24/03/2017 (final)
University of Amsterdam
Faculty of Economics and BusinessMsc. Business Administration - Marketing Track Supervisor: dhr. R.E.W. (Roger) Pruppers
Student name: Vladislav Florea Student number: 11206799
Managing the feedback effect in co-branding
initiatives: a multiple co-branding perspective.
Statement of originality
This document is written by Student Florea Vladislav 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
Table of Contents Page Title Page Table of Contents Abstract 1. Introduction 1 2. Main Body 10 2.1 Brand Alliances 11
2.1.2 Advantages and Disadvantages of Brand Alliances 12
2.2 Customer Perception 15
2.2.1 Consumer Based Brand Equity 15
2.2.2 Customer Brand Knowledge 18
2.2.3 Mapping Consumer Brand Associations 19
2.3 Associations in Brand Alliances 21
2.3.1 Associations in Brand Extensions and Feedback Effect 23
2.4 Brand Fit 26
3. Conceptual Model and Hypothesis development 29
3.1 Conceptual Model 29
3.2 Hypothesis Development 32
4.1 Sample 38
4.2 Measures 39
5. Data and Analysis 40
5.1 Brand Associations 40 5.2 Data Analysis 43 5.2.1 Results part 1 50 5.2.2 Data analysis 2 52 5.2.3 Results part 2 61 6. Discussion 62 7. Conclusion 66 7.1 Theoretical Contribution 68 7.2 Managerial Contribution 70
7.3 Limitations and future study 71
8. Bibliography 74
Appendix A. Survey for low fit co-branding (Tesla & Toyota) Appendix B. Survey questions related to Toyota
Appendix C. Survey questions related to Apple
Appendix D. Tesla and Apple co-branding initiative treatment
Abstract: The following thesis is set to explore multiple co-branding initiative as means of reducing
the negative spillover effect from a bad co-branding initiative. The topic is studied through the
perspective of customer brand associations, perception of fit, and brand equity. Consequently, the
current thesis draws attention to brand alliances, customer based brand equity, brand concept mapping
and brand fit. The methodology used in this study was a mixed approach which combined establishing
brand associations by the means of qualitative research with a study group and an experimental design
in the form of questionnaire. The results support a number of already existing research about the
importance of brand fit, but also suggests that brand attitude plays a role in perceived brand fit.
Furthermore, the research argues that a good co-branding initiative can combat negative spillover
effect from a bad co-branding initiative, presumably through reassuring main brand’s strong positive
associations.
Finally, recommendations for future research are proposed to account for some of the limitations
1.INTRODUCTION
1.1 The Golden Age of Brands
The presence and the role of brands in the society became more accentuated, thus at the moment we are in, as Fournier had put it, “Golden Age of Brands”. (Fournier et al., 2007, p. 781). The fierce competition and globalization has led to development of industries in which branding plays a significant role (Burnaz & Bilgin, 2011), and companies aim to improve their brand equity and as a result their position in the market (Ahn, Kim, Forney, 2009).
In search for added value, companies often use various branding strategies which involve collaboration with other brands. Therefore, by leveraging brand equity companies appeal to branding strategies like co-branding in order to overcome some of the hurdles on the company’s way (Desai & Keller, 2002) for example through acquiring a greater exposure, decreasing cost of promotional activities (Spethmann & Benezra, 1994), or make changes in brand’s associative network (Broniarczyk & Alba, 1994). As a result, this collaboration between brands changes some elements in the relationship between the brand and the consumer.
For example, brands like Beats and Dolby Audio provided numerous partnerships to various companies and products, from smartphones to laptops (Jennings, 2011; Gladstone, 2012), such as “HTC Sensation XE” which is advertised together with Beats by Dre sound technology (Westaway, 2011) and Lenovo laptops advertised as having Dolby Audio systems
inside (Dolby, 2017).
Consequently, such brand would be acting as a testament for the consumer of quality of a component or the entire product, as a symbol of a certain brand image and as a marketing ploy (AMA, 2007; Kapferer, 1992).
Spethmann and Benezra mention that because there is an increase in the importance of branding, companies actively search for opportunities to collaborate with other brands, thus trying to increase the strength of their own brand (1994). Even in the 90’s the brand
collaborative activities had achieved 40% growth per year (Spethmann & Benezra, 1994).
The existence of collaborations between brands imply that there might be a significant chance of mutual benefit for both companies, at least at conception of this partnership. Indeed, studies confirm that Brand Alliances, like Cause-Brand Alliances, could result in an
enhancement of favourable attitudes toward the brands (Lafferty, Goldsmith & Hult, 2004). Additionally, partnering with a local well known brand could help companies enter foreign markets (Voss & Tansuhaj ,1999). Thus, the plan behind Brand Alliances is based in hopes of spillover effect from the partner brand. Certainly, it was proven that pairing with a high customer based equity brand, results in increased customer based brand equity of the partner as well (Washburn, Till & Priluck, 2004). Furthermore, the spillover could be of exposure, consumers, but most interestingly of associations (Keller, 2013). More specifically, Keller states that when two brands are promoting their brand alliance, their brand nodes are becoming associated together in the minds of the consumer, thus creating secondary associations for each brand (1993).
However, many scholars worn about the potential consequences of having a brand alliance which is perceived negatively by the the consumers, or creating associations with a negatively perceived brand (Keller, 2013), or even having a partner which engages in negative
behaviour (Votola & Unnava, 2006). Moreover, scholars point out that the attitude towards the co-branding and its consequential spillover effect is highly dependant on how favourably the co-branding partner is perceived by the consumers (Suh & Park, 2009).
Consequently, a company might want to engage in co-branding partnership for various lucrative benefits, however there might be a significant risk of a negative feedback effect on its brand. As for example, Lego and Shell had set a partnership in 2011, which was followed by a backlash from companies like Greenpeace (Maxwell, 2014). Similarly, BenQ had a negative feedback effect from a brand alliance with Siemens, which was struggling at the time (Bluemelhuber, Carter & Lambe, 2007).
Hence, it would be interesting to explore if there is a way to engage in a partnership and combat the potential negative effect emerging from the alliance through an additional partner brand.
1.2 The Research Gap and Problem Definition
Though clearly, the positive effects of brand alliances have been researched at quite a great extent, not all circumstances receive a great deal of attention. As such, studies on brand alliance have not studied very meticulously the negative effect of co-branding, mostly ignoring brand associations when talking about fit between the brands (ex.: Simonin and Ruth, 1998). On the other hand, many scholars in the field of brand extension have argued
strongly for the importance of brand associations as determinant of fit (Keller, 1993; Keller, 2013; Schnittka, Sattler & Zenker, 2012).
Moreover, as suggested in the article by Simonin and Ruth (1998), it would be compelling to analyse the feedback effect in the case of multiple co-branding initiatives. As a matter of fact, multiple co-branding initiatives are quite prominent in the “Golden Age of Brands”, with smartphones like HTC co-branding with brands like Beats by Dre and Android (Westaway, 2011), and laptops like Medion co-branding with Dolby Audio and NVIDIA (Medion, 2017).
In the case of multiple co-branding initiative, the brands which fit better might have a
different impact strength on the feedback effect, than the worst fitting brands. Consequently, a multiple co-branding initiative might help cancel some of the negative feedback effect
resulted from being associated with one of the brands involved in that multiple co-branding.
From the perspective of brand associations this thesis speculates that co-branding with a brand with antonymous associations, would result in a negative feedback effect, as seen in the case of poor image fit co-branding. Furthermore, the aforementioned effect could have been reduced through having an additional alliance with a brand whose associative network mostly reassures and/or compliments the salient main brand associations.
For example, would Beats damage its brand by partnering with a brand associated with cheap technology? On the other hand, would Beats by Dre also partnering with Sennheiser reassure consumers of their audio system quality, thus canceling the negative feedback effect?
As an example, the research done by Milberg, Park and McCarthy (1997) is closest to the literature gap the current thesis is set to explore. In the study of Milberg, Park and McCarthy (1997), the scholars analysed the negative feedback effect when engaging in brand extension, and ways in which such an effect could be mitigated. The scholars also mention the
importance of measuring brand attitudes and belief, and doing so prior to measuring other variables.
The thesis is set to research the impact of alliance initiatives on main brand image. The underlying assumption is that brand associations fit can determine the success of brand alliances and that the negative effect can be mitigated through having another alliance initiative.
Therefore, the problem statement in this case could be formulated as such: Is there a negative feedback effect on the main brand when it engages in a poor fit co-branding initiative? Can the potential negative feedback effect from a poor-fit co-branding initiative be mitigated through engaging simultaneously in a high-fit co-branding initiative?
Consequently, this effect must be researched and analysed through corresponding literature. More specifically, the research will investigate if the above-mentioned effect might be driven by compatibility of associations between brand maps, fit between companies and brand equity.
In order to tackle the aforementioned problem statement there are certain literature fields which needs to be addressed and researched first.
First of all, one should establish the basis of brand alliance within the thesis framework. The next step would be to present some of the current findings related to the effect of brand
alliances on the main brand, the partner brand and brand alliance itself. In addition, customer based brand equity has to be discussed, since the customer plays a central role in the
determination of what effect the brand alliances have. Furthermore, since associative networks play a significant role in the mind of the consumer, theories which focus on the cognitive aspect in branding and associations have to be explored and applied.
Finally, since arguably the extent of overlap between brand networks depends on their fit, one should take into account the brand fit between the companies, thus Brand Fit literature would also be addressed.
The sub-questions therefore are as follows:
- What are the advantages and risks when engaging in a brand alliance ?
- What is customer-based brand equity and what role does it play in consumer’s attitude towards brands ?
- What role do brand associations play in formation of consumer’s brand equity and attitude towards the brand ? How to establish those brand associations ?
- What role do brand associations play in brand alliances ? How important is the fit in such marketing activities ?
Though this thesis will mention literature related to brand extensions and other types of branding strategies, it will use them only to explore their effect on the associative network of the consumer. Thus, the primary focus of this research will be brand alliances, more
specifically ingredient based brand alliances.
Moreover, though the associative network will be one of the most important aspect of this research, it will only be addressed within the context of business related studies, not as much psychological studies. Indeed, the associative network will be discussed mostly in order to explain Customer Based Brand Equity, and in order to acquire the means to explain certain research findings.
1.4 Theoretical Contributions
Though there has been devoted a significant number of research in the field of Brand Alliances, most of them look into the added benefit of such collaboration without giving much attention to the limits and negative effect of such branding strategy. Thus, this thesis will explore some of this issues, while also tying them together with the cognitive
associations literature which is often discussed in brand extension literature (Loken & Roedder John, 1993; Keller & Aaker, 1993; as cited by Milberg, Park and McCarthy, 1997). Consequently, this would hopefully result in a deeper understanding of the role of associative networks in the Consumer Based Brand Equity and Brand Alliance Evaluation.
By exploring the topic of Brand Alliance initiative and the aforementioned literature fields through associative networks, this thesis will allow the further elaboration of our
understanding of the relationship between brands and customers. Moreover, the current thesis could potentially bridge the gap between Brand Extension and Brand Alliance literature in
terms of customer cognition, thus giving a new perspective to such branding strategies. Furthermore, by exploring the topic of multiple co-branding initiative new insight might be untailed about how such marketing activities affect consumer perception about a brand. Finally, such exploratory study could enable further research of brands within the cognitive domain and through anti-positivist paradigm.
1.5 Managerial Contributions
Considering the magnitude of the role of Brand Alliances in the Business world, companies invest significantly in creating collaboration which will make sense strategically. In search of best partner brands, the company might overlook some aspects which could potentially hurt its own brand. The overlooked aspects could result in loss of potential revenue and dilution the brand image. Hence, by looking into the aforementioned aspects, the current thesis could potentially help managers determine whether creating a brand alliance would carry some risks of brand image distortion along with the potential benefits. More specifically, this study will help managers understand what role play attitudes towards brands and the associative network of brands in the customer's mind, when deciding to engage in co-branding activities. Thus, if core assets within brand’s associative network and equity would be degraded because of an alliance, this could also damage the brand image. As such, the managers could answer the key question: Could the brand alliance impact the brand image in such a way, as to degrade it and can an additional brand alliance cancel the negative effect ?
Thus, the current study will help answer if there are conditions when partnering with other brands might damage the company’s brand image, what role the core brand associations play, and what could be a solution to such a negative effect..
1.6 OUTLINE
The second chapter of this thesis will explore the Brand Alliance literature and the findings related to the effects of such branding strategy, thus setting up the context of the study. Additionally, it will discuss the Customer perception of brands, the Customer Based Brand Equity and Associative Networks, in order to set up the framework through which the topic will be further investigated. Furthermore, within this chapter Brand Fit literature will be discussed, thus potentially finding a moderator for this study.
After the main body, will follow a methodology chapter which will discuss how the research would be conducted and the theories of research method used. The fourth chapter would be a findings chapter, which will present the key findings related to the topic of the thesis and characteristics of the research itself. Next would follow the fifth chapter, which will discuss the previously mentioned findings through appropriate theories introduced in the main body.
Finally the thesis would contain a conclusion chapter which will sum up the thesis so far and will examine the possible implications of its findings.
2. Main Body
AMA (2007) defined brand as “A name, term, design, symbol, or any other feature that identifies the seller’s good or services as distinct from those of other sellers’”. Brands in this sense are a bundle of multiple associations in our memory about a specific company and its products (Brown, 1992), thus whenever we encounter a brand name it acts as a cue for our memory. Consequently, through the symbolic nature of the brands customers are able to easily identify and distinguish various products (Keller, 1993). Moreover, brands can be perceived as a representation of values and identity of the consumer (Kapferer, 1992).
Therefore, it can be understood that there is an implied contract between the company and the consumer, through which the consumer both buys a product and invests in the image the brand portraits.
Considering that for customers brands are a significant means through which they make buying decisions, branding is utilised to create durable mental structures and associations (Keller, 2013). Despite Brands trying to remain consistent in their image, in search for added value, companies often use various branding strategies which involve manipulation of brands associative network. One of such branding strategies is collaboration with other brands (Keller, 2013; Gwinner & Eaton, 1999). Considering the aforementioned role of brands in consumers’ mind, there surely are certain risks involved when partnering with another brand. As such, one risk could be of no longer corresponding with the image and associations the company once had for the customers. Moreover, the mere act of partnering with another brand
for the betterment of the product, could be communicating some downside of the company. Consequently, this topic will be expanded upon further in this thesis.
2.1 Brand Alliances
Brand managers are consistently on a lookout for new opportunities to increase the Brand’s equity and its competitiveness, and to make use of this opportunities they apply various branding strategies (Ahn, Kim & Forney, 2009).
Companies can leverage brand associations and brand equity through linking a brand with other brands (Desai & Keller, 2002; Keller, 2013). In order to achieve such link, brands are partnering to create a collaborative product and/or are doing marketing in a combined way (Rao, Qu & Ruekert, 1999). The aforementioned type of branding strategy is known as Co-branding, as well as multi-branding, joint branding, brand bundling and brand alliance (Keller, 2013; Besharat, 2010). Moreover, Ingredient Branding can also be included in this category, since it involves a collaboration between a brand of component necessary for a product of another brand (Kotler & Pfoertsch, 2010). Consequently, Brand Alliance can be defined as a combination of two or more brands over a period of time, made through association of some distinguishable assets (Rao and Ruekert, 1994).
Brand Alliances are by no means a new branding strategy nor a rare and unimportant one. For example, Betty Crocker has been creating alliances since 1961, McDonald’s and Disney had a 10 year alliance based on exclusive rights agreement, and over 16 major global Airlines created the Star Alliance (Keller, 2013). Moreover, there are multiple number of jointly branded credit cards and computer software, and such collaborative branding activities
received impressive growth rates (Simonin & Ruth, 1998; Spethmann & Benezra, 1994, as cited in Simonin & Ruth, 1998). Additionally, over the past years the marketplace has seen an increase in collaborations both within and among industries (Ahn, Kim & Forney, 2010).
As a result, the interest to brand alliances has not been omitted by the academic domain as well (Voss & Gammoh, 2004; Uggla, 2004; Simonin & Ruth, 1998; Keller, 2013). Scholars suggest that Brand Alliances work under the assumption that each party has some
complementary assets, attributes or capabilities which in turn would allow the alliance to be perceived positively by the customers (Heslop, Nadeau, O'reilly, & Armenakyan, 2013). Additionally, through such an alliance companies would be able to achieve a great output based on synergy (Rao & Ruekert, 1994). Moreover, scholars noted that Brand Alliances entail not just a mere transfer of brand meaning but a“conjunction and elaboration of meaning from brand to brand with category meanings underneath the involved brands” (Uggla, 2004, p. 106).
2.1.2 Advantages and Disadvantages of Brand Alliances
Brand Alliances could potentially bring certain benefits to partner brands. One of such benefits could be the increased awareness and reach of the brand resulting from access to additional customer base made available through the partner brand (Leuthesser, Kohli, & Suri, 2003). Furthermore, through brand alliances partner brands can gain access to certain assets and capabilities (Heslop, Nadeau, O'reilly, & Armenakyan, 2013) otherwise
Moreover, scholars recognised that Brand Alliances carry a potential to leverage brand equity (Voss & Gammoh, 2004; Washburn, Till & Priluck, 2004), assert some new attributes (Keller, 2003) and improve brand image (Voss & Gammoh, 2004). Furthermore, Uggla (2004) suggests that besides image transfer brand alliances could potentially strengthen associations with specific attributes, benefits, product categories and quality. As such, a company like Lunchables, provider of packed lunches for kids, in order to combat criticism of serving unhealthy food, collaborated with a company which possesses the attribute it needed, in Lunchables’ case - Dole, a fruit company (Schultz, 2011). Moreover, to maintain a more healthy image, in 2016 Lunchables collaborates with Capri Sun, natural juice company (Lunchablesparents, 2016). Consequently, it can be argued that Lunchables was able to increase its brand equity and improve its brand image through using associations and attributes of other brands. Indeed, scholars recognised that brand alliances provide opportunities to alter brand specific associations (Broniarczyk & Alba, 1994; as cited in Simonin & Ruth, 1998).
Therefore, the advantages of Brand Alliances are not only tangible, like gaining access to new expertise, equity and revenue, but can also include the potential of expansion of brand
meaning and access points (Keller, 2013).
Interestingly, co-branding allows for a brand to act as a symbol of quality of a certain part of attribute (Rao & Ruekert, 1994). Relating back to the Lunchables example, the partner brands of fruits and juices in the Launchables communicate a certain guarantee of quality and their responsibility over specific attributes. Nevertheless, the symbol aspect of brands is quite subjective to customers perception. As such, even when consumers are made aware that
branded attribute has no added benefit over a non-branded attribute, they still prefered it over the counterpart (Carpenter, Glazer & Nakamoto, 1994). Thus, it might be the case that customers perceive added benefit from branded attributes which extends past direct benefit from the product itself. Consequently, the above given statement might create an impression that partnering with any brand is by default better than partnering with none. However, there might be some disadvantages implicated in the formation of brand alliances as well.
The Disadvantages of Brand Alliances, respectively, would consist of less control, less brand focus, more risk of dilution of brand equity and so on (Keller, 2013). As stated by Rao and Ruekert (1994), by creating an alliance brands are opening themselves up to potential detrimental effect resulting from a less favourable perception of the brand alliance. Simonin and Ruth (1998) confirm that the spillover effect onto the partner brand is dependant on the evaluation of the brand alliance. Thus, whether the consequential brand attitude will be damaged or improved, and the extent of such effect depend on the way the alliance is evaluated by the consumers and the fit between companies (Simonin & Ruth, 1998). Seeing that when creating a brand alliance companies are giving up some of their control over tangible and intangible assets, like some of the products attributes and the image of the collaborative product, their brands are being exposed to the additional risk mentioned above. Moreover, there are additional potential issues related to brand dilution which could arise (Uggla, 2004). Brand dilution could arise from the poor fit between the companies engaged in the alliance or the link with an extensive number of various brands (Uggla, 2004).
Furthermore, it might be the case that the borrowed associations offer only a short-term leveraging opportunity.
The way in which Brand Alliances influence consumer perception of the alliance, the product, and each brand individually is of great importance. Specially considering that the associations between brands can be transferred by the means of Brand Alliance (Dickinson & Barker, 2006). Moreover, the way in which the co-branding is done to influence customers is also important. For example, the impressions of the co-branded product heavily depend on the header brand (Park, Jun & Shocker, 1996), and though the spillover effect exists, it also depends other factors related to how customers perceive the fit between brands (Simonin & Ruth, 1998). Consequently, the way in which customers perceive and evaluate brands and some of the branding strategies needs to be addressed. Especially considering that
Co-branding is creating associative links between brand related nodes in customers’ minds (Keller, 1993). Thus, the Next chapter will cover Customer Brand Equity, Brand Associative networks and Feedback effect.
2.2 Customer Perception
2.2.1 Consumer Based Brand Equity
As previously stated, brand alliances can influence brand equity (Voss & Gammoh, 2004; Washburn, Till & Priluck, 2004; Keller, 2013). Moreover, arguably all marketing activities are directed at influencing and making use of brand equity (Yoo & Donthu, 2001; Aaker, 1991; Keller, 1993). However, it is important to establish what constitutes brand equity and why it is relevant to the business world.
Aaker (1991) states that brand equity is the combination of all the positive and negative aspects attached to the brand, which influence the value of the product or service.
Therefore, good brand equity has the power to influence the prosperity of a company (Yoo & Donthu, 2001; Keller, 2013), though increase in customer willingness to pay (Keller, 2013), consumer satisfaction (Aaker, 1991), customer decision confidence (Aaker, 1991), future profits (Srivastava & Shocker, 1991; Ambler, 1997), cash flow, marketing success (Ambler, 1997), and so on. Moreover, Aaker suggests that brand equity is relative to the value
perceived by the customers in relation to a product or service of a brand (1991).
Consequently, customers and their perception of brand with its benefits and disadvantages play a crucial role in brand equity determination and as a result in brand alliance success measurement.
Customer Based Brand Equity is “the differential effect that brand knowledge has on consumer response to the marketing of that brand” (Keller, 2013, p. 69). Furthermore, Keller states that the main ingredients of CBBE are the experience and all associations related to a brand a customer has, which by taking the form of feelings, opinions and beliefs, form the power of a brand. The CBBE translate into multiple benefits like: improved loyalty and performance perception, less vulnerability to competition, customer resilience to price increase, etc (Keller, 2013). Certainly, as seen in the case of Hitachi and GE television sets,
though the product was the same, Hitachi was able to sell double the amount and at greater prices because of its brand equity (Berry, 1988; as cited in Keller, 2013).
Though brand equity clearly is indispensable equity for branding purposes, the issues related to calculation of brand equity arises. Therefore, one should look at the dimensions of brand equity, customer based brand equity and the ways such equity can be measured.
Aaker (1991) proposed to calculate brand equity through a number of dimensions,
consisting of: brand awareness, brand associations, perceived quality, brand loyalty, and other brand assets. While the brand loyalty reflects the likelihood of repeated purchase from a brand, the perceived quality reflects the brand’s level of promise satisfaction, and the other brand assets represent trademarks, patents and such; the other two dimensions are strongly related to customer perception. Moreover, the two dimensions are components within Keller’s “brand knowledge” which as Keller (2013, p. 71) suggests, “is the key to creating brand equity”. Consequently, brand knowledge in the customer minds needs to be discussed in more detail.
2.2.2 Customer Brand Knowledge
Customers hold multiple memory nodes and concepts, in their memories, which through being connected with other concepts create an association network (Quillian, 1966). Such concepts, though abstract in their nature, are the building blocks of belief and attitude
creation. Moreover, the beliefs and attitudes are formed when a person receives new
information in the form of stimuli, which is than evaluated and integrated within the existing cognition (Anderson, 1981; as cited in Simonin & Ruth, 1998). Consequently, Keller (1993, p.2) puts forward the model of “associative memory network” which indicates that people create associations with brands, also knowns as “brand knowledge”. Thus, the brand acts as a node linked to various knowledge about set brand, which in this case could be described as concepts. Furthermore, brand associative networks play a role in the buying decision of the customer. As a result, the strength of the brand node links to other nodes and concepts will determine which of them will be activated or ignored.
Within the aforementioned context brand awareness represents the strength of a brand in the memory of a person (Aaker, 1991; Keller, 2013). Consequently, brand awareness is made up of recognition and recall, and can be measured as brand identification under various conditions (Keller, 2013). The importance of brand awareness can be seen through its translation into numerous cognitive benefits. Through high brand awareness customers will more likely include the brand in their decision consideration set (Nedungadi, 1990), thus having less chances of being ignored by the customer (Nickerson, 1984; as cited in Keller, 2013). Moreover, improved awareness increases the chances of being selected especially under low-involvement (Hoyer & Brown, 1990), and can expand brand associative network (Keller, 2013).
On the other hand, brand image is just as important, if not more, as brand awareness.
Indeed, though awareness ensures that a brand node is created, the ensuing associations linked to that node are what determined what customers perceive (Keller, 2013). Aaker (1996) postulates that brand image is what represents a brand value and uniqueness to the customer, and it influences significantly the way in which brand equity can be leveraged . Consequently,
the associations related to the brand image in the mind of the consumer can include product and non-product related attributes, functional and experiential benefits, and attitudes (Keller, 1993). Moreover, the associations in the mind of the consumer can differ in terms of
favorability, strength and uniqueness (Keller, 2013; Keller, 1993). Furthermore, associations in the minds of customers are context dependant, thus their value vary from one situation to another (Miller & Ginter, 1979).
Multiple scholars have realised the importance of associations in the topic of brand image and brand equity (Keller, 1993; John, Loken, Kim & Monga, 2006; Schnittka, Sattler & Zenker, 2012; French & Gareth, 2013). Consequently, methods of establishing brand associative networks needed to be developed in order to evaluate brands from the consumer cognitive perspective.
2.2.3 Mapping Consumer Brand Associations
It is established that most consumer mapping techniques consist of three stages (John et al, 2006). The stages are: eliciting brand associations from consumers, establishing the
connections between associations, and finally producing a consensus map.
One of first methodologies of brand association mapping was Zaltman Metaphor Elicitation Technique (ZMET), also known as consumer mapping (Zaltman & Coulter, 1995). Such methodology is based on establishing the brand associations through the means of qualitative approach with subsequent link establishment by the means of interview. In contrast,
“analytical mapping” presented by Henderson, Iacobucci and Calder (1998; as cited in John et al, 2006) the associations are disclosed by consumer, while the links between associations are
retrieved through a network algorithm. However, both aforementioned methods have quite significant hurdles. The ZMET approach requires great effort of data collection though the interviews, while the Analytical mapping requires advanced statistical knowledge (John et al, 2006). In order to overcome such hurdles, John, Loken, Kim and Monga (2006) have
proposed the Brand Concept Maps approach.
BCM methodology uses the concept maps approach, which in the previous studies has not cover adequately the aggregation of consensus maps stage (John et al, 2006). Under BCM approach, after establishing the salient associations, each participant selects the associations reflective of own brand perception and then creates a BCM which corresponds with that perception. Since the links between associations are made in a simple ranking system, consequential development of consensus map based on frequency does not constitute a
difficult task. Moreover, in order to reflect consumer favorability of certain associations a new dimension has to be included in the BCM method, namely favorability within associative network (Schnittka, Sattler & Zenker, 2012). Such addition to BCM methodology would thus cover the favorability of brand image mentioned in the works of Keller (1993; 2013).
Furthermore, to elaborate upon the favorability Schnittka, Sattler and Zenker’s BCMV approach evaluate the judgement of each association and the importance of each association in purchase situation. As a result, the BCMV method adds the context of purchase, which as already established effects the perceived value of brand associations (see Miller & Ginter, 1979). Thus, arguably the BCMV will be more reflective of actual purchase intention of the consumers, than the BCM method. Hence, the current thesis will implement some of the concepts found in the BCMV methodology.
Rao and Ruekert (1994) described Brand Alliance as a combination of two or more brands over a period of time, made through association of some distinguishable physical or
symbolical assets. The associations of symbolical assets and the evaluation of the physical ones correspondingly are to be perceived by the customers and potential customers.
Consequently, companies have to take into account how the customers form their associative networks about the brand and how they evaluate brands and their alliances.
The greater the similarity between two brands engaged in a brand alliance, the better the fit (Collins & Loftus, 1975), which might be the case because the associations consumers have are transferred between brands which engaged in the brand alliance (Dickinson & Barker, 2006). Consequently, the added benefit from an alliance partner which is consistent with the company’s brand could be explained through categorization theory (Aaker & Keller, 1990; Fiske & Pavelchak, 1986). According to the aforementioned theory, whenever a human encounters new information it will be fit within an already existing cognitive schema (Fiske & Pavelchak, 1986). Moreover, if the new information is consistent with the cognitive schema, it will more likely be recalled. As a result, the better the partner brand fits in the already existing attitudes and beliefs about the main brand, the more likely their alliance will be perceived cohesively and positively (Ahn & Sung, 2012).
Keller (1993) has noted that the Least Mean Square connectionist theory interprets the human memory not simply as a shared nodes system, but as well as a dominant association predictability system. Thus, though certain connections in the memory might be shared, the probability of a recall of a node is dependant on its dominant ranking. Consequently, in the
context of brand alliances, it might be the case that a stronger brand might be taking over some of the associations from the partner brand.
The least mean squares learning rule was proposed first by Widrow and Hoff (1960), and it was later used to interpret the connectionist networks of the human learning (Bower & Heit, 1992). The adaptive networks model which implements least mean squares has proven to be predicting quite accurately human behavior in learning, including difficulty and choice prediction. Furthermore, the model states that whenever there are multiple cues, they will be competing among themselves. Consequently, the overlapping elements within the associative network, in order to reduce the error only cues with stronger associative strength will
dominate.
Moreover, Suh and Park (2009, p.1) have noted that “ high-favorability partner brand may do more harm than good to a moderate-favorability host brand’s new product because the partner brand does not block the activation of counter-arguments after co-branding is terminated”. Therefore, while collaborating with other brands a company can certainly put its brand associations at risk.
2.3.1 Associations in Brand Extensions and the Feedback Effect
In the brand extension literature, as well as in brand alliance literature scholars acknowledge the brand associative network in the mind of the consumers. Fiske and Pavelchak (1986) in their works talk about the brand schema, which reflects all the knowledge an individual has about a certain brand. Additionally, Weber and Crocker (1983) put an emphasis on the role of beliefs in the mind of the consumer and how they change in the case of brand extension.
Consequently, when a parent brand is introducing a new extension consumers are compelled to redefine what the brand means to them (Buday, 1989). Furthermore, the differentiation of processes involved in extension is made based on the consistency with parent brand image (Park, McCarthy & Milberg, 1993). The assimilation process occurs when the brand
extension does not divert from the core of the parent brand. On the other hand, when there is an inconsistency between the extension and the parent brand accommodation occurs, thus modifying consumer attitudes and brand schema (Weber & Crocker, 1983).
Brand extensions and their success can have a variety of negative effects on parent brands. Based on the points of parity and difference companies might create an extension which will take sales from the parent brand, resulting in cannibalization (Keller, 2013). Moreover, some companies might be extending into multiple products and categories, which will obscure the brand's original category and attribute identification (Keller, 2013). Consequently, the
diminishing identification with the original product category also diminishes brand awareness (Morrin, 1999). Furthermore, brand extensions can potentially hurt parent brand’s brand image and dilute its brand meaning (Keller, 2013; Loken & Roedder, 1993). Though scholars acknowledge that negative effects in brand extensions are possible, this effect is believed to be understudied (Milber, Park & McCarthy, 1997; Lane & Jacobson, 1997).
In the scholarly literature there has been a few studies which took into account negative feedback effect. Keller and Aaker (1992), for example, examined how product similarity within brand extension could affect the parent brand evaluation. In their research the similarity was based on features of products, and the results showed that there was no
1997) also did not find evidence of feedback effect when accounting for the product similarity and negative information of extension performance.
However, in the study of Park, McCarthy and Milberg (1993) the scholars have found a negative feedback effect while also adjusting for product similarity and brand image beliefs. Furthermore, Loken and Roedder (1993) have mentioned that consistency between attributes of parent brand and its extension play a significant role in post extension brand beliefs evaluation. More specifically, the scholars have discovered that when brand extensions attributes are less typical of the family brand attributes, this dilutes the parent brand image beliefs. Moreover, Milberg, Park and McCarthy (1997) have confirmed that product category dissimilarity can have a significant impact on feedback, especially when there are salient product associations in the parent brand. Additionally, when the attribute information of the brand extension is dissimilar from the image beliefs related to the parent brand, brand extension strategy could also result in a negative feedback effect.
Consistent with this narrative is the research on brand attitude reciprocity effect, which proved through need for cognition (NFC) and relational attribute cues that feedback effect exists in case of evaluation of an extension and evaluation of parent brand (Lane & Jacobson, 1997).
From the aforementioned scholars and research, it is clear that feedback effect does in fact occur, and associations might play a significant role in this process. Indeed, Broniarczyk and Alba (1994) have noted that brand specific associations of parent brands have a highly significant impact on brand extension evaluation, and a strong moderator effect on brand affect and product category similarity. Thus, supporting Keller’s (2003) narrative that brands should be perceived as networks of knowledge in the form of attribute, attitude and benefit associations in the mind of the customers.
Simonin and Ruth (1998) point out that the studies within brand extension literature are mostly concerned with how consumers perceive and associate product categories, while the aspect of brand combinations are left out. The scholars also suggest that brand associations were often left out in the research on brand extensions. As a result, Simonin and Ruth included brand attitude in their research on brand alliance (1998).
Some scholars in the brand extension literature refer to the collection of brand associations as brand image (Dwivedi, Merrilees & Sweeney, 2010). The brand image associations cover product associations, brand name, reputation and many more. However, studies which address particular associations are quite complex in terms of research since they include mapping a brand associative network through multiple interviews, pre-testing, analysis etc. (Zaltman & Coulter, 1995; John et al, 2006). Moreover, results of experiments which deconstruct brands into associations don’t necessarily account properly for generalisation of the results. On the other hand, brand image as researched by Dwivedi, Merrilees and Sweeney (2010) is not deconstructing the brand associations map, instead the focus is on the impact the brand extension initiative has on the attitude toward parent brand.
It is worth noting that Aaker and Keller (1990) concluded that level of fit determines if positive associations will be transferred or negative ones will be formed. Moreover, Romeo (1990) concluded that the parent brand will less likely be affected by an unsuccessful
extension if both are in dissimilar product categories. Thus, reinforcing the opinion that brand extension’s success is dependant on the degree of fit and consistency with its parent brand.
2.4 Brand Fit
Though brand equity is a monumental element for a successful co-branding activity, there are multiple other elements which are also very significant (Keller, 2013, pg. 271). Keller
suggests that brand awareness, good quality of associations and positive beliefs are a impacting positively a co-branded strategy; however, the logical fit is the most imperative element for its success. In unison with Keller other scholars confirm through their studies that the evaluation of a brand alliance is quite dependant on the fit between the partner brands engaged in the collaboration (Simonin & Ruth, 1998; Park et al, 1996; Pruppers et al, 2007). Indeed, the work of Aaker and Keller (1990) confirms that level of fit determines what changes will happen to one's brand associations.
Accordingly, elements which have a record of impacting the brand alliance are found within the Brand Fit literature. This field incorporates a multitude of dimensions of fit, and they should be addressed in order to create a balanced outlook on the current research.
“Brand alliances can have several bases of fit; category fit, brand associations, consumer goals, culture, product user, product usage and self-representation” (Park et al., 1996; Simonin and Ruth, 1998; Washburn et al., 2000; Motion et al, 2005; Bouten et al., 2011; Lanseng and Olsen, 2012; as cited in Oeppen & Jamal, 2013, p.2).
Consequently, in the scholarly research authors use the terminology of “product fit”, however it often refers to different bases of fit. Pruppers et al. (2007) refer to product fit as the fit between the parent brands and the extension product. Whereas, Simonin and Ruth (1998) perceive product fit only through the prism of product categories of the partner brands. Nevertheless, Aaker and Keller (1990), Park et al (1996) and Pruppers et al (2007) all
acknowledge that the fit between partner brands, and parent brand and their extension should be taken into account. Indeed, Simonin and Ruth (1998) confirmed that the better the product fit, the higher the perception of the alliance.
In order to asses the level of fit there are certain dimensions which have to be measured. Fit can be measured through product based fit though: category complementarity, category substitutability, category diversification ability (Aaker & Keller, 1990), brand concept consistency (Park et al., 1991), context similarity (Broniarczyk & Alba, 1994).
On the other hand fit can be measured through brand to brand fit through: attribute salience complementarity, performance complementarity (Park et al., 1991), category complementarity (Simonin and Ruth, 1998) and brand concept consistency (Pruppers et al., 2007; Simonin and Ruth, 1998).
Multiple scholars suggest that the fit between partner brands is determined at a great extent by the brand image fit (Gwinner & Eaton, 1999; Pruppers et al., 2007). Which could
potentially be because consumers can retain abstract information, such as brand image, easier and they base their evaluation on information easily available in their cognition (Pruppers et al., 2007). Moreover, Gwinner and Eaton (1999) convey that the fit between the images of partner brands has a significant role in brand alliance and companies have to evaluate this dimension carefully, or risk negative spillover effect.
Nevertheless, scholars stretch the importance of product based fit, since it can potentially clarify to consumers the reasons behind the brand alliance of brands with complementary attributes (Pruppers et al., 2007). Moreover, product categories are part of brand associative network in the minds of the consumers, which are also important in the evaluation (Pruppers et al., 2007) and in case of a spillover effect (Keller, 1993).
Interestingly enough, though Dwivedi, Merrilees and Sweeney (2010) notice an effect of parent brand on both the extension and the fit, a similar effect was not tested in the work on co-branding by Simonin and Ruth (1998). Moreover, though fit was considered important in both of the aforementioned studies, its effect on parent brand image change were not tested in the case of brand alliances (Simonin and Ruth, 1998).
Consequently, the level of fit is an important variable to be measured when conducting a research on co-branding, and will be used for the research in this thesis.
3. Conceptual Model and Hypothesis development
3.1 Conceptual Model
In order to answer the problem statement of this thesis first it has to be established if the feedback effect on the attitude towards the main brand exists in the context of co-branding, and what variables impact the change in that attitude. Consequently, it would be possible to predict not only how would the consumers perceive the co-branding initiative, but also what would be the impact on the main brand image.
The conceptual model in Figure 1. was developed by combining the conceptual models proposed in the articles of Simonin and Ruth (1998), and Dwivedi, Merrilees and Sweeney (2010). On one hand, the variables within brand alliance mentioned by Simonin and Ruth (1998) would be used in the current research. On the other hand, the relationship between the variables would be similar to Dwivedi, Merrilees and Sweeney (2010), since their model implies that variables are interdependent.
The conceptual model designed by Simonin and Ruth (1998) considered the attitudes towards the two brands involved in brand alliance as two independent variables, each influencing their own post-attitude toward the brand and the attitude toward the brand alliance. On the other hand, in the work of Dwivedi et al (2010), the authors considered attitude toward the brand as the only independent variable in the model, and it would have an effect on all the other variables within the model like: perceived fit, brand extension attitude and parent brand attitude change. Moreover, the authors argue for the impact of perceived fit not only on the
extension, but also on the change in the attitude towards the main brand. In the framework of brand alliance, a comparable relationship could be expected between perceived brand fit and parent brand attitude change.
Therefore, in the current conceptual model the attitudes towards each individual brand will be considered as separate independent variables which have the potential to influence the
co-branding attitude, as in the works of Simonin and Ruth (1998), as well as the attitude change towards the main brand. Furthermore, since the primary focus of the study is the perspective of a brand which decides to engage in co-branding, the perspective of partner brand attitude change will be left out of the study. Consequently, only the change in the attitude toward the main brand will be measured.
At the same time, the current conceptual model also considers that each brand attitudes also influence the perceived fit, while the perceived fit influences the the attitude towards the post co-branding initiative of the main brand, similarly to the conceptual model of Dwivedi et al (2010).
In order to test the second part of the problem statement, another conceptual model is required. To test if low fit co-branding has a stronger negative feedback effect on the main brand than the high and low fit co-branding. NPS scores would be used as measures of attitude towards the brand and its brand equity. The reason for choosing NPS as a measurement of brand attitude is based on the fact that it is a easy and straightforward approach. NPS scores can be easily adapted for any product or service, are very simple to measure, and are already used by many companies.
The second variables within the model is Co-branding fit. The fit in this case would be based on compatibility of brand associations and response of the pre-test group.
Figure 2. Conceptual Model nr.2
Main Brand Evaluation - is the attitude the respondents have towards the focal brand - Tesla, and is a independent variable.
Partner Brand Evaluation - is the attitude the respondents have towards the secondary brand - Toyota and/or Apple.
Brand Fit - is the opinion of the respondents regarding how well the main and the partner brand fit together, and is a mediating variable.
Co-Branding Evaluation - is the respondent's attitude towards the co-branding, and is a mediating variable.
Main Brand Attitude Change - is the measurement of the change of the main brand attitude after the co-branding initiative, and is a dependant variable.
All this variables are measured by adapted questions from the research by Dwivedi, Merrilees and Sweeney (2010), Martinez and Chernatony (2004), Tauber (1988) and Bousch and Loken (1991).
3.2 Hypothesis Development
As previously mentioned, brand equity represents a combination of all positive and negative associations a consumer has in relation to a brand (Aaker, 1991). Moreover, Keller (2013) states that customer based brand equity is a combination of the experience and associations toward a brand, and is responsible for how the consumers respond to a brand’s marketing activities. The aforementioned associations are often found to have the ability to influence how a brand extension is evaluated (Aaker & Keller, 1990; Keller, 2003), hence the positive attitude towards a brand is transferred onto the brand extension which bears the same brand name (Broniarczyk & Alba, 1994). This is further supported by the idea that brands act as testimonials for quality of a product (Kapferer, 1992).
Finally, in the research by Simonin and Ruth (1998) prior brand attitudes have proven to directly influence the attitude towards the brand alliance.
Since the current research is an adaptation of Dwivedi et al (2010) within the co-branding field, the results of Simonin and Ruth (1998) might not be replicated. Nevertheless, a similar type of relationship is expected between attitude towards the brand and the co-branding. Therefore, the first two hypotheses are representing the relationship between prior brand attitudes and the co-branding initiative attitude:
H1: Main brand evaluation has a positive effect on the attitude toward the co-branding.
H2: Partner brand evaluation has a positive effect on the attitude towards the co-branding evaluation.
Park, Milberg and Lawson (1991) came to the conclusion that when consumers are assessing the level of fit within an brand extension their judgement is influenced by various factors related to brand concept consistency, such as unique image associations. The scholars suggest that the combination of consumers perception of attributes and benefits are partially
responsible for how they will also perceive the fit between the parent brand and the extension. Moreover, scholars have demonstrated that consumer knowledge can have an impact on brand extension and the perception of fit, since the mere placement of parent’s brand name on the product creates the association in the mind of the consumer (Gronhaug, Hem and Lines, 2002; as cited by Dwivedi et al, 2010). Additionally, Dwivedi et al (2010) were able to prove the existence of the aforementioned effect in brand extension literature. On the other hand, within the brand alliance literature the relationship between the attitude toward a brand and brand image fit was not studied meticulously. Consequently, the following is hypothesized within the current co-branding initiative research:
H3: Main brand evaluation has a positive effect on the perceived fit between the brands.
H4: The attitude toward the partner brand has a positive effect on the perceived fit of the co-branding brands.
Since brands are represented in the minds of consumers as a series of brand specific associations (Aaker, 1991; Keller, 2013), this associations play an integral role when consumers encounter joint branding (Broniarczyk and Alba, 1994; as cited in Simonin and Ruth, 1998). Moreover, if the consumers perceive the brand images of companies engaged in joint branding activities as being inconsistent with one another, they would more likely appeal to attributional type of decision making (Folkes, 1988; as cited in Simonin and Ruth, 1998). Therefore, scholars suggest that if there is no appropriate fit between the brands of the companies engaged in brand alliance, they are more likely to experience a negative spillover effect (Gwinner and Eaton, 1999).
Consequently, the level of fit between the brands has a significant impact in the evaluation of their brand alliance (Aaker and Keller, 1990; Simonin and Ruth, 1998). However, since the way the fit is measured in the current study is broad, as per Dwivedi et al (2010), the relationship between the variables has to be tested within the co-branding setting. Hence, the following hypothesis will be tested within the current framework:
H5: The better the fit between the companies, the higher is the co-branding evaluated.
Brand associations play a significant role in attitude formation (Aaker and Keller, 1990; Gwinner and Eaton, 1999), therefore it is a logical assumption that consumers can change their attitude toward the brand if it has associated itself with a new product or brand. As , multiple scholars within brand extension literature confirm that attitudes towards brand extensions have an effect on parent brand attitude (Martinez & Chernatony, 2004; as cited in Dwivedi et al, 2010; Keller and Aaker, 1992) and this is expected to bear truth also in the case of brand alliance.
attitudes toward the brand alliance have an effect on the attitude toward the brands involved in the alliance. Therefore, the following hypothesis is proposed:
H6: Attitude toward co-branding initiative has a direct effect on main brand attitude change.
As stated previously, perceived fit plays an important role in the creation of consumer
attitudes toward brand extension (Aaker and Keller, 1990), and brand alliances ( Simonin and Ruth, 1998). Nevertheless, fit is a very subjective variable, and each consumer forms their own perception of fit (Murphy & Medin, 1985; as cited in Dwivedi et al, 2010). Moreover, Dwivedi, Merrilees and Sweeney argue based on the works of scholars like Gürhan-Canli and Maheswaran (1998), Martinez and Chernatony (2004), and Sheinin (1998) that the perception of fit influences consumer’s attitude toward the parent brand (as cited in Dwivedi et al, 2010). The scholars state that the aforementioned effect functions by the means of strengthening or diluting the primary associations and beliefs about a brand, based on consumer’s perception of fit. Consequently, depending on the perception of fit, consumers will form new
associations which they will attribute to the parent brand.
Considering that Dwivedi et al (2010) have proven that this relationship between fit and feedback to the parent brand, the following hypothesis was set to be tested for the co-branding initiative:
H7: Brand fit has a direct effect on main brand evaluation change.
Studies on brand alliances has proven that brand alliances can influence the attitude towards the brands involved in the alliance (Simonin and Ruth, 1998). However, the scholars have only tested the two brand alliances. Therefore, the current research will analyse the effect
multiple co-branding initiatives have on a brand’s image. Considering that brands in the minds of consumers made up of various associations (Keller, 2013), and that each association has a particular strength. Considering the previously mentioned role of fit, the current thesis assumes that partner brands not only can influence the attitude towards the main brand through the means of co-branding, but also assumes that there might be a distinction in the strength of influence between a low and a high fit brand. Moreover, since scholars suggest that attributes are transferred from one brand to another during brand alliance (Rao et al. 1999), it can be assumed that when there are a high and a low fit co-branding partner, their combined good and bad associations would be transferred to the main brand. In contrast, the low fit co-branding partner alone would not have as many positive transferable associations to mitigate the negative feedback effect from its bad brand associations.
Consequently, the following hypothesis is proposed:
H8: Low fit co-branding initiative has a stronger negative feedback effect on
respondents attitudes towards the main brand, in contrast with a simultaneous high and law fit co-branding initiative.
4 Methodology
The current research was divided in two segments: establishing associations, evaluation of co-branding. This methodology was adapted from the studies of Miller and Ginter (1979), and Schnittka, Sattler & Zenker (2012).
First part of the research was a qualitative exploratory study, done with a control group of nine student, in order to establish three well-known brands which will be used for different treatments. During this phase the control group was questioned in order to asses brand compatibility and the chances of possible co-branding. Moreover, each individual within the control group had to establish the associations of five brands: Tesla, Apple, Toyota, Microsoft and Nike. It should be noted that the aforementioned brands were selected based on the list of most valuable brands (Badenhausen, 2016), under the assumption that this brands are also most wide known brands. After establishing the associations, the subjects had to rank the associations and establish the favorability of the associations, as per BCMV method described by Schnittka, Sattler & Zenker (2012). Consequently, the scenario of co-branding initiatives was based on the level of associative fit between the brands, as perceived by the students in the exploratory session.
The next stage was, quantitative study, with the data collection procedure performed through the use of surveys. One of the studies which published its process of measuring variables related to fit was the study done by Dwivedi, Merrilees and Sweeney (2010) in their
holistic research. Consequently, their study was used as a guidance for developing the survey, with some questions being slightly changed to reflect the current research topic of
co-branding. The survey was designed in Qualtrics with the fictional co-branding initiatives, and was administered digitally to student at the university of Amsterdam. The brands in the survey were presented in a similar manner though same size logos and same format, as can be seen in Appendixes A, B and C. This was done in order to eliminate additional variables. Co-branding initiatives were announced through the same message, and with an image
reassuring the validity of the statement, as seen in Appendix D and E. The questions related to evaluation of fit and attitude towards the co-branding were also formulated in the same
manner, just changing the name of the company, see Appendix F.
4.1 Sample
The target population of this study will be the current university business students at
University of Amsterdam. As a result, the research will be as close as possible to the study of Dwivedi, Merrilees and Sweeney (2010). However, in contrast, the current study will also include female respondents, thus accounting for possible gender differences. The target sample size is of 100 student, since the study aims to get over 30 respondents for each treatment. Moreover, the sampling technique administered for this study will be
non-probability sampling. The sample will be selected based on the email list of students attending business seminar course at University of Amsterdam, which doesn’t include people outside the University of Amsterdam, nor all the students enrolled at the business course at this university. On the other hand, collecting data through this method requires less resources,
is more manageable and has a tolerable margin of error. Additionally, in order to increase the response rate an incentive will be offered. The incentives will be given to randomly selected respondents.
4.2 Measures
The measures included in the questionnaire will ask respondents about their gender (nominal variable) and their age (ratio variable). The measures for the next items in the survey will be based on a 7-point Likert scale. The anchors for the Likert scale will be from strongly disagree (1) to strongly agree (7).
As for the items, they will be based on previous research done in the field of feedback effect and brand extension. Therefore, Main Brand Image items will be based on Martinez and Chernatony (2004) (E.g.: The X brand has personality), Fit will be based on Tauber (1988) (E.g. According to me the co-branding initiative seems logical), Brand Extension Attitude will be based on Bousch and Loken (1991) (E.g. I am favourably disposed towards the co-branding initiative).
The Attitude Change item will be based on Dwivedi, Merrilees and Sweeney (2010) (E.g. My admiration toward the main brand would become), where anchors are set as “negative
attitude” (-3), “same attitude” (0), and “positive attitude” (+3). In total 18 items will be adopted from previous studies.
The constructs chosen for this study have been checked for their reliability by Dwivedi, Merrilees and Sweeney (2010), and all the constructs had Cronbach’s Alpha level above 0.70, ranging from 0.75 to 0.80.
5. Data and Analysis
5.1 Brand Associations, Pre-testing
Firstly, brands had to be selected and brand associations had to established in order to assess compatibility between brands. This was done as part of an assessment of the brand equity for the testing phase and preliminary assessment of fit between the brands. During the pre-testing phase a small control group of 9 people was questioned in order to select brands for the testing phase of the research. This step was necessary in order to account for brand familiarity and plausibility of co-branding between the selected brands. The underlying assumption was that if the co-branding is more realistic the survey answers would more likely reflect the true attitude of the respondents.
The brands which were selected for possible co-branding were based on the list of most valuable brands of 2016 (Badenhausen, K., ed., 2016), since it was assumed that those brands were more well known across the population. During this phase the group was simply asked to select the brands they are most familiar with. As a result, the current research tried to account for brand familiarity and selected only the brands which were well known by all 9 respondents. Afterwards, in an informal setting the group was asked to pair up brands they assume are more likely to collaborate in the future. The respondents selected the following brands: BMW, Toyota, Tesla, Apple, Nike, Microsoft; which were paired between themselves in one way or the other. From the aforementioned list, Tesla was often paired with Toyota and
Apple. Since Toyota and Apple brands are different in their value proposition, the three brands: Tesla, Toyota and Apple were selected for the next phase of the pre-testing.
In the next phase, the same 9 respondents were asked to establish the brand associations of Apple, Tesla and Toyota. The associations mentioned by the member of the pre-test group were ranked on their strength and favorability; as suggested by multiple previously mentioned scholars (Schnittka, Sattler & Zenker, 2012; Miller & Ginter, 1979; John et al, 2006). The ranking was done by each individual through placing the associations in descending order of association strength on a piece of paper. Additionally, the respondents had placed the
favorability next to each associations, when they saw fit. The favorability was ranked on a 7-point Likert scale, from -3 (very unfavourable) to +3 (very favourable).
Afterwards, in order to consolidate the ranking it was compared with the ranking of brand associations from “brandtags” website, a platform where one word associations of various brands are collected from random respondents (Brandtags, 2016). As a result, the subjective associations mentioned by respondents were left out. Furthermore, some of the core brand associations mentioned by less than 50% of the respondents in the pre-testing phase, were kept since they were one of the core associations on Brandtags website. Based on the brandtags list and the responses from the pre-testing group, the cut off point for being
considered a strong associations is 4 mentions, while a strong favorability or unfavorability is considered anything greater than | ±1.5 |.
Consequently the associations which were most prevalent can be seen in the Table 1 and Table 2, below.
