Different or the same? : the accentuation of unique associations into evaluation techniques of brand concept maps

Different or the Same?

-

The Accentuation of Unique Associations

into Evaluation Techniques of Brand

Con-cept Maps

MASTER THESIS

Patrick Best

Student number: 10824448

June 29, 2015

Supervisor: drs. Jorge Labadie MBM

Co-Supervisor: dhr. drs. R.E.W. (Roger) Pruppers

M.Sc. Business Administration

This document is written by Student Patrick Best who declares to take full responsi-bility 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.

A special word of thanks goes to all people who supported me during the last ten months. First, I would like to express my gratitude to my supervisor, Jorge Labadie, for his support throughout the entire thesis process. Our inspiring discussions showed me the diversity of branding and encouraged me to work continuously on this last step toward finishing my aca-demic career.

I would also like to thank my co-supervisor, Roger Pruppers. Even though his advice during the plenary meetings was valuable and useful, a far bigger impact on this thesis was provided by the knowledge transfer and the introduction of several interesting and applicable research articles during the enthusiastically guided Consumer Behavior classes.

I would also like to thank all participants who have taken the trouble to align brand maps in the course of this study. Without their cooperation, this work could not be brought to an end. Last but not least, I want to thank my family for their ongoing support and instilling confi-dence in me.

Brand concept maps by John et al. (2006) represent an approach to vividly illustrate consum-ers' brand associative networks. In order to further enhance this method of brand image re-search, Schnittka et al. (2012) developed the brand association network value (BANV) which quantifies the overall network value of brand maps as constituted by consumers. Thereby the metric succeeded in proving its positive relationship to consumers' brand familiarity, pur-chase intention, and overall brand attitude.

However, the basis for calculation of the BANV builds on the assumption that all tions within customers' brand associative networks are to be treated equally. While associa-tions which are product category related are likely to be shared with competitors, brand-specific associations are widely seen as requirement for differentiation and the generation of sustainable competitive advantages. However, if brand specific associations are perceived as unfavorable they might also be the source of competitive disadvantages for brands. In this research it was tested whether special weightings for unique associations can enhance a met-ric for the overall evaluation of consumers' brand association networks (as they are pictured by brand concept maps) in terms of its nomological and predictive validity. By using two brands out of the product category of cars, the amplitude of a unique association's weight was set in dependence to its respective favorability, as it is assumed that points of differences can both be a chance and a risk to brands competing in one marketplace.

The findings of this work revealed that brand-specific associations enjoy at least in some product categories an accentuated role in consumers’ brand association networks, as the new metric was able to prove significantly stronger positive relationships to consumers’ purchase intention and overall brand attitude. Thus, Schnittka et al.'s call for a more differentiating weighting of associations (2012) was met and criticism on the inclusion of uniqueness as a criterion for the buildup of customer-based brand equity (Romaniuk and Gaillard, 2007) de-bilitated.

Table of content

1 Introduction ... 4

1.1 The importance of associations for brand value ... 4

1.2 Lack of knowledge ... 5 1.3 Problem Definition ... 6 1.3.1 Problem statement ... 6 1.3.2 Sub questions ... 8 1.3.3 Delimitations ... 9 1.4 Contribution ... 10 1.4.1 Theoretical contribution ... 10 1.4.2 Managerial contribution ... 11 1.5 Structure/outline ... 11

2 The value of brands ... 13

2.1 Associations ... 13

2.2 Customer-based brand equity (Keller, 1993) ... 14

2.2.1 Customer-based brand equity pyramid ... 15

2.2.2 Brand image ... 16

2.3 Criteria for customer-based brand equity ... 18

2.3.1 Strength of associations ... 18

2.3.2 Favourability of associations ... 19

2.3.3 Uniqueness of associations ... 20

2.4 Brand value chain ... 23

3 Consumer mapping ... 26

3.1 Advanced Brand Concept Maps / BANV ... 26

3.2 Category-specific and brand-specific associations: Strategic Concept Maps (SCM) ... 32

3.3 Adjusted weighting of the new BANV2 formula ... 34

3.4 Hypotheses ... 37

4 Method ... 40

4.1 Stimulus development ... 40

4.2 Investigation of brand image ... 42

4.3 Advanced Brand Concept Maps... 44

4.4 Identifying unique associations ... 47

4.5 Calculation of BANV and BANV2 and investigation of relationships ... 48

5 Results ... 50

5.2 Reliability ... 50

5.3 Computing scale means... 51

5.4 Correlation analysis ... 51

5.4.1 Study one: Opel ... 52

5.4.2 Study two: Tesla ... 55

5.4.3 Overall ... 58 5.5 Regression analysis ... 61 5.6 Consensus Maps ... 62 6 Discussion ... 69 6.1 Findings ... 69 6.1.1 Brand familiarity ... 70 6.1.2 Purchase intention ... 71 6.1.3 Brand attitude ... 71 6.2 Theoretical implications ... 72 6.3 Managerial implications ... 74 6.4 Limitations ... 78

7 Summary and Conclusion ... 81

References ... 85

List of figures and tables

Figure 1 Customer-based brand equity pyramid (Keller, 1998) Figure 2 The Brand Value Chain (Lehmann and Keller, 2003)

Figure 3 Advanced Brand Concept Map, example (Schnittka et al., 2012) Figure 4 BCM Example (Volkswagen Beetle) shown to participants Figure 5 Brand consensus map Opel

Figure 6 Brand consensus map Tesla

Table 1 Intercorrelations of BANV{Eaj, Saj, Iaj, Laj}, BANV{Eaj, Saj, Iaj}, BANV{Eaj, Saj} with the validation criteria brand familiarity, purchase intention, and brand attitude by Schnittka et al. (2012)

Table 2 Conversion table for the BANV2 metric Table 3 Conversion table for the control metric Table 4 Descriptive statistics for study one (Opel) Table 5 Full correlation matrix study one (Opel) Table 6 Descriptive statistics study two (Tesla) Table 7 Full correlation matrix study two (Tesla) Table 8 Full correlation matrix across both studies

Table 9 Testing of independent variables (brand familiarity, purchase intention, brand attitude) against BANV2

Table 10 Regression model overview

Table 11 Measures for core associations, study one (Opel) Table 12 Measures for core associations, study two (Tesla)

List of abbreviations

ABCM – Advanced Brand Concept Maps (Schnittka et al., 2012) BANV- Brand Association Network Value (Schnittka et al., 2012) BCM – Brand Concept Maps (John et al., 2006)

1

Introduction

1.1

The importance of associations for brand value

The value of the currently most valuable brand in the world, Apple, is according to the Best Global Brands 2014 report (Interbrand brand consultancy, 2015) specified with 118.863 bil-lion US dollars. Even though rankings for brand value such as the one of Interbrand try to express their findings in monetary values, brand equity goes beyond the border of pure finan-cial performance: another market based asset brands can entail is consumers’ knowledge about a particular brand which is stored in their memory and used in actual purchase situa-tions (Romaniuk and Gaillard, 2007).

In contrast to financial based methods of brand equity assessment, consumer based approach-es do focus on measuring the value created in customer’s minds which in turn are based on the associations a brand elicits (Aaker, 1991). Given the assumption that prior marketing pro-grams have a significant influence in building brand associations, profound knowledge about what has been created in consumer’s minds may be one of a firm’s most valuable assets for improving their marketing productivity (Keller, 1993) and thus can provide valuable infor-mation regarding what is going wrong or well in the public perception of a brand.

The role of brand associations for customer’s brand perceptions have been subject of investi-gation in various influential marketing research articles. Only if a brand is able to communi-cate relevant product meaning and positive feelings, it is likely that consumers start to build a relationship with it. At the final stage of such relationship building processes, customers may become loyal to a particular brand and therefore less likely to be lured away by competitors (Keller, 1998). However, if the image of a brand is weakened or has not adapted to dynamic market changes over time, customer's loyalty will most likely decrease. Therefore the man-agement of brand related associations stands central for brand managers whose goal it is to

proactively steer their product or service through the very complex environment of interde-pendent customer perceptions and changing market conditions.

1.2

Lack of knowledge

The procedure of Brand Concept Maps (BCM) as developed by John et al. (2006) represents an approach for graphical illustration of consumers’ brand association networks. By plotting and interconnecting brand specific associations elicited by participants, most common attrib-utes and benefits as perceived by respondents are summarized and suitable to serve as an at-tempt for creating a holistic picture of consumers’ brand association networks. Thereby indi-vidual brand maps for each respondent build the starting point for conducting an aggregated brand map, the so-called consensus map.

In order to incorporate associations’ favorability as one of Keller’s criteria for the build-up of customer-based brand equity (CBBE), Schnittka, Sattler and Zenker (2012) expanded the BCM model by adding the so-called evaluation stage. Thereby participants are requested to assess both the relative favorability and the relative importance for each elicited association, before researches finally develop a metric, the so-called brand association network value (BANV). The BANV is an attempt to numerically express the overall favorability of an indi-vidual brand associative network (Schnittka et al., 2012). As every single association gets evaluated in terms of four different dimensions, the process of this so-called Advanced Brand

Concept Maps (ABCM) provides information regarding each single associations’ perceived

favorability, as well as about the overall favorability of a brand association network.

With this method of quantification a big step was taken towards measurability and hence ef-fective evaluation of associative networks and their inherent constituents. However, a pitfall of the Advanced Brand Concept Maps (ABCM) approach by Schnittka et al. (2012) was de-tected in the course of this work: the calculation method of BANV considers all associations

elicited by a brand. Thereby brand specific associations are granted the same weight as asso-ciations which can be attributed to the respective product category and thus many other com-petitors. As Schnittka et al. (2012) state in their limitations, unique associations (as they are able to reflect a brand’s unique selling proposition in customers’ minds) should be granted more influence than compared to those which can be attributed to the respective product cat-egory and thus to most of a brand’s direct competitors. Especially in the case of unique and predominantly unfavorably perceived associations the previous BANV methodology exhibits a significant weakness. Since the metric is composed based on a multiplicative function and all associations are adding to each other, all of them contribute to a higher BANV value. As brand specific and unique associations which are perceived as unfavorable by the majority of consumers can be seen as particularly critical for a brand’s brand image, this kind of summa-tion contradicts the original intensumma-tion of the metric. To date, there is no quantitative evalua-tion method for brand associaevalua-tion networks established that takes the accentuated role of brand specific associations into account.

1.3

Problem Definition

Brand maps as they have been developed by John et al. (2006) and advanced by Schnittka et al. (2012) are an attempt to vividly illustrate consumers’ brand association networks. Since only little attention has been paid to quantitative evaluation methods for brand maps, the BANV metric can be seen as first approach to both assess consumer brand maps and accord-ingly brand association networks in general. By investigating the calculation basis of the BANV metric, new ways should be revealed how to additionally incorporate uniqueness of associations as one of Keller’s prerequisites for the buildup of customer-based brand equity.

Keller's other two requirements for CBBE building associations (strength and favorability of associations) are already taken into account in both the graphical presentation of the Ad-vanced Brand Concept Maps, as well as in the BANV metric. As the importance of both strength and favorability of brand associations can be seen as largely undisputed, uniqueness of association has served as starting point for a scientific controversy in marketing research. So have Romaniuk and Gaillard (2007) empirically shown that customers with higher prefer-ence for one brand do not hold more unique associations than those with lower preferprefer-ence. Their criticism of the inclusion of uniqueness as a decisive factor for the development of CBBE is a central thread running through the work of the Australian Ehrenberg Institute to which both authors belong to. Though, they stand in conflict to other studies that emphasize the need for differentiation such as Krishnan's (1996) in which he empirically proofed unique associations within a given product category (USPs) as indicator for brand equity. It is im-portant to mention that both Sharp (2011) and Romaniuk and Gaillard (2007) support their arguments (against unique associations as being a factor contributing to brand performance) by referring to studies which exhibit an inherent bias toward fast-moving consumer goods (fmcg) in their research setup (Wilson, 2011). Those categories (like energizer drinks or yo-ghurts) are not likely to put consumers under high-involvement decision-making settings, which due to Keller (1993) strengthen the crucial role of strong, favorable and unique associ-ations for the buildup of brand equity.

Under the above-mentioned aspect a possible extension of the ABCM approach in favor of unique associations has to be conducted under special conditions: a new approach that incor-porates a prominent position of unique associations within the overall assessment of brand network favorability has to empirically prove its added value compared to the original BANV metric. As this kind of proof has to go along with the “success” of the brand in consumers’ brand perception, purchase intention and overall brand attitude are suggested to serve as

measures for evaluating both the BANV and BANV2’s suitability to express the overall fa-vorability of brand association networks.

Summarizing the key starting points for studying brand association networks and brand con-cept maps as stated in this chapter, the research question of this work is stated as follows:

To which extent can unique associations enhance a metric for assessing brand concept maps

and accordingly brand association networks in terms of improving its relationship with

con-sumers’ brand familiarity, purchase intention and overall brand attitude?

1.3.2 Sub questions

In order to deliver an answer for the main research question, several sub questions are ad-dressed to ensure that the work is based on a solid theoretical background. The concept of customer-based brand equity needs to be explained, since it is the starting point for figuring out how associations influence the building of strong brands. Based on a sound understanding of how strong brands are created by strong, favorable and unique associations, brand associa-tion networks as they are displayed in brand concept maps are investigated for fulfilling all of the above mentioned requirements by Keller (1993). Thus, peculiarities in brand associative networks should be examined and serve as origin for the managerial contribution.

The sub-questions therefore are as followed:

- What is customer-based brand equity and how does it influence people’s relationships with brands?

- How do brand associations determine people’s relationships to brands and which kind of associations differ strong brands from problem brands?

- What kind of issues or peculiarities can be identified in the associative networks of the investigated brands?

1.3.3 Delimitations

As previous investigations of brand concept maps (John et al., 2006; Till, Baack and Water-man, 2011; Schnittka et al., 2012) have examined brand equity based on a customer-centric approach, namely customer-based brand equity (Keller, 1993), this work also aims to investi-gate brand value from this point of view. Thus, a clear distinction from financial evaluation methods of brand equity is made. Nevertheless, by introducing the concept of brand value

chains (Keller and Lehmann, 2003) the relationship of customers’ mindset to a brand’s

finan-cial performance will be illustrated and thus ultimately make a reference to finanfinan-cial equity. Due to the fact that several brand associations within a customer’s brand associative network relate to its product category, the resulting BANV figures are meaningful for comparisons of brand or services within the same category or different customer segments of the same brand, but expressly not for comparisons of brands from different product categories. Hence the re-sulting BANV figures do not strive to serve as a benchmark for investigations of brands from other categories.

As the consideration of unique associations has to be made dependent on the investigated product category, a higher weighting for unique associations compared to shared associations automatically excludes so-called generic brands and their corresponding product categories from consideration. Generic brands are considered as brands whose brand name is being used as a synonym for the original product category the brand is operating in (Low and Blois, 2002) and thus permit no separation of brand specific- and category specific associations.

1.4

Contribution

1.4.1 Theoretical contribution

The BANV metric is a first attempt to reflect consumers’ evaluation of brand association networks (both for single associations and the overall network). The quantitative approach enables managers for critical consideration over time and across different segment- and user groups. Thus Keller's call for a methodology for comprehensive representation of consumer brand association networks (1993) has been met and the method is expected to serve as a suitable tool for studying brand association networks.

The aim of this work is to investigate the calculation basis of the BANV for incorporating factors which are assessed as having a significant impact on brand equity. Whether unique-ness of associations can make a contribution to the calculation model or not has to be tested. For this purpose, a conceptual extension of the BANV calculation model will be delivered and the new BANV2 metric introduced. Based on the results of this work, a confirmation or debilitation regarding the criticism of Romaniuk and Gaillard (2007) and Sharp (2011) on uniqueness of associations as requirement for CBBE will be provided.

Based on Keller’s conceptualization of CBBE and the criteria for associations to contribute to CBBE, this study presents an in-depth analysis of a currently largely untapped marketing research tool. Since the method has only barely been applied in practice yet, there have been no investigations for its applicability regarding identification of severe issues in customers’ brand perception. As the evaluation stage of Schnittka et al. (2012) delivers promising start-ing-points for a systematic and quantitative evaluation of each single association within a brand association network, this work aims to enhance experience and knowledge on how struggling brands can be investigated for their key disruptive associations.

1.4.2 Managerial contribution

From a brand manager's point of view, the results of this work can be of practical use and hence considerable value. If it is possible to develop a metric which can reflect changes in consumers’ brand maps, brand managers could make use of it as a performance measurement scale which indicates a brand's current state of public perception. Such a monitoring system is likely to increase a companies’ ability to be “close to the market” with its market sensing instruments. In-depth knowledge of associations and their embedding in associative networks can assist managers to proactively develop measures which are adjusted to customers’ cur-rently perceived brand perceptions. Since both the BANV, as well as the in the course of this work introduced BANV2 metric, permit comparisons within a specific product category, a strategic evaluation of a brands' current positioning against key competitors should be pro-vided. As Schnittka et al.’s call for a metric which incorporates higher weightings for brand specific associations (2012) is followed by introducing the BANV2 metric, the identification of such brand specific and at the same time category specific associations is a necessary step. Once a company knows which features and associations are currently unique or untapped in a certain product category, attempts to differentiate from competitors can be derived. Further-more, an assessment of associations in terms of favorability can be of particular interest when associations are revealed that once have been valued by customers, but due to a dynamic market context and temporal progress have changed over time towards the negative.

1.5

Structure/outline

Following this chapter, the literature review of this work is placed in chapter 2. Thereby the concept of customer-based brand equity (Keller, 1993) as conceptualization for explaining consumers’ relationships with brands is explained in sub-chapters 2.2 until 2.3. In order to connect the findings and concepts relating to a customer-centric approach of brand equity to a

brand’s actual financial performance, chapter 2.4 illustrates Lehman and Keller’s (2003) con-cept of brand value chains.

Consumer mapping methods as graphical approaches to represent consumers’ brand associa-tion networks are introduced in chapter 3. Thereby a conceptual extension of the BANV met-ric as defined by Schnittka et al. (2012) is provided and an alternative metmet-ric, the BANV2 introduced. In chapter 4 the hypothesis of this work are presented, before in chapter 5 the methodology used to set up the study for evaluating the potential of BANV2 is explained. In logical order the corresponding results are presented in chapter 6 and discussed in detail in chapter 7. Finally, the most important findings of this work are drawn in the conclusion.

2

The value of brands

As the identification of problem brands requires knowledge about how brands can succeed or fail in creating value for consumers, it is important to understand how brands are processed in customer’s minds and how they contribute to the business goals of a firm. Brand equity is what describes this value of a brand and is considered in general as "the set of brand assets

and liabilities linked to a brand that add to or subtract from the value provided by a product

or a service" (Aaker, 1991, p. 15). While under this definition also financial-based brand

eq-uity is implicitly included, the focus in the course of this work is placed on a customer-based approach, the so-called customer-based brand equity (CBBE) by Keller (1993). The follow-ing paragraph discusses customer-based brand equity and explains different types of associa-tions which are prerequisite for its development.

2.1

Associations

Both Aaker (1991) and Keller (1993) build their models of brand equity based on brand asso-ciations staying central and as the most important key components of consumers’ mindsets. Associations are seen as basic construct of human’s semantic memory structure (Anderson, 2013). Consisting of a set of nodes and links, semantic memory or knowledge is considered as a dynamic system in which stored information (conceptualized by nodes) can be connected via links to other entities of stored information (Anderson, 2013). As they are connected by links with each other, an activation of one node can spread to another one. The more often two events occur together, the stronger consolidates the link between the nodes. Ultimately the strength of the links determines the extent of this spreading activation and thus prescribes which kind of information is likely to be retrieved from memory (Keller, 1993). According to this, the associative network memory model conceptualizes brand knowledge as a node of a brand in customer's memory which is linked to several associations (Keller, 1993). For the

graphical illustration of associative networks, nodes (representing stored information) are represented by circles, while links between information nodes are shown as lines. Consider-ing associative networks of brands, the respective brand note stands central in such a network and is either connected directly to other nodes of stored information, or via loops of interme-diary information nodes.

As associations stand central in the process of creating brand equity (Aaker, 1991; Keller, 1993), they enjoy high attention by practitioners whose intention it is to align marketing strategies and campaigns specifically to the associations perceived by a majority of the rele-vant target group. Associations are seen as basic components for creating a brand’s image in consumers’ minds (Aaker, 1996), and have the potential to create beneficial attitudes and feelings toward a brand (Keller, 1993). Also for the consideration and overall strategy of brand extensions, associations determine whether certain product categories are exploitable or not for specific brands (Aaker, 1991). Due to their high importance for branding, Aaker calls associations loftily “the heart and the soul of the brand” (1996, p. 8).

2.2

Customer-based brand equity (Keller, 1993)

The presumably most significant development of brand equity concepts was delivered by Keller (1993). Customer-based brand equity (CBBE) is defined as “the differential effect of

brand knowledge on consumer response to the marketing of a brand” (Keller, 1993, p. 8).

This means that brands with strong CBBE are superior in triggering favorable consumer re-sponses (as a reaction to one element of their marketing-mix) compared with other brands who offer the same product or service and using the same element of the marketing-mix. CBBE represents this marginal value difference of the superior brand in comparison with an un- or other-branded competitive product or service. Thus, CBBE represents a clear competi-tive advantage and ultimately results in higher sales, larger profit margins, lower costs, and bigger market shares (Keller, 1993).

2.2.1 Customer-based brand equity pyramid

Keller describes the development of customer-based brand equity as a hierarchical process. Thereby four different brand building blocks are distinguished, with so called brand salience standing on the bottom and thus representing the most fundamental prerequisite for climbing up the CBBE-pyramid (figure 1). Keller uses the term salience in order to specify recognition and awareness in customer’s minds. Only when brands are recognized, they can start to build meaning to consumers. Brand salience is determined by the accessibility of brand-related thoughts when thinking about a certain product category (recall), or by the fact how fast a brand can be detected (for instance when seeing it at the point of sales). Serving as first re-quirement, salience is located on the level of identity (“Who are you?”).

Figure 1 Customer-based brand equity pyramid (Keller, 1998)

Once a certain level of brand awareness is achieved in customer's minds, the next step in climbing the pyramid is designated by the level of meaning ("What are you?"). By leaving the state of pure recognition, brands start to transfer important information regarding what they stand for (Bremer, 2010). Depended on whether the considered brand can be described as a functional or symbolic brand, performance- or image-related associations are elicited. Thereby functional brands are getting evaluated by their performance, whereas symbolic

brands are used to express oneself to others (Bhat and Reddy, 1998). Since functional brands are mainly perceived through rational evaluation criteria, and symbolic brands based on emo-tional routes, the pyramid is divided in the middle. Performance and judgments are consid-ered as part of the rationale route and stand on the left side, while imagery and feelings repre-sent the emotional route standing on the right side.

Whereas the level of identity was related to brand awareness, the second level is based on brand image and perceptions as reflected by brand associations held in consumer memory (Keller, 1993). At this stage associations determine basic requirements by which have the potential to be connected to other nodes of information. Thus, this level provides the basic foundation for all other processes and steps that are accessible for a brand, in order to build relationships with its individual consumers.

At the third level of the hierarchy, judgments (about the performance of functional brands) and feelings (about the inner imagery of symbolic brands) are developed. They are based on associations and represent consumer's personal response to them ("What about you?").

The final stage of the CBBE pyramid is designated by the resonance between the brand and the customer. Thereby information of brand image are directly linked customers’ self-perception (“What about you and me?”) and thus represents the highest desirable state a brand can achieve. Once arrived at this ultimate stage of the CBBE pyramid, customers are characterized by high commitment toward the brand, having high brand loyalty and therefore are more willing to spend money in terms of repetitive sales (Keller, 1993).

2.2.2 Brand image

Keller (1993, p. 3) defines brand images as “perceptions about a brand reflected by the brand

associations held in consumer memory”. Thus, the meaning of a brand is not determined by

classify the connected nodes (associations) due to their role for the emergence of brand equi-ty, Keller distinguishes favorabiliequi-ty, strength and uniqueness of associations as key dimen-sions for brand knowledge. Before considering these three key dimendimen-sions more in detail, different types of associations as defined by Keller (1993) are briefly introduced.

Types of brand associations are classified into attributes, benefits, and attitudes. Thereby attributes describe a product’s or a service’s descriptive features which characterize them. By further sub-dividing attributes into product related attributes (features which are inevitable for performing the product’s or service’s basic functions) and non-product related attributes (external aspects such as price information, packaging/product appearance information, user imagery, and usage imagery), Keller (1993) covers a wide range different variants of attrib-utes. While product-related attributes and non-product related attributes like price infor-mation and packaging/product appearance inforinfor-mation are quite easy to overlook, the sub-categories user imagery and usage imagery represent areas which are often hidden behind abstract classification schemes and therefore are more difficult to identify.

Benefits represent the second type of associations. They describe the personal value a

con-sumer receives from the product or service attributes. They also are further sub-classified into

functional benefits, experiential benefits, and symbolic benefits. Functional benefits represent

the intrinsic advantages of a product or service and usually correspond to the product-related attributes (Keller, 1993). Experiential benefits relate to customers’ personal triggered by the use of the product or service. Thereby in experiential needs such as sensory pleasure or cog-nitive stimulation are satisfied (Keller, 1993). Finally, symbolic benefits refer to the extrinsic value of a product or service and aim to satisfy primarily social needs like self-expression or social approval.

Finally, Keller defines brand attitudes as third type of associations and relating to the overall assessment of a brand. A widely accepted approach to measure brand attitudes was delivered by Fishbein and Ajzen (1975) via a multi-attribute model. Thereby attitudes are conceptual-ized as a multplicative function of salient beliefs a consumer has about a brand (i.e. the extent to which consumers perceive that a brand has certain attributes or benefits) and the evaluative judgments of those beliefs (i.e. how favorable these attributes or benefits are) (Keller, 1993). However, Petty and Cacioppo (1986) have shown that attitude can also be the result of more inconsiderate decision making processes, such as simple heuristics and decision rules.

2.3

Criteria for customer-based brand equity

Now that attributes, benefits and attitudes were presented as the different types of associa-tions, the three decisive criteria for associations to build customer-based brand equity are discussed in the following. These are the strength, favorability, and uniqueness of

associa-tions. It should be noted that these, are also part of the brand image. Since all three criteria

are necessary to build CBBE, a lack of one of those criteria might create serious problems for a brand’s brand image und thus can be the origin for issues that make up a struggling brand.

2.3.1 Strength of associations

The first criterion for a brand to develop CBBE is to have associations with an appropriate

strength. As already described in section 2.1, brand associations can be described as all nodes

that are connected with the brand node in a brand associative network. Thereby not all asso-ciations come equally easy to mind. The thickness of the links between two elements in the classic associative network model conceptualizes how strong these two elements are connect-ed with each other in consumers’ memory and thus how likely it is that they are accessconnect-ed together (Anderson, 2013). The formation of related associations is created through observa-tion of two elements occurring at the same time or in the same context. How often and under what conditions a person records these two elements together impacts the strength of their

relation. Therefore, the degree of repetition affects directly how easy the connected infor-mation will be retrieved from memory in the future (Keller, 1993).

According to Keller (1993), the strength of brand associations can be increased when prod-uct information are processed carefully and they relate to already existing brand knowledge. Whether a customer processes product information more in-depth or not depends on his val-ues and assessment of personal relevance of the product. Thus, from the perspective of a brand manager it is of high importance to address the question of how to make the product for customers relevant. According to Walker and Olson (1991), attributes and benefits are most relevant for consumers when they are connected to their terminal values (e.g. social recognition or happiness). Thereby attributes and benefits need not necessarily be connected directly to the terminal values, but can also via a multi-step process of abstraction which des-ignate Walker and Olson as "means-end chain".

Further Keller (1998) notes that associations that are built on indirect means such as adver-tisement or other forms of marketing campaigns are not characterized as having the same strength as those that have been created via direct experience. Thus, Bremer (2010) concludes that companies rather should focus on the deliverability of product promises, than on extend-ed marketing communications.

2.3.2 Favorability of associations

Dacin and Smith (1994) consider the favorability of associations as the most important com-ponent for measuring brand equity, since it has an empirically proven impact on the relevance of attributes (Keller, 1993). This means that attributes which have been evaluated by con-sumers as very favorable are most likely to be evaluated as very relevant as well. Marketing campaigns succeed in creating favorable associations when consumers are getting convinced that the product attributes and benefits contribute to the satisfaction of their needs and wants (Keller, 1993). However, favorable associations also need to be successfully delivered by the

product, since brand associations relate to expectations regarding its potential performance (Bremer, 2010).

For the consideration of problem brands favorability issues in brand association networks are of particular importance. Negative media buzz and/or bad personal experiences can lead to unfavorable brand associations and may become difficult to control (Bremer, 2010). As the term favorability can cover both positive and negative information, it can be also the source for a negative brand image. In psychology and consumer behavior it is widely accepted that positive and negative associations are not of the same impact. Peeters and Czapinsky (1990) have studied the influence of positive and negative information and empirically proven a greater impact of negative information compared to an equally intense positive stimulus. From a business point of view, this bias is an ominous potential, since due to the negativity effect a long and arduous constructed positive brand image can be destroyed much fast faster time than it took to build it.

2.3.3 Uniqueness of associations

In section 2.1, the structure of associative networks was discussed. Associations refer to a connection between two or more entities of information in human memory. Since the com-pounds in associative networks can be very numerous it stands to reason that some associa-tions may be shared by several brands. As brands which are competing with each other nor-mally belong to one and the same product category, many (product category specific) asso-ciations will be shared by them. Since the set of assoasso-ciations which are unique for a brand directly corresponds to its positioning in consumer’s minds, Krishnan (1996) considers unique associations as an indicator for brand equity.

In order to differentiate from competitors, unique associations are seen as key requirement for positioning within a certain product category or competitors from other categories (Keller, 1993). Since positioning is elementary in marketing strategy for gaining and retaining a

sus-tainable competitive advantage, brand managers need a rich understanding of which of the brand associations can be attributed to the brand exclusively, and which are shared with com-petitors (Keller, 1998). Thereby the points of difference can distinguish differentiation criteria in a brand association network which enable the brand to leave the boundaries of its product category. As points of difference describe a brand's differentiating assets in consumer's asso-ciative networks, they are not constricted to be exclusively unique. In order to clearly differ-entiate from competitors also strong and effective associations are required that make cus-tomers believe that certain attributes and benefits are exclusively provided by this specific brand and not by any other competitors. Keller (1998) defines three criteria for sustainable points of differences. First, they have to be desirable (based on the customer’s personal rele-vance). Second, they need to be superior compared to competitive offerings. Finally, the promised attributes and benefits have to be deliverable and proofed by actual product perfor-mance.

In order to be allocable to a particular product category, a brand needs associations which are shared with competitors (Krishnan, 1996). Points of parity describe such shared associations within a specific category and facilitate the assignment and categorization of brands. Without knowledge about the associated product category, consumers may have difficulties in pro-cessing the meaning of the brand. Keller (1998) distinguishes two different types of points of parity. The first one is denominated as category points of parity and incorporates all associa-tions which can be used to assign a brand to a specific product category. Since points of pari-ty can be determined by technological equipment lines, trends and general developments, they can change over time. So benefits and attributes which were once unique for a certain brand may have been used to differentiate from competitors in the past. However, in the course of time they may have been imitated and thus transformed from points of difference into points of parity. This process is exactly what Keller describes as the intention behind the

second type, competitive points of parity: by taking away competitive advantages through imitation, brands can bridge the gap to competitors who have been in the lead. Thereby they can strengthen their own position and weaken those of the competitors at the same time. Both concepts illustrated the importance of unique brand associations within one product category as they are widely seen as indispensable for gaining and achieving sustainable competitive advantages in marketing research (Broniarczyk and Gershoff, 2003; Porter, 1976).

However, uniqueness of associations as Keller’s last requirement for CBBE-building associa-tions have served as starting point for a controversy among marketing researchers. In particu-lar, members of the Australian Ehrenberg Institute such as Sharp (2011) and Romaniuk and Gaillard (2007) conducted empirical studies in which they have shown that in many product categories such those of soft drinks, computers, or mineral water unique associations do not have an impact on a brand’s performance in terms of market share and choice preference. As a conclusion of these findings, the authors expect that consumers’ perceptions of uniqueness are either temporary or quickly usurped by competitor activity (Romaniuk and Gaillard, 2007). Accordingly, the most significant managerial implication derived from their findings is, that brand managers should focus on providing cues which make the brand allocatable to its specific product category. Since unique associations might hinder consumers from retriev-ing the brand when thinkretriev-ing about its category, a focus on the points of parity could ensure sufficient availability in terms of consumers’ potential to retrieve the brand from memory. To support their conclusion, Romaniuk and Gaillard (2007) argue that even brands as Coca-Cola or Sony were incorporated in their investigations and it turned out, that uniqueness was not a major contributor to their large brand equity (when brand equity was operationalized as mar-ket share). Reviewing the book of Sharp (2011), which makes use of the studies of Romaniuk and Gaillard (2007) as argumentation basis for doubting the role of unique associations as necessary prerequisite for CBBE, Wilson (2011) notes an inherent bias toward fast-moving

consumer goods (fmcg) brands in their research setup and warns against challenging too many well-established marketing foundations at once.

2.4

Brand value chain

As already indicated in chapter 1, the investigation of brand value in the course of this work relates to a customer-centric perspective. In order to understand the impact of brand value in customers' mindsets on the economic success of a company, the concept of brand value chains (Lehmann and Keller, 2003) an illustration of how intangible customer appreciation forms the basis for the actual financial performance of a company.

The brand value chain as defined by Keller and Lehmann (2003) is based on the premise that brand value ultimately resides with customers. This means that starting point for the build-up of brand value is represented by marketing program investments which aim to affect custom-ers’ mindsets. As described in chapter 2.2, a customer’s mindset reflects knowledge and emo-tions about a brand, which in turn subsequently create outcomes that determine a brand’s performance in the marketplace. As the aggregated quantities purchased by consumers, and prices that they were willing to pay are crucial indicators about a brand’s expected financial incomes in the future, these data (besides other factors such as replacement cost and acquisi-tion price) are used for assessing the shareholder value in general and brand equity in particu-lar (Lehmann and Keller, 2003). Thus, the four value stages in the model are represented by a brand’s marketing program, customers’ mindset, brand performance and shareholder value.

Figure 2 The Brand Value Chain (Lehmann and Keller, 2003)

However, Lehmann and Keller assume that there are further factors (“multipliers”) moderat-ing the value transfer between each stage. The first multiplier moderatmoderat-ing the value transfer between marketing program investments and customer mindset is designated as program

quality multiplier. As the effectiveness of marketing campaigns is dependent on several

fac-tors, the sole amount of financial investment committed to the program does not ensure suc-cess in terms of brand value enhancement (Lehmann and Keller, 2003). Though, a cam-paign’s clarity will influence whether consumers interpret the conveyed meaning as intended by marketers. Also the campaign has to transfer relevance to consumers in order to attain their serious consideration. For the scope of this work it is particularly noteworthy that the authors also include the uniqueness of marketing programs as decisive factor for assessing its effectiveness, as it would determine how differentiating it is from other campaigns. Finally, the consistency of the program may enhance or decrease the value creation by its fit with past campaigns and developments in customers’ brand perceptions.

The crucial key dimensions of the customer mindset are brand awareness, brand associations (representing strong, favorable and unique attributes and benefits of the brand), brand atti-tudes, attachment to the brand, and brand activity as the extent to which customers use the brand as starting point for interactions with others (Lehmann and Keller, 2003). The as just mentioned order of dimensions represents the hierarchy by which the value creation process is developing in customers’ mindset. Connected via the marketplace conditions multiplier (which relates to competitive actions, customer profiles, and the support of distribution chan-nels and intermediaries in general) the brand performance points out how customers respond in the given marketplace. Economic variables such as price premiums, price elasticities, mar-ket share, cost structure and ultimately profitability measure how the brand is performing in its competitive environment.

However, the brand performance is not the exclusive determinant for assessing the share-holder value. Financial analysts also have to consider external factors such as general market dynamics, the brand’s growth potential and its risk profile. The moderating role between brand performance and shareholder value is called investor sentiment multiplier and incorpo-rates all above mentioned factors. Finally, the shareholder value is assessed based on the course of value creation and its stages and multipliers as stated in the brand value chain mod-el. Thus, the brand value chain model illustrates how customer-based brand equity as intangi-ble assets located in customers’ mindset has strong implications for a brand’s concrete finan-cial valuation.

3

Consumer mapping

“Due to their fundamental importance, measurement of brand associations is at the centre of

brand management” (Till, Baack and Waterman, 2011, p. 93).

In chapter 2 the basis for construction, composition and relevance of brand associative net-works were presented. In this chapter, methods are introduced which enable brand managers to get deep insights into consumers' brand associative networks. Thereby the focus is on ana-lyzing single associations, as well as overall associative networks of brands. A first approach for analyzing associative networks, the network analysis, was provided by Henderson, Iacobucci and Calder (1998). John et al. (2006) build on their approach and provide with the so-called Brand Concept Maps (BCM) a method that aims for graphically illustrating con-sumers’ brand associative networks. Thereby, respondents rely on a given set of associations and are asked to align their individual brand map by connecting their chosen associations with each other and the brand node staying central in each map.

3.1

Advanced Brand Concept Maps / BANV

Dacin and Smith (1994) consider the favorability of brand associations as the most important part of building brand equity. Also Krishnan (1996, p. 393) points out that “a strong brand

should focus on consistently achieving net positive associations”. The more surprising is the

fact that favorability has enjoyed inadequate consideration in the previous BCM approach. Schnittka et al. (2012) took the opportunity to expand the traditional BCM approach by inte-grating a further stage in the codes of procedure: an evaluative judgment for every associa-tion.

The evaluative judgment of respondent j for association a (Eaj) is queried by a 7-point Likert scale and relates to the favorability of associations. By entitling each association neutral and then adding the phrase “is good” (e.g. “The customer support by <brand name> is good”)

both positive and negative evaluations are possible (i.e. 1=totally disagree; 7=totally agree) (Schnittka et al., 2012). However, since not all associations are relevant for a purchase deci-sion (Keller, 1993), participants also assess the respective importance (Iaj) of each associa-tion by a 7-point Likert scale (e.g. "A good customer support is an important criterion for choosing a mobile network provider"). In the consensus map evaluative judgment and im-portance are both graphically displayed. The imim-portance of the association for a specific pur-chase decision is reflected in the size (diameter) of each circle around the associations (the larger, the more important). Thereby the average of all individual assessments is used as the decisive criterion. For the evaluative judgment of each brand association, also the average of all participants' indications is used and thereby determines the darkness of each association circle (Schnittka et al., 2012). Figure 3 shows an example of an Advanced Brand Concept Map by illustrating the consensus map of participant’s associations regarding the vehicle type

Golf by Volkswagen. According to the map, the associations usualness and prominence are

evaluated as highly applicable and favorable. However, as particularly important for the pur-chase decision of the vehicle, predominantly the associations quality and reliability have been specified.

In a next step, Schnittka et al. (2012) introduced a new measurement metric to quantify con-sumer’s overall brand association networks regarding its overall favorability for consumers: the so-called Brand Association Network Value (BANV) is calculated based on the multi-attribute model by Fishbein and Ajzen (1975) which suggests that brand attitudes can be seen as a multiplicative function of cognitive beliefs consumers have about a brand and the evalua-tive judgments of those beliefs (Schnittka et al., 2012). To determine the BANV of an indi-vidual consumer j (BANVj), first all evaluative judgments (Eaj) of all associations m (as direct indication of the favorability of an associative network) are summed up. As activations of nodes within an associative network (i.e., brand associations) are expected to jump over and activate nodes which are directly connected to them, strong links between associations in-crease the probability of reciprocal activation (Keller, 2008). Thus, Fishbein and Ajzen (1975) propose a weighting of each association’s evaluation by its strength of linkage to the brand node itself or to the next superordinated node within an associative network. According to that, all evaluative judgments are multiplied with their perceived strength of connection (Saj) to the brand node or the next superordinated association (7 = triple line connection; 4 = double line connection; 1 = single line connection). If the association is connected to several objects, the average is used.

In order to further enhance the predictive power of the model, Schnittka et al. (2012) also added a factor for the importance of each association in a relevant purchase situation (Iaj), as well as one for the distance of each association to the brand node (Laj): As indicated by Keller (1993), not all associations are equal important for consumer’s choice in a purchase situation. Even though there might be associations which are assessed as very favorable, they do not inevitably have to be of strong influence for consumer’s purchase decisions. As it is empiri-cally proven that the inclusion of such weights for importance can increase the predictive

power of multi-attribute models (Lehmann, 1971), the authors added it as further calculation component of the BANV.

The last component of the model is represented by a weight regarding the relative position of each association (Laj). As information nodes in associative networks vary in their pathway to the brand node and can either be connected directly to the brand node, or via the loop way of other associations, the impact of associations which are associated very close to the brand is expected to be higher. The layer Laj is thereby called association level and also converted to a 7-point Likert scale (7 = first layer and directly connected with the brand circle; 4 = second layer and indirectly connected via one other association with the brand circle; 1 = third layer and indirectly connected via two other associations with the brand circle).

Formula 1

BANV

=





m

a 1

E

S

I

L

Formula 1 shows how the BANV for an individual consumer j is composed. By comparing the predictive power of the BANV metric with more simplified models (considering less components), it was empirically proven that each component contributes to the nomological and predictive validity of the metric, since the strongest correlations between BANV and de-pendent variables like overall brand attitudes or purchase intention have been observed when all four components were included (Schnittka et al., 2012). In addition, it was shown that the increasing degrees of correlations are not to be referred to random deviation, but express sig-nificant differences.

Table 1 Intercorrelations of BANV{Eaj, Saj, Iaj, Laj}, BANV{Eaj, Saj, Iaj}, BANV{Eaj, Saj} with the validation criteria brand familiarity, purchase intention, and brand attitude by Schnittka et al. (2012)

The managerial implications of the Advanced Brand Concept Maps and the BANV are of high utility. As the original BCM approach of John et al. (2006) does not consider favorabil-ity of associations, no implications about customers’ appreciation of individual associations can be derived. By the introduction of the evaluation stage, strategic brand management is enabled to identify critical brand associations within the brand map based on empirical find-ings. As their closeness to the brand and interdependencies to other associations are also graphically displayed, brand managers are empowered to proactively focus on areas within associative networks which are assessed as unfavorable or stay in conflict to the brand’s de-sired positioning in the marketplace. The introduced procedure for calculating the BANV provides a metric for quantitative evaluation of associative networks on an individual level (for each consumer j). Thereby developments over the course of time can be captured and deliver useful information regarding the effectiveness of conducted marketing campaigns. Furthermore, different customer groups can be compared to each other. In order to enhance

consumers’ brand images and thereby also the BANV of a brand, marketing managers can take measures to link favorable nodes with each other, strengthen existing linkages or try to displace unfavorable associations out of the network. As the set of associations within a net-work is determined by a considerable extent by the brand’s product category, the ABCM / BANV procedure is particularly suitable for comparisons of customer segments and/or com-peting brands within one category. This is not a limitation, but an advantage for the compara-bility of direct competitors. From this angle, the uniqueness of associations is essential to determine the brand’s actual position within one product category.

As the BANV considers favorability (Eaj,Iaj) andstrength (Laj,Saj) of associations, two of Keller’s criteria for building customer-based brand equity are incorporated. Although the third dimension, uniqueness, was also named as considered in the formula, but the explana-tion for its influence remains very vague: “(…) the multitude of brand specific associaexplana-tions

within a network m might provide a proxy for the level of uniqueness of brand associations

within a network” (Schnittka et al., 2012, p. 10-11). As one can see, the sheer amount of

as-sociations shall give conclusions about the uniqueness of individual brand asas-sociations or even the overall associative networks of brands. This is clearly a strong simplification and a pitfall of the ABCM method and the associated BANV metric. This approach is not defined by objective measures, neither it seems logical that a high amount of associations inevitably has to lead to a unique association network. Also in reverse, few brand-specific associations cannot be regarded as a sufficient condition for a not-unique associative network. A tie to this pitfall is made in the course of this work and aims for a first approach to incorporate unique-ness of associations as Keller’s third requirement for CBBE in the evaluation of brand maps and brand-associative networks.

3.2

Category-specific and brand-specific associations: Strategic Concept

Maps (SCM)

As already mentioned, the method of the Advanced Brand Concept Maps and the calculation of the BANV are based on assessments and weightings that are determined to a considerable extent by the product category in which a brand operates. For example, it is easy to imagine that brands from the product category of children's toys create a richer and better evaluated set of associations than those of a generic mineral water brand. For this reason, the traditional BANV method is particularly suitable for comparisons of customer segments of a single brand or comparisons within a certain product category (Schnittka et al., 2012). To further improve the insights within a specific product category and in order to benchmark with rele-vant key competitors, the methodology of Strategic Brand Association Maps was developed by Till, Back and Waterman (2011). Since this methodology delivers interesting approaches in considering associations which are exclusively attributable to the brand (as opposed to those of its product category), the main idea should be briefly described:

The procedure of Strategic Brand Association Maps is conducted with four to six brands of the same product category. Thereby the strength of associations is measured by a response latency task which records participants’ speed of response for every given association (Till, et al., 2011). It is interesting to note that Keller already called for an approach which is based on response times in 1993. In the next step the participants assess every association with regard to favorability, relevance and uniqueness by using a 7-point Likert scale. The dimensions favorability and relevance correspond to the evaluative judgment and the importance of the BANV method.New is that respondents are prompted to assess associations in regard to their uniqueness.

Furthermore, the authors consider two more dimensions of brand association networks: the

effectiveness of advertising activities by Krishnan (1996); and second, the respective

rele-vance of each brand association, in order to ensure that brand managers can focus on

associa-tions which are meaningful for customers. However, the most important change that is made in the Strategic Brand Association Maps is the creation of two different brand maps: first, the

strategic brand association map is conducted as consensus and aggregation of all

partici-pant’s associative networks and includes all frequently mentioned associations. However, the second map is characterized by the elimination of associations that are shared with other brands of the product category. The resulting core brand essence map is thus clearly arranged and particularly stands those associations in the centre that illustrate the points of difference of the examined brand. As a result, a clearly arranged brand map is provided with focus on core-elements of a brand’s positioning in public perception. In order to be identified as a member of a certain product category, every brand needs associations which it shares with its competitors (Krishnan, 1996). However, too large overlaps with other brands can complicate or even totally impede a clear positioning and differentiation within a given product category. In order to build up a brand image which is hard to imitate and thus less vulnerable against competitive actions, associations which distinguish a brand from competitors need to be strengthened and defended.

As already mentioned in the introduction of this work, individual empirical studies such as from Romaniuk and Gaillard (2007) conclude that unique associations do not deliver added value for customers and therefore marketing campaigns should focus rather on category-specific features. Thereby they stand in contrast to the assumptions and basic principles of influential marketing researchers like Aaker (1991), Keller (1993) or Krishnan (1996). A cru-cial factor that can explain the different viewpoints in this context is to be considered the product category. As product categories exist in which awareness and availability are the most important drivers of a brand’s market performance, uniqueness not compulsory need to

be a source of competitive advantages (Sharp, 2011). However, differentiation can build market based assets which last until competitors copy them (Sharp, 2011). Thus, product cat-egories with high innovation density and high degree of differentiation represent those mar-kets in which uniqueness is likely to be a source of crucial importance.

3.3

Adjusted weighting of the new BANV2 formula

As described in the previous chapters, the concept of the ABCM procedure does not suffi-ciently consider uniqueness, one of Keller’s prerequisites for building customer-based brand equity. However, the concept of Strategic Brand Association Maps by Till et al. (2011) en-counters this gap and puts the focus on illustrating consumer brand associative networks which are adjusted from product-category specific associations and thus strongly emphasize brand-specific associations. This approach should be covered by the calculation of the brand associative network value (BANV), but without deleting category-specific associations at all. As proposed in the limitations of Schnittka et al. (2012), a stronger weighting for brand-specific associations, relative to those which are product-category related, are to be made since unique associations differentiate the brand from competitors and are crucial for the build-up of USPs. Thereby first unique associations have to be indentified and distinguished from the set of all associations of a brand. As the procedure of Till et al. (2011) starts with comparisons of four to six brands of one category, the identification of unique and shared associations can be conducted in the beginning. Another option is to fall back on prior free association data. Once unique associations are distinguished from shared ones, Schnittka et al.’s call for higher weights of brand specific associations (2012) can be conducted.

As adjustments in the BANV metric should only be made when unique associations are in fact relevant for consumers, it is contemplated that only those unique associations are weighted more heavily which possess an at least moderate degree of importance (Iaj). Adjust-ed to the 7-point Likert scales, a minimum score of four is definAdjust-ed as decisive criterion about

whether an unique association should be granted extra weight. If consumers do assess a brand specific association as particular important, its favorability is the last determinant to define the strength of the weight. Thus, the influence of unique associations is set in dependence to their respective favorability, since it is assumed that the perceived degree of favorability is crucial for determining the criticality of associations distinguishing the brand from competi-tors. For example, a unique and differentiating association which is assessed as highly favor-able by the majority of consumers is seen to deserve a particular high weight when calculat-ing participants’ overall brand network value. In contrast, brand specific and unfavorable associations are regarded as having the potential to decrease the overall network value in par-ticular. Thus, points of differences between a brand and its competitors are seen as providing both opportunities (when favorable) and risks (when unfavorable). Distinguishing between both conditions reflects an entirely new approach in the overall evaluation of consumers’ brand maps and associative networks.

As both negative and positive assessments regarding favorability are possible on the 7-point Likert scale, unique and brand specific associations which are evaluated as very favorable should add particular value to the BANV, whereas unfavorable and unique associations can be seen as notably critical and thus should diminish a brand’s overall network value. This ability to diminish the overall network value (in case of brand specific and unfavorable asso-ciations) represents another extension of the original BANV metric, as in the old calculation principle also highly unfavorable associations added to the overall network value. Based on the same principle, an additional positive weighting of associations is possible, given they are considered as unique and as particular favorable by the majority of respondents.

Formula two shows how the additional weights for unique associations (U) contribute to the overall network evaluation. When association aj has a minimum score of four regarding its relevance for purchase decision (Iaj), its favorability is decisive for determining the final