Social Media and Cheese Business Case for the Slankie Brand

(1)

Social Media and Cheese

Business Case for the Slankie Brand

Royal Friesland Campina

Master Thesis, MSc Marketing Management University of Groningen

Faculty of Economics and Business Department of Marketing

January, 2012

Marije de Jong

studentnumber: 1608843 Ondiep Zuidzijde 2 3551 BW Utrecht +31-6 13 45 43 02 marije.dejong@hotmail.com

Supervisors University of Groningen

Dr. J.A. Voerman & MSc. L. de Vries

Supervisor Royal Friesland Campina

(2)

2

PREFACE

This thesis is the final step in completing my Master of Marketing Management of the University of Groningen. It has been a journey writing this report and I am very happy with the result.

My special thanks first of all goes to my supervisor Dr. Liane Voerman of the University of Groningen, for her advice, helpful comments and constructive feedback during the writing process. I would also like to thank MSc. Lisette de Vries, for her feedback in the final phase. Furthermore, I want to thank Royal Friesland Campina for the opportunity to write my thesis for the department of Cheese Specialties and for providing me the information that was necessary to complete this report. I especially want to thank Dhr. Frank de Win for his support and his faith in me during my time as a Marketing intern. With regard to the research process, I am very thankful for the work of Direct Research, who helped me to set up and sent out the questionnaire and who provided me the data. Last but not least, I would like to show my gratitude to my parents, who enabled me to study.

(3)

3

ABSTRACT

The rise of Internet has caused major changes in the media technologies of today. Especially social media have gained competitive advantage during the last years, which has increasingly diversified consumer-to-consumer interactions and which forced companies to redesign their strategy into a more customer-centric approach. It is more and more important to involve the consumer in the total business process, which is referred to as “engagement” in the world of social media. The aim of this study is therefore to provide an insight into the factors that have an influence on the engagement level of consumers, with the brand-consumer interaction level as the central aspect and with social media as the main tool. Based on scientific literature and findings of previous studies, three independent variables (e.g. consumer involvement with a brand, consumer attitude towards a brand and consumer involvement with social media) and various communication characteristics (e.g. appropriateness of social media application, relevancy of content, regularity of interaction, intimacy of interaction and credibility of interaction) are expected to have an influence on this brand-consumer interaction level. Furthermore, the brand-consumer involvement level with a brand is expected to mediate the relationship between the personal relevancy of a brand and the perceived risk of purchase for a brand on the brand-consumer interaction level. The model is tested for the Slankie brand of Royal Friesland Campina, which can be defined as a low-involvement (cheese) brand, since consumers perceive cheese brands to be quite similar. With the results of the study, an advice to the marketing department of this brand can be given, on whether social media tools should be used to increase the engagement level of their consumers and if so, how these tools should be used. The problem statement of this thesis can therefore be defined as follows:

“Should social media tools be used to increase the engagement level of consumers with low-involvement products and if so, how should these tools be used?

Quantitative data is collected by means of an online survey, which is distributed among 203 current users of the Slankie brand. The results show that the consumer involvement level with a brand and the personal relevancy of a brand have a significant effect on the brand-consumer interaction level. In line with this, full mediation of the consumer involvement level for the effect of the personal relevancy of Slankie on the brand-consumer interaction level is supported. The relevancy of the content that is presented also has a significant on the brand-consumer interaction level. The attitude consumer have towards the brand has however a negative effect. The other independent variables do not have a significant effect.

The results of the questionnaire also show that the majority of the Slankie consumers do not want to have an online dialogue with the brand at all, although their activity on social media is high. If the brand does want to use social media in their marketing (communication) strategy, Facebook or the website of the brand are the most appropriate tools to use.

(4)

4

– INTRODUCTION

For this thesis, a research will be conducted to determine whether social media can be effective in the marketing- and communication strategy of Slankie Cheese on the Dutch Gouda cheese market. The strategy of this brand is more and more focused on the emotional and involving aspect of cheese, instead of the functional part of this product. This chapter will start with some background information about cheese, social media and the Slankie brand. Thereafter, the problem statement and the research questions will be stated. The chapter will end with an outline of the structure of the thesis.

§ 1.1 Cheese

Gouda Cheese can be seen as a dairy product that people buy and consume on a daily basis. Especially in the Netherlands, cheese is part of the daily life of people. It has a penetration rate of almost 100% (GfK Nederland, 2011), which means that almost every consumer on the market buys and consumes cheese. Although most Dutch people have a passion for cheese, they do not know much about the product. Besides, the differences between cheeses in the supermarket are considered to be small and with the extended choice that is offered, cheese is bought out of habit and convenience. Just like for many other products and services, the only difference is price (Copernicus Marketing Consulting, 2000 in: Roberts and Alpert, 2010) Therefore, cheese is more and more becoming a commodity product on the market. Next to this, cheese is a non-durable product, mainly communicated with rational argument-based appeals. Since the involvement level of consumers for cheese is quite low, providing arguments and reasons for usage seems to be the best and most appropriate way to send messages (Fennis & Stroebe, 2010). The communication strategy that Slankie Cheese followed in the past years has been in line with this, since the provided information has been functional and rational. It mainly considered argument-based appeals to inform consumers about the brand and about the benefits and healthiness of cheese. Communication about the brand has mainly been to instead of with the consumer, making it only one-way traffic, and the main goal has always been the creation of brand awareness.

(7)

7

§ 1.2 Social Media

In the last few years, major changes have occurred in the technological area, especially when it comes to media technologies. This all started with the entrance of the PC in 1980 and the emergence of Internet in 1985, which provided firms with new ways to conduct business, to communicate ideas and to exchange information (Bell & Loane, 2010). The Internet itself has evolved continually in the last decade and especially social technologies gained competitive advantage, which are referred to as “Web 2.0 technologies” (Bughin & Chui, 2011). These technologies have changed the way consumers and firms work and interact with each other, since it allows for close collaboration, co-operation and networking (Bell & Loane, 2010). Especially social media, which can be defined as “a group of Internet-based applications that build on the ideological and technological foundations of Web 2.0, and that allow the creation and exchange of User Generated Content” (Kaplan & Haenlein, 2010), has gained competitive advantage in the marketplace. It is a medium that is global, open, transparent, non-hierarchical, interactive, and real time (Dutta, 2010) and it includes a variety of online information sharing formats, of which social networking sites (SNSs) like Facebook and micro blogging sites like Twitter are used increasingly (Kaplan & Haenlein, 2010). According to Forrester Research, 75% of Internet surfers used social media in the second quarter of 2008, while this was only 56% in 2007. This growth mainly comes from teenagers and members of Generation X (consumers in the age of 35 till 44 years) (Kaplan & Haenlein, 2010).

So, social media is about relationships and connections between people and organizations (Nair, 2011), in which the active participation of consumers is facilitated (Scott & Orlikowski, 2010). It allows for timely and direct contact between the firm and consumers, at relatively low costs and with high efficiency levels (Kaplan & Haenlein, 2010). Social media allows every consumer to be an online commentator, reviewer and publisher in the market place (Smith, 2009) and because of that, the customer-brand relationship has changed (Edelman, 2010). As a result, many of the traditional marketing strategies and organizational structures have become obsolete and are more and more redesigned into online marketing practices to align with this new relationship (Edelman, 2010). It is important for companies to move away from talking through traditional mass media towards listening and conversing through social media with their consumers (Smith, 2009). Since consumers expect companies to be present at all times (Nair, 2011), companies should respond to this consumer-trend in order not to stay behind in the market. Despite the acknowledged business benefits, calculating the hard numbers or defining whether social media really contributes to a business strategy is difficult, however necessary to justify the spending of time and money on these activities.

§ 1.3 Research Relevance

(8)

8

consumer and the brand will be accomplished (Dorsch et al., 1998). A high engagement level first of all has a direct positive effect on the buying intentions as well as the actual future purchase behaviour of consumers (Bloemer & Lemmink, 1992). One of the most widely used and accepted measures for this value is the Customer Lifetime Value (CLV) (Kumar et al., 2010). Besides, with a higher level of engagement, consumers are intrinsically motivated to generate positive word of mouth, which influences already acquired as well as prospective consumers. This Customer Influencer Value (CIV) is of high importance for firms, since it lowers their acquisition and retention costs and it increases the share-of-wallet of already acquired consumers (Kumar et al., 2010). Lastly, the value of engaged consumers can relate to their Customer Referral Value (CRV) or their ambassadorship, which will also reduce the acquisition costs for companies (Kumar et al., 2010). Therefore, engaged consumers will ultimately become loyal ones, who will bring in revenue growth and profitability at the same time (Heskett et al., 2008).

Since the effects and even the usage of social media as a marketing tool for a dairy product like cheese have not been researched thoroughly, it would be very interesting to research this category of consumer products as well. Above all, it should be researched whether social media tools would be useful for a dairy product like cheese to get consumers more involved and to make them more engaged, as there is no satisfying answer to this question yet. Since the interaction level between brands and consumers on social media can be seen as the central aspect of the engagement level of consumers (Kumar et al., 2010), this construct will be used as the central aspect throughout this thesis. Further elaboration will follow in paragraph 2.2.2.

§ 1.4 Problem Statement & Research Question(s)

To decide whether and how social media tools should be used to increase the engagement level of consumers with low-involvement products, the following problem statement is formulated:

“Should social media tools be used to increase the engagement level of consumers with low-involvement products and if so, how should these tools be used?

To be able to answer this problem statement, the following research questions have to be answered: - What does the relationship between companies and consumers look like and how has it

changed during the years?

- How is consumer engagement constructed?

- How do consumer engagement and brand-consumer interaction relate to each other? - What are the antecedents of the brand-consumer interaction level?

(9)

9

§ 1.5 Structure of the Thesis

(10)

10

2 – THEORETICAL FRAMEWORK

This chapter shows how the relationship between companies and their consumers have changed, due to the rise of social media, and the impact of this on marketing strategies. Besides, the concepts of “consumer engagement (CE)” and “brand-consumer interaction level” are outlined, together with its antecedents. Hypotheses are formulated and a conceptual model is presented as well.

§ 2.1 The Brand – Consumer Relationship

Before the rise of technology in communication channels, it has always been assumed that a large number of potential brands were narrowed down in the buying decision process of consumers and that eventually one brand was chosen (Edelman, 2010). After the purchase was made, the brand relationship focused on the usage of the product or services. To facilitate and influence the buying decision, companies used traditionally paid-media push marketing at a few well-defined points along the way in their marketing strategy, to build brand awareness, drive brand consideration, and ultimately inspire purchase (Edelman, 2010). However, Word of Mouth (WOM), which can be defined as “the act of exchanging marketing information among consumers” (Katz & Lazarfeld, 1955 in: Chu & Kim, 2011) has always been one of the most influential channels of communication in the marketplace in the formation of consumer opinions, as well as in post-purchase product perceptions (Bone, 1995). It was more effective in influencing the consumer purchase than advertising or personal selling (Katz & Lazarfeld, 1955 in: Gruen et al., 2006).

(11)

11

increasingly diversifies the consumer-to-consumer (C2C) interactions and because of that, companies should see these interactions in a broader perspective, beyond just the verbal communication of WOM (Libai et al., 2010). Social networking sites (SNSs) are increasingly used to facilitate eWOM and especially the collaborative and social characteristics of these sites have changed the consumer-to-consumer conversations (Chu & Kim, 2011) and makes the society increasingly networked (Verhoef et al., 2010). This user-generated content has become a mass phenomenon which threatens established business models (Hennig-Thurau et al., 2010) and can have a big impact on the reputation of a brand or company (Fisher, 2009). Despite this impact however, the content, timing and frequency of consumer conversations of social media are often beyond manufacturers’ and retailers’ control (Mangold & Faulds, 2009), which is in contrast with traditional marketing communications that were generated and spread in the past by media companies and publishing houses (Fisher & Reuber, 2011).

All these changing and evolving media scenarios and consumer profiles have forced companies to redesign their strategy into a brand strategy that focuses on the customer, which has resulted in (marketing) approaches that are more and more customer-centric. The management of customer relationships has become a top priority for many companies (Verhoef et al., 2010). Social media changes the way of communication in the marketplace and companies should respond to this by actively start listening and conversing with their consumers, instead of just talking and telling their story (Smith, 2009). The non-transactional behaviour of consumers should be managed properly (Verhoef et al., 2010) and social media should be used as a strategic instrument (Nair, 2011). Companies should use online communities to help uncover insights, to generate and validate new concepts and to create conversations with their customers (Maddock & Vitn, 2009). They must learn from and collaborate with these consumers, in order to meet their needs and to be able to create value (Prahalad & Ramaswamy, 2000).

In summary it can be concluded that it is all about involving the consumer in the total business process. In the world of social media, the word “engagement” is used to refer to this consumer involvement, which allows companies to engage in social interactions on a scale that was not possible before (Fisher & Reuber, 2011) and which should be seen as an opportunity.

§ 2.2 Consumer Engagement (CE)

(12)

12

§ 2.2.1 Dimensions of Consumer Engagement

Consumer engagement can be explained with the dimensions purpose, valence, form, scope and the nature of impact (Van Doorn et al., 2010). These dimensions will be explained in the following paragraph.

In order to understand the concept of consumer engagement, it is first of all important to know the reason and purpose for engagement (Van Doorn et al., 2010). To whom the engagement is directed, whether it is planned or not and the goals of the consumer all together form this consumer engagement purpose. When the reason for consumer engagement is decided upon, the engagement itself can be expressed in a positive or negative way, which refers to the valence of the concept. Consumers can use blogs or online reviews for example to express their engagement and the way they think or feel about things. The way consumer engagement is expressed can have consequences for a firm, financially as well as non-financially (Van Doorn et al., 2010). The third dimension of consumer engagement is the form or modality of it, which refers to the different ways it can be expressed. At the most basic level it concerns the different types of resources, for example time and money, consumer may utilize to express their engagement. The last dimension of consumer engagement is the scope of it, which can relate to time or geography. The temporal reach refers to whether consumer engagement is temporally momentary or ongoing and the geographic reach can be local or global reach (Van Doorn et al., 2010).

Overall, the nature of impact of consumer engagement can be different in every situation and it can be expressed in several ways. The immediacy of the impact relates to how quick the engagement behaviour has an effect on the constituents, whereas the intensity refers to the level of change that is accomplished within the target group. The breadth of impact is synonym for the reach of it and the longevity tells something about the time the engagement behaviour lasts (Van Doorn et al., 2010). § 2.2.2 Brand-Consumer Interaction as part of Consumer Engagement

(13)

13

§ 2.3 Antecedents of Brand – Consumer Interaction

There are various antecedents of the interaction level between brands and consumers. In order to simplify the model and to get a clearer overview, they are bundled into different areas for this thesis. In the following paragraphs four of these areas are outlined, which are characteristics of brand-consumer interaction (§2.3.1.), brand-consumer attitude towards a brand (§2.3.2), brand-consumer involvement with a brand (§2.3.3) and consumer involvement with social media (§2.3.4.).

§ 2.3.1 Characteristics of Brand – Consumer Interaction

There are various characteristics that are related to the interaction level between brands and consumers. In this thesis, the appropriateness of the medium that is chosen for interaction, the relevancy of the content that is presented and the intimacy as well as the regularity of the interaction are considered to be the most important ones. Besides, the credibility of the brand that is interacting is an important characteristic of the level of interaction between brands and consumers. These five characteristics will further be explained in this paragraph.

Consumer engagement reflects the motivational investment of consumers in the interaction with a brand and it seems that a two-way interaction process is essential in reaching this engagement (Kumar et al., 2010). Companies should actively start listening and conversing with their consumers, instead of just talking and telling their story (Smith, 2009), which makes the old fashion one-way traffic approach obsolete. By encouraging consumers to participate, a mutual dialogue is stimulated and interaction accomplished (Ahuja & Medury, 2010). This reflects the shift from a good-centred to a more service-centred orientation (Vargo & Lusch, 2004), in which customers are seen as proactive co-creators. The digital environment, to which social media applications belong, is used as a primary tool to establish this (Gambetti & Graffigna, 2010). It should be noted however, that in choosing an online application, the target group should be taken into consideration, since different applications attract different people and it is important for brands or companies to be active wherever their consumers are active (Kaplan & Haenlein, 2010). Besides, some applications are more effective to establish a two-way interaction process than others and are therefore more appropriate to encourage the interaction between brands and consumers. Especially virtual online communities, like Hyves and Facebook, and Weblogs provide great opportunities for interaction and participation, whereas applications like the creativity works-sharing site Youtube or micro blogging site Twitter are better to entertain the crowd or to generate traffic (Gambetti & Graffigna, 2010; Ahuja & Medury, 2010; Mangould & Faulds, 2009). No matter which social media applications are chosen to establish a dialogue with the consumer, it is important that all these applications are aligned with each other.

Hypothesis 1: The higher the appropriateness of the social media application that is chosen, the

higher the interaction level between the brand and the consumer.

(14)

14

information and content about brands and products that are relevant to them, their desire to interact increases as well as their level of engagement (Mangold & Faulds, 2009). Furthermore, regular and meaningful brand encounters should be created and content should be kept up-to-date. When the moments of contact are repeated, consumer interest can be stimulated and participation can be achieved. This will strengthen the relationship between the consumer and the brand or company, which will increase their level of engagement (Ahuja & Medury, 2010).

Hypothesis 2: The more consumer-relevant the content of interaction is, the higher the interaction

level between the brand and the consumer.

Hypothesis 3: The more regular the interaction between the brand and the consumer is, the higher

the interaction level between the brand and the consumer.

Blogs are also used to induce consumers to participate, by providing feedback and by posting comments, which are easy to read and follow by others. The biggest benefits for users are therefore social as well as informational, since content can be read and posted (Ahuja & Medury, 2010). By focusing on these possibilities of blogs, the behavioural activation component of engagement is activated (Gambetti & Graffigna, 2010), which brings in great opportunities for companies or brands. This way of participation helps to build trust between the brand and the consumer and it contributes to the feeling of consumers of being part of a community (Mangold & Faulds, 2009). This can only be accomplished however, when communication is informal and intimate and focuses on the relational dimension between the two parties (Ahuja & Medury, 2010). The illusion of having face-to-face interaction, which can be referred to as Para-Social Interaction (PSI), is important here (Horton & Wohl, 1956 in: Colliander & Dahlén, 2011). Informality and intimacy are also important in producing content with the consumer, to explore their attitudes and expectations. This co-production is a result of decision making by consumers that reflects their preferences (Etgar, 2008). Platforms are used for interaction, to exchange know-how, and it is possible that value that might be overlooked will be unlocked (Gruen et al., 2006). This feedback is referred to as Customer Knowledge Value (CKV) (Kumar et al., 2010) and this knowledge capital is useful in segmenting consumer bases and in the formulation of strategies (Ahuja & Medury, 2010). It is also possible that know-how will be unlocked in the conversation between consumers.

Hypothesis 4: The more informal and intimate the interaction is, the higher the interaction level

between the brand and the consumer.

In general, blogs can be seen as a form of eWOM, which is described in § 2.1. In WOM, the listener puts a lot of weight in the sender of the message (Dichter, 1966 in: Colliander & Dahlén, 2011) and especially the credibility of the source is of high importance for persuasion. Consumers are more likely to respond to published items when they perceive the sender to be credible and unbiased (Colliander & Dahlén, 2011). This holds for the conversation between brands and consumers as well as for consumer-to-consumer interactions.

Hypothesis 5: The higher the credibility of the interacting brand, the higher the interaction level

(15)

15

§ 2.3.2 Consumer Attitude towards a Brand

Consumers are faced with a tremendous amount of advertising messages and brand encounters per day (Roberts & Alpert, 2010). Together with the fact that people do not have a lot of time, purchasing decisions are being made in 2.6 seconds on average (Fisk, 2006 in: Roberts & Alpert, 2010). Next to this, the world is increasingly competitive, connected and commoditised and the society is multi-tasking (Gambetti & Graffigna, 2010). All of this makes it hard, however highly important for brands to stand out from the crowd, which can be done when great customer experiences are delivered. In the interaction with the consumer, it is important to communicate a strong value proposition in a consistent way, which is constituted by a myriad of things. Not only the psychical aspects like the features, reliability and timeliness of the product are important; the emotional elements as being friendly and welcoming as well as the usage of humour should not be forgotten (Roberts & Alpert, 2006). By using for example surprises or novelties that are in some way unexpected, consumer emotions can arise and engagement can be enhanced (Wang, 2006). Furthermore, a message should be consistent across all media and companies should make sure one style and voice is used to spread the message. In the communication about a brand or product, different types of appeals can be used to reach consumers and to get messages across. By using argument-based appeals, advertisements focus on the usage of the product, whereas with affect-based appeals, consumer emotions and feelings are targeted. The advertising strategy that is chosen depends on the type of product that is advertised, which can either be experiental or (non-)durable, and the involvement level of the consumer (Fennis & Stroebe, 2010).

The way consumers feel about a brand or product is an important part of their total experience with this specific brand (Roberts & Alpert, 2010) and especially brands or products that elicit positive affect can draw engagement and therefore interaction (Wang, 2006). When consumers’ overall attitude towards the brand or product is positive, it is likely they want to establish a dialogue with the brand, in order to have the same brand experience again.

Hypothesis 6: The more positive the attitude of the consumer towards the brand, the higher the

interaction level between the brand and the consumer.

§ 2.3.3 Consumer Involvement with a Brand

Next to the communication characteristics and the attitude of consumers towards a brand, the involvement level of the consumer with a brand or product is also an important antecedent to take into consideration (Hollebeek, 2011). This subject has been discussed thoroughly in literature and it has been used to refer to at least five distinct concepts (e.g. ego involvement, commitment, communication involvement, purchase importance and response involvement) (Muncy & Hunt, 1984). Furthermore, it can relate to products, advertisements, messages, programmes, situations and behaviours (Sridhar, 2007). Since consumer involvement covers so many different areas and since there is still no agreement about this topic among researchers, some choices about this subject have been made for this thesis, with regard to the definition of the concept as well as for the antecedents that are related to it.

(16)

16

and psychical effort in consumer decision making (Beharrell & Denison, 1995 in: Sridhar, 2007), which is determined by product attributes and the relatedness of the product to the needs and wants of the consumer (Finn, 1983). Different people can perceive the same brand or product differently and everyone can have different levels of involvement (Zaikowsky, 1985), which is determined by characteristics of a person (Zaikowsky, 1986). Certain product classes may be more or less central to an individual’s life, i.e. automobiles or soft drinks for example (Traylor, 1981).

When the involvement level of consumers increases, the amount of attention and comprehension that is devoted to a brand or product increases as well (Celsi & Olson, 1988). Consumers who are more involved are more active processors of cognitive information about a brand or product (Ray et al., 1973 in: Swinyard, 1993). This relates to “response involvement” that is mentioned above, which is defined as “the complexity of cognitive and behavioural processes in the consumer decision process” (Houston & Rothschild in: Muncy & Hunt, 1984). Consumers who are more involved with a product will choose a brand more thoughtfully and are more likely to communicate about it with others. The level of involvement therefore influences the decision making of consumers as well as their communication behaviours (Michaelidou & Dibb, 2008).

Hypothesis 7: The higher the involvement level of the consumer for a brand, the higher the

interaction level between the brand and the consumer.

§ 2.3.3.1 Antecedents of Consumer Involvement with a Brand

There are various antecedents of the involvement level of consumers with a brand. In this thesis, “personal relevancy” is mentioned as the first antecedent of consumer involvement, conform the theory of Zaikowsky (1985). The “perceived risk of purchase” is the second antecedent of consumer involvement (Sridhar, 2007).

Personal Relevance of a Brand

In the academic literature about consumer involvement it is often confused with the personal relevancy of a brand. Some researchers use the words synonymously; others clearly make a distinction between the two. In this thesis, personal relevancy is an antecedent of consumer involvement, conform the theory of Zaikowsky (1985). It is defined as “the fit between a brand or product and the consumer” (Ephron, 2005) and something is personally relevant when consumers perceive it to be self-related or in some way instrumental in achieving their personal goals and values (Celsi & Olson, 1988). Brands or products speak to consumers’ interest when they are high in utility or relevancy. When there are no significant differences among brands, product features or prices are chosen that fit the personal goals of the consumer (Wang, 2006).

(17)

17

system of the consumer, fits this type of personal relevance (Ostrom & Brock, 1968 in: Muncy & Hunt, 1984). When personal relevance for a brand or product is determined by intrinsic sources, consumers have probably developed a preference for a brand or product and a relationship between the two has emerged. Since this will more likely lead to engaged consumers, it will be this type of personal relevance that is referred to throughout this thesis.

Hypothesis 8: The higher the personal relevance of a brand for the consumer, the higher the

involvement level of the consumer with that brand.

Perceived Risk of Purchase for a Brand

The perceived risk that is related to a brand or product also has an effect on the involvement level of consumers (Sridhar, 2007). It is defined by Schiffman and Kanuak (2002 in: Sridhar, 2007) as “the uncertainty that consumers face when they cannot foresee the consequences of their purchase decisions”. It covers both the uncertainty in purchase and as well as the consequences of mispurchase and it can be decomposed in several types of risk (e.g. financial, social, performance risk etc.) (Sridhar, 2007). The price of a brand or product is commonly used as an indicator of (financial) risk (Laurent & Kapferer, 1985). Consumers are more involved when a brand or product with a relatively higher price is bought, since the risks of mispurchase are higher (Rothschild, 1979 in: Laurent & Kapferer, 1985). This also holds for durable products, to which consumers can get stuck for a long time. Decisions about durable goods are in general more highly involving decisions (Traylor, 1981).

Hypothesis 9: The higher the perceived risk of purchase for a brand, the higher the involvement level

of the consumer with that brand.

§ 2.3.4 Consumer Involvement with Social Media

As is mentioned in paragraph 2.3.3, consumer involvement can relate to products, advertisements, messages, programmes, situations and behaviours (Sridhar, 2007). In line with this, it can be assumed that consumers can also be involved with social media. Although it is more and more integrated into the lives of people, some consumers have a larger interest in the possibilities of these applications than others and they will probably use it on a more regular basis in their online interaction with other consumers or brands. The involvement that consumers have with social media can therefore be an important antecedent of the brand-consumer interaction level.

Hypothesis 10: The higher the involvement level of the consumer with social media, the higher the

(18)

18

§ 2.4 Conceptual Model Brand – Consumer Interaction Level

By combining the hypotheses that are described in the previous paragraphs, the relations between the antecedents and the interaction level between the brand and the consumer can be graphically displayed. The conceptual model that is presented below will give insights of these linkages and it shows how the various aspects are related. The characteristics that are closely related with the interaction between the brand and the consumer are taken together.

(19)

19

3 – RESEARCH METHOD

Before the hypotheses can be tested, the type of research that will be used will first be explained. Thereafter, the population and sample and the data analysis will be outlined.

§ 3.1 Type of Research

In order to test the hypotheses that are stated in the previous chapter, a descriptive research design will be used, which can be characterised as pre-planned and structured. With the usage of surveys, numerical data will be gathered, which makes the study a quantitative research (Malhotra, 2009).

§ 3.2 Population and Sample

Slankie Cheese is a consumer brand of Royal Friesland Campina that belongs to the department of Cheese Specialties. This department is part of the division Cheese, Butter and Milkpower and together with the departments Consumer Products Europe, Consumer Products International and Ingredients, it forms the total international company of Royal Friesland Campina. The Slankie brand is a local dual, which means it is only branded on the Dutch cheese market, on which pieces, slices and grater are offered, next to cheese spreads. On the total Dutch cheese market, it has a market share of 1,5%, while on the market for cheese spreads, Slankie covers 23% of the market (Nielsen, 2011). The brand positions itself as a product for weight management, which is mainly consumed by consumers in a functional way, to solve a problem. Consumers who care about their weight and people who are dieting are therefore the biggest target group for the brand. This will mainly be women in the age of 30 till 49 years old. The questionnaire will be randomly sent to the target group of the Slankie brand. The researcher wants to gather at least 150 respondents who are current users of the brand on the Dutch market.

In total, 203 respondents have filled in the questionnaire for Slankie; mainly women with kids living at home or women without any children, who are highly educated (e.g. HBO or university) and have an average or above average income level. The average age of the women is 40,71 years. Looking at the characteristics and different backgrounds of the respondents it can be said that it is a representative sample of the target group of Slankie. A total description is given in table 1, 2 and 3.

Household Situation Slankie target group Percentage

Single 14,6%

Single, with kid(s) living at home 6,9%

Single, with kid(s) not living at home 0,9%

Living together with roommates 0,0%

Living together/married, no kids 18,9%

Living together/married, with kid(s) living at home 55,8% Living together/married, with kid(s) not living at home 1,3%

Living with parents/family 0,4%

Other 1,3%

(20)

20

Education Level Slankie target group Percentage

Primary School 0,0% LBO/VBO 7,3% MAVO/VMBO 14,6% HAVO/VWO/GYMNASIUM 8,2% MBO 14,2% HBO 36,9% University 15,5% Other 3,4%

Table 2: Education Level Slankie target group

Average Income Slankie target group Percentage

Less than average 11,8%

About average 25,6%

Between average and 2x average 27,1%

More than 2x average 10,8%

Want to keep it to myself 24,6%

Table 3: Average Income Slankie target group

§ 3.3 Data Analysis

(21)

21

indicate the validity of the answers that are given by the consumers. A pre-test of the questionnaire is first of all done by the researcher of this thesis.

§ 3.3.1 Measurement Model

The questionnaire that is sent to the target group of the Slankie brand starts with some general questions about the respondents’ usage of social media applications. After these questions are answered, the respondent is asked about the different constructs that are presented in the conceptual model and the corresponding hypotheses. This model consists of four independent variables and one dependent variable. The independent variables “characteristics of interaction”, “consumer attitude” and “consumer involvement with social media” are expected to have a direct effect on the dependent variable “brand-consumer interaction level”. Next to this, it is assumed that the independent variable “consumer involvement level with Slankie” mediates the effect of the personal relevancy of Slankie and the perceived risk of purchase for the brand on the brand-consumer interaction level. To measure the constructs, validated scales are used from earlier academic research (table 4). In order to test the effect of the communication characteristics on the brand-consumer interaction level, a snapshot of the current Slankie Hyves is used in the questionnaire that is distributed among Slankie users. Besides, some content is created to investigate whether there is a difference between various ways of communication.

The independent variables will be measured by multiple items, with corresponding questions. These questions are used as input for the questionnaire. The questions without reference are formulated by the researcher. Constructs Hypotheses and Questions Source CHARACTERISTICS OF INTERACTION

Appropriateness of Social Media Application

Relevancy of Content Regularity of Interaction Intimacy of Interaction Credibility of the Brand

H1: Q31 H2: Q32 H3: Q36 H4: Q33 H5: Q34 H5: Q35, 36

Self developed; showing three types of communication expressions. Mishra, Umesh & Stem Jr., 1993 Mohr & Sohi,1995

Self developed. Mohr & Sohi, 1995

MacKenzie & Lutz’ Writer Credibility three-item Scale, 1989

CONSUMER ATTITUDE H6: Q9 Self developed.

CONSUMER INVOLVEMENT LEVEL H7: Q10, 11 Zaichkowsky’s Personal Involvement

(22)

22

Personal Relevancy

Perceived Risk of Purchase

H7: Q12 H8: Q13 H8: Q14, 15, 16 H9: Q17, 18, 19, 20, 21, 22

Mittal & Lee’s Level of Involvement Scale, 1988

Zaichkowsky’s Personal Involvement Inventory Scale, 1994

Laurent & Kapferer’s Product-Class Involvement Scale, 1985

Sridbar’s Perceived Risk Scale, 2007

BRAND-CONSUMER INTERACION LEVEL Q23, 24

Q25

Q27

Fisher, Maltz & Jaworski, 1997 Gruner & Homburg’s Co-Creation Theory, 2000

Fisher, Maltz & Jaworski, 1997

Table 4: Used metrics to test hypotheses

§ 3.4 Plan of Analysis

The data that is gathered by the questionnaires will be put into SPSS, in order to analyse whether the independent variables have an effect on the dependent variable “brand-consumer interaction level”. The results of this analysis can be found in chapter 4.

In paragraph §3.4.1, the Cronbach’s Alpha’s for Slankie will be given, in order to decide whether the various questions that have been asked in relation to a construct can be taken together. In paragraph §3.5.2, the multiple regression analysis will be explained.

§ 3.4.1 Cronbach’s Alpha’s for Slankie

(23)

23

decide whether the separate questions can be taken together to represent the total construct. The various Cronbach’s Alpha’s for each construct are presented in table 5.

Construct Cronbach’s Alpha

Personal Relevancy of Slankie 0,978

Perceived Risk of Purchase for Slankie 0,841

Involvement Level for Slankie 0,956

Credibility Slankie Hyves 0,965

Credibility Slankie Facebook 0,963

Credibility Yoga Website 0,966

Interaction Level Slankie Hyves 0,934 Interaction Level Slankie Facebook 0,926 Interaction Level Yoga Website 0,921

Table 5: Cronbach’s Alpha’s for the constructs

As can be seen in table 5, the Cronbach’s Alpha’s are high enough for every construct, which means that the internal consistency reliability of the set of items is satisfactory and that they can be taken together to represent the separate constructs. In further analyses, the average of the four questions that are asked in relation to the personal relevancy of Slankie is for example used to represent this construct (i.e. question 13+14+15+16 / 4).

§ 3.4.2 Multiple Regression Analysis for Slankie

After it is decided whether the summated scales are reliable or not, a regression analysis will be done to decide whether the independent variables have an effect on the dependent variable. Since the conceptual model includes one dependent (e.g. brand-consumer interaction level) and multiple independent variables, multiple regression analysis is appropriate to use (Malhotra, 2009). The relationship between the independent variables and the dependent variable can be summarized with a regression line, which represents the best-fitting line through the data points (Malhotra, 2009).

The basic regression equation can be written as follows:

Y = β0 + β1 * X + e1 Y = dependent variable X = independent variable Β0 = intercept of the line Β1 = slope of the line e1 = error

(24)

24

hypothesized that the involvement level of consumer with the brand as well as with social media and the attitude consumers have towards Slankie will have an effect on this interaction level.

The multiple regression analysis that will be performed for the general brand-consumer interaction level (BCIgeneral) recognises the following model:

BCIgeneral = β0 + β1 (CIL) + β6 (ATT) + β7 (CIS)

Where:

CIL consumer involvement level Slankie ATT consumer attitude towards Slankie CIS consumer involvement level social media

Besides, a multiple regression analysis will be done after the communication expressions on Hyves, Facebook and the website about yoga have been presented, to indicate the total interaction level consumers indicate. The appropriateness of the medium that is used, the relevancy of the content that is presented, the intimacy of the interaction and the credibility of the brand in its communication are included here, together with the variables that are mentioned before. The regularity of interaction cannot be included in these analyses, since only one example is shown and rated each time, while regularity involves multiple measures in time. Hypothesis 3 can therefore not be tested in this model. By measuring the intended interaction level of consumers with Slankie at two points in time (before and after the communication expressions have been shown), it can be decided what the general interaction level of consumers is and whether showing the expressions has an effect on this level.

The multiple regression analysis that will be performed for the total brand-consumer interaction level (BCItotal) recognises the following model:

BCItotal = β0 + β1 (APP) + β2 (REL) + β3 (INT) + β4 (CRED) + β5 (CIL) + β6 (ATT) + β7 (CIS)

Where:

APP appropriateness of social media application REL relevancy of content

INT intimacy of interaction CRE credibility of interaction

CIL consumer involvement level Slankie ATT consumer attitude towards Slankie CIS consumer involvement level social media

(25)

25

the dependent variable. The closer the number is to 1, the greater this effect is. The separate coefficients will be used to decide upon the level of influence of each variable and the significance number tells whether the level of influence is significant for the model that is tested. Throughout this thesis, a significance number of .05 will be held (Malhotra, 2009).

In general, multiple regression analysis can be complicated by the presence of multicollinearity, which is the intercorrelations among the predictors that are used in the model (e.g. de independent variables) (Malhotra, 2009). The higher the multicollinearity in the regression, the less precisely the partial regression coefficients can be estimated and the more difficult it is to assess the relative importance of the independent variables in explaining the variation in the dependent variable. If the value of tolerance is less than 0,1, so when the variance inflation factor (VIF) is higher than 10, the collinearity is problematic and one of the independent variables can better be excluded from the multiple regression analysis (Field, 2000).

The values of tolerance and variance inflation factors for the involvement level of consumers with Slankie and the general brand-consumer interaction level are first of all presented in table 6 and 7.

Constructs Tolerance VIF

Personal Relevancy of Slankie 0,881 1,127

Perceived Risk of Purchase for Slankie 0,881 1,127 Dependent Variable: Involvement Slankie

Table 6: Multicollinearity Involvement Level Slankie

Constructs Tolerance VIF

Consumer Attitude towards Slankie 0,663 1,507

Consumer Involvement Level Slankie 0,651 1,535

Consumer Involvement Social Media 0,978 1,023

Dependent Variable: Interaction Slankie

Table 7: Multicollinearity Brand-Consumer Interaction Level Slankie

As can be seen in table 6 and 7, there is no multicollinearity between the independent variables in these multiple regression analysis and they can therefore be analysed together. The results can be found in chapter 4 of this thesis.

(26)

26

Communication Characteristics Slankie Hyves

Tolerance VIF

Appropriateness of Medium 0,281 3,557

Relevancy of Content 0,182 5,497

Intimacy of Interaction 0,178 5,607

Credibility of the Brand 0,125 8,018

Consumer Involvement Slankie 0,633 1,579

Dependent Variable: Interaction Slankie Hyves

Table 8: Multicollinearity Brand-Consumer Interaction Level Slankie Hyves

Communication Characteristics Slankie Facebook

Dependent Variable: Interaction Slankie Facebook

Table 9: Multicollinearity Brand-Consumer Interaction Level Slankie Facebook

Communication Characteristics Yoga Website

Dependent Variable: Interaction Yoga Website

Table 10: Multicollinearity Brand-Consumer Interaction Level Yoga Website

(27)

27

correlation analyses for the three communication expressions (Appendix 2) show that the credibility characteristic mainly correlates with the other communication characteristics (e.g. appropriateness of the medium, relevancy of the content and intimacy of the interaction). This is not very surprising, since the communication characteristics are closely aligned with each other and consumers probably rate the expressions in an integrated way.

Since the credibility characteristic for the website about yoga is higher than the value of tolerance, it should officially be excluded from the multiple regression analysis and a separate regression analysis should be performed. However, since the number of multicollinearity is only a little too high and since it is preferred to keep this independent variable in the multiple regression analysis, in order to compare the three multiple regression analyses in a consistent manner, this characteristic will not be excluded. All the variables will therefore be included in the multiple regression analyses for the three communication expressions, of which the results can be found in chapter 4 of this thesis.

§ 3.4.3 Mediation Analysis Consumer Involvement Level Slankie

As can be seen in the conceptual model, the influence of the personal relevance of Slankie and the perceived risk of purchase for the brand on the brand-consumer interaction level is expected to pass through the consumer involvement level. This relation is referred to as a mediating effect, which is created when a third construct (consumer involvement level) intervenes between other related constructs (personal relevance and perceived risk vs. brand-consumer interaction level) (Baron & Kenny, 1986). This effect is present when there are significant correlations between the constructs and there is complete mediation when the mediating construct (consumer involvement level) explains the relationship between the other constructs. When some of the relationship is explained, there is partial mediation. In order to test for mediation by means of multiple regression analysis, a series of steps apply (Baron & Kenny, 1986).

Step 1: establish that the necessary individual relationships are statistically significant:

 The personal relevancy of Slankie and the perceived risk of purchase for the brand on the brand-consumer interaction level: establishing the direct relationship;

 The personal relevancy of Slankie and the perceived risk of purchase for the brand on the consumer involvement level for Slankie: establishing that the mediator is related to the input constructs;

 The consumer involvement level for Slankie on the brand-consumer interaction level: establishing that the mediator has a relationship with the outcome construct.

The relationships recognise the following models:

(28)

28

Where:

PR personal relevance of Slankie

RIS perceived risk of purchase for Slankie

Step 2: estimate a regression model adding the mediating variable CIL to the direct relationship from

personal relevancy of Slankie and perceived purchase for the brand to brand-consumer interaction level. The extent of mediation is assessed as follows:

 If the relationship between personal relevance of Slankie and perceived risk of purchase for the brand and brand-consumer interaction level remains significant and unchanged once consumer involvement level is included in the model, mediation is not supported;

 If the relationship between personal relevance of Slankie and perceived risk of purchase for the brand and brand-consumer interaction level is reduced but remains significant when consumer involvement level is included in the model, partial mediation is supported;

 If the relationship between personal relevance of Slankie and perceived risk of purchase for the brand and brand-consumer interaction level is reduced to a point where it is not statistically significant once consumer involvement level is included in the model, full mediation is supported.

The multiple regression analysis wherein the mediating variable “consumer involvement level” is included recognises the following model:

BCI = β0 + β1 (PR) + β2 (RIS) + β3 (CIL)

(29)

29

4 – RESULTS

In this chapter, the data of the 203 Slankie respondents that is gathered with the questionnaire will be analysed in a quantitative way. In this questionnaire, questions about social media and the usage of it have first of all been asked. Thereafter, respondents had to indicate their involvement level with Slankie and their attitude towards the brand. In the second part of the questionnaire, the communication expressions on Hyves, Facebook and the website about yoga have been presented, which respondents have ranked on the different communication characteristics. They have also indicated their intended level of interaction with Slankie after each expression. At the end of the questionnaire, demographic information of the respondents is gathered.

§ 4.1 Social Media and Slankie

The results for the activity of the respondents on social media show that that 4,4% does not have a social media account at all, 32,1% indicates a (very) low activity, 13,3% is in the middle and 50,2% indicates a (very) high activity on social media, which is a little more than half of the respondents. In table 11 is presented what the level of activity is for the social media applications that are mainly used.

Social Media Application Active Usage Passive Usage

Facebook 30,5% 41,2%

Hyves 16,5% 20,1%

Youtube 4,6% 47,5%

Twitter 8,8% 13,4%

LinkedIn 6,1% 11,3%

Table 11: Activity on Social Media by Slankie respondents

The other social media applications (e.g. blogs, Flickr, Foursquare and MySpace) are the least often used by the respondents, passively as well as actively.

In table 12 is presented what the main reasons for consumers are to use social media, together with the social media applications that are most often used.

Usage of Social Media Percentage Facebook Hyves Twitter

To get information about brands 32% 61,3% 30,6% 19,4%

To communicate about brands 22,7% 86,4% 34,1% 29,5%

To communicate with brands 21,6% 78,6% 28,6% 23,8%

Table 12: Usage of Social Media by Slankie respondents

(30)

30

§ 4.2 Mediation Analysis Consumer Involvement Level Slankie

In the conceptual model, it is assumed that the consumer involvement level for Slankie mediates the effect of the personal relevancy of Slankie and the perceived risk of purchase for the brand on the brand-consumer interaction level. In order to decide whether this mediating effect is indeed present, three multiple regression analyses should be performed.

First of all, there should be a direct relationship between the personal relevancy of Slankie and the perceived risk of purchase for the brand on the brand-consumer interaction level. The results of this multiple regression analysis can be found in table 13.

Brand-Consumer Interaction Model Slankie

p-value of the model 0,000

F-value 47,363

R-square 0,321

Constructs B Significance

(Constant) 0,635 0,001

Perceived Risk of Purchase for Slankie 0,190 0,005

Table 13: Mediation Analysis Brand-Consumer Interaction Level Slankie

As the results of this analysis show, the model is overall significant with an F-value of 47,363 (p=0,0000) and a direct relationship between the constructs is established, since the personal relevancy and the perceived risk of purchase both have a statistically significant effect on the brand-consumer interaction level. The regression line for this analysis can be written as follows:

BCI general = 0,635 + 0,414(PR) + 0,190(RIS)

It should now be decided whether the independent variable “consumer involvement level for Slankie” is related to the input constructs “personal relevancy of Slankie” and “the perceived risk of purchase for the brand”, as is stated in hypothesis 8 and 9. In order to test these two hypotheses, a multiple regression analysis is executed, of which the results are presented in table 14.

Consumer Involvement Level Slankie

F-value 1309,776

R-square 0,929

(Constant) 0,213 0,004

Perceived Risk of Purchase for Slankie -0,002 0,946

(31)

31

As can be seen in table 14, the multiple regression is overall significant with an F-value of 1309,776 (p=0,000) and the consumer involvement level is highly decided by the two variables. However, the personal relevancy of Slankie mainly influences the consumer involvement level. This variable has a big and significant effect in this model. As can be seen, the B for this variable is very close to the R-square of the total model, which means that the personal relevancy of and the consumer involvement level for Slankie are closely aligned with each other, as is mentioned in paragraph 2.3.3.1. The effect of the perceived risk of purchase is highly insignificant and should therefore not be included in the regression line. In conclusion it can be said that hypothesis 8 can be accepted, while hypothesis 9 should be rejected.

The corresponding regression line of this multiple regression analysis can be written as follows:

CIL = 0,213 + 0,953(PR)

The last thing that should be decided is whether effect of the personal relevancy on the brand-consumer interaction level is mediated by “brand-consumer involvement level for Slankie”, which can be analysed with a regression analysis. The results are presented in table 15.

F-value 83,231

R 0,541

(Constant) 0,944 0,000

Table 15: Mediation Analysis Consumer Involvement Level Slankie

The results in table 15 show that the model is overall significant with an F-value of 83,231 (p=0,000) and that there is a significant effect of the consumer involvement level on the brand-consumer interaction level.

The corresponding regression line of this regression analysis can be written as follows:

BCI general = 0,944 + 0,471(CIL)

(32)

32

F-value 42,639

R-square 0,299

(Constant) 0,951 0,000

Table 16: Total Mediation Analysis Slankie

As can be seen in table 16, the multiple regression is significant with an F-value of 42,639 (p=0,000). Besides, the results of the mediation analysis show that the relationship between the personal relevancy of Slankie and the brand-consumer interaction level is reduced to a point where it is not statistically significant once consumer involvement level is included in the model. Therefore, full mediation is supported in this model. This result is not unexpected, since table 14 already showed that the personal relevancy for Slankie and the consumer involvement level for the brand are closely aligned.

The corresponding regression line of this regression analysis can be written as follows:

BCI general = 0,951 + 0,252 (PR) + 0,225(CIL)

In further analyses, the consumer involvement level will be the only variable that will be included, as it is not possible to include personal relevancy as well in the multiple regression. It should however be kept in mind that, through consumer involvement, personal relevancy has an effect on the brand-consumer interaction level.

§ 4.3 Effect of Consumer Attitude and Consumer Involvement Level for Slankie

In the conceptual model, hypotheses 6 and 7 concern the effect of the attitude as well as the involvement level of the consumer for the Slankie brand on the general interaction level between the brand and the consumer. Besides, hypothesis 10 assumes that the involvement level of consumers for social media has an effect on the general interaction level with the brand as well. These hypotheses can be tested with a multiple regression analysis, of which the results are presented in table 17.

F-value 27,614

(33)

33

(Constant) 1,025 0,000

Consumer Attitude towards Slankie -0,035 0,620

Table 17: Multiple Regression Analysis Brand-Consumer Interaction Level Slankie

The results in table 17 show that the model is overall significant with an F-value of 27,614 (p=0,000). The effect of the three independent variables on the interaction level between the brand and the consumer is however not very big. It seems that the tendency of consumers to interact with the Slankie brand is high already and next to this, the involvement level of the consumer has a significant and big effect on the interaction level with the brand. Hypothesis 7 can therefore be accepted. The results also show that the attitude of the consumer has the opposite, however insignificant effect and that the involvement level of consumers with social media has a small and insignificant effect on the interaction level with the Slankie brand. Hypothesis 6 and 10 should therefore be rejected. The corresponding regression line can be written as follows:

BCI general = 1,025 + 0,487(CIL)

§ 4.4 Brand-Consumer Interaction Level for Slankie

In the questionnaire, some general questions have first of all been asked about the dialogue between consumers and the Slankie brand. Thereafter, the three communication expressions for Slankie have been presented, after which various questions have been answered and statements have been rated by the respondents.

With the general questions, the 203 respondents have indicated what they expect from a dialogue with the brand, as well as how often they want to interact and by which channel. These results are presented in figure 2, 3 and 4.

(34)

34

Figure 2 shows that consumers mainly want to receive information about Slankie, whether it is about the products the brand offers or the campaigns that are executed.

In figure 3 below it can be seen that most of the respondents (62%) do not want to have a dialogue with Slankie at all. If they do want to have a dialogue, it would be with a regularity of less than 1 to 2 times per six months.

Figure 3: Regularity of Slankie-Consumer Dialogue

The consumers who do want to have a dialogue with the Slankie brand indicate that they would like to establish this dialogue mainly with the usage of the website of the brand, as figure 4 shows. It is interesting to note that “other” is also often chosen by the respondents and that a dialogue with the usage of e-mail is the most often mentioned as the best way to establish this.

Social Media and Cheese Business Case for the Slankie Brand