Business-to-millennial branding : an exploratory research on the brand knowledge and brand valuation of highly educated Dutch millennials (age 20-30) for IBM and Google

(1)

Business-to-Millennial Branding

An exploratory research on the brand knowledge and brand valuation of highly educated Dutch millennials (age 20-30) for IBM and Google

Abstract

Brands are considered one of the most important intangible assets that firms nowadays possess. As the value of a brand, or the “brand equity”, positively influences consumers’ perceptions and (buying) behaviors, the identification of factors that build brand equity represents a central priority for academics and marketing managers. In this study the brand equity concept is researched in a practical business-to-business environment as part of an internship at IBM, whereas competitor Google is used as a benchmark. Brand equity and the recently developed CRUSH-model (acronym for coolness, realness, uniqueness, self-identification and happiness) are tested through a quantitative survey for a specific target group: Dutch, highly educated millennials between the age 20 of and 30 (N = 119), as they can be seen as potential (future) employees, clients, and partners. Results confirm relationships between all the CRUSH-variables and the overall brand valuation. Moreover, the perceived realness or authenticity of a brand shows to be a predictor of this valuation, however the other CRUSH variables do not. As Google scored better than IBM on all of the brand knowledge and brand valuation scales, a specific advice for IBM is provided on how to improve their brand equity with millennials. In further research the possible relationship between the usage of business-to-consumer products and brand valuation should be examined, as this could be one of the explanations for the differences between Google and IBM.

Keywords: millennials, generation Y, brand knowledge, brand equity, CRUSH-model, IBM, Google,

branding, business-to-business

Name: Judith van Dellen Student number: 10192506

Date: June 29th 2016

Supervisor: Dhr. drs. ing. A. C. J. Meulemans Program: BSc Business Administration

(2)

Statement of Originality

This document is written by Student Judith van Dellen 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)

1. Introduction

1.1 Brand equity

With the trend toward digital globalization and commoditization of products and services, brands are considered as one of the most important intangible assets that firms nowadays possess (Keller & Lehmann, 2006). In both consumer markets and business-to-business markets, brands serve the general purpose of facilitating the identification of businesses as well as differentiating them from the competition (Anderson & Narus, 2004 in Kotler & Pfoertsch, 2007, p. 358). As the value of a brand, or the “brand equity”, positively influences consumers’ perceptions and (buying) behaviors, the identification of factors that build brand equity represents a central priority for academics and marketing managers (Buil, Chernatony, & Martinez, 2008).

1.2 Context of the study: Internship at IBM

To research the brand equity concept in a practical environment, I did an internship as part of this thesis at American multinational technology and consulting corporation IBM (short for International Business Machines). As this company is acknowledging the increasing importance of branding and especially with regard to highly educated millennials (age 20-30), I was assigned by the IBM BeNeLux Brand Manager to find out what the brand equity of IBM currently is like and on what areas the company could improve to assure a positive image in the eyes of so called millennials or Generation Y. Particularly this generation is important to IBM, because (highly educated) millennials can be seen as the potential decision makers of the (near) future as employees, clients or partners. Even though IBM is one of the largest IT firms worldwide with about 400,000 employees and 174 countries to be operating in, many of the Dutch millennials may have heard of IBM, but do not seem to know much about the company and its operations. IBM’s business mix has shifted by exiting commoditizing markets such as PCs, hard disk drives and DRAMs and focusing on higher-value, more profitable markets such as

(6)

business intelligence, (big) data analytics, security, cloud computing, and the newest focus: cognitive business. The fact that IBM does not produce business-to-consumer products, such as the PC, anymore, makes it extra challenging to change the brand awareness of millennials. Furthermore, IBM BeNeLux is non-advertising, so that alternative marketing ways to increase brand equity are required.

To be able to compare the brand equity of the company to some standard, well known multinational technology company Google serves as a benchmark. Benchmarking an existing brand is an important step of developing a more formal approach to branding, to be able to establish where the firm is strong and where it needs support (Michell, King, & Reast, 2001). With the shifted focus of IBM to cognitive business, Google is one of the most important competitors.

1.3 Generation Y

The generation IBM wants to focus on is a difficult one to target. That is because millennials are full of contradictions: they have thousands of songs and photos on their iPhones, and yet vinyl and polaroids are in again. They order pre-cooked meals, and at the same time create little herb gardens on their balconies. They combine Primark and H&M with exclusive brands, but they also want to buy authentic, organic, local, less global (Van den Bergh & Behrer, 2011). They are constantly using the latest and fanciest technologies, but leave for backpack trips around the world with nothing but flip flops and a Lonely Planet. In other words, Generation Y is a unique and influential consumer group whose behavior is often discussed, but not fully understood (Valentine & Powers, 2013). Heavily influenced by technology and the Internet, this cohort has evolved differently from previous generations, making it a challenging group to target (Lester, Forman, & Loyd, 2006). Therefore, increased interest in identifying aspects of Generation Y that differentiate them from previous generations has been called for.

(7)

Recently Van den Bergh and Behrer (2011; 2016) developed the so called CRUSH-model, which focuses completely on generational characteristics of millennials and is an acronym for coolness, realness, uniqueness, self-identification, and happiness: the five key elements that Gen Y’s most favorite brands all share (Van den Bergh & Behrer, 2016). This model could be an addition to the more traditional brand equity models, such as those of Keller (1995), as it is fully dedicated to millennials and is also applicable for business-to-business brands.

1.4 Problem Definition and Research Questions

The central question of this research is: How well do the highly educated Dutch millennials (age 20-30) know IBM (compared to Google) and what can IBM do to improve its brand equity? To be able to answer this question, the brand equity and CRUSH scores of IBM will be examined and compared to those of Google. This leads to the next question: Which of these variables influence overall brand valuation of the millennials? Also the influence of demographic variables gender, age, and education type on the valuation of the two brands are investigated, resulting in the sub questions: Is there a difference in brand equity between education types (alpha/beta/gamma)? What is the effect of age and sex on the overall brand valuation?

1.5 Theoretical Contributions

First of all, this study contributes to the generational characteristics of millennials, more specifically to highly educated, Dutch millennials, as the factors that influence their overall brand valuation are identified. Another theoretical contribution of this research is the business-to-business approach of brand equity. Consumer based factors like perceived product quality and buying intentions of traditional brand equity models for measuring the strength of brands, such as those of Aaker (1991) and Keller (1993), may not be fully applicable to the branding of

(8)

business-to-business companies, since these companies have a specialized nature of marketing and purchasing (Glynn, 2012, p. 666). For instance, in business-to-business companies the buyer is often not the end user; the complexity of the buying process often involves groups of individuals in the firm including buying committees; the value of transactions is much higher; there are fewer customers; and there is an emphasis on longer-term corporate relationships (Glynn, 2012, p. 673). To expand the brand equity model with variables that are also relevant to business-to-business companies and in particular to millennials, the CRUSH-model (Van den Bergh & Behrer, 2016) is included. As this is still a very new model, testing it will contribute to the validity of it.

1.6 Practical Contributions

This research provides valuable information for both IBM and Google about their brand equity for a specific target group and thus they will be able to deduce what aspects they should focus on when they want to build or increase brand equity with Dutch, highly educated millennials. The distinction between education types will provide the opportunity to know which specific students or alumni to target. Furthermore, in the discussion a specific advice to IBM is given based on the brand knowledge and CRUSH-scores.

1.7 Structure

This thesis is structured in six chapters. First, the most important literature on business-to-business companies, brand equity, millennials, and the CRUSH-model are reviewed in the theoretical framework. Chapter 3, the methodology, is concerned with the research design, conceptual model, measures, and procedure of this study. Next, the sample, reliabilities and results are described in chapter 4. To conclude, the contributions and limitations of this research and a detailed managerial advice for IBM are provided in chapter 5.

(9)

2. Theoretical Framework

In this section first some important characteristics of business-to-business companies are discussed, since they differ from business-to-consumer companies on several aspects. Then the most important literature on brand equity is reviewed, including the models of Aaker (1991) and the upon Aaker’s model conducted model of Keller (1993), which is composed of brand awareness and brand image. The third paragraph includes a discussion of generational segmentation and the currently existing research on millennials, including the CRUSH-model (Van den Bergh & Behrer, 2016) and its five elements.

2.1 Business-to-Business (B2B) Brands

Branding has been a key topic in marketing in recent years (Buil et al. 2013). Brands can act as a magnet to attract new customers, as a “Hallmark card” reminder to ensure that customers continue to think about the firm, and to improve the relationship between firm and customers (Zeithaml, Lemon, & Rust 2001). However, many brand marketing texts assume a business-to-consumer (B2C) perspective, even though many of the best global brands, such as Microsoft, IBM, and GE are business-to-business (B2B) brands (Interbrand, 2016). B2B refers to business that is conducted between companies, rather than between a company and individual consumer. Much of the brand image in business literature is founded on consumer research, and as a result, many of the inferences in B2B literature are the same as in consumer literature (Glynn & Woodside, 2009, p. 276). The disadvantage of applying a B2C brand perspective to B2B brands is that the specialized nature of business marketing and purchasing is sometimes ignored (Glynn, 2012, p. 666). B2B purchasing differs from end-consumer buying in many respects: the value of transactions is much higher, the complexity of the buying process often involves groups of individuals in the firm including buying committees, there are fewer customers, and the buyer is often not the end user (Glynn, 2012, p. 673). Another key difference between B2B and B2C

(10)

marketing is the emphasis on longer-term corporate relationships and not a relationship off transactions (Glynn, 2012, p. 673). Furthermore, businesses buy things out of ‘derived demand’, to facilitate the production of goods and services, in contrast to the direct demand of consumers that buy goods and services to satisfy their wants (Brennan, Canning & McDowell, 2014, p. 13-14). Furthermore, in industrial markets, the company itself is often the brand; but in consumers markets, the emphasis is usually on the products or a limited group of them (Bendixen, Bukasa, & Abratt, 2004).

However, according to Bendixen et al. (2004) several studies (Hutton, 1997; Gordon et al., 1993; Bendixen et al., 2004) have proven that brand equity also exists in the business-to-business sector. Buyers are willing to pay a price premium for their favorite brands, to extend the brand’s halo to other product lines and recommend the brand to others. Brand equity therefore plays a very powerful role in influencing business customers, especially in those markets where products or services are difficult to differentiate based on quality features (Mudambi et al., 1997). The brand equity concept is discussed in more detail in paragraph 2.2.

2.2 Brand Equity

Brand equity emerged as a concept in the 1980s, as it became apparent that the purchase price paid for many firms largely reflected the value of their brands, or in other words it reflected their “brand equity”. The clear implication of these transactions was that brands are one of the most important intangible assets of a firm (Leone, Rao, Keller, Luo, Mcalister, & Srivastava, 2006). As defined by the American Marketing Association, a brand is a "name, term, design, symbol, or any other feature that identifies one seller's good or service as distinct from those of other sellers". The words “any other feature” are an important addition to the original 1960 definition, as these words allow for intangibles, such as image, to be the point of differentiation (Wood, 2000). Keller (2003) for example refers to brands as “a certain amount of awareness,

(11)

reputation and prominence in the marketplace”; and Zeithaml et al. (2001) define brand equity as: “the customer’s subjective and intangible assessment of the brand, above and beyond its objectively perceived value”.

Since the emergence of the concept, brand equity has aroused intense interest among marketing managers and business strategists from a wide variety of industries (Aaker & Biel, 2013), resulting in many different proposed academic and industry models of branding and brand equity. These different models do however share certain basic premises about brand equity: that the power of a brand lies in the minds of consumers and what they have experienced, learned, and felt about the brand over time, and that brand equity can be thought of as the “added value” endowed to a product in the thoughts, words, and actions of consumers (Leone et al., 2006).

In one of the first brand equity models (Aaker, 1991) brand equity is defined as a set of assets and liabilities linked to a brand, its name, and symbol, that add to or subtract from the value provided by a product or service to a firm and/or to that firm’s customers. This set of assets and liabilities consists of five components: (1) brand loyalty, (2) brand awareness, (3) perceived quality, (4) brand associations, and (5) other proprietary assets (e.g., patents, trademarks, and channel relationships) as visualized in Figure 1.

Figure 1. Aaker’s Brand Equity Model (1995).

Building upon Aaker’s definition, Keller (1993) conducts the customer-based brand equity conceptual model. In this model, customer-based brand equity is defined as the differential effect of brand knowledge on consumer response to the marketing of the brand. Brand knowledge

(12)

does not just refer to the facts about the brand, but to all the thoughts, feelings, perceptions, images, experiences, and so on that become linked to the brand in the minds of customers (actual or potential customers, individuals or organizations). All of these types of information can be thought of in terms of a set of associations to the brand in customer memory (Leone et al., 2006). Furthermore, Keller defined two components of brand knowledge (see Figure 2): brand awareness (relating to brand recall and recognition performance by consumers) and brand image (referring to the set of associations linked to the brand that consumers hold in memory), which are discussed in the next paragraphs.

Figure 2: Dimensions of Brand Knowledge.

2.2.1 Brand Awareness

The first concept of Keller’s Brand knowledge model is brand awareness, which formally refers to the ability to recall and recognize a brand (Hoeffler & Keller, 2001). Brand recognition

(13)

is the ability of the consumer to confirm prior exposure to the brand; brand recall is the unaided retrieval of the brand from memory. Brand awareness is thus more than just knowing the brand name and having previously seen the brand; it also involves linking the brand—the brand name, logo, symbol, and so forth—to certain associations in memory. Brand awareness is a necessary condition for building brand equity. Firms have the opportunity to deepen the emotional tie with customers and strengthen the customer’s associations and attitudes toward the brand only after the initial brand awareness has been created (Zeithaml et al., 2001). The top level in brand awareness, which is brand name dominance, is achieved when only a certain brand is mentioned during a brand recall task (Aaker, 1996). For example, when the only brand that comes to mind when asking about car navigation would be TomTom. In that case people may even use the brand name as a generic trademark when they are in fact talking about a product itself instead of the brand. When the brand can easily be recalled or recognized, the depth of brand awareness is strong. The breadth of brand awareness on the other hand refers to the range of purchase and consumption situations in which the brand comes to mind (Hoeffler & Keller, 2001). Ideally, a brand would have both depth and breadth of brand awareness.

2.2.2 Brand Image/Brand Associations

The second part of brand knowledge according to Keller (1993) is brand image, which refers to the associations consumers have in their memories about a brand. The term brand image is therefore used interchangeable with brand associations (Hoeffler & Keller, 2002). Different definitions of the concept of brand associations exist, but they are quite similar. According to Aaker (1991), brand associations are the category of a brand's assets and liabilities that include anything linked in memory to a brand. Keller (1993) defines brand associations as informational nodes linked to the brand node in memory that contain the meaning of the brand for consumers. Both definitions can be simplified as “what consumers link in their memory to a certain brand”.

(14)

Brand associations are important to marketers and to consumers. Marketers use brand associations to differentiate, position, and extend brands, to create positive attitudes and feelings towards brands, and to suggest attributes or benefits of purchasing or using a specific brand. Consumers use brand associations to help process, organize, and retrieve information in memory and to aid them in making purchase decisions (Aaker, 1991, pp. 109-113). To create brand equity, it is important that the brand has some strong, favorable, and unique brand associations (Keller, 1993).

2.3 Generational Segmentation

This study is focused on the brand equity of a specific generation: millennials. The way youth socialize, build relationships, shop and make career choices is heavily affected by the era they have been raised in (Van den Berg & Behrer, 2001). This is supported by generational theory, which posits that generational cohorts share life experiences that cause them to develop similar attitudes and beliefs (Meriac, Woehr, & Banister, 2010). Even though there is some debate to the validity of cohort segmentation, e.g. predicting behaviors based upon dates of birth as opposed to life-stage and lifestyle segmentation (Noble & Schewe, 2003), several researchers have identified specific characteristics that validate these generational cohorts. The most important findings regarding Generation Y are discussed in this chapter, but not before briefly mentioning their ancestors.

2.3.1 The Silent Generation, Baby Boomers, and Generation X

Three generations prior to Generation Y, the Silent Generation covers adults born from 1928 to 1945. Their ‘silent’ label refers to conformist instincts and contrasts with the noisy anti-establishment offspring: the Baby Boomers, who mark the years after the Second World War (1946-1964) (Van den Bergh & Behrer, 2016, p. 8). The Baby Boomers grew up in an era of

(15)

economic growth and full employment. The austerity of the Silent Generation was replaced by technological advancement and increasing freedom and leisure time. As a result of the commercial launch of birth control pills, a much smaller Generation X (also known as the Post-Boomers) evolved, consisting of people who were born form 1965 to 1979. As Xers began their career in a recession with much downsizing of the workforce, they adopted the work ethic and focus of the Baby Boomers but were more individualistic and pessimistic (Van den Bergh & Behrer, 2016, p. 7).

Table 1 shows the unique and distinctive characteristics of the Silent Generation, Baby Boomers, Generation X and Y, as uttered by themselves. Although Generation X also cites technology as their generation’s source of distinctiveness, this is only done so by half the amount of the Gen Yers, who have fused their social lives into technology (Van den Bergh & Behrer, 2016, p. 290). For Baby Boomers, work ethic is the most prominent identity claim; for the Silent Generation it is the Second World War and the Depression that makes them stand apart.

Table 1 Generational Characteristics Silent Generation (1928 – 1945) Baby Boomers (1946 – 1964) Generation X (1965 – 1979) Generation Y (1980 – 2000)

1 WWII/Depression (14%) Work ethic (17%) Technology use (12%) Technology use (24%) 2 Smarter (13%) Respectful (14%) Work ethic (11%) Music culture (11%) 3 Honest (12%) Values/morals (8%) Conservative (7%) Liberal/tolerant (7%) 4 Work ethic (10%) ‘Baby boom’ (6%) Smarter (6%) Smarter (6%)

5 Values/morals (10%) Smarter (5%) Respectful (5%) Clothes (5%) Note. From Pew Research Center, January 2010, in: Van den Bergh & Behrer, 2016, p. 9.

(16)

2.3.2 Generation Y

Even though researches differ on when the demographic cohort Generation Y exactly starts and ends, they mostly use birth years ranging from the 1980s to 2000. It is argued that Generation Y has evolved differently from previous generations, as they are heavily influenced by technology and the Internet (Lester, Forman, & Loyd, 2006). As consumers, millennials (members of Generation Y) are characterized as: more educated, more materialistic, and more spontaneous. They are also known for giving more emphasis to the ‘immediacy and instant gratification’ involved in the purchasing process; having higher expectations for goods and services; desiring greater ‘connectedness’ with peers and purchase influencers – including retailers and service providers; being more technology savvy; using the Internet more for product-information search and purchase; being more skeptical of advertising and media; and being more socially conscious than other consumer cohorts (Cauley, 2006; Cone Inc., 2006; Jayson, 2006; Krotz, n.d.; Loroz, 2006; National Retail Federation, 2006; Noble et al., 2009 in Hyllegard, Yan, Ogle & Attman, 2010). Millennials have also been described as being more affluent, more self-sufficient, more individualistic, more brand loyal, and more tolerant than members of other cohort groups (Farris, Chong, & Danning, 2002; Krotz, n.d.; Morton, 2002; Noble, Haytko, & Phillips, 2009; O’Donnell, 2006 in Hyllegard et al., 2010). Furthermore, driven by advances in digital and mobile technology, millennials have the ability to participate in product development and marketing processes. Therefore it is argued that this generation wants to actively participate, co-create, and most importantly, be included as partners in the brands they love and that they even almost expect that companies should want to seek their opinion (Fromm & Garton, 2013). All these characteristics form a good base to search further for factors that drive millennials in their valuation of brands. Van den Berg and Behrer (2016) did so by trying to provide insights into the consumer psychology and behavior of millennials. Their most important findings will be discussed next.

(17)

2.3.3 CRUSH-model

Van den Berg and Behrer (2016) argue that traditional models measuring the strength of a brand are often too focused on brand equity or brand image (2016, p. 47), such as the models of Aaker and Keller that were discussed earlier on. Therefore, they wanted to identify new factors that are important for the evaluations of brands, or “brand leverage”. Based on approximately 5.000 stories of 14- 29-year-olds across different regions of Europe (the UK, Germany, France, Spain, Sweden, the Netherlands and Belgium) on their most and least favorite brands, they identified five denominators through text-mining techniques, which they subsequently validated globally by a quantitative survey (N = 4.065) (Van den Berg & Behrer, 2016, p.45-46). The five key elements that Gen Y’s most favorite brands according to this research all share and that together form the acronym “CRUSH”, are: coolness, realness, uniqueness, self-identification, and happiness.

Before deciding that coolness was a good word to use for the first element, the concept “cool” and several synonyms were investigated by Van den Bergh & Behrer, as youth slang is continuously changing and generations tend to linguistically differentiate themselves from their predecessors (2016, p. 58-59). For millennials, “cool brands are attractive and appealing brands that are popular in their immediate social circle and bring a sense of novelty, surprise, or originality” (Van den Bergh & Behrer, 2016, p. 99). Some of the most important archetypical characteristics predicting the coolness of a brand are: trendy, high status, clean reputation, successful, creative, fun, cheerful, and own style (p. 60).

The second element of the CRUSH model, realness (or brand authenticity), is a key aspect that discerns long-term winning brands from fads. Authenticity is defined as “the quality of being of an established authority or being genuine, not corrupted from the original, or truthfulness of origins” (Van den Bergh & Behrer, 2016, p. 102). People today are discontented with commercial existence and lack faith in marketing, with almost everything in their lives seeming to be

(18)

contrived (Napoli, Dickinson, Beverland, & Farrelly, 2014, p. 1096). Consumers are therefore demanding products that reflect the renewed desire for what is authentic. Thus, it is important that authenticity claims capture the experiences, expectations and desires of the proposed target and reflect their prevailing values and beliefs (Molleda, 2010 in Napoli et al., 2014, p. 1096). With Generation Y, authenticity is attained in another way than the traditional approach of claiming origin, heritage or history. The modern interpretation of authenticity is being honest to yourself as a brand (the brand DNA), to youngsters (transparency) and to society (CSR) (p. 128).

Uniqueness is the third element of the CRUSH model. The unique selling proposition is one of the oldest core marketing principles and according to Van den Bergh and Behrer it is still one of the drivers of choice (2016, p. 131). A clear positioning based on a sustainable and unique brand DNA will increase impact. The uniqueness of a brand, of more specifically the perceived uniqueness of a brand, improves brand image and stimulates brand conversations or buzz among Generation Y (p. 131).

Self-identification with the brand can be defined as the extent to which a brand is seen as a mirror of one’s own passions, interests, and lifestyle (Van den Berh & Behrer, 2016, p. 150). In other words: when a brand or company is capable of getting closer to the lives of youth, in a way that millennials feel it is part of their lives, it will improve both its short- and long-term market success. It therefore has a very strong direct effect on brand leverage. Furthermore, identification is also influencing brand image and conversations about the brand. Brands should reflect the diverse lifestyles of millennials, as a better understanding of their identity construction will make a brand fit in with youths’ lives while embracing diversity.

The fifth and final element of the CRUSH model is happiness, which according to Van den Bergh and Behrer seems to have the largest impact on brand leverage (2016, p. 177). Popular youth brands know how to leverage from positive emotions and avoid arousing negative ones. Making millennials happy will make them feel a stronger emotional attachment to a brand.

(19)

3. Methodology

As the most important theories regarding brand equity, millennials, and the CRUSH-model are reviewed, this chapter is concerned with the methodology of the research. First the research design and conceptual model are discussed, followed by the data collection and measures of the variables. Finally the procedure including the used statistical tests is explained.

3.1 Research Design and Conceptual Model

The design of this research is exploratory, to find out how well the highly educated Dutch millennials (age 20-30) know IBM compared to Google and what IBM can do to improve its brand equity. An important question that can lead to an advice for IBM is: Which variables are most important to the overall brand valuation? Also the influence of demographic variables gender, age, and education type on the valuation of the two brands are investigated, resulting in the sub questions: Is there a difference in brand equity between education types (alpha/beta/gamma)? What is the effect of age and sex on the overall brand valuation? The conceptual model is shown in Figure 3.

(20)

3.2 Data collection

A questionnaire (see Appendix A) was composed and distributed online using the Qualtrics Survey Software tool, which also allows for filling out the survey on mobile devices and thus increases the accessibility of the survey. The collection of respondents was based on volunteer sampling due to self-selection. The questions were all translated to Dutch, to prevent ambiguities that would result from having to master a foreign language. Before distributing the survey, it was checked by four persons in order to guarantee the comprehensibility of the translated questions with a pilot test. For example, the question “Do you know what IBM stands for?” was adjusted to “Do you know what the brand IBM stands for?” because one of the test subjects found it confusing if the abbreviation “IBM” was meant or the company itself. Also, the open question in which respondents were asked to name 5 IT companies, was changed from forced to requested response, since one of the test subjects was not able to name 5 companies.

After conducting the survey, a request and link to fill out the survey followed by the requirements of age (20-30) and education (HBO/WO) were shared through the Facebook page and Whatsapp groups of the researcher as well of an HBO colleague, and on the Thesis course page of Blackboard (University of Amsterdam). Data was collected for a period of one week from the 23rd of May 2016 onward. The survey started with a short introduction page on which the respondents were thanked for participating and (again) notified about the requirements. Furthermore, it was emphasized that anonymity was guaranteed and also it was mentioned that the estimated time it would take to complete the survey would be 5 minutes.

3.3 Measures

In this paragraph the different measures of all the variables, to be brand knowledge (brand awareness and brand image, the CRUSH variables, overall valuation, and finally the demographic variables are discussed.

(21)

3.3.1 Brand Knowledge

As discussed in the theoretical framework, Brand Knowledge consists of Brand Awareness (Brand Recall and Recognition) and Brand Image (Brand Associations) (Keller, 1993). For the most part the scale of customer-based brand equity of Yoo and Donthu (2001) was used, since this is one of the most widely adopted customer-based brand equity measurements. However, some alterations had to be made. For example, two of the four elements, namely brand loyalty and perceived brand quality, are not included, since these are focused on business-to-consumer products. The exact Brand Knowledge scales are discussed next.

3.3.1.1 Brand Awareness

Brand Awareness consists of two parts: Brand Recall and Brand Recognition. The first question of the survey was an open question to measure Brand Recall: ‘Which five information technology (IT) companies are you most familiar with?’. At this point the respondents do not know yet that the survey will contain questions about IT companies (IBM and Google). There were five open text fields, but respondents were also able to go to the next question if they were not able to mention five (or any) companies, after confirming they were indeed not able to mention five companies. Afterwards, the Google and IBM answers were rated from zero points (not mentioned at all) to five points (the first mentioned company).

After the open question, a section with questions on Google was followed by the same questions on IBM. The first of these questions, another measure of brand recall, was: “I am aware of Google/IBM”, for which participants were asked to rate themselves on a five-point Likert scale, ranging from “1 (strongly disagree)” to “5 (strongly agree)”. This statement was part of the scale of Yoo and Donthu (2001) as well as of Chen, Yeh and Jheng (2013). The latter study revises the measurement of brand awareness (BA) from existing studies (Buil et al., 2008; Aaker, 1996).

(22)

3.3.1.2 Brand Image/Brand Associations

The variable Brand Associations was measured through three items, again formulated as statements on which the participants were asked again to rate themselves on a five-point Likert scale, ranging from “1 (strongly disagree)” to “5 (strongly agree)”. These three statements are based upon the research of Yoo and Donthu (2001) to quantitatively measure brand associations: “Some characteristics of Google/IBM come to mind quickly”, “I can quickly recall the symbol or logo of Google/IBM”, “I have difficulty in imagining Google/IBM in my mind” (reversed scoring).

3.3.1.3 Overview Brand Knowledge Variables

An overview of the brand knowledge variables is provided, as in the rest of this study the variable names are used. BRC stands for brand recall, BRG for brand recognition, and BAS for brand associations.

Brand Knowledge Brand Awareness

BRC1 Which five IT companies are you most familiar with? (open question) BRC2 I am aware of the brand Google/IBM.

BRG1 I know what the brand Google/IBM stands for.

BRG2 I can recognize Google/IBM among other competing brands. Brand Image

BAS1 Some characteristics of Google/IBM come to mind quickly. BAS2 I can quickly recall the symbol or logo of Google/IBM.

(23)

3.3.2 CRUSH-model

The CRUSH-scores of the two brands (Google and IBM) were based on the five elements of the model: coolness, realness, uniqueness, self-identification, and happiness (Van den Bergh & Behrer, 2016). Participants were asked again to rate the extent to which they agreed to five statements on a five-point Likert scale, ranging from “1 (strongly disagree)” to “5 (strongly agree)”. The statements were: “Google/IBM is a cool brand”, “Google/IBM makes me feel happy”, “Google/IBM is a real/authentic brand”, “Google/IBM is a unique brand”, and “Google/IBM is a brand I can identify with”.

3.3.2.1 Overview CRUSH variables CRUSH variables

CRUSH1 Google/IBM is a cool brand.

CRUSH2 Google/IBM is a real/authentic brand. CRUSH3 Google/IBM is a unique brand.

CRUSH4 Google/IBM is a brand I can identify with. CRUSH5 Google/IBM makes me feel happy.

3.3.3 Overall Valuation

To measure the overall valuation of the brands, respondents were asked to rate Google and IBM on a 10 point scale. This was the last question regarding the companies.

3.3.4 Demographic variables

Three demographic variables were included: age, gender, and education type. Participants were asked to pick their age from a list of options ranging from 20 until 30, since these were the ages considered ‘millennials’. The gender variable was dichotomous with the values ‘female’ or ‘male’. For the education type, participants could choose from 15 popular studies or add their education if it was not part of the list.

(24)

3.4 Procedure

After the data was collected, first, the answers to the open Brand Recall question were transmitted into numeric scores, ranging from 0 (not mentioned at all) to 5 (first mentioned IT-company). Second, the answers to the question “I have difficulties imagining Google/IBM in my mind” were reversed. Also a new variable was conducted, categorizing all the different education types into three groups: alpha, beta, or gamma. Furthermore, the average Brand Knowledge and CRUSH scores were calculated into new variables.

For the data analysis IBM SPSS Statistics version 23 was used. In the next paragraphs is discussed what statistical tests were performed.

3.4.1 Correlations

Nonparametric measure Spearman’s rank correlation coefficient (ρ) was used to assess the statistical dependence between the individual Brand Knowledge and CRUSH variables, and between the CRUSH scores and the overall valuation of IBM and Google, because normal distribution can not be assumed because of the Likert scales.

3.4.2 Comparing Google and IBM

To test whether the mean Brand Knowledge and CRUSH scores of Google and IBM differ significantly, the non-parametric Wilcoxon signed-rank test was performed, as normal distribution can not be assumed because of the Likert scales.

3.4.3 Education Distribution (alpha/beta/gamma)

A Chi-squared test of independence was performed to examine the relation between the education types (alpha, beta, and gamma) and brand knowledge, CRUSH-scores, and overall valuation of both Google and IBM.

(25)

3.4.4 The Effects of Sex on Overall Valuation Scores

To test the effect of sex on the overall valuation scores of both IBM and Google, the nonparametric Mann-Whitney U test was performed, since there is no assumption of normal distributions and there are two groups (male or female).

3.4.5 The Effects of Age on Overall Valuation Scores

To test the effect of age on the overall valuation scores of both IBM and Google, the nonparametric Kruskal-Wallis test by ranks was used, since there is no assumption of normal distributions and there are more than two age groups (ranging from 20 to 30).

3.4.6 Regressions

Multiple linear regressions were performed to predict participants’ overall brand valuation scores for Google and IBM based on their CRUSH values.

(26)

4. Results

This chapter provides an overview of the most important results of the research (the additional SPSS output can be found in the appendices). First, a description of the sample is given. Second, the reliabilities of the scales are discussed. Then in the third paragraph, the most important descriptive statistics and correlations of the brand knowledge and CRUSH variables for both Google and IBM are mentioned and compared to each other. In paragraph 4.4, the education distribution is discussed, by comparing the relation between the education types (alpha, beta, and gamma) and the brand knowledge, CRUSH-scores, and overall valuation. The next two paragraphs include the effects of sex and age on the overall valuation scores. Finally, the chapter ends with the regression results, of how the CRUSH-values predict the overall brand valuation.

4.1 Sample

A total of 119 responses have been taken into consideration for data analysis. No outliers were identified, due to the multiple choice setup of the survey. The sample consisted of Dutch women (63%) and men between the age of 20 and 30 (M = 23.53, SD = 2.48). Of these respondents, the majority studied Economics and Business (27%), followed by Psychology (10%), Law (8%), and Media, Information and Communication (8%). 27% of the respondents’ studies were categorized as ‘other’. As the total population of Dutch, highly educated millennials between the age of 20 and 30 was estimated to be 629.771 based on numbers of demographic figures of Central Agency for Statistics Netherlands (Centraal Bureau voor de Statistiek)1

, a 8.98% margin of error was achieved with a 95% confidence level (N = 119).

(27)

4.2 Reliabilities

In this paragraph the Cronbach´s alpha of the Brand Knowledge and CRUSH scales are discussed (see Appendix B). For the brand knowledge scale, the open brand recall question was not included, since this is a different type of scale (recalling a brand from the mind as opposed to rating from strongly agree to disagree on statements about those brands) and so this is a different construct. Thus, the Brand Knowledge scale contained 6 items and is considered highly reliable (α = .884) for IBM, but not very reliable for Google (α = .628). It was decided not to remove the (reversed) Brand Associations 3 item (“I have difficulties in imagining Google/IBM in my mind”) from the scale even though this would increase Cronbach’s alpha (from .884 to .897 for IBM, and from .628 to .666 for Google), because the scale contains so few items, which could also explain a lower Cronbach’s alpha (Field, 2013, p. 2356).

For the CRUSH-model the scales include 5 items and are considered reliable for both IBM (α = .774) and Google (α = .745). Despite a higher Cronbach’s alpha of Google after removing the Happiness variable (α = .770), this variable was not removed again because it is such a big part of the CRUSH-model and the scales exists of few items.

4.3 Descriptive Statistics and Correlations 4.3.1 Brand Knowledge

In this section, all the Brand Knowledge descriptive statistics and correlations are discussed. First the results of the open Brand Recall question are addressed. Then all the other Brand Knowledge results for Google, followed by those for IBM, and finally a comparison between the two companies.

(28)

4.3.1.1 Brand Recall: Familiarity with IT Companies

For the first Brand Knowledge variable, Brand Recall 1 (‘Which five IT companies are you most familiar with?’), Apple scored the highest (N = 97) of the five most mentioned companies, then Microsoft (N = 82), Google (N = 51), IBM (N = 45), and finally Samsung (N = 24) as shown in Figure 4. By assigning scores to the answers ranging from 0 (not mentioned at all) to 5 (first mentioned IT-company), Google scored an average of M = 1.479 (SD = 2.033) which was better than IBM with an average of M = 1.118 (SD = 1.718).

Figure 4. Pie chart of the five most recalled IT brands.

4.3.1.2 Brand Knowledge Google

The other six Brand Knowledge scores for Google had an average between 3.546 and 4.647 on the 5-point Likert scale as shown in Table 2, with an overall average Brand Knowledge score of 4.206 (SD = .450) when combing the scales. This indicates that the participants on average agreed to be familiar with the brand. Nonparametric measure Spearman’s rank correlation coefficient (ρ) was used to assess the statistical dependence between the individual Brand Knowledge variables. Several significant correlations were found, indicating a statistical dependence, as shown in Table 2. Brand Associations 1 shows a weak to moderate correlation with all the other variables.

Apple Microsoft Google IBM Samsung

(29)

Table 2

Means, Standard Deviations and Correlations for Google Brand Knowledge

M SD 1 2 3 4 5 6 1. BRC1 1.479 2.033 2. BRC2 4.378 0.748 .129 3. BRG1 3.546 1.015 .178 .340** 4. BRG2 4.429 0.619 -.039 .412** .205* 5. BAS1 4.059 0.751 .265** .299** .338** .384** 6. BAS2 4.647 0.497 .105 .458** .208* .337** .331** 7. BAS3 4.177 0.830 .132 .092 .068 .143 .385** .174

4.3.1.3 Brand Knowledge IBM

For IBM, the other six Brand Knowledge scores (besides the already mentioned Brand Recall 1) had an average between 2.706 and 2.429 on the 5-point Likert scale, as shown in Table 3. This indicates that the participants on average were not familiar with the brand IBM. The only variable on which participants rated higher than neutral, was Brand Associations 2, indicating some familiarity with the logo or symbol of IBM. The Spearman’s rank correlation coefficients indicate statistical dependence between all the six Brand Knowledge variables for IBM (see Table 3), of which the strongest between Brand Recognition 1 and Brand Recall 2 ρ = .757 (p < .001) and between Brand Associations 1 and Brand Recognition 2 ρ = .749 (p < .001).

(30)

Table 3

Means, Standard Deviations and Correlations for IBM Brand Knowledge

M SD 1 2 3 4 5 6 1. BRC1 1.118 1.718 2. BRC2 2.706 1.152 .455** 3. BRG1 2.370 1.080 .386** .757** 4. BRG2 2.840 1.269 .325** .621** .611** 5. BAS1 2.319 1.057 .288** .632** .633** .749** 6. BAS2 3.269 1.418 .393** .704** .560** .633** .598** 7. BAS3 2.429 1.154 .279** .537** .505** .478** .519** .404**

4.3.1.4 Brand Knowledge: Comparing Google to IBM

A Wilcoxon Signed-Ranks Test indicated that all the Brand Knowledge scores of Google were statistically different from the IBM scores (p < .001), except for the Brand Recall 1 variable (see Appendix C). As shown in Figure 5, Google scored much higher on all the variables than IBM. This indicates that participants are more aware of Google, they better know what the brand Google stands for, they can recognize Google better among other competing brands, characteristics of Google come to mind more quickly, they can recall the symbol or logo of Google more quickly, and they have less difficulties in imagining Google in their minds.

(31)

Figure 5. Comparison Brand Knowledge of IBM and Google

4.3.2 CRUSH

In this section, all the CRUSH descriptive statistics and correlations are discussed: first the results for Google, then the results for IBM, followed by a comparison between the two.

4.3.2.1 CRUSH Google

The CRUSH scores for Google ranged from 3.269 to 3.773 on the 5-point Likert scale, with an average of 3.590 (SD = .701). Except for Uniqueness with Happiness, significant correlations between all the CRUSH variables were found, indicating a statistical dependence between the variables (see Table 4). An overall valuation score of 7.983 (SD = 0.823) was measured for Google on a scale from 1 to 10. All the CRUSH scores correlated significantly with this overall valuation score, of which Coolness the strongest ρ = .464 (p < .001).

1 2 3 4 5 Associations 3 Associations 2 Associations 1 Recognition 2 Recognition 1 Recall 2 Recall 1 IBM Google

(32)

Table 4

Means, Standard Deviations and Correlations for Google CRUSH-Variables

M SD 1 2 3 4 5 1. Coolness 3.773 0.828 2. Realness 3.647 0.860 .496** 3. Uniqueness 3.672 0.966 .401** .484** 4. Self-identification 3.269 0.980 .593** .426** .363** 5. Happiness 3.647 0.829 .350** .261** .105 .285** 6. Overall valuation 7.983 0.823 .464** .456** .257** .406** .404** 4.3.2.2 CRUSH IBM

The CRUSH scores for IBM ranged from 2.286 to 3.135 on the 5-point Likert scale, with an average of 2.780 (SD = .924). Significant correlations between all the CRUSH variables were found, indicating a statistical dependence between the variables (see Table 5). For IBM an overall valuation score of 5.924 (SD = 1.1341) was measured on a scale from 1 to 10. All the CRUSH scores correlated significantly with this overall valuation score, of which Realness the strongest ρ = .463 (p < .001).

Table 5

Means, Standard Deviations and Correlations for IBM CRUSH-Variables

M SD 1 2 3 4 5 1. Coolness 2.714 0.772 2. Realness 3.135 0.769 .182* 3. Uniqueness 2.983 0.748 .267** .644** 4. Self-identification 2.286 0.783 .449** .196* .297** 5. Happiness 2.517 0.780 .632** .205* .373** .539** 6. Overall valuation 5.924 1.341 .289** .463** .427** .282** .218*

(33)

4.3.2.3 CRUSH: Comparing Google to IBM

A Wilcoxon Signed-Ranks Test indicated that all the Google CRUSH scores were statistically different from the IBM scores (p < .001) (see Appendix C). As shown in Figure 6, Google scores better than IBM on all CRUSH levels. This indicates that participants think that the brand Google is cooler, more real/authentic, and more unique than IBM, that they can identify more with Google than IBM, and that Google makes them feel more happy than IBM. A difference was that Realness was more correlated to the overall valuation score for IBM, as this was Coolness for Google.

Figure 6. Comparison CRUSH-scores of IBM and Google

4.4 Education Distribution (alpha/beta/gamma)

A Chi-squared test of independence was calculated comparing the relation between the education types (alpha, beta, and gamma) and brand knowledge, CRUSH-scores, and overall valuation of both Google and IBM. For the most part, these scores did not differ by education

1 2 3 4 5 Happiness Self-identification Uniqueness Realness Coolness IBM Google

(34)

type (see Appendix D). However, four variables did show significant differences between the alpha, beta, and gamma students, as discussed next.

First, a significant interaction was found between education type and the Brand Recall 2 variable (‘I am aware of the brand’) for Google (χ2

(6) = 12.759, p = .047). Beta students scored relatively much lower on the brand awareness of Google (66.7% agreed or strongly agreed with the statement), compared to alpha students (90%) and gamma students (96.2%).

Second, a significant interaction was found between the education types and the Brand Recognition 1 variable for Google (‘I know what the brand Google stands for’) (χ2

(8) = 16.778, p = .033*). Gamma students were more likely to know what the brand stands for (60.5% agreed or strongly agreed with the statement) compared to alpha (30%) and beta students (27.8%).

Third, a significant interaction was found between the education types and the Brand Recognition 2 (‘I can recognize Google among other competing brands’) (χ2

(6) = 12.759, p = .047*). Alpha students had relatively more difficulties to recognize Google among other competing brands (80% agreed or strongly agreed with the statement) compared to beta students (94.5%) and gamma students (98.9%).

Fourth, a significant interaction was found between the education types and the Brand Associations 2 variable for IBM (‘I can quickly recall the symbol or logo of IBM’) (χ2

(8) = 16.612, p = .034*). Beta students were less likely to be able to quickly recall the symbol or logo of IBM (22.2% agreed or strongly agreed with the statement) compared to alpha students (70%) and gamma students (68.2%).

4.5 The Effects of Sex on Overall Valuation Scores

A Mann-Whitney test indicated no significant evidence that there is an effect of sex on the overall valuation scores of both Google (M = 7.983), U = 1629.00, p = .822 and IBM (M = 5.924), U = 1367.00, p = .112 (see Appendix E).

(35)

4.6 The Effects of Age on Overall Valuation Scores

The nonparametric Kruskal-Wallis test by ranks indicated no significant difference between the age groups (20-30) for the overall valuation scores of both Google and IBM (see Appendix F).

4.7 CRUSH-values as Predictors for Valuation Google

A multiple linear regression was calculated to predict participant’s overall brand valuation for Google based on their CRUSH values. A significant regression equation was found (F(5,113) = 10.762, p < .000), with an R2

of .323. Participants’ predicted valuation of Google is equal to 5.384 + .247 (Realness) and .247 (Happiness), where Realness and Happiness were measured on a scale from 1 to 5. Participants’ overall brand valuation for Google increased with .247 for one point on both the Realness and Happiness scales. Both Realness and Happiness were significant predictors of the overall brand valuation, but Coolness, Uniqueness, and Self-Identification were not.

4.8 CRUSH-values as Predictors for Valuation IBM

A multiple linear regression was also calculated to predict participant’s overall brand valuation for IBM based on their CRUSH values. A significant regression equation was found (F(5,112) = 10.754, p < .000), with an R2

of .324. Participants’ predicted valuation is equal to 5.1.886 + .517 (Realness), where Realness was measured on a scale from 1 to 5. Participants’ overall brand valuation for IBM increased with .517 for one point on the Realness scale. Realness was a significant predictor of the overall brand valuation, but Uniqueness, Self-Identification, and Happiness were not. Coolness only indicated to be a marginal predictor of .331 (p = .071).

(36)

5. Discussion

The aim of this study was to find out how well the highly educated Dutch millennials know IBM (compared to Google) and on what facets IBM could improve, by testing the brand equity and CRUSH scores. In this chapter first the theoretical implications of the research are discussed, followed by the practical implications. Next an advice for IBM is provided: What should IBM do to improve its brand equity with the highly educated Dutch millennials? To conclude, limitations and suggestions for future research are discussed.

5.1 Theoretical Implications

According to the CRUSH model, coolness, realness, uniqueness, self-identification and happiness are the five elements that Generation Y’s favorite brands all share. This indicates that there should also be a relationship between these elements and the overall valuation of brands. Indeed, this study showed statistical dependence between the individual CRUSH scores and the overall brand valuation for both IBM and Google; however the strength of these relationships was moderate. Moreover, according to Van den Bergh and Behrer, out of all the CRUSH scores, happiness is supposed to have the largest impact on brand leverage (2016, p. 178). Unexpectedly however, for IBM only realness was a significant predictor of the overall brand valuation (and coolness marginally). For Google happiness was a predictor together with realness, but coolness, uniqueness, and self-identification were not. The results or existing literature did not provide an explanation for these findings, so further research would be necessary to find out if there for example is a difference between B2B and B2C brands, between lower and higher educated millennials, or between Dutch millennials and millennials with another nationality.

One finding regarding the brand knowledge model, is that the lower brand knowledge of IBM possibly resulted in higher correlations between the IBM brand knowledge variables compared to Google. This could be explained by the fact that IBM scores low on all the brand

(37)

knowledge variables, whereas having some knowledge about a brand could result in differentiating more between answers. For example, being able to recall the brand logo, but not knowing what the brand exactly stands for, compared to never have heard about a brand and disagreeing with all the statements.

5.2 Practical Implications

By the investigation of the brand equity and CRUSH scores of Google and IBM, this study has provided insights on what Dutch millennials value in brands, which could thus be useful for brand managers when creating and devising brand strategies aimed at this target group. Based on the results regarding the relationship between the five CRUSH variables and the overall valuation of brands, managers (or at least managers of business-to-business brands) could decide on focusing especially on the realness and happiness of their brand to improve the brand equity with Dutch millennials, as these were the most important predictors.

In the current experience economy, the realness or authenticity of brands is becoming more important. The consumers' quest for authenticity will drive marketers to reassess their strategies. According to Van den Bergh and Behrer the modern interpretation of authenticity includes being honest to yourself, to youngsters, and to society (2016, p. 128). This implies that brands should not try to be something that they are not.

Even though happiness was only found to be a predicting variable for Google’s overall brand valuation and not for IBM’s (only correlated), according to Van den Bergh and Behrer it is the most important element of the CRUSH model (2016, p. 178). To improve happiness, managers could try emotional branding, as Generation Y is an emotional consumer generation, which is reflected in their shopping behavior and brand preference (Van den Bergh & Behrer, 2016, p. 208). This could be achieved by maximizing the sensory appeal as well as bringing the brand alive through experiences in events. Gamification, or “the use of game thinking and

(38)

mechanisms to engage users in a certain behavior” (Van den Bergh & Behrer, 2016, p. 201) like the Magnum Pleasure Hunt could also be an approach. Since emotions and happiness spread easily, an intelligent use of social media marketing can boost the feelings of Generation Y (Van den Bergh & Behrer, 2016, p. 208).

In the next paragraph specific practical implications for IBM are discussed. The brand manager could use this advice in the brand strategy to reach millennials.

5.2.1 Advice for IBM

This study has showed that Google does a much better job with the highly educated, Dutch millennials in terms of branding than IBM. Comparing the mean brand knowledge scores of IBM and Google indicated that Dutch, highly educated millennials are more familiar with Google than with IBM (as expected since Google was chosen as a benchmark). The millennials are on average not aware of IBM, they do not know what the brand IBM stands for, they can not recognize IBM among other competing brands, characteristics of IBM do not come to mind more quickly, they can only slightly recall the symbol or logo of IBM, and they have difficulties in imagining IBM in their minds, whereas all the opposite is the case for Google. This implicates that IBM in the first place needs to get millennials acquainted with the brand.

Since IBM BeNeLux does not have high budgets for advertisements, introducing the brand to millennials could perhaps be done at events, festivals, and universities. As there are so many applications of cognitive business, it could be applied to many different disciplines. For example, by going to music festivals and giving suggestions to millennials for new bands or DJ’s based on their Spotify accounts, or using Chef Watson at food festivals to create recipes based on favorite ingredients. Another way is to collaborate with universities to let students work on an IBM business case (such as during the UvA Business Lab), so that students, which are highly educated millennials, would have to familiarize with the brand and its operations. Being active

(39)

and visible on student or business events and career fairs is also a good way to increase brand awareness.

Another (low cost) channel to reach millennials is the Internet, as it was rated very high in importance as a source of advertising information the generation (Valentine & Powers, 2013). IBM could for example increase the number of and improve the content of its posts on Facebook. According to Van den Bergh and Behrer, brands on social networks should behave like friends connecting with them, not just like distant brands, but they should not try to act as a friend in the traditional offline form; rather as one of the passive friendships that exist and develop in online social networks (2016). As the IBM Netherlands Facebook page is not very active and only has a little over 1.400 likes, posting interesting and inspiring videos, articles and links should be the first step. As word-of-mouth marketing is proven to be effective amongst millennials because they trust their friends’ opinion (Van den Berg & Behrer, 2016), having friends liking the brand (page) could cause a snowball effect to increase IBM’s brand awareness.

Besides improving the brand awareness, IBM should focus on its brand image as the company scored low on all the CRUSH variables: Dutch millennials do not think IBM is cool or unique, they feel neutral about IBM being real or authentic, they can not identify with IBM and the brand also does not make them feel happy. As realness was the highest predictor for IBM’s brand valuation, this should be the first focus of attention. As mentioned with the general practical implications, brands and thus IBM should make sure it stays true to the brand and does not try to be something that it is not by imitating or faking.

Coolness was a moderate predictor for IBM and according to the literature an important factor for millennials, so IBM should also try to improve this. “The trick is to keep your marketing campaigns up-to-date without losing your brand’s authenticity” (Van den Bergh & Behrer, 2016, p. 90). The most important sources to find out what is cool, are social media, peers, TV, magazines, advertising and music festivals. Cooling a brand includes constant innovation,

(40)

exclusive offerings and advertising and promotion on cool channels or in cool environments (Van den Bergh & Behrer, 2016, p. 129). One way for IBM to improve its coolness is to collaborate with other cool brands. For example, as the use of Snapchat is increasing under (young) millennials, IBM could do something with cognitive photo filters. Or with a popular alcoholic beverages brand, like Bols, for which IBM Watson could be used to suggest cocktail mixes based on flavor preferences. Google also uses this approach, for example by working together with Levi’s as they introduced a cognitive jacket (Project Jacquard), which makes it possible to weave touch and gesture into any textile using standard, so that everyday objects such as clothes can be transformed into interactive surfaces. IBM should however not try to imitate these ideas, as authenticity is very important to the Dutch millennials as discussed in the previous paragraph. This is also why it is difficult to assert uniqueness when most innovations are copied so quickly. “Due to the overload of choice that Generation Y is confronted with, it is more skeptical of new products than ever” (Van den Bergh and Beher, 2016, p. 131).

As there were no differences found between males and females or the different ages regarding their brand knowledge and valuation, this study does not imply that IBM should focus on one of these particular groups more than the other. However, in terms of education type, beta students seemed to have the lowest brand knowledge on IBM since they were not able to recall the IBM logo, whereas the majority of alpha and gamma students were. Therefore, perhaps IBM should focus most on this group. However, due to the low percentage of beta students in the sample, further research on this field would be desirable to confirm these findings. In the next paragraph other suggestions for future research are discussed.

5.3 Limitations and Future Research

There are limitations of the research that should be considered, such as the use of a non-random student sample. Since the survey was (also) distributed through Facebook and Whatsapp,

Business-to-millennial branding : an exploratory research on the brand knowledge and brand valuation of highly educated Dutch millennials (age 20-30) for IBM and Google