control variables were included regarding gender, age, occupation and income level. No measurement checks were included. With the pre-test it was already established that the manipulations were successful. Furthermore, including measurement checks in the experimental design of the main study can alter the results. It is possible that people judge the extensions differently when they are first confronted with the manipulation checks, when the order is reversed this could also have an effect on the outcomes (Hauser et al., 2018). However a disadvantage of not including manipulation checks is that respondents of the main study could have different opinions than the respondents of the pre-test.
Table 6: Measures experiment design
Extension evaluation (Riley et al., 2015)
(1=strongly disagree, 7=strongly agree) 1: I like the extension.
2: The extension is favourable. 3: The extension is appealing.
35
4 Results
In the results section an analysis of the gathered data will be provided. First the responses are described. Following that an overview is provided of the different steps that needed to be taken to prepare the data for analysis. Thereafter, the actual analysis takes place and the hypotheses will be tested.
4.1 Response analysis
As mentioned before, respondents were partially self-selected with the other part being
random strangers approached in a University library. 127 responses were collected, however 7 were not finished, making the total of surveys completed 120. The responses were almost evenly distributed amongst the conditions. For the product category extension; the high fit (smart tv) – owner, low-fit (vacuum cleaner) – owner, high fit (smart tv) – non-owner and low fit (vacuum cleaner) – non-owner all received 30 responses. For the vertical line extension; owner – high fit (iPhone light) 30, owner – low fit (iPhone simple) 30, non-owner and low fit (iPhone simple) 29, non-owner high fit (iPhone light) 31. Respondents were randomly
assigned to each condition. However, at the end of the data collection some respondents were selected on purpose for certain conditions to make sure every condition had the same number of respondents. For an ANOVA it is important to have the same group sizes, because
otherwise the homogeneity of variance could be affected. However, a disadvantage is that it makes it a partially quasi-experiment, which lowers the internal validity because not every respondent was randomly assigned to a condition.
Of the sample, which was used both in the vertical line extension model and the product category model, 66 persons were males (55%), 45 were female (45%). 114 people (95%) had an age between 18 and 29, the rest belonged to other age groups. Furthermore, 116
36
(96,7 %) were students, and 115 (95,8%) belonged to a low income group. So, the sample is representative of a student population.
4.2 Data preparation
In order to analyse the data, several steps had to be taken. First a frequency check was performed to examine if there were any missing data, however this was not the case. Second, the three questions that were used to assess the product category extension evaluation, had to be transformed into one variable. The same applies for the vertical line extension evaluation. There were no counter-indicative questions, so the variables didn’t have to be recoded. Furthermore, respondents saw only one scenario for both the product fit and image fit. For example high product fit and high image fit, and not low product fit and low image fit. For the scenario they saw they received a 1 in SPSS; for the scenario they didn’t see no value was given. So, manually a 0 was put into these places. In this way the variable could be analysed, where a 1 represented a high fit and a 0 a low fit.
4.3 Pre-analysis
Since the independent variables used in this study are categorical and the dependent variable is continuous, a factorial ANOVA is used to analyse the data. For an ANOVA there are several assumptions that need to be fulfilled. Firstly, the data have to be normally distributed. Secondly, there should be homogeneity of variances. Finally, there should be independence of observations. According to the central limit theorem when a sample is equal to or greater than 30 a normal distribution can be expected. Since the sample used in this thesis is greater than 30 the first assumption of ANOVA is likely to be fulfilled. In order to be sure a histogram was drawn in SPSS for both the product category extension model and the vertical line extension
37
model which had the shape of a bell curve, providing additional evidence that indeed there the data are normally distributed. The second assumption, homogeneity of variance, is assessed in the next subchapter. Independence of observations can also be assumed, since for each of the two models used all the four conditions consisted of different respondents.
Reliability checks demonstrated that the measurement items were consistent. There was only one question used that consisted of numerous items. However, it was used in both the model for the product category extension as that for the vertical line extension. For the first Cronbach’s alpha was 0,94, thus the scale has high reliability. The corrected-item total correlation is higher than 0,3 for all items, this indicates that all items have a good correlation with the total score of the scale. Furthermore, deleting any of the items doesn’t raise
Cronbach’s alpha, so the reliability would not be improved. For the vertical extension the Cronbach’s alpha was 0.96. In this case too all the items had a good correlation with the total score of the scale, and deleting any items doesn’t improve Cronbach’s alpha.
4.4 Hypotheses testing
Before the hypotheses could be tested, it was first necessary to run Levene’s test, to examine the homogeneity of variance, one of the assumptions for a factorial ANOVA. For the product category extension model the test was not significant (p = 0,051 > 0,05), so homogeneity of variances could be assumed. For the vertical line extension model, the p-value was 0,268 > 0,05 so again homogeneity of variances could be assumed.
First of all the vertical line extension model was analysed (table 7 and 9). The results show a significant main effect of image fit on vertical line extension evaluations, F(1, 116) = 129.56, p < 0.05, η2 = 0.53. With the mean for low fit 3.03 and for high fit 5.09. There was no significant main effect of ownership on vertical line extension evaluations, F(1, 116) = 1, p = 0.32, η2 = 0.01. However, there was a significant interaction effect between ownership and
38
image fit on vertical line extension evaluation, F (1, 116) = 27.49, p < 0.05, η2 = 0.19. As can be seen from figure 1, owners indeed have higher vertical line extension evaluations in a high image fit situation, but lower ones in a low image fit situation. A simple main effect analysis shows that the mean difference between owners and non-owners in a high image fit situation is 0.77 and significant with p < 0.05. For the low image fit situation the mean difference between owners and non-owners is 1.13 with p < 0.05, so also significant. Thus, hypothesis 1 is supported. Additionally figure 1 shows that for both owners and non-owners evaluations are higher in a high fit situation compared to a low fit situation.
Next the product category model was analysed (table 8 and 10). There was a significant main effect of product fit on product category extension evaluation, F(1, 116) = 91.53, p < 0.05, η2 = 0,44. With the mean for low fit 3.66 and for high fit 5.34. There was also a significant main effect of ownership on product category extension evaluation F(1,116) = 23.95, p < 0.05, η2 = 0.17. With the mean for owners 4.93 and for non-owners 4.07.
Furthermore, there was a significant interaction effect of product fit and ownership on product category extension evaluation F(1, 116) = 3.96, p < 0.05, η2 = 0.03. Figure 2 illustrates that owners have higher product category extension evaluations than non-owners in a high product fit situation, however this difference is less pronounced in a low product fit situation. A simple main effect analysis demonstrates that the mean difference between owners and non-owners in a high product fit situation is 1.21, p < 0.05, thus significant. In the low product fit situation the mean difference is 0.51, with p <0.05, thus also significant. So, hypothesis 2 is supported. Additionally, figure 2 shows that both for owners and non-owners evaluations are higher for high product fit compared to low product fit.
39 Figure 1: Vertical line extension evaluation
40 Table 7: means vertical line extension
Mean (SD) N
Owner – high fit 5,48 (0,93) 30
Owner – low fit 2,46 (0,95) 30
Non-owner – high fit 4,71 (0,89) 31
Non-owner – low fit 3,60 (1,17) 29
Table 8: means product category extension
Mean (SD) N
Owner – high fit 5,94 (0,75) 30
Owner – low fit 3,91 (1,15) 30
Non-owner – high fit 4,73 (0,92) 30
Non-owner – low fit 3,40 (0,98) 30
Table 9: factorial ANOVA output vertical line extension
SS DF MS F η2 Sig. ownership 0,99 1 0,99 1 0,01 0,32 Fit 127,43 1 127,43 129,56 0,53 0,00 Ownership*fit 27,04 1 27,04 27,49 0,19 0,00 Error 114,09 116 0,98 Total 2259,78 120
41
Table 10: factorial ANOVA output product category extension
SS DF MS F η2 Sig. ownership 22,25 1 22,45 23,95 0,17 0,00 Fit 85,01 1 91,53 91,53 0,44 0.00 Ownership*fit 3,68 1 3,68 3,96 0,03 0,049 Error 107,74 116 0,93 Total 2645,67 120
5 Discussion
This section of the thesis is used to put the results in the right perspective. It starts with a summary of the goal of this research. After that a thorough discussion of the results is provided. The theoretical and managerial implications will be addressed. Finally, the limitations of this study and the options for future research will be discussed.
5.1 Summary
Goal of this research design is to assess how ownership moderates the effect of image fit on vertical extension evaluation and the effect of product fit on product category extension evaluation. It is proposed that for a vertical line extension in the case of high image fit owners will have higher evaluations than non-owners, however in the low image fit context owners will have lower evaluations than non-owners. Self-brand relations are put forward as an explanation. Owners are likely to have stronger self-brand relations and this makes them evaluate a high image fit extension more. However, when the image is incongruent in a low image fit situation, their self-brand relations are threatened, resulting in lower evaluations for owners compared to non-owners.
42
However, it is proposed that in a product category extension context different results will appear. In a high product fit situation owners will have higher evaluations than non-owners, but in a low product fit situation the difference will be less pronounced. Motivation theory is used as an explanation. Owners will engage in defensive reasoning, to keep their existing opinions about the brand intact. So owners are less sensitive to the negativity associated with a low product fit extension. However, they are not entirely immune, so the difference in evaluations between owners and non-owners is less pronounced in a low fit situation.
5.2 Discussion
In this paragraph the results of this study will be placed in perspective, starting with a thorough discussion of the vertical line extension analysis and ending with an assessment of the results of the product category extension analysis.
The results of this study add to the existing scientific knowledge in numerous ways. Firstly, the vertical line extension literature is relatively thin compared to the subject of product category extension. Especially the topic of fit, one of the most commonly researched subjects in the product category extension literature, has received a low amount of attention.. The little research that has been done on this topic is not clear-cut. Since a vertical line extension is launched in the same product category as the original product, fit in this regard refers only to image fit. For a vertical line extension the image fit is always based on the similarity of the extension and the original product in terms of price and quality. Some studies showed that high image fit, so an extension with a comparable price and quality level, led to higher evaluations compared to low image fit (Desay and Hoyer 1993). However, other studies gave opposite results (Liu, 2006; Musante, 2007). The results of this thesis show that fit also plays an important role in a vertical line extension, and that this topic deserves more attention from scholars. Furthermore, this study shows that a high image fit vertical line
