Enter the authentic side of strategic cooperation : seeking the role of Brand [Alliance] Authenticity on consumers’ evaluations of co-branding : an explorative investigation

(1)

Master Thesis

Enter the authentic side of strategic cooperation:

Seeking the role of Brand [Alliance] Authenticity

on consumers’ evaluations of co-branding

An explorative investigation

Author: Paul Grandjean

Student number: 11664436

Supervisor: Dhr. R.E.W. (Roger) Pruppers

MSc. Business Administration - Marketing Track

(2)

Statement of originality

This document is written by Paul Grandjean who declares to take full responsibility for the contents of this document.

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

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

(3)

Abstract

Co-branding, the founding of lasting partnerships by two brands via the launch of new conjointly developed products, has become a notorious leveraging strategy. Its perks as well as drawbacks are well-known within the academic literature with most scholars tending to agree that reason for the success or failure of such items is principally due to how good or bad consumers perceive the match-up/fit between the constituents’ brand images. That is, it whether tarnishes or heightens their evaluations of the overall alliance. However, a minority of authors argue that the cognitive approach is too exclusive on its own. Alternatively, the premise of this study relies on emerging views which consider the cultural lens of branding. Building on top topical accounts of (composite) brand extensions and brand authenticities, we introduced the concept of Brand Alliance Authenticity that integrates both notions. An initial experiment proclaimed that our constructed model with conditional processing of moderated mediation represents a viable option to assess the perceived legitimacy of brand alliances and also predict if it will encounter resistance or adoption. Our findings show that when two partners are highly authentic, they enjoy better prospects compared to two lowly or one highly with one lowly authentic pairing, as the coming together in the form of the co-branded product is perceived to be more genuine. Results of a second experiment suggest that these effects are somewhat stronger for a pair based on the integrity dimension of brand authenticity. Theoretical, managerial implications and future research are discussed.

(8)

1 | P a g e

Introduction

1 The challenge of co-branded offerings

1.1 Bases of fit, quality and consumer responses

Brands represent the most valuable intangible asset for companies since the management of a brand’s core features substantially helps to achieve a sustainable differential competitive advantage by creating a specific position in consumers’ memory maps, which engenders an important source of value (Keller & Lehmann, 2006). Nonetheless, it has become popular for firms looking at further growth opportunities to engage in portfolio diversifications of their own using their brand name to reach other product classes, as well as working closely with other brands to design separate compelling products. These brand leverage strategies are respectively brand extensions and co-branding. The latter has received surging attention from scholars in recent years as a kind of brand alliance, especially since it is stated to provide the best return on investments although Helmig, Huber, and Leeflang (2008) concede that empirical demonstrations are scarce in essence. The central tenet on which co-branding relies is the process of cognition and affect transfer between seemingly unrelated entities, simultaneously combined into one new offering. Ideally, this leads to a prominent overlapping network of positively reinforcing brand associations, flowing from the original constituent brands into the composed one and vice-versa (D. James, 2005; C. W. Park, Jun, & Shocker, 1996; Washburn, Till, Priluck, & Boughton, 2015). An example is Senseo, the pioneering filter pad machine developed conjointly by the Dutch brands Philips and Douwe Egberts back in 2001. The allied competences of their respective expertise in electronic home appliances

(9)

2 | P a g e

and coffee gave birth to a well-received ‘coupled’ brand, still sold in many countries worldwide (Bouten, Snelders, & Hultink, 2011). While this illustration exalted a fine chemistry, it turns out that in practice only an estimated one out of five brands’ pairing attempts are successful (Van den Bergh, 2014).

The literature on brand alliance is broad but concentrates around the outside-in consumer-based brand equity stance as duly noted in Gammoh and Voss (2011). Previous research in this area of interest has highlighted congruity in distinct variations and at multiple hierarchical levels of abstraction [brand fit but also product fit, category fit etc.] as the necessary component that allows to understand and prevent failure of cooperative projects (Ahn, Kim, & Forney, 2009; Czellar, 2003; Völckner & Sattler, 2007). Moreover, this – according to Tauber (1993), ‘blindly’ embraced notion of perceived fit - tends to have implications on desired branding outcomes of each party in regards to their bottom lines; such as attitude towards the individual partner brands, willingness to pay a premium price, word of mouth and purchase intentions of the cobranded item (Belén del Río, Vazquez, & Iglesias, 2001; Helmig, Huber, & Leeflang, 2007). Consequently, unintended aversive feedback may arguably hinder a brand’s reputation, even dilute its image by spill over if the partnerships in which it willingly takes part simply do not ‘click’ (Baumgarth, 2004; Lafferty, Goldsmith, & Hult, 2004; Simonin & Ruth, 1998) – in others words when low-fit situations at the brands’ level occur. That is because evaluations of fit from the consumers’ point of view must eventually stay coherent after a cobranded product has been made available to them. Nowadays, it is an issue mainly discussed either in terms of similarity or complementarity of characteristics, which can serve to frame and prime contextual cues in order to convince consumers which would

(10)

3 | P a g e

facilitate their assessment of the allies’ collaboration (Levin & Levin, 2000; Swaminathan, Gürhan-Canli, Kubat, & Hayran, 2015).

1.2 Gap: limits of the brand fit clash

Fit in terms of consistency of the external brand concept – “brand-unique abstract meanings (e.g., high status) that typically originate from a particular configuration of product features (e.g., high price, expensive-looking design, etc.) and a firm's efforts to create meanings from these arrangements (e.g., "the relentless pursuit of perfection" by Lexus)” (C. W. Park, Milberg, & Lawson, 1991, p. 186) – provides quite a comprehensive explanation for when brand stretching works out or not. However, Johan Lanseng and Erling Olsen (2012) obtained opposite results for symbolic versus functional brand alliances as compared to the influence exerted on consumer evaluations in the context of brand extensions. Some moderating / intervening variables or other ways of measuring brand fit could reduce the negative effects coming from low fit perceptions to the extent that it becomes irrelevant in certain circumstances (Dens & De Pelsmacker, 2016; Milberg, Sinn, & Goodstein, 2010; Riley, Charlton, & Wason, 2015).

Strong evidence suggests that the fairly novel construct of brand extension authenticity (hereafter BEA) developed in Spiggle, Nguyen, and Caravella (2012) goes beyond both the relevance and similarity aspects perceived by consumers. In effect, BEA significantly deviates from the classic cognitive categorization theory since it is normative. In addition, it then provides hints onto why consumers favourably accept or not an entirely new product from existing brands in a more intuitive manner depending on their internal consistency (remaining true to its tenure) rather than external consistency (claiming to be something else via brand position). This distinction is capital as Beverland (2018) further expresses in his latest book.

(11)

4 | P a g e

That might be the missing piece of the puzzle where obvious mismatches at first glance, such as an H&M partnership with designers like Alexander Wang, which scores rather poorly on both factors of product feature similarities and especially [external] brand concept consistency ended up selling successfully all in all ( https://blog.hubspot.com/marketing/best-cobranding-partnerships). How about if the alliance itself was perceived as being “authentic”? That is the empty space that we address and are willing to fill.

1.3 Problem definition and statement

Building on the key insight that if consumers can appraise an extension on the relatively better and complementary basis of its constant legitimacy and contiguity thanks to BEA rather than via regular fit perceptions (Spiggle et al., 2012), would it act in the same manner when two brands are integrated through one dual branded offering? What happens if a brand perceived as authentic is involved in a commercial partnership with an inauthentic one and their combined influence on the common marketed offer’s genuineness? To date, prior investigations looking at the impact of the concept of [perceived] brand authenticity in a co-branding setting are, to our knowledge, very few if not inexistent.

The purpose of this paper is to shed light on these increasingly pressing questions as it is advanced according to this stream of thought that authenticity has risen to become the highest preferential criterion of brand choice over and above quality (Gilmore & Pine, 2007). A series of reports issued by the Cohn & Wolf1 communications agency (WPP group) supports the trend, stating that 63% of people globally indicated they would privilege to buy from a brand they see as displaying authenticity in that it is reliable, respectful and real. In order to establish a smooth linkage from brand extensions towards brand alliances about authentic

1

(12)

5 | P a g e

matters, we shall consider a couple of cases of branded components in products since it represents an optimal sort of co-branding in the spectrum between the two strategic approaches (Kotler & Pfoertsch, 2010). The interesting story of Betty Crocker (cake mix) and Hershey’s (chocolate candies) notably is a platform to disentangle the prowess of cultural meanings out of just brand images since in nature it harbours a functional as well as symbolic positioning (Askegaard & Bengtsson, 2005). Consequently, the following interrogation emerged:

Does Perceived Brand Authenticity of co-brands affect consumers’ evaluations of brand alliance, and what is the role of Brand Alliance Authenticity?

Multiple sub-questions will help to guide the resolution of this endeavour:

 What are the main differences and similarities between the brand extension and co-branding strategies?

 Why is brand authenticity and related cultural meaning(s) crucial to consider within a branding alliance in place of categorization?

 Which aspects support the process by which consumers evaluate a brand’s internally consistence?

1.4 Theoretical value

Several implications are expected. As discussed in the preceding paragraphs, there is a general consensus of why products from brand alliances embrace either approval or resistance in the markets or how consumers interpret and experience brand associations transfer amongst co-brands largely due to the appropriate fit. Often though, discoveries are somehow contradictory depending on contingencies. While signaling theory for example pretends that augmenting quality perceptions is the main goal in pursuing the quest for differentiation thanks to

(13)

6 | P a g e

commercial partnerships (Kirmani & Rao, 2000), there are various attractive reasons to leverage and call upon an ally doing business in other branches (Leuthesser, Kohli, & Suri, 2003). This thesis will bring some clarifications by contributing to both the co-branding and brand authenticity fields. Effectively, it checks the impact of authentic cues in a modified perceptual framework based on the cultural stance underscored by Brand Extension Authenticity. Furthermore, it is an enterprise that has also been acknowledged by branding guru and authority Keller (2016) himself, therefore being encouraged to follow-up in this unusual direction. The model will include and discuss inputs around the multidimensionality of perceived brand authenticity in relation to what it presently consists of, in an attempt to improve it. Finally, our research tries using non-fictitious brands to form [dis]symmetric partnerships so that our arguments can be generalized (Desai & Keller, 2002; Van der Lans, Van den Bergh, & Dieleman, 2014).

1.5 Managerial utility

Our study can be useful for strategic brand managers and those involved in marketing affairs. The information from our conceptualized hypotheses can allow to see whether principles of ‘cultural branding’ can be relied upon within a co-branding context in order to evoke the protection of “myths” attached to an iconic/authentic brand (D. B. Holt, 2004). Indeed, it would be reckless to engage in a co-branding alliance if this threatens a brand’s DNA which signifies “more than relatively fixed arrangements of associative nodes and attributes” (Brown, Kozinets, & Sherry Jr, 2003, p. 31).

1.6 Structure

In the next sections, the backbone to this research will be laid down. First, chapter two gives a brief overview of the nuances in what separates or/and links brand extensions and [diverse

(14)

7 | P a g e

sorts] of co-branding (such as ingredient branding mentioned above). It also references relevant findings of previous studies regarding the driving pillars of fit (product feature similarity and brand concept consistency) as well as the effect on consumers’ evaluations drawn from the core literature(s). The brand authenticity construct is then conceptualized in the light of its peculiarities and detailed in-depth within chapter three. These sub-units will provide the foundation and lead to generating our hypotheses in terms of the connections and expected directional relationships between them in chapter four. This will be followed in chapters five and six by two studies, enumerating the principal methodological phases before reporting their findings. Chapter seven concludes on giving concrete responses to the focused inquiry, its implications and proposes a non-exhaustive quest agenda.

(15)

8 | P a g e

Brand Extensions VS Co-Branding Alliance

2 From divergence to commonalities

This second chapter focuses on outlaying what are co-branding and brand extensions in more details. Therefore, it first highlights differentiating aspects, yet strives to describe connected theoretical themes between them, notably about consumer evaluations. After all, both follow the logic of brand equity leverage which allows a master brand’s induced equity to increase by building secondary associations from external entities (Keller, 1993, 2003).

2.1 Defining these types of leveraging strategies

2.1.1 Brand extension

A brand extension consists in the use of an established brand name into another product category than its original one. By extending a brand this way, companies can diminish the risk and related costs of building a new brand from scratch, while being able to broaden the range of their business activities in multiple markets. Especially if they are of high quality then there is plenty of room to stretch (Keller & Aaker, 1992). In order to dismiss any confusion, we add that – defined as such within our research – it refers to the exact term of brand franchise or category extensions which differentiates from other types of extensions [line extension, flanker and new products] as illustrated in figure 1 (Tauber, 1981). The initial chocolate bar Mars that was

(16)

9 | P a g e

transformed into an ice cream version is an example of such extension. Thus, brand extension seems to be limited to the study of single brands belonging to corporate groups / firms that in their vision are somehow motivated to end up generating financial profits by deriving added brand value from consumers’ perceptions (Keller & Lehmann, 2003).

2.1.2 Co-branding

From a managerial perspective, co-branding is often compared to a temporary or time-bound marital union because organizations commit around a determined goal – mostly about achieving an opportunity for growth. Indeed, in perhaps its simplest form it is “a co-operation between two or more brands with significant customer recognition, in which all the participants' brand names are retained. It is of medium to long term duration and its net value creation potential is too small to justify setting up a new brand and/or legal joint venture” (Blackett & Russell, 2000, p. 163). This adopted definition remains in line with the three main criteria identified by Besharat (2010) from the subsequent academic body of research in that co-branding is generally understood as taking the shape of 1) a consented non short term collaboration, wherein 2) the names of involved parties are displayed on the alliance package, product or logo and 3) entails the launch of a commonly conceived offering under those flags [‘its primary objective’; p.1241]. That implicitly excludes bundling strategies that have multiple individual products from various brands in one pack (Sheng & Pan, 2009).

It is important to notice that getting involved in a co-branding strategy does not necessarily equals monetized incentives nor always implies to enter a parallel class of goods or services. Moreover, it has the spanning ability to gather partners who compete across diverse customer-centred industries as well as in other contexts (institutions, B2B, cause-related marketing, events sponsorship, humanitarian support programs and so forth…) (i.e. Ahn et al., 2009; Uggla, 2006). For example, Pampers [baby care] and Unicef [charity] have been joining

(17)

10 | P a g e

hands in a regional sensitization campaign for over a decade now. Not only is their coming together mutually beneficial on paper, it allows to contribute to society by indirect donations but could potentially hurt the co-brands, considering Dickinson and Barker (2007). A reason is that such an initiative tends to raise cunning questions in regards to the true motivation behind it (Hawkins, 2011; Vanhamme, Lindgreen, Reast, & Van Popering, 2012).

2.1.3 The special case of Brand alliances embedded in brand extensions

Figure 2 summarizes the differentiation of both concepts as determined above. Also, it clearly shows that at times the separation line between a classic brand extension and co-branding becomes blurred.

Figure 2: Co-branding and brand extension (excerpt from Helmig et al., 2008)

The scope of this research is situated in the upper right part of the schema. It draws on the blended strategic approach that Park and colleagues (1996) call composite brand extensions

(18)

11 | P a g e

(CBE). It is a complex situation when existing brands decide to ally in order to create an innovative jointly developed value proposition inside a new product-market and brand it by juxtaposing their names. The conclusive remark of the study of K. Thompson and Strutton (2012, p. 23) brings support to this aspect : “when targeting brand extensions toward a divergent product category, co-branding is a promising strategy for ensuring that brand extension efforts enjoy greater prospects for success”. In other words, lone brands should primordially seek to express, besides improving quality signals via the establishment of a partnership because a reputed ally provides believable assertions and with it increased credibility, notably regarding unseen or outspoken performances (Rao, Qu, & Ruekert, 1999; Rao & Ruekert, 1994).

CBE somewhat resembles but diverges from the idea of ingredient branding as introduced in Norris (1992; 1993) since the former assumes that the parent brand and co-brand keep a certain control over, at least coordinate in sync their manufacturing process and marketing, possibly via shared resources or assets [e.g. budget agreed upon, packaging design, distribution channel, media selection, production plant..]. Regarding the latter, it is presented as an efficient way for ‘merging’ together products which might preferably be benchmark performers elsewhere in the marketplace. In short, ingredient branding “essentially comes down to a new product acquiring consumer awareness by the integration of a well known branded component” that could be borrowed from a supplier in the value chain notably (Arthur Rooney, 1995, p. 53). For example, the Intel microprocessor chips integrated inside computers which are ‘owned’ by another primary brand (Leuthesser et al., 2003). Nonetheless, it can happen that a parent brand chooses to self-brand its own secondary attributes, such as recently Samsung which devised laundry machines with Eco-Bubble clean

(19)

12 | P a g e

technology & AddWash door opening. This proprietary exploitation actually stands for an attractive, plausible option too aside from vertical integration (Desai & Keller, 2002).

2.2 Key influential factors

Tauber (1988) and C. W. Park, Jaworski, and Maclnnis (1986) amongst other seminal branding academics argued in the eighties that, in the pursuit of gaining extra brand equity points at the expense of competition, a vital consideration for brands is to avoid dispersing themselves in the perceptivity of their broader audience of present and prospective customers. Until today, the optimal requirements that they proposed are the ones receiving the utmost scrutiny from modern research around both the topics of brand extensions and co-branding.

2.2.1 Perceived fit

Uggla (2004) opts for ‘psychological congruency’ when defining fit. For that matter, the paramount role of perceptual fit has been intensely documented after the publication of D. A. Aaker and Keller (1990) whose contributions are actually very insightful also for understanding co-branding evaluations (Owen James, 2006). Empirically, fit has been found to be the highest significant direct antecedent determining the success of a brand extension (Völckner & Sattler, 2006). Besides, fit is the constant number one independent variable that delivers such impacting force within holistic frameworks that demonstrate overall reciprocal effects. For instance, it is the most influent indirect driver of change in attitudes towards the core brand from consumers’ evaluations of its extension (Dwivedi, Merrilees, & Sweeney, 2010).

Typically, fit in brand extensions is conceptually formulated and operationalized in terms of either the relationship of associations between the parent brand’s initial category and the one it extends towards, or relevance between the brand image and the category of its product

(20)

13 | P a g e

extension [ability to accommodate]. Invariably, both are rooted in cognitive categorization approaches (Spiggle et al., 2012). Murphy and Medin (1985, p. 297) criticize categorization theories, which suppose that every person makes sense of phenomena in the same manner, because these rely too simply on attribute matching and/or similarities. As a consequence, the authors deem them “insufficient to explain conceptual coherence and the richness of conceptual structure” if utilized as a sole basis for assessing fit in consumers mind herein (c.f. Tauber, 1993 thoughts = triviality, overdependence). Basically, it has been observed that an inference process can take place based on respondents’ sentiment and judgements. The expectation in the case of functional brand elicitations is that an extended brand naturally embodies a brand-specific association of the already existing marketed product. For instance, fresh breath for a toothpaste parent brand is almost automatically transferred along to the extension’s category chewing gum (Broniarczyk & Alba, 1994).

Though no precise nor definite theoretical description regarding fit exists in the vast brand extension literature, in co-branding there is a supplementary layer added on top namely the evaluation of the compatibility between the partner brands (Helmig et al., 2008; K. Thompson & Strutton, 2012). This cannot be done at a concrete level alone because in the context of brand alliances two or more brands with their set of meaningful nodes and bundle of mental links are intertwined (Simonin & Ruth, 1998). As written by Hadjicharalambous (2002), it results that in the case of co-branding “the major effects of fit of each of the constituent brands are indirect. The consumers' evaluation of the extension takes place, first, by assessing the degree of overall fit between the cobrand and the extension. In turn, the overall fit is determined by considering simultaneously the fit of each of the constituent brands with the extension and the compatibility between the brands, irrespective of the product featured in the

(21)

14 | P a g e

extension”. In the remainder of our study we shall only consider standard coupling cases as evidence suggests that there is little to none incremental benefit gleaned from inviting a third party in the marriage (Voss & Gammoh, 2004).

2.2.2 Brand Concept Consistency

If a commonly used base characterizes the resemblance of product features as underlined in the previous sub-section, Brand Concept Consistency (BCC) is another useful one to appraise fit. BCC prescribes that the impressions of the brand formed by consumers should not be altered much over time. Otherwise it results in confusion, which in turn negatively affects their evaluations of extensions, rubbing off to the focal brand by feedback effect (C. W. Park et al., 1986; C. W. Park et al., 1991). Brand concept is a variant denomination referring to the term brand image which has been around since the mid-twentieth century. With several definitions, Dobni and Zinkhan (1990, p. 118) assembled all of these as follow in order to help capture its principal demarcations:

a. Brand image is the concept of a brand that is held by the consumer. b. Brand image is largely a subjective

and perceptual phenomenon that is formed through consumer interpretation, whether reasoned or emotional.

c. Brand image is not inherent in the technical, functional or physical

concerns of the product. Rather, it is affected and molded by marketing activities, by context variables, and by the characteristics of the perceiver.

d. Where brand image is concerned, the perception of reality is more important than the reality itself.

It is acknowledged that a brand leaves an image classified as whether being made of more symbolic or functional beliefs (Low & Lamb Jr, 2000). But it can also be experiential benefits

(22)

15 | P a g e

which appeal to consumers’ senses, accentuating hedonic and indulgent or conspicuous consumption advantages (Kapferer, 2012; Keller, 2013). Nevertheless, even if different brand concepts are apparently perceived by consumers as being mutually exclusive approaches, Bhat and Reddy (1998) observed that positioning a brand on both (e.g. Nike) can work too, thus suggesting that they are not per se antagonist strategies.

Again, in brand alliance settings, the notion of fit regarding two different brand concepts other than a functional-functional concept combination applies with a small twist since the argument is that “no single readily identifiable basis, such as shared attributes or analogies, exists for consumers to evaluate the fit levels of one functional and one expressive brand or two expressive brands” (Johan Lanseng & Erling Olsen, 2012, p. 1112). The evaluation of co-branding initiatives therefore often depends on the efforts of cognitive elaboration required to find correspondence between the two different entities (Gammoh, Voss, & Chakraborty, 2006). On the one hand, it is relatively more difficult to find a logical connection if Nespresso (status oriented) and Ben & Jerry’s (fun) were to develop and promote jointly a hypothetical Iced-coffee pistachio flavoured milkshake or an odour remover such as Febreze having a Chanel number 5 perfume scent. On the other hand, a Danone Activia yoghurt (digestion) with Lipton tea (relax) would supposedly make sense already. Notwithstanding, Rao (1997) argued that possible misfit constraints should not deter the creation of an alliance between brands, suggesting that fit can be worked through if the partners manage to form convenient heuristics that may combine two brands which are visibly idiosyncratic.

2.2.3 Alternative drivers of Fit

As Wason and Charlton (2015) indicate, the recent literature discussions and enquiries converge towards demystifying the regular bases of brand fit. Studies imply that they are extendable beyond consumers’ [categorical] interpretation, as they somewhat link back and

(23)

16 | P a g e

forth to the parent brands. Such considerations might give additional substance to integrated branding opinions in which all characteristics that make a brand what it wants to stand for are somehow intra-related to one another (the brand networks system; Franzen & Moriarty, 2009; the consumer-psychology model of brands; Schmitt, 2012). Thus, it reflects on a double-sided mirror in the case of co-branded products. Noteworthy, alliance attitude scores are higher when two brands appear to match in terms of their human-like traits (D. O. James, Lyman, & Foreman, 2006). In a setup using a novel technique for this area [Bayesian nonlinear structural equation model] – inspired by how dating sites find and recommend corresponding or compatible online profiles - it transcends that coherence in the extrinsic brand personality dimensions of Jennifer L Aaker (1997) is sufficient to foster positive consumer feedback on the overall brand alliance. “In sum, it is as if partners should look the same on the outside (e.g., Sophisticated = feminine and upper-class; Ruggedness = masculine and Western), but different on the inside (e.g., Sincerity = honest and genuine; Competence = reliable and responsible), although the latter is less important”. Moreover, the authors note that combinations between two brands described as Exciting always yielded positive consumer responses (Van der Lans et al., 2014, p. 563). Eventually, brand identity – the salient and aspirational foundation of a brand (D. A. Aaker, 2012; Keller, 2001) - is a promising predictive construct to rely on for assessing fit (e.g. Viot, 2011).

To conclude, there is no closed road to utilize another focused predisposition than pure brand image fit (Bouten et al., 2011; Singh, P. Kalafatis, & Ledden, 2014). This thesis builds on those less traditional views by examining how authenticity could be insightful to take into consideration for brands that ally with peers.

(24)

17 | P a g e

Authenticity in Branding

3 A socio-cultural perspective

Within the former chapter, the scope of our preliminary investigation has been posited in the realm of launching an item with co-brands. Chapter three portrays the milestones that constitute and justify our chosen angle. It intends to show the advantages of a new paradigm in which co-creation of brand meanings by consumers in order to safeguard their distinctive individuality is becoming of crucial importance (c.f. Allen, Fournier, & Miller, 2008 for a comprehensive overview).

3.1 Brands as semiotic entities

McCracken (1986) had planted the seeds for a very competing and non-conventional view of branding that started blooming a decade afterwards [in the late 1990’s], then really matured past the millennium. Fellow endorsers of the Consumer Culture Theory tradition as referred to by Arnould and Thompson (2005) argue that brands carry with them markers of cultural meanings. According to major influencers, in any culture – whether national, popular, religious or organizational - lie values (the hard nucleus) as well as normed practices (relatively malleable parts conveyed by symbols, heroes and rituals) (Hofstede, Hofstede, & Minkov, 2010, pp. 8-9). However, it is those softer rules of the game of which the persistent cultural meaning cannot be readily understood by outsiders unless they are willing to learn then assimilate them (see also authentic "folklore" in Bendix, 2009). Originating in the anthropology field, cultural meaning can be described as “the shared cognitive-emotional state that results when the mental

(25)

18 | P a g e

structures of a group of people respond to typical objects and events in their wo rld” (Strauss & Quinn, 1997, p. 15).

Particularly, it is about these brands whose aura has enduring resonance for their buyers rather than a monotonous commercial transaction and from there strong emotions can be observed which define a special consumer-brand bond. The main point is: brands are not only mere utilitarian commodities or just a mean to an end, as they transcend through consumption with the cultural capital that they embody which in turn guides consumer behaviours (Solomon, 1983) [accentuated in the work of French sociologists Pierre Bourdieu and Roland Barthès]. At least, not many are able to produce engaging stories, specific ‘ceremonies’ or enjoyable experiences that have the potential to transform into unwavering frenzy among certain groups of people (Batra, Ahuvia, & Bagozzi, 2012; Brakus, Schmitt, & Zarantonello, 2009; Fournier, 1998). These celebrated brands even engender active exchange and conversations inside brand communities (Muniz & O'guinn, 2001), further reinforced within [counter] subcultures with their own ideologies that brand managers are advised to pay close attention to, monitor and anticipate nowadays especially with the advent of digital media (D. Holt, 2016; D. B. Holt, 2002; Leigh, Peters, & Shelton, 2006).2 An illustrative distinction between the more ‘mythic and intimate’ vision of branding in comparison to the current mainstream semantic lexicon of brands is provided in figure 3.

2

A brand such has Apple could be seen as fitting most descriptions with the p opular and active

(26)

19 | P a g e

Figure 3 : Brands & Lovemarks, as Roberts (2005) presents it

Some authors emphasize that the terminology behind a brand concept [2.2.2], as an abstract subcomponent of brand image, fails to capture this deeper signification of brands. Instead, cultural meanings would be embodied either in brand personality attributes or more so in concern with the wider brand identity chunk (Jennifer Lynn Aaker, Benet-Martinez, & Garolera, 2001; Xiao & Hwan Lee, 2014). Avis, Aitken, and Ferguson (2012), contrary to Covington (2012), think that metaphors in research surrounding brand personification as well as brand relationships are mostly not realistic enough and therefore warn that they may as such embed a grey area when formulating theoretical repositories. Knowing that, we consider herein that the socially constructed brand perceived authenticity (Peterson, 2005) is reaching its peak, competent to bring altogether diverging argumentations as we put forward in the next paragraphs.

(27)

20 | P a g e

3.2 Perceived Brand Authenticity

3.2.1 What is PBA?

This nascent, yet evolving concept, remains undeservedly ignored by a majority of investigators whilst authenticity can be traced till the beginning of the 7th century (Yuan, Liu, Luo, Nguyen, & Yang, 2014). PBA rose into the brand management landscape in an answer to the postmodern materialistic society together with consequent growing ill-at-ease feelings evocated by consumers’ uncertainties in search again for truthful objects of consumption that help them build and/or perpetuate their self-identities (Elliott & Wattanasuwan, 1998; Escalas & Bettman, 2005; Mittal, 2006). Hence, it has been defined succinctly as “the subjective evaluation of genuineness ascribed to a brand by consumers” (Napoli, Dickinson, Beverland, & Farrelly, 2014, p. 1091). Morhart, Malär, Guevremont, Girardin, and Grohmann (2015) identified that PBA consists of four key dimensions as “the extent to which customers perceive a brand to be faithful toward itself (continuity), true to its customers (credibility), motivated by caring and responsibility (integrity) and able to support customers in being true to themselves (symbolism)” (p. 203), all of which are shaped by existential, iconic and indexical cues.

Although we are aware that other authors attempted to conceptualize consumer-based brand authenticity as being a similar construct comprising for example heritage, quality commitment and sincerity in Napoli et al. (2014) or naturalness, originality, reliability and continuity as well in Bruhn, Schoenmüller, Schäfer, and Heinrich (2012), the entailed symbolism aspect of a brand’s perceived image at large does seem to have a substantial prevalence. Namely as recognized by Cian (2011) for activating the cultural synthesis type of consumers’ representations (e.g. social representation, Jodelet, 2003). In fact, the notion of brand authenticity is on par with proponents of the alternative inside-out orientation of branding in

(28)

21 | P a g e

which a consumer’s superficial desires are relatively much less rated since it starts internally with employees, directors and/or back to the founders’ vision where the product is honed (Burmann, Jost-Benz, & Riley, 2009; De Chernatony, 1999; Moulard, Raggio, & Folse, 2016). Additionally, it would seem that PBA exploits “a porous boundary between consumer fantasy and subjectivity, on one hand, and consumer perceptions of reality and objectivity, on the other” (Grayson & Martinec, 2004, p. 306). Thus, this means that the task of managers might be close to successfully aligning the promises and behaviours of a brand’s projected essence / extended [versus core] identity, however carefully preserve its centrally-oriented [versus peripheral] strong, favourable and wherever applicable unique symbolic associations stored in consumers’ memory (Ghodeswar, 2008; Keller, Sternthal, & Tybout, 2002; Michel & Donthu, 2014; Till, Baack, & Waterman, 2011). Eggers, O’Dwyer, Kraus, Vallaster, and Güldenberg (2013, p. 346) stress that brand authenticity in this sense “is about doing what is said, being totally clear about where the company has come from, and who it is today. When a brand’s communication rhetoric becomes disconnected with stakeholders’ actual experiences, they will no longer trust the brand”. Besides, these qualitative considerations were proven in an empirical manner within another recent study (Schallehn, Burmann, & Riley, 2014).

3.2.2 Why is perceptive authenticity greatly interesting for brand alliances? Research findings in the co-branding discipline are not uniform. For instance, some compelling results suggest that high-equity established brands “do not benefit from co-branding strategy for a new product in terms of consumers' attitudes and quality perceptions” (Besharat, 2010, p. 1247). This contradicts earlier studies pointing almost to the reverse (McCarthy & Norris, 1999; Washburn, Till, & Priluck, 2000, 2004).

Conversely, we believe the generic effect demonstrated by Simonin and Ruth (1998), supported in the replications of Baumgarth (2004), that familiarity of the target group with the

(29)

22 | P a g e

brands plays an eloquent role may not apply by the same standard mechanisms when the varying degree of authenticity is taken into account because of today’s climate. First, a brand’s credibility plus integrity regarding its purpose in terms of a combination of ‘says what it does and does what it says’ has never been more pertinent to avoid distrust nor address increasing demands for corporate business transparency. Second, the social demonstrance component of a brand for a lot of common categories arrives at the forefront besides its risk reduction function [quality, recalled awareness]. Third, too much fame without robustly actualized cultural meanings could expose the brand to uncontrollable antinomy like Lonsdale (white extremely racist skinhead groups3), which is maybe worse than dissonant company-emitted claims. Finally, there is almost no denial that a non-negligible proportion of consumers tend to be attached to brands that prominently contribute to improving their lives as individuals, instead of having simplistic positive attitudes towards them or wondering if the brands in question that they freely choose to consume are well-known to the rest of the world (J. Aaker, Fournier, & Brasel, 2004; Blackshaw, 2008; de Swaan Arons, Van den Driest, & Weed, 2014; Erdem, Swait, & Valenzuela, 2006; Fischer, Völckner, & Sattler, 2010; C. J. Thompson, Rindfleisch, & Arsel, 2006; Thomson, MacInnis, & Park, 2005; Whan Park, MacInnis, Priester, Eisingerich, & Iacobucci, 2010). These trends are observable when looking at some large discrepancies between the latest published rankings of the worldwide Interbrand’s brands equity/utility http://interbrand.com/best-brands/best-global-brands/2017/ranking/ and Cohn & Wolf’s 2017 Brand Authenticity index http://authentic100.com/. Samsung and Adidas occupy the 6th and 55th positions respectively

3

https://www.independent.co.uk/news/world/europe/lonsdale-faces-ban-over-neo-nazi-associations-6105955.html

(30)

23 | P a g e

in terms of most valuable brands, but on the other list Adidas finds itself in the top 10 (number 6) of most authentic brands while Samsung even misses the top 40 (number 43).

The next chapters detail the design, methods, procedures and instruments used to answer the research question and test our range of hypotheses. The thesis investigation is broken down into two sub-studies. The first one considers authenticity as a single measure construct, thus allying brands that differ in terms of overall authentic perceptions (Chapter 4). The second study utilizes specific various dimensions of perceived authenticity in order to seek if one dimension might potentially matter more than another (Chapter 5).

(31)

24 | P a g e

Study I

4 Global effects of authenticity in co-branding

Chapter Four is dedicated to the specification of variables and formulation of working hypotheses. As a reminder, this study seeks to investigate the effects of authenticity on consumer evaluations of an alliance. This work resorts to the Brand Extension Authenticity instrument featured in the introduction of this report. Spiggle et al. (2012, p. 968 emphasis in italics added) define BEA as the “consumer’s sense that a brand extension is a legitimate, culturally consistent extension of the parent brand”. Similarly, the concept of Brand Alliance Authenticity in this study is the extent to which a co-branded offering is genuinely perceived by consumers in that it can properly and seamlessly encompass both parent brands’ intrinsic meanings. In other words, staying true as well as simultaneously providing clear brand core meaning within the alliance strategy; sort of as a relentless token for its justification (Beverland, 2018).

4.1 Conceptual model and hypotheses development

4.1.1 Antecedent: Brand authenticity mix

This paper has an exploratory character as it indirectly tests “fit” between co-brands given the concordance or incoherence in terms of degrees of brand authenticity. We therefore mark a departure from the description of brand identity (BI) fit offered by Xiao and Hwan Lee (2014, p. 1242) where BI fit is the “perceived congruence between two brands’ cultural meanings” while considering a singular focal brand in their article. However, this conceptualization relies principally on the interpreted philosophies of the corporations behind the brands based on what transpires from their arguments and sources. In contrast, PBA already implicitly links

(32)

25 | P a g e

these to consumers’ identification project since it incorporates elements of brand personality and brand relationship to find coherent narratives relevant to their inner selves (Ahuvia, 2005; Fritz, Schoenmueller, & Bruhn, 2017) as argued in point 3.2.1. Furthermore, it is not because cultural meanings appear dissimilar that they can’t forcibly complement one another. After all, a brand’s cultural meaning is a complex dialectical, heterogeneous, dynamic and sometimes paradoxical force (Brown et al., 2003). For example, high congruence in our research would represent two highly authentic brands, but it is not assumed to be also the case for two poorly authentic brands which are paired together. We casted this perhaps ambiguous prediction compared to the vision relying on perceptual fit’s and brand concept consistency’s theoretical accounts because evidence from studies suggests that the credibility of the constituent brands [one of the partially captured PBA’s dimensions] has a significant, positive and direct impact on the credibility of the co-brand object itself (Aghdaie, Dolatabadi, & Aliabadi, 2012). Low authenticity fit shall therefore stand for a partnering between one highly authentic and one significantly lesser authentic brand, regardless of which category they belong to. Moreover, we hereby consider the more/less authenticity approach (relative value-driven rather than only extremes of either authentic or inauthentic). Brands range on a perceived brand authenticity continuum as shown by Napoli, Dickinson-Delaporte, and Beverland (2016) who derived distinctive levels within clusters referred to as Novice (low PBA means), Apprentice (moderate PBA), Professional (high PBA) and Master brands (very high PBA). In order to reduce the number of conditions to the possible ones mentioned above, we determine as an arbitrary threshold lesser authentic to pertain to the Novice [ideal] or Apprentice and highly authentic belonging to Professional and Master groupings. We are confident that it is enough for fulfilling our purpose. Consequently, hypotheses 1a/b become:

(33)

26 | P a g e

H1a: Brand Alliance Authenticity is higher when partner brand A is [highly] authentic than when it is not.

H1b: Brand Alliance Authenticity is higher when partner brand A is [highly] authentic than when it is not.

In line with interacting evidence from Aghdaie et al (2012) we propose hypothesis 2 as follow:

H2: The effect of an individual brand’s authenticity on Brand Alliance Authenticity (H1) is stronger when the other partner brand is authentic than when it is not.

4.1.2 Brand Alliance Authenticity (BAA) as mediator

We postulated in section 2.1 that co-branding and brand extensions can be interpreted as a strategic move in another direction for the individual partner brands. From a cultural perspective of extending brands, it follows that the co-brands might unequivocally “compromise their core and essential nature - being constant and true to themselves - and thus their internal consistency” which does jeopardize subsequent evaluations of the newly created ‘entity’ bearing the name (Spiggle et al., 2012, p. 969). Transposing this observation to brand alliances, it means that the collaborative product should be a consistent and legitimate cultural outgrowth of its constituent brands. Otherwise, it would negatively impact consumer responses. This marketing offer ought to be worthwhile of the bloodlines brought together as somewhat the same way it works for royal dynasties’ succession: an uncontested, rightful heir for the King’s throne instead of any random pretender that would have to try pleasing its subjects (Yuan et al., 2014). Moreover, Alexander (2009) argues that downplaying commercial motives in a co-branding context - represented in one of the factors composing the initial BEA - is a fundamental building block at the ‘front’ stage in the authentication

(34)

27 | P a g e

process of an alliance between two partners. We expect Brand Alliance Authenticity to be the intervening variable by which the joint effects of co-brands PBA flow towards alliance evaluation within hypothesis 3:

H3: The effect(s) of the individual brands’ perceived authenticity on alliance evaluations is mediated by BAA, in the sense that BAA positively influences Alliance Evaluations.

4.1.3 Moderation of Self-Authenticity

Liao and Ma (2015) remarked in interviews that people undergo difficulties to articulate properties of authenticity in objects and persons. Yet, they found that consumers who strive for authenticity have a propensity to consume authentic brands or products in a deliberate manner, remain loyal to them and discard imitations (ones perceived as artificial). Drawing on Self-Determination Theory, authentic functioning has been defined as the “unimpeded operation of one’s true- or core-self in one’s daily enterprise” (Kernis & Goldman, 2006, p. 344) which largely aligns with the cultural context view of one’s authenticity advanced by Erickson (1995). This denotes that consumers displaying authentic behaviours ascribe greater importance towards authenticity in general. Regarding applications within the field of Perceived Brand Authenticity in particular, results strongly amend that consumers rather low in self-authenticity express some notable indifference towards authentic brands (Morhart et al., 2015). By extension, this could also apply after having determined the authenticity of a co-branded product. Therefore, the model includes Self-Authenticity as a path moderator in order to test for potential boundaries for the effect of Brand Alliance Authenticity on Alliance Evaluation.

H4: The effect of brand alliance authenticity on alliance evaluations are stronger when consumers are high in Self-Authenticity then when they are low in SA.

(35)

28 | P a g e

4.2 The visualized research framework

(36)

29 | P a g e

5 Methodology (study 1)

5.1 Research design

The conceptualized model described in chapter Four will be hereby studied. The objective of this study is threefold: 1) to determine the extent to which an alliance becomes perceived as authentic by looking at the individual and interacting influence of the authenticity degrees of the co-brands that compose it, 2) verify that authenticity of the alliance itself mediates these effects towards alliance evaluations and 3) if the relationship between brand alliance authenticity and alliance evaluations may somewhat be influenced by consumers’ level of self-authenticity. The design employed a seemingly 2 x 2 full factorial experiment where the ‘general’ perceived brand authenticity of alliance partners (High versus Low) in each category (A/B) is manipulated, subsequently creating four different brand alliances which correspond to the number of experimental conditions. Indeed, a set of four brands – i.e. two per category - were selected presenting a differing degree of PBA.

Partner Category B Highly Authentic (Brand 3) Lesser Authentic (Brand 4) P ar tne r C at egor y A Highly Authentic (Brand 1) HA + HA HA + LA Lesser Auhentic (Brand 2) LA + HA LA + LA

(37)

30 | P a g e

5.2 Stimuli development

The experimental set used real brands in order to increase external validity of the findings and easier for studying evoked imagery that consumers may attach to certain brands (Van der Lans et al., 2014).

5.2.1 Qualitative pre-test

To start finding appropriate stimuli, eight individuals were approached and presented with two possible co-branding innovations. The product needed to realistically incorporate properties of potential partners from two categories so that their relative contribution would not be unbalanced, thus avoiding a normal categorization process but also trying to keep overall perceived fit optimal in our specific case (Y. W. Park, Son, & Kim, 2016). Option 1 was the alliance of an Automotive (Car) and Technology (IT) brand that would create an advanced hoverboard. Option 2 consisted of pieces of party clothing (dress and shirts) embedded with fiber LED’s made by a Technology and Fashion brand. Their preference went for the first collaboration in terms of appeal, plausibility and sense-making which were the required voting criteria.

In the second phase of this stage, each of the same participants was asked to decide on suitable partners per category, mentioning at least one brand that they consider highly authentic and one other lowly authentic to them. An underlying requisite was that they ought to be arguably familiar and of good quality. The definition of perceived brand authenticity from Morhart et al. (2015) was read to them for better comprehension of what the abstract PBA concept entails. For IT brands, Apple, Philips and Microsoft were most often cited as authentic, while Lenovo and Huawei were the principal inauthentic enumerations on the other hand. Regarding automobile manufacturers, Ferrari, Tesla and BMW (mentioned three times)

(38)

31 | P a g e

were the top-of-mind authentic ones with Volvo almost too. Presumed inauthentic car brands were mainly Volkswagen and Renault, both suggested twice and others were e.g. Mini and Toyota. Acknowledging the difficulty to identify authenticity (Liao & Ma, 2015), ten of the most relevant brands across the categories were selected for the quantitative pre-testing besides the product as to be sure of having satisfying degrees of variation in every potential pair.

5.2.2 Quantitative pre-test

This step of the process allowed either confirming or rejecting the assumptions obtained via the qualitative discussions in regards to the likely co-branded product without actual brand names and the separate brands at play.

Firstly, the displayed alliance product itself must be realistic, appealing and plausible whilst demonstrating a rather equal contribution of the partners (unbranded yet) from the diverse categories. The questionnaire consisted of several 7 point-Likert scales ranging from strongly disagree (1) to strongly agree (7) and was taken by 30 respondents. A standard text was developed highlighting some key features accompanied by a photo to see how the result would look like. Realism was assessed with the statement “This co-branded offering is realistic”; plausibility with an item derived from D. A. Aaker and Keller (1990) “It makes sense that a brand from category A and a brand from category B work together to create such a product”. Appeal – which denotes product involvement (Zaichkowsky, 1985) – was adapted into “The suggested product is appealing to me” found also in Desai and Keller (2002). The affirmation “As far as I can tell, the two types of partner brands would equally contribute to this alliance” tested neutrality about the added-value in the collaboration since the manipulation should be driven mostly by the authenticity experienced within the brands

(39)

32 | P a g e

themselves instead of any unintentionally perceived role division of the allies into i.e. host and modifier (Leuthesser et al., 2003; Uggla, 2004). The results of this phase are reported in Table 1. All controls met the pre-requisites elicited above. One sample t-tests heightened confidence about the outcome as the scores were significantly above the mid-point of their respective indicator [4; p-values < .05]. Although standard deviations relative to the means are spread, the unique mode and median were both at 6 except the median of contribution at 5, which still reduces any existing risk. Consequently, the hoverboard was deemed appropriate to carry on with in the main study.

Table 1

Means and Standard Deviations on the Measurements for the Criteria of the Co-branded offer N= 30 Realism Equal contribution Appeal Plausible

M (SD) 5,23 (1,38) 4,70 (1,49) 5,13 (1,53) 5,17 (1,58)

Secondly, the appropriateness of the brands’ initial PBA ratings had to be tested quantitatively as well. The same 30 people were randomly assigned to four brands (two per category) with a battery of items on 7 point Likert scales (20 per brand) where they could strongly agree (7) or strongly disagree (1) to every statement encountered. This partial randomization was set up to counterbalance hindrances that might happen when performing repeated measures tasks such as practice and boredom effects (Field, 2013). Nevertheless, it ensured that all single brands had a minimum rotation of ten to twelve viewings. As a matter of fact, subsequent analysis of the pre-test considered them as non-dependent groupings. Preferably, our respondents would have more than sufficient knowledge of the brands via the modified statement “I am familiar with this brand” utilized by Simonin and Ruth (1998) and have a positive outlook towards them measured by the item: “This brand offers great quality” (D. A. Aaker & Keller, 1990). The most important aspect is the variable we wish to manipulate, namely General

(40)

33 | P a g e

Authenticity (G.A.). It comprised three self-devised items drawn from the literature on brand authenticity: that it is perceived as genuine, original (opposite negation phrasing: “The word ‘artificial’ does not apply to this brand”) and of course simply very authentic to someone. In case this initiative would be deemed unsuccessful, full previously validated subscales (Morhart et al., 2015) representing PBA in terms of continuity (four items e.g. “A timeless brand”), integrity (four items e.g. “A brand with moral principles”), credibility (three items e.g. “An honest brand”) and symbolism (four items e.g. “A brand that adds meaning to people’s lives”) were included in order to decompose the subjective driving compartments. In effect, only the logos and names of the brands were shown in the pre-test. Table 2 sums up the univariate descriptive statistics of this phase. The results were encouraging however some detected issues needed particular attention. Ferrari had an extremely high perceived quality (M = 6.55, SD = .52) compared to the nearest other brands with a difference in means of .5 points and more, both categories confounded. Due to the likely exclusivity attributed to its luxury and pricey sports cars, Ferrari was left out of further considerations. Similarly in the other direction, with the ‘worst’ quality appraisal within the presented set, Renault (M = 4.09, SD = 1.30) was dropped. A proposition for a replacement presumed low in authenticity was Skoda (cited once in the qualitative pretest) which was assessed quantitatively with a separate pool of eleven respondents. One sample t-tests revealed that the majority of the remaining brands satisfied preliminary conditions in regards to quality and familiarity with p-values < .05, except: Lenovo (both familiarity and quality, p > .05), Huawei (near the edge for familiarity’s significance) and Skoda (familiarity). Even if quality perceptions can strongly advocate how authentic a brand actually is (Bruhn et al., 2012; Napoli et al., 2014), the selection should be, as much as possible in our situation, primordially based upon notable differences/similarities for the General Authenticity measure. Familiarity on the other hand

(41)

34 | P a g e

could be later used as a covariate for hypotheses testing. Nonetheless, a disclaimer telling respondents to respond to the brands’ questions according to their intuition based on provided details (regardless of whether they are familiar with them) will be added in the main study. A little worry occurred in that the reliability of the G.A. scale could be jeopardized with a score under the .70 benchmark of goodness (α = .66). A look at the corrected inter-item total correlations showed that one of the items was below the advised .30 estimate and Cronbach’s Alpha would increase to .76 if it were gotten rid of, which is substantial enough to consider herein. This is the reason why Table 2 reports two G.A. averages, one with all items and the other without the statement “The word ‘artificial’ does not apply to this brand”. In the main study, this problem will be addressed with the item reformulated in an understandable and less confusing way into: “To me, this brand does NOT give the impression of being a fake”. We could remark that G.A. scores mostly got a bit inflated but not for Skoda (decrease), nor Lenovo (stayed constant) which were kept on this ground. Indeed, these two brands - as somewhat expected - had the lowest levels of G.A. though they did not score significantly below the mid-point of the scale, neither significantly past it within one-sample t-tests (p > .05). It was decided to portray them as ‘copycats’ into the main study in order to again try bring those scores downwards. For selecting their nemesis, a rule of thumb was resorted to: a) both the G.A. (2) as well as G.A. (3) means scores shall be at least equal to 5 and b) their sum ought to be highest in each of the respective category in order to obtain maximum variation. In this respect, the only IT candidate that stood out on (a) is Apple. Taking into account fulfillment of (b), Tesla came just above BMW for car-makers [10.79 versus 10.54]. For these brands, the decision was to provide an additional short section emphasizing their authentic strengths to heighten their absolute authenticity scores. Independent samples t-tests infirmed that the choice was right regardless of which G.A. (2 & 3) value was compared. Tesla’s and

(42)

35 | P a g e

Apple’s G.A. means did not differ significantly from one another, Lenovo’s and Skoda’s neither (all p-values > .05). However, there was a significant difference between Apple’s and Lenovo’s G.A. means and between Skoda’s and Tesla’s G.A. means too (all p-values < .05).

Table 2

Means and Standard Deviations on the Measurements for the Criteria of the Brands Car Fam. Qual. G.A. (3) G.A. (2) Symb. Cred. Integ. Cont.

Renault 5.73 (.47) 4.09 (1.30) 4.21 (,81) 4.36 (.71) 3.64 4.48 4.43 5.11 Tesla 5.73 (1.74) 5.91 (1.14) 5.06 (1,18) 5.73 (1.21) 5.39 5.27 5.25 4.93 BMW 6.17 (.84) 6.08 (1.00) 5.17 (1,02) 5.37 (1.21) 4.67 5.33 5.12 5.96 Volkswagen 6.17 (.72) 5.83 (.58) 4.47 (1,40) 4.54 (1.70) 4.02 3.72 3.75 5.83 Ferrari 5.73 (.65) 6.55 (.52) 5.18 (,82) 5.82 (.56) 5.02 5.51 4.70 6.16 Skoda 4.91 (2.26) 4.91 (.94) 4.33 (1,28) 4.27 (1.23) 4.27 4.73 4.18 4.64 Tech Microsoft 6.75 (.45) 5.08 (1.16) 4.64 (1.07) 5.04 (1.36) 4.19 4.75 4.33 5.46 Lenovo 4.25 (2.00) 4.50 (1.17) 4.00 (.94) 4.00 (1.00) 3.90 4.25 3.92 3.96 Philips 6.17 (.72) 5.67 (.78) 4.69 (.70) 4.96 (.89) 4.33 5.17 4.65 5.50 Huawei 4.92 (1.83) 5.00 (1.13) 4.42 (.87) 4.50 (1.13) 4.42 4.72 4.23 3.96 Apple 6.31 (.48) 6.08 (1.04) 5.00 (1.21) 5.23 (1.05) 4.88 5.00 4.66 5.42 M (SD). The mean scores for the PBA dimensions are hereby reported solely for indication.

G.A. (3) refers to general authenticity on three items, G.A. (2) with two best items.

A two-way ANOVA (Table 3 and Graph 1) was conducted against G.A. (3) to support these predictions and prove the suitability of the stimuli for the experiment. In order for the manipulation to succeed, there should be a main effect of levels of authenticity (dummy coded Low/High) between partners but no significant main effect of category (Car/IT) nor significant interaction effect. Levene’s test statistic was non-significant, F(3, 43) = .705, p >

(43)

36 | P a g e

.05, meaning that the assumption of homogeneity of variances was not violated across the various group combinations. There was a significant main medium to large effect size of authenticity level on G.A. scores, F(1, 43) = 6.495, p < .05, η² = .131 which does attest that the presumed highly authentic brands score significantly higher on the G.A. measure relative to the lowly authentic brands. There was a non-significant very low main effect size of category, F(1, 43) = .338, p > .05, η² = .008. There was a non-significant negligible interaction effect size between category and level of authenticity, F(1, 43) = .162, p > .05, η² = .004. To conclude, the brands that have been chosen checked all the boxes to be utilized in the definitive experimental design.

Table 3

Univariate ANOVA results. Dependent variable: general authenticity (G.A.)

Effect df F p ηp² Authenticity (AUT) 1 6.495 .014* .131 Category (CAT) 1 .338 .564 .008 AUT*CAT 1 .162 .689 .004

*Significant at the p < .05 level

5.3 Methods of the grand study

The core study used a simplified 2 (constituent brand from category A: Highly vs lesser authentic) x 2 (constituent brand from category B: Highly vs lesser authentic) between-subjects full factorial experimental setting. Thus, a total of four manipulations were created.

Enter the authentic side of strategic cooperation : seeking the role of Brand [Alliance] Authenticity on consumers’ evaluations of co-branding : an explorative investigation

Master Thesis