Local Celebrity Endorsement337633/FULLTEXT01.pdf · celebrity endorsement but lack sufficient financial capital, which conducts the purpose. Further, the research questions are discussed

J Ö N K Ö P I N G I N T E R N A T I O N A L B U S I N E S S S C H O O L JÖNKÖPING UNIVERSITY

Local Celebrity Endorsement - Can You Go Far by Staying Close?

Bachelor Thesis within Business Administration

Author: SARA EKBERG 860303-1629 LINN MELLGÅRD 860120-7429 MAGDALENA MICKO 870323-7480 Tutor: HAMID JAFARI Jönköping JUNE 2010

i

Acknowledgements The authors of this thesis would like to acknowledge the following people, who have been indispensable to us during the research process: First of all, we would like to thank our awesome tutor Hamid Jafari who has guided and supported us in times of good and bad. Without him, this thesis would not have been possible. We would like to thank Erik Hunter for taking time with us and providing us with inspiration. We would also like to thank Johan Larsson for guiding us through the SPSS jungle. Finally, we would like to thank the students who participated in our survey and our fellow students for their moral support and constructive feedback.

Jönköping International Business School June, 2010

Sara Ekberg Linn Mellgård Magdalena Micko

ii

Bachelor Thesis within Business Administration Title: Local Celebrity Endorsement

- Can You Go Far by Staying Close?

Authors: Sara Ekberg Linn Mellgård Magdalena Micko

Tutor: Hamid Jafari

Date: Jönköping, June 2010

Keywords: Celebrity Endorsement, Local Celebrities, Well-known Cele-brities, Celebrity Capital, Reputational Capital, Marketing Communication, Communication Effectiveness

Abstract Background: Celebrity endorsement consists of a written or spoken

statement from a publicly known person, proclaiming the benefits of some product or service. Previous research on celebrity endorsement has proved it very successful in pro-moting brands or companies. Even though many marketing strategies exist, it can be especially effective for newly started ventures to apply celebrity endorsement in their approach. While celebrity endorsement may be a good way to over-come weaknesses, such as liability of newness and lack of le-gitimacy, new ventures often cannot afford to implement this strategy. Therefore, an option to this might be local cele-brity endorsement.

Purpose: This thesis investigates the impact of local and well-known

celebrity endorsement on communication effectiveness.

Method: The effects of using local celebrity endorsement has been

investigated by combining theories drawn from previous re-search with new insights gained from performing quantita-tive surveys on a sample of 240 participants.

Conclusion: Local celebrity endorsers are perceived more trustworthy and emotionally involved in the endorsement process than well-known celebrity endorsers. Local celebrities and well-known celebrities are perceived equally expert, attractive and capable of transferring meaning to the endorsed product. Local celebrity endorsement is overall more effective than well-known celebrity endorsement in communicating the endorsement message. It can therefore be a suitable tool for newly started ventures that cannot afford to employ more expensive, well-known celebrity endorsement in their quest for gaining quick reputational capital.

iii

Table of Contents

1 Introducing Celebrity Endorsement ......................................... 1

1.1 Background to Study: Celebrity Endorsement, A Good but Expensive Idea ......................................................................................... 1

1.2 Problem Discussion ................................................................................ 2

1.3 Purpose: Investigating Communication Effectiveness of Local and Well-known Celebrity Endorsement ................................................... 3

1.4 Research Questions ............................................................................... 3 1.5 Delimitations ........................................................................................... 3 1.6 Definitions of Concepts ........................................................................... 3

1.7 Thesis Outline ........................................................................................ 6

2 Theoretical Framework ............................................................. 7

2.1 Choice of Theory .................................................................................... 7 2.1.1 The Importance of Reputational Capital ................................................. 8 2.1.1.1 Reputational Capital for Newly Started Ventures ...........................................................8 2.1.1.1.1 Liability of Newness ........................................................................................................8 2.1.1.1.2 Lack of Legitimacy ..........................................................................................................9 2.1.1.1.3 Lack of Financial Capital .................................................................................................9 2.1.2 Celebrity Endorsement in Depth ........................................................... 10 2.1.2.1 Risks Associated with Celebrity Endorsement ............................................................ 11 2.1.2.1.1 Controversial Celebrity Endorsers ............................................................................... 12 2.1.2.2 The Hefty Price Tag of Celebrity Endorsement ........................................................... 12 2.1.2.3 Celebrity versus Non-Celebrity Endorsers .................................................................. 13 2.2 Choosing the Right Celebrity ................................................................ 13 2.2.1 The Source Models .............................................................................. 13 2.2.1.1 Source Credibility Model .............................................................................................. 14 2.2.1.1.1 Trustworthiness ............................................................................................................ 14 2.2.1.1.2 Expertise ...................................................................................................................... 14 2.2.1.2 Source Attractiveness Model ....................................................................................... 14 2.2.2 Comments on the Source Models ........................................................ 14

2.2.3 The Meaning Transfer Model: Cultural Meaning and the Celebrity Endorser ................................................................................................. 15

2.2.4 Emotional Involvement ......................................................................... 16 2.2.5 Summary of Models .............................................................................. 17

3 The Local Endorser: Why it should Work.............................. 18

3.1.1 Celebrities with Benefits ....................................................................... 18 3.1.1.1 Trustworthiness ............................................................................................................ 19 3.1.1.2 Expertise ...................................................................................................................... 19 3.1.1.3 Attractiveness .............................................................................................................. 19 3.1.1.4 Emotional Involvement ................................................................................................ 20 3.1.1.5 Meaning Transfer ......................................................................................................... 20 3.2 Communication Effectiveness of Local and Well-known Celebrity

Endorsement .......................................................................................... 20

4 Methodology ............................................................................ 22

4.1 Inductive or Deductive Approach .......................................................... 22 4.1.1 Chosen Method: Deductive Approach .................................................. 22 4.2 Qualitative versus Quantitative Research Approach ............................ 23 4.2.1 Intensive and Extensive Data Collection .............................................. 23 4.2.2 Choice of Method: Quantitative Approach ............................................ 23

iv

4.3 Research Strategy ................................................................................ 23

4.3.1 Survey Design ...................................................................................... 24 4.3.1.1 Independent and Dependent Variables ....................................................................... 24 4.3.2 Choosing the Right Celebrity ................................................................ 24 4.3.2.1 Matching the Celebrities with Suitable Products ......................................................... 26 4.3.3 Designing Survey Questions ................................................................ 26 4.3.3.1 Chosen Survey Questions to Measure Communication Effectiveness ....................... 26 4.3.3.1.1 Pilot Study .................................................................................................................... 28 4.3.3.2 Likert Scale, Also Called the Summative Scale ........................................................... 28 4.4 Survey Procedure ................................................................................. 28

4.4.1 Background .......................................................................................... 28 4.4.2 Celebrity Advertisements ...................................................................... 29 4.4.3 Participants ........................................................................................... 29 4.5 Validity and Reliability ........................................................................... 29

4.5.1 Increasing Internal Validity ................................................................... 30 4.5.1.1 Reliability Tests ............................................................................................................ 30 4.5.2 Increasing External Validity .................................................................. 30 4.5.2.1 Multiple Experiments.................................................................................................... 30

5 Findings and Analysis ............................................................ 32

5.1 Data Analysis ....................................................................................... 32 5.2 Overview of Findings ............................................................................ 33

5.2.1 Comparison of Mean Values ................................................................ 34 5.2.2 Hypothesis 1: Credibility ....................................................................... 35 5.2.3 Hypothesis 2: Attractiveness ................................................................ 36

5.2.4 Hypothesis 3: Emotional Involvement................................................... 37 5.2.5 Hypothesis 4: Meaning Transfer ........................................................... 38

5.2.6 Hypothesis 5: Communication Effectiveness ........................................ 39 5.3 Analysis in Accordance with Theoretical Framework............................ 42

5.3.1 Credibility .............................................................................................. 42 5.3.2 Attractiveness ....................................................................................... 43

5.3.3 Emotional Involvement ......................................................................... 43 5.3.4 Meaning Transfer ................................................................................. 43 5.3.5 Communication Effectiveness .............................................................. 44

5.4 Analysis of Celebrity Endorsement Today ............................................ 45

6 Conclusion .............................................................................. 46

7 Limitations, Suggestions for Future Research and Final Thoughts ............................................................................. 47

7.1 Limitations ............................................................................................ 47 7.2 Suggestions for Future Research ......................................................... 47 7.3 Final Thoughts ...................................................................................... 48

References ................................................................................... 49

Appendix 1: Source Credibility Scale (Ohanian, 1990) ............. 54

Appendix 2: Survey Introduction and Questionnaire (Swedish) ...................................................................................... 55

Appendix 3: Survey Introduction and Questionnaire (English) ....................................................................................... 59

v

Appendix 4: Advertisement Leila Lindholm (Including English example) ......................................................................... 63

Appendix 5: Advertisement Rachael Ray ................................... 65

Appendix 6: Advertisement Niclas Adler ................................... 67

Appendix 7: Advertisement Jan Björklund ................................ 69

Appendix 8: Advertisement Daniel Karlsson ............................. 71

Appendix 9: Advertisement James Blunt ................................... 73

Appendix 10: Advertisement Nikita Johnson (Local) ................ 75

Appendix 11: Advertisement Nikita Johnson (Well-known) .......................................................................................... 77

vi

List of Tables Table 1.1 -Disposition of Thesis 6

Table 2.1 -Summary of the Theoretical Models Covered 17

Table 4.1 -The Selected Celebrities for the Study 25

Table 4.2 -How Research Questions are Measured in the Survey 27

Table 5.1 -Mean Values for Local and Well-known Celebrity Endorsers 34

Table 5.2 -Statistical Results for Trustworthiness 36

Table 5.3 -Statistical Results for Expertise 36

Table 5.4 -Statistical Results for Attractiveness 37

Table 5.5 -Statistical Results for Emotional Involvement 38

Table 5.6 -Statistical Results for Celebrity-Product Association 39

Table 5.7 - Statistical Results for Image Transfer 39

Table 5.8 - Statistical Results for Communication Effectiveness 40

Table 5.9 - Statistical Results for Purchase Intentions 41

Table 5.10 -Statistical Results for Attitude towards the Advertisement 41

List of Figures

Figure 2.1 -Published articles from Google Scholar ranging between 1972 and 2009 7

Figure 2.2 -Meaning movement and the endorsement process 15

Figure 2.3 -The latent emotional involvement dimension and corresponding items 16

Figure 3.1 -Proposed model to test communication effectiveness 21

Figure 5.1 -Diagram showing the survey participants‟ age distribution 33

Figure 5.2 -The participants‟ gender distribution in the local celebrity group 34

Figure 5.3 -The participants‟ gender distribution for well-known celebrity group 34

Figure 5.4 -Diagram presenting the mean values of communication effectiveness 40

List of Pictures Picture 2.1 -Johnny Cash 10 Picture 2.2 -Tiger Woods 12

Local Celebrity Endorsement Ekberg, Mellgård and Micko, 2010

1

1 Introducing Celebrity Endorsement

This chapter introduces the background of this study and guides the reader into the subject of celebrity en-dorsement. The problem section discusses the dilemma posed on newly started ventures who wish to employ celebrity endorsement but lack sufficient financial capital, which conducts the purpose. Further, the research questions are discussed in depth, and the chapter concludes with some words on delimitations and the thesis outline.

1.1 Background to Study: Celebrity Endorsement, A Good but Expensive Idea

New venture creation can today almost be seen as mesmerizing as rushing for gold was 1848-1855. The thought of becoming rich over night is capturing the masses, and striving to become an entrepreneur is probably more popular today than ever. Once a new venture is started however, it generally faces some difficulties that will not be overcome overnight. Stinchcombe (1965) introduces the argument that new organizations suffer from liability of newness, which means the higher propensity of younger organizations to die. This is com-posed of their dependence on the cooperation with strangers, low levels of legitimacy and inability to effectively compete with established organizations. Delmar and Shane (2004) further develop the concept of new ventures‟ lack of legitimacy. They propose that firms can reduce the hazard of organizational death and facilitate the transition to other organiz-ing activities by undertaking activities to generate legitimacy.

In order to survive, new ventures need to try to overcome the weaknesses of liability of newness and lack of legitimacy (discussed in section 2.1) as soon as possible. While many strategies to overcome these exist, one way to gain more attention, visibility and generate legitimacy is to use celebrity endorsement. The power of celebrity endorsement is captured in the following quote:

“Your client, whether they are an athlete or an actor or an actress, has intangible assets: a name, a reputa-tion, a credibility and an image. All of those attributes may be combined into something that could be made into a brand.” (Brian Dubin, quoted in Towle, November 18, 2003)

The celebrity endorser, as defined by McCracken (1989, p.310), is “any individual who en-joys public recognition and uses this recognition on behalf of a consumer good by appear-ing with it in an advertisement.” The use of celebrity endorsement in marketing communi-cation activities has risen remarkably. Solomon (2009) states that 20 percent of American advertisements feature celebrities. Even though celebrity endorsement in some cases can have negative consequences for the firm (Andersen, 1983; Marchand, 1985; Mehta, 1994; Ainsworth, 2007), most researchers agree that celebrity endorsement can, to differing de-grees, benefit the firm by increasing communication effectiveness (Kamen, Azhari, & Kragh, 1975; Friedman & Friedman, 1979; Atkin & Block, 1983; McCracken, 1989; Silvera & Austad, 2004; Forehand & Perkins, 2005). McCracken (1989), states that the attractive and likeable qualities of the endorsers are transferred to the products being endorsed. Cele-brities also create and maintain attention and achieve high brand name and advertisement recall rates (Friedman & Friedman 1979). Given that there is a good match between the endorser, product, and target audience, celebrity endorsers are more effective than non-celebrity endorsers in generating attitudes towards advertising and the endorsed brand, in-creasing purchase intentions, and increasing sales (Erdogan 1999).


2

However, a major problem for new venture start-ups that has been ignored in this context is their lack of financial capital. Naturally, a lack of financial capital translates into problems of financing large celebrity endorsements. The price of endorsement contracts are not al-ways disclosed, but usually average between $200,000 and $500,000 (Johnson, 2005) and can reach up to $30 million (Talmazan, 2009). Even if large celebrity endorsements simply cannot be afforded, smaller celebrity endorsements may be. With this in mind, it can be of interest to investigate whether or not newly started ventures, with lack of financial capital, can benefit from investing in smaller celebrity endorsement. As it is argued later on, this thesis proposes that local celebrity endorsement can be at least as effective as well-known celebrity endorsement in overcoming liability of newness and lack of legitimacy. In this context, local celebrity endorsement can be generated by using celebrities famous in a smaller geographical region, a certain market, industry, community or other niche.

1.2 Problem Discussion

The field of celebrity endorsement already enjoys academic attention. However, a subfield has been detected which is yet to be researched. New ventures often suffer from liability of newness and lack of legitimacy (Stinchcombe, 1965; Delmar & Shane, 2004). Therefore, they find themselves in the need to create fast reputational capital (defined in 1.6). New ventures urgently need to create attention, interest, respect and demand. This is important today, when the environment of competing for attention is more severe than ever. For newly started ventures, this environment is even tougher due to the lack of adequate mone-tary capital (Schaefer, 2006). Therefore, it could be of interest to investigate whether or not the wanted attention can be gained effectively, without large sums of money.

One way for businesses to create reputational capital is through the use of celebrity en-dorsement (Hunter, Burger & Davidsson, 2008). The aim is to transfer some of the celebri-ty‟s image and brand to the business (McCracken, 1989). However, the problem is since newly started ventures often lack sufficient financial capital, they cannot afford to use ex-pensive celebrity endorsement. This thesis tries to solve this problem by investigating if lo-cal, and therefore cheaper, celebrity endorsement may be as effective as well-known celebr-ity endorsement. Until now, research has focused on other themes. There are extensive studies covering whether or not celebrity endorsement is beneficial (Agrawal & Kamakura, 1995) and cases where celebrity endorsement can have negative effects on the brand (Shimp & Till, 1998; Farrell, Karels, Montfort, & McClatchey, 2000). Other studies focus on what determines the communication effectiveness of the endorsement (Hovland & Weiss, 1951; Hovland, Janis & Kelly, 1953; Kahle & Homer, 1985; McGuire, 1985; McCracken, 1989). However, no study so far focuses on the theme of this thesis. There-fore, the study is conducted by combining general theory of celebrity endorsement with the performance of a study consisting of widely distributed surveys with manipulated variables.

This study is significant since the specific topic has not been researched previously, and it may be of great value for new ventures that find themselves in the described situation. If it is found that local celebrity endorsement is as effective, this can open up many opportuni-ties for new ventures, as well as other financially constrained firms, to explore.


3

1.3 Purpose: Investigating Communication Effectiveness of Local and Well-known Celebrity Endorsement

This thesis aims to investigate whether a local, and therefore a lot less expensive celebrity, might be an as effective endorser as a well-known celebrity, given the right contextual set-ting.

By researching the effect which local and well-known celebrity endorsement can have on the communication effectiveness of newly started ventures, the authors wish to make an important theoretical contribution.

The key research purpose of this thesis is to:

Investigate the impact of local and well-known celebrity endorsement on communication effectiveness.

1.4 Research Questions

This thesis investigates whether or not local celebrity endorsement can be as effective as well-known celebrity endorsement. In order to fulfill this, the following research questions are developed:

1. Can local celebrity endorsers be viewed at least as credible as well-known celebrity endorsers? 2. Are local celebrity endorsers perceived equally attractive as well-known celebrity endorsers? 3. Can local celebrities be perceived at least as emotionally involved in the endorsement process as well-

known celebrities? 4. Can local celebrities be as capable of transferring meaning to the endorsed product as well-known

celebrities? The research questions are further discussed and developed along with related literature and formed hypotheses in sections 2.3 and 2.4.

1.5 Delimitations

The focus of this thesis is the effectiveness of local celebrity endorsement, used by newly started ventures that target local markets. Due to time constraints, the authors were re-stricted in their choice of method. More specifically, the survey is only made available in eight versions. This number could have been higher if more time had been available. Also due to time constraints, the base of the study is restricted to Jönköping, Sweden and the participants are all located within the Jönköping area. Additionally, because the authors had financial limitations, the advertisements presented in the surveys are made by the authors, not a graphic designer.

1.6 Definitions of Concepts

This part presents the frequently used concepts according to occurrence, to facilitate com-prehension of the thesis.


4

Endorser: There are different types of endorsers; celebrities, experts, and typical consumers (Fried-man & Friedman, 1979). The celebrity endorser is discussed in the definition below. The expert is an individual or a group that has acquired superior knowledge concerning the en-dorsed product class. This expertise is gained through education, training or experience (Friedman & Friedman, 1979). Friedman and Friedman (1979) argue that the typical con-sumer, who is an average person, also can act as an endorser. This person is assumed to represent an individual in everyday life, with no previous experience of the product class. An endorser can also be called a spokesperson, human brand, et cetera. Basically, it is someone who expresses their support toward a brand or a product. Celebrity Endorser:

A celebrity is in general someone who is widely known for something they have achieved. The celebrity can, for example, be a sports figure, actor or entertainer (Friedman & Fried-man, 1979). It can be anyone who is recognized by the public (McCracken, 1989). As mentioned above, a celebrity is someone who is acknowledged by the public. This fa-mous individual will, as an endorser, use their recognition and transferring it to a product or a service through advertisements (McCracken, 1989). By using an endorser, companies hope to use the endorser‟s personality and personal values to gain attention from the public (Knott & James, 2004).

Local Celebrity:

The definition of the word local is according to Merriam-Webster Online Dictionary “primarily serving the needs of a particular limited district”. The local celebrity can be fa-mous within a society, niche, industry or market. Examples of local celebrities can be news reporters, politicians, religious leaders, popular bloggers or famous radio DJ:s. It can also be a chef at a well-known restaurant, an anchorman or a social figure. An example of a local celebrity is the locally known singer from Uddevalla, Sweden; Lasse Matilla. A more well-known celebrity in this example is the Swedish singer Timbuktu, who is widely known all over Sweden. Another example of a local and a well-known celebrity is Tina Nordström and Martha Stewart. The first mentioned is a chef who is famous in the Swedish cooking world. Martha Stewart is famous in the same niche, but in an international perspective.

An endorser is someone who “…recommend[s] (as a product or service) usually for financial compensation” (Merriam-Webster Online Dictionary).

One definition of a celebrity is someone who is in “the state of being well known” (The Oxford Pocket Dictionary of Current English).

A local celebrity is in this thesis defined as someone who is famous within a region, city or culture.


5

Well-Known Celebrity: It can, for example, be an internationally known movie star, singer, athletic star or enter-tainer. This type of celebrity often has a large influence on the public. A well-known celeb-rity receives extensive attention in the media and can be known globally. Some parts of the public can even be obsessed with the internationally well-known celebrities (Maltby, Day, McCutcheon, Martin & Cayanus, 2004). Reputational Capital:

Reputational capital is the perceptions and associations made by the consumers about the company (Suh & Amine, 2007). Reputational capital can also be known as „corporate per-sonality‟ and „corporate reputation‟ (Suh & Amine, 2007). By having a good reputation, the firm stabilises customer demand for its services or products (Jackson, 2004). Celebrity Capital:

Celebrity capital is also known as celebrity equity (Louie, Kulik & Jacobson, 2001). Com-panies hope to transfer the celebrity‟s favourable awareness and associations to the brand or product. Since newly started companies often lack reputational capital (Stuart, Hoang & Hybels, 1999; Delmar & Shane, 2004) they can use the reputation of others, such as celeb-rities, to improve their reputation (Rao, 1994; Yiu & Lau, 2008). When using celebrity capi-tal a company desires to increase its market position by strengthening its brand. The inten-tion of using celebrity capital is to create positive reactions and attitudes among the target audience. It is also used to increase attention and transfer the celebrity‟s image to the prod-

uct and the character of the good (O Mahony & Meenaghan, 1998). Communication Effectiveness:

The desired communication outcome is often to prompt action, in this case increased pur-chase intentions (Drogan, 2006). In this thesis, communication effectiveness refers to how successful an endorser is in communicating the message of the advertisement and brand to the public. If the message is successfully conveyed, it should translate into positive attitudes toward the brand, and increased purchase intentions. Communication effectiveness can therefore, from the company‟s side, most easily be measured by increased purchases. This could be done, for example, by comparing revenues before and after an advertisement campaign.

A well-known celebrity is defined as someone who is famous worldwide or nation-wide.

Communication effectiveness can be described as the degree to which the desired communication outcome is achieved (Drogan, 2006).

Jackson (2004) defines reputational capital as an intangible asset that is intended to generate profit in the long run.

Celebrity capital consists of a celebrity‟s past behaviour, reputation, public awareness, their favourability and personality (Hunter et al., 2008).


6

1.7 Thesis Outline

Table 1.1 Disposition of the Thesis

Chapter 1- Introducing Celebrity Endorsement

Introduction of the background and a pres-entation of the celebrity endorsement sub-ject. The problem section discusses the di-lemma posed on newly started ventures who wish to employ celebrity endorsement but lack sufficient financial capital, which con-ducts the purpose. Further, the research questions are discussed in depth, and the chapter concludes with some words on de-limitations and thesis outline.

Chapter 2- Theoretical Framework

Presentation of the theoretical framework, from which the empirical findings are ana-lyzed. The section aims to describe the dif-ferent models available to predict communi-cation effectiveness of celebrity endorse-ment. Furthermore, a summary of the mod-els used has been conducted in order to provide a good overview.

Chapter 3- The Local Endorser: Why it should Work

Presentation of the authors own proposi-tions and formed hypotheses.

Chapter 4- Methodology

Presentation of the research process of this thesis. The method chosen for the research is elaborated and a motivation of how the survey is structured and why, is presented. In extent, the participants of the study are presented and the validity of the survey is discussed.

Chapter 5- Findings and Analysis Presentation of the empirical findings and analysis, in accordance to the theoretical framework.

Chapter 6- Conclusion Conclusion of the output from the analysis of the research questions and fulfillment of the purpose of the thesis.

Chapter 7- Limitations, Suggestions for Future Research, and Final Thoughts

Presentation of the critique of the study and method, as well as further research sugges-tions by the authors. To finish the chapter, additional insights that were learned during the process of writing the thesis, but not di-rectly connected to the purpose of this re-search, are presented.


7

2 Theoretical Framework

In this section the theoretical framework is presented, from which the empirical findings are analyzed. The section aims to describe the different models available to predict communication effectiveness of celebrity en-dorsement. Furthermore, a summary of the models used has been conducted in order to provide a good over-view. Celebrity endorsement, as it is known today, was first introduced in the early 1970s. Since then, it has been applied as well as researched extensively. This graph shows that the inter-est of celebrity endorsement has increased and gained more attention during the last ten years.

Figure 2.1 Published articles from Google Scholar ranging between 1972 and 2009, collected by the authors using the keyword „celebrity endorsement‟

Due to its popularity, the research on celebrity endorsement has focused on numerous amounts of different themes which are brought up later in the chapter. The following sec-tions guide the reader through the general topics of celebrity endorsement, describe the dif-ferent models applied and introduce the concept of local celebrity endorsement; which has not been academically researched and tested before.

2.1 Choice of Theory

There are many theories employed in celebrity endorsement research. The theories used in this thesis are chosen since they are well known and often used in the area of celebrity en-dorsement. There are some similar theories that overlap each other, such as the match-up hypothesis and the meaning transfer model. The match-up hypothesis is used in this thesis to complement the statements of the meaning transfer model and employed in the process of matching the celebrities with the respective products for the survey.

The following sections present the theory chosen to illustrate the problem statement of this thesis, namely that newly started ventures who wish to use celebrity endorsement as a way to gain reputational capital, most often cannot afford it.

0

10

20

30

40

19

72

19

75

19

78

19

81

19

84

19

87

19

90

19

93

19

96

19

99

20

02

20

05

20

08

The History of Celebrity Endorsement Research Between 1972 and 2009

No. of articles


8

2.1.1 The Importance of Reputational Capital

It is crucial for companies to be well known and to have a good reputation in order to be able to compete with other firms. According to Jackson (2004), reputation is everything in today‟s society. Often, before deciding which company to invest in or from what company to buy products, presumptive buyers tend to gather information about the competing firms. However, sometimes they find it difficult to gather enough information. In these cases consumers usually compare the different companies‟ reputations to make their deci-sion process easier. Jackson (2004) defines reputational capital as an intangible asset that is intended to generate profit in the long run. Petrick, Scherer, Brodzinska, Quinn and Ainina (1999) argue that it is time consuming to build a reputation, especially for newly started companies.

Even though the process of building a reputation is time consuming, a firm‟s reputation is very fragile. Damaging the reputation can be easily and quickly done (Petrick et al., 1999). An example of such an incidence is when the car company Toyota used the celebrity Mi-chael Irvin from Dallas Cowboys as their endorser. In 1996, Irving was caught by the po-lice with prostitutes and drugs. This action affected Toyota remarkably, since the company had planned to use Irving in a series of advertisements worth $500,000. Only thirteen Toyota dealerships survived this event. Toyota did not only lose trust from the public, the company also lost the money it had invested in the Dallas Cowboys star and had to find a new alternative for Irving, which meant even more expenses (Lane, 1996).

Reputational capital measures the trustworthiness, reliability, credibility, responsibility and accountability of a company (Petrick et al., 1999). An effective way to gain reliability is to guarantee service, innovation and quality. Reputational capital is an important factor, and when used properly it can help companies in the process of obtaining great benefits and competitive advantage (Petrick et al., 1999; Jackson, 2004). Other companies will consider firms with a positive reputational capital that stress its fair play and honesty, as a potential strategic partner. Furthermore, Jackson (2004) states that good reputation often results in customer loyalty. By having a good reputation a firm stabilises customer demand for the firm‟s services or products.

2.1.1.1 Reputational Capital for Newly Started Ventures

For newly started firms it is important to build a good reputation fast (Székely & Sabot, 1997). A company cannot start from scratch and expect to be able to sell its stock or bonds to the public right away. The company needs to gain customer trust and prove itself trust-worthy and reliable. This can take years to obtain and is not easily done (Székely & Sabot, 1997).

2.1.1.1.1 Liability of Newness

Newly started companies suffer a higher risk of failure than more established and skilled firms. New businesses will most likely lack the expertise that these firms posses (Stinch-combe, 1965). Moreover, established firms usually have complementary assets organised, that newly started organisations do not have (Teece, 1986). As mentioned previously, a firm‟s reputational capital is often a factor that determines if a firm is seen as a potential business partner. This is disadvantageous for newly started firms since they initially lack re-putational capital (Delmar & Shane, 2004; Stuart et al., 1999). Newly started ventures do not have any existing relationships with suppliers or customers, which makes it difficult for them to compete with established firms (Stinchcombe, 1965). Since newly started ventures are not seen as liable as established firms, it is essential for them to build an awareness of


9

being legitimate. If they do not, it is difficult to endure the competition from established companies (Delmar & Shane, 2004). Newly started ventures should quickly develop rou-tines within the organization to be able to compete with other firms (Delmar & Shane, 2004).

2.1.1.1.2 Lack of Legitimacy

Stinchcombe (1965) argues that legitimacy is a social judgment of appropriateness, desira-bility and acceptance. Stinchcombe also states that legitimacy makes it possible for compa-nies to gain access to other resources. Legitimacy can be used in a strategic way to help companies increase resources and growth (Zimmerman & Zeitz, 2002). Gaining legitimacy helps newly started firms to overcome the liability of newness (Stinchcombe, 1965). Re-searchers argue that legitimacy is one of the most important factors for new venture sur-vival. Zimmerman and Zeitz (2002), state that legitimacy is present in every company that has survived its start-up.

Legitimacy is described as the equivalence between norms, values, and expectations of so-ciety, and performance of the organization (Dowling & Pfeffer, 1975; Ashforth & Gibbs, 1990). To gain legitimacy a newly started company must prove itself desirable by engaging itself in things seen as legitimate; such as respecting rules, undertaking values and norms in its field and demonstrating competence (Shepherd, 1999). Zimmerman and Zeitz (2002) believe that fewer companies would fail if they managed to acquire, build and use legitima-cy even before the start-up. Hunter et al. (2008) claim that when using a celebrity endorser, the company might gain legitimacy through the celebrity‟s reputational capital. As celebrity endorsers are well known and often seen as trustworthy, this might be reflected in the new-ly started venture. By using celebrity capital they can be perceived as a more legitimate and reliable company by potential business partners (Hunter et al., 2008). People are usually not willing to purchase products where they lack knowledge. Even if some people do know about the products, they are skeptical towards newly started firms since they lack credibili-ty. Customers do not know whether the firm will deliver products with good quality or not (Glynn, Motion, & Brodie, 2007).

2.1.1.1.3 Lack of Financial Capital

Statistics state that only two-thirds of new ventures survive more than two years, and 44 percent stay alive for more than four years (Schaefer, 2006). Schaefer (2006) describes sev-en pitfalls for newly started firms; 1) You start your business for the wrong reasons, 2) Poor management, 3) Insufficient Capital, 4) Location, Location, Location, 5) Lack of Planning, 6) Overexpansion, and 7) No website. Number three on this list is insufficient capital, which shows the importance of funding a newly started venture.

According to Cooper and Gimeno-Gascon (1991), there is a relationship between the initial amount that is invested in the newly started firm and the performance of the firm. Out of eight former studies that investigated this relationship, six of the cases performed better when the initial capital was higher. For example, more financial capital will allow a firm to buy more assets. Cooper and Gimeno-Gascon (1991) also state that higher initial financial capital will increase the possibilities of survival, since the owner can afford to make mis-takes and have time to figure out solutions to problems that might occur. A common rea-son for the lack of financial capital is an underestimation of the money needed. This might force the end of a newly started venture. Funds to cover the first couple of years may be needed, since many newly started ventures take a year or two before the business can sur-


10

vive on its own profits (Schaefer, 2006). If the new venture wishes to invest in a marketing strategy, such as celebrity endorsement, it is essential to have sufficient financial capital since it is expensive (Talmazan, 2009).

2.1.2 Celebrity Endorsement in Depth

The phenomenon of celebrity endorsement is widely used today. As mentioned previously, Solomon (2009) states that celebrity endorsement is present in 20 percent of all American advertisements. However, what many people do not know is that celebrity endorsement has been present since the nineteenth century. At that time, the celebrities used for en-dorsements were royalties. Nevertheless, the intentions regarding the strategy of achieving recognition and increasing value to the product through customers‟ social desires, were the same as today (Almquist & Roberts, 2000).

Due to a strong development of the media and an increase in number of famous people, celebrity endorsement has grown. One major event in the beginning of the 1970s which is said to have changed advertising strategy was when Johnny Cash appeared in Amoco (also known as American) Oil‟s commercials. This was the first time a person from the enter-tainment industry, who was known nationwide, endorsed a company. Amoco Oil knew be-fore it chose Johnny Cash as endorser that there were risks, since motorists already op-posed Cash no matter of his prior popularity and familiarity. Nevertheless, Amoco Oil de-cided that the benefits from using Cash were more than the negative attention he con-veyed. As Amoco Oil expected, soon after Cash‟s first commercials the company received hundreds of letters, which were mainly negative. However, the marketing management de-partment believed in their choice and was confident that if only the motorists got familiar with Cash, their negativity would decline (Kamen et al, 1975).

1

Since its beginning, when used correctly celebrity endorsement is proven to be advanta-geous (Friedman &Friedman, 1979; McCracken, 1989; Silvera & Austad, 2004; Forehand & Perkins, 2005). An important aspect of this marketing strategy that must be considered wisely is the selection of endorser. Friedman and Friedman (1979) performed a study where three endorser types were compared; celebrity, expert, and typical consumer. They surveyed the different endorsers with different products to design the most effective match up. The conclusion made by the study was that celebrities work best, with some exceptions.

1 Picture: Johnny Cash, retrieved 2010-0318, from http://zishan.com/Amoco/

Picture 1 Johnny Cash1


11

For instance, if the product is complex and high in financial, performance or physical risk, it is better to choose an expert endorser.

Another strategy that can be used by companies when deciding on whom to endorse its products is the match-up hypothesis (Kamins, 1990; Misra & Beatty, 1990; Till & Busler, 1998). The match-up hypothesis concerns the fit between the endorser‟s image and the product or brand. The hypothesis suggests if the fit between the endorser and product is successful, the endorsement will be more effective. A fit between these two also implies that the celebrity endorser‟s image matches the product‟s characteristics (Kamins, 1990; Misra & Beatty, 1990, Till & Busler, 1998). Many studies claim that different attributes of the celebrity will affect the outcome of an endorsement. Kamins (1990) performed a study in relation to the physical attractiveness of the celebrity endorser to establish if an attractive endorser is more effective than an unattractive endorser. Kamins (1990) concluded that physically attractive celebrity endorsers are more effective in a particular product group (especially vanity related products). For other types of products there is no difference of influence between the physically attractive and unattractive celebrity endorser.

Celebrity endorsement is effective, especially if the intention is to increase awareness (Alm-quist & Roberts 2000). Nevertheless, the effectiveness may vary due to positive and nega-tive variables that must be considered as trade-offs when choosing the strategy (Kamen et al., 1975). One positive aspect of celebrity endorsement is that positive feelings towards the celebrity endorser may be transferred to the endorsed product. Another positive effect is the possibility of free advertisement through the celebrity in question and their spotlight, when they appear in other contexts (Silvera & Austad, 2004). When reviewing prior re-search it is safe to say that celebrity endorsement is a successful concept when used proper-ly and when the endorser is matched-up with a product well-suited for them.

2.1.2.1 Risks Associated with Celebrity Endorsement When a company uses celebrity endorsement, there are certain risks that should be consi-dered. These risks are described in this section.

When a celebrity endorses a product or brand, a link between the two is formed that relates the image of celebrity on to the product or brand and vice versa (Andersen, 1983). If nega-tive information is revealed about the endorser it might reflect negatively on to the brand the celebrity is endorsing and transfer the negativity from the celebrity toward the brand (Andersen, 1983).

If the celebrity „over-shines‟ the endorsed product, the targeted customers will only recall the spokesperson, not the product or brand itself. This is called the vampire effect where the “celebrities suck the life-blood out of the product dry” (Evans 1988, as cited in Erdogan, 1999, p. 303). A different risk connected to the just mentioned problem is if a celebrity en-dorses multiple products (Marchand, 1985). This may confuse audiences, who might have trouble connecting the celebrity with a particular brand, hence ruining the purpose of cele-brity endorsement. Another risk that is impossible to control is controversial behavior by a spokesperson that might damage the image of the endorsed brand (Marchand, 1985).

There are recent studies regarding the effects arising when negative information about a ce-lebrity endorser is revealed (Ainsworth, 2007). Ainsworth (2007) tested to which extent young customers are affected by negative information concerning a celebrity endorser. The study suggests that companies choosing celebrity endorsement need to worry more about the strength of association between a product or brand and the celebrity, rather than the negative information that can be revealed about the celebrity. Ainsworth (2007) continues


12

to argue that customers may not be turned off from a product or brand simply if a celebrity endorser is involved in a controversial incident. However, the nature of the controversy will have an impact on the consumer base; rape and murder accusations are such occur-rences that definitely would turn customers away from a product or brand.

2.1.2.1.1 Controversial Celebrity Endorsers

There are few possibilities for companies to control the behavior of their endorsers, ac-cording to Marchand (1985), and there are many examples of celebrity endorsements that have gone terribly wrong. One endorsement incident that definitely could not be controlled was when James Garner, who was a spokesperson for beef, had a heart attack resulting in a triple bypass (Trout, 2009). This was probably not the message that the beef company wanted to portray. Other endorsement deals that went down the drain involved Kobe Bryant. Bryant was endorsing McDonalds, Sprite, and Nutella when he was charged with sexual assault (Trout, 2009). Once again, probably not the message these companies wanted to portray.

Celebrity endorsement can also fail since there is no assurance that the spokesperson stays faithful to the company. One example of this is when the former tennis star Martina Hingis promoted an Italian sneaker company that also made tennis-gear. This endorsement was working smoothly until Martina Hingis sued the Italian company claiming that the gear it provided caused her injuries (Trout, 2009). A current major endorser who has made a lot of money from endorsement contracts is Tiger Woods (Farrell & Van Riper, 2008). The massive endorser and golf-professional, Tiger Woods, was recently involved in a sex-scandal where it was revealed that he had cheated on his Swedish wife, Elin Nordegren. When the scandal was exposed there were discussions on how his troublesome personal life would affect his lucrative endorsement deals (Gregory, 2009; Talmazan, 2009). The risks the companies face from staying with Woods could escalate if other scandals unwrap involving the golf champion and have dreadful consequences for the companies.

If the choice of implementing celebrity endorsement in the marketing strategy is made, Till and Shimp (1998) argue that there is no possibility for the companies to control the celebri-ty‟s future behavior. Any negative news regarding the celebrity might decrease the appeal of the celebrity, hence transfer negative feelings towards the endorsed brand. Furthermore, the risk is greater for newly started firms rather than well-known established brands (Till & Shimp, 1998).

2.1.2.2 2The Hefty Price Tag of Celebrity Endorsement

As stated throughout this thesis, celebrity endorsement is a phenome-non that is increasing. However, the average compensation paid to spokespersons is also increasing (Agrawal & Kamakura, 1995). Some celebrities, such as Tiger Woods, today earn more money as a spokes-person than in their original profession (Farrell, 2008). Even though the average profitability from the usage of celebrity endorsement is high, the original investment is costly for the company (Agrawal & Kamakura, 1995). Depending on the celebrity, the endorsement could cost millions of dollars over a period of many years. The infamous

2Picture: Tiger Woods, retrieved from http://mcsearcher.wordpress.com/2009/09/13/tiger-woods-waal-

papers/

Picture 2 Tiger Woods2


13

model Linda Evangelista made a remarkable statement regarding the payment she de-mands: "I won't get out of bed for less than $10,000 a day" (Supermodelguide.com). Another celebrity with a high price tag for taking part in commercials is Christy Turlington. She endorsed Maybelline‟s products for a mere $3 million, which only involved twelve days of work a year (Supermodelguide.com). Tiger Woods, who earns $100 million a year through endorsement contracts, certainly has a high price on his endorsement skills. To mention a few, his deals include a $30 million contract with Nike, $20 million with Accen-ture, and $15 million with Gillette (Talmazan, 2009). Studies have been conducted to test whether the large investment of celebrity endorsement is a wise decision or not, and have shown both positive and negative results. However, the average outcome is that investing in celebrity endorsement is a profitable strategy (Agrawal & Kamakura, 1995).

2.1.2.3 Celebrity versus Non-Celebrity Endorsers

Even though there has been many studies proving the effectiveness of celebrity endorse-ment, given that the match between celebrity and product is successful (Kamen et al., 1975; Friedman & Friedman, 1979; Atkin & Block, 1983; McCracken, 1989; Silvera & Austad, 2004; Forehand & Perkins, 2005), there are studies questioning the use of celebrities in ad-vertising (Mehta, 1994; Almquist & Roberts, 2000). These studies question whether a cele-brity endorser is better than a non-celebrity endorser. If a company creates a character, typ-ically with an anonymous model, to represent its products in advertisements, it can have to-tal control over the endorser. The company can freely assign preferred characteristics de-sired for the brand. If a celebrity is chosen as endorser, the match-up with the product is what will make or break the advertisement (Tom, Clark, Elmer, Grech, Masetti & Sandhar, 1992). There are researches in both directions; however, the majority supports celebrity en-dorsement rather than non-celebrity endorsement. Mehta (1994) suggests that there may be more focus on the celebrity rather than the endorsed product or brand. Atkin and Block (1983) suggest in their study that celebrity endorsement is more effective on teenagers; however, celebrities in advertisements are apprehended as more competent and trustwor-thy than non-celebrity endorsers in all age groups. The study also confirms that a more positive reaction will be present towards an advertisement starring a celebrity than an iden-tical advert with a non-celebrity. The celebrity endorser is described as strong, interesting, effective, and important. Other studies show that celebrity endorsers attract more attention to an advertisement since there is a massive stream of messages for consumers to filter (Kamen et al., 1975).

2.2 Choosing the Right Celebrity

When firms search for a celebrity to endorse their brand, their aim is for the communica-tion effectiveness to be as high as possible. Choosing the right celebrity is the first crucial step in the endorsement process, which makes it important to do so with care (Kamins, 1990; Misra & Beatty, 1990). Previous research has established various models that provide the tools to predict communication effectiveness of a possible endorsement (Hovland & Weiss, 1951; Hovland, et al., 1953; Kahle & Homer, 1985; Ohanian, 1990; Erdogan, 1999; Hunter et al., 2008; Hunter, 2009). These models can guide firms in the process of finding and choosing the right celebrity that matches their brand. This section will summarize the source models, the meaning-transfer model and the concept of emotional involvement.

2.2.1 The Source Models

„The Source Models‟ refer to two models; namely source credibility model and source at-tractiveness model. These models originate from the study of communications (Ohanian,


14

1990), but have later become common models applied in the field of celebrity endorse-ment. Both models aim to determine under which conditions the source, in this case the endorser, is persuasive and influences attitude and opinion change (Erdogan, 1999; Hunter, 2009).

2.2.1.1 Source Credibility Model

The source credibility model implies that potential consumers are more open to persuasion when the source, in this case the endorser, presents itself as credible (Hovland, et al., 1953). The source credibility model demonstrates that the communication effectiveness of a mes-sage depends on the perceived level of trustworthiness and expertise of the endorser (Hov-land & Weiss, 1951; Hovland et al., 1953). Hovland et al. (1953) assume that if communica-tors are perceived trustworthy and show expertise, they are seen credible, and therefore persuasive.

2.2.1.1.1 Trustworthiness

Trustworthiness refers to the perceived honesty, integrity and believability of the endorser. Ohanian (1990) maintains that the value of trustworthiness can be capitalized by employing endorsers who are alleged as being honest, reliable, sincere and dependable.

2.2.1.1.2 Expertise

Expertise is defined as the ability of the source to make truthful statements. Expertise con-sists of the possession of knowledge, experience and skills (Erdogan, 1999). The value of expertise can be capitalized by employing endorsers who are alleged as being experienced, knowledgeable, qualified and skilled (Ohanian, 1990).

2.2.1.2 Source Attractiveness Model

Although attitude change may often be related to physical attractiveness of a celebrity (Kahle & Homer, 1985), attractiveness in this sense does not only refer to physical appear-ance. It includes all characteristics that a consumer may view as attractive in an endorser. These may be characteristics such as intellectual skills, personality properties, lifestyle, or athletic ability (Erdogan, 1999). Research has shown that attractive endorsers are more ef-fective communicators than unattractive endorsers. The McGuire model states that the ef-fectiveness of the message depends on the familiarity, similarity and likeability of the source (McCracken, 1989). McGuire (1985) assumes that if communicators are perceived to hold these attributes, they are seen attractive by consumers, and therefore persuasive. Familiarity is here defined by McGuire (1985) as knowledge of the endorser, similarity as the seeming resemblance between the endorser and the consumer, and likeability as affection for the endorser as a result of their physical appearance and behavior.

2.2.2 Comments on the Source Models

The source models have been researched and confirmed in many studies (Hovland & Weiss, 1951; Hovland, et al., 1953; Kahle & Homer, 1985; Erdogan, 1999; Hunter et al., 2008; Hunter, 2009) and it may be safe to say that the communication effectiveness of ce-lebrity endorsement activities to some extent relies on the credibility and attractiveness of the source. However, these models have been criticized (McCracken, 1989 & Hunter, 2009) for not portraying all the elements that determine communication effectiveness. In the light of this, the alternative meaning transfer model and the concept of emotional involvement will be introduced to complement the source models.


15

2.2.3 The Meaning Transfer Model: Cultural Meaning and the Celebrity Endorser

As already established, the effectiveness of celebrity endorsement can be explained in many different ways. According to the meaning transfer model as developed by McCracken

(1989) in criticism to the source models, the effectiveness of celebrity endorsers depends on the cultural meanings which they encompass and the meanings they bring to the en-dorsement process. The model both shows how meaning is transferred from celebrity to product and from product to consumer (see figure 2.2).

Figure 2.2: Meaning movement and the endorsement process, adapted from McCracken (1989)

Celebrity endorsers differ in gender, age, status, personality and lifestyle types (McCracken, 1989). These characteristics combine to produce a myriad of different meanings available for the marketer to employ in the endorsement process. The celebrity world has all types to offer; such as the careless young partygoer Paris Hilton, the mature philanthropist Angelina Jolie, the „crazy slob‟ Jack Black and the unreliable heartbreaker Charlie Sheen. By this, McCracken (1989) makes a point that the celebrity world contains more than simply „credi-ble‟, „attractive‟ and „expert‟ individuals. McCracken states that the best endorsements stem from successful meaning transfers.

In practice, the meaning transfer process begins with the advertiser choosing what mean-ings to attribute to the product (McCracken, 1989). Basically any meanings can be decided, depending on the marketing plan. Well created advertisements succeed in conveying these meanings to the consumer. Celebrity endorsers contribute to this process by letting their own cultural meanings be ascribed to the product, which is later transferred to the con-sumer. It is important to point out that celebrities are much more successful in delivering meanings to the product, than the anonymous model that otherwise might be used to bring meaning to an advertisement (Kamen et al., 1975). Celebrities can convey their meanings with more precision and can offer their ranges of already well-known personality and life-style meanings that an anonymous model cannot. It is important that the advertisement is created to show similarities between the celebrity and product, so that these meanings can be easily absorbed and accepted by the consumer (McCracken, 1989).

To conclude the discussion of the meaning transfer model, it should be noted that the model does not disregard the source model concepts of credibility, expertise and attrac-tiveness. It merely shifts the debate and considers credibility not in terms of the manner in which celebrities communicate a message, but the manner in which celebrities communicate a meaning. This also suggests that a celebrity may be extremely credible for transferring some meanings, but may be much less credible for other meanings (McCracken, 1989).


16

Emotional Involvement

Dedicated

Thrilled

UsesLikes

Enthused

Passionate

2.2.4 Emotional Involvement

So far, it has been expressed that the communication effectiveness of a celebrity endorser depends on its trustworthiness, expertise (Hovland & Weiss, 1951; Hovland et al., 1953), attractiveness (McGuire, 1985) and meaning transferred to the product (McCracken, 1989). „Emotional Involvement‟, as suggested by Hunter (2009), is an additional concept to also be categorized as a measure of credibility. “Perceived Emotional Involvement measures in-ferences being made about the endorser‟s attitude towards the product and attitude to-wards working with the product” (Hunter, 2009, p.44). Hunter proposes that higher per-ceived emotional involvement increases the communication effectiveness of celebrity en-dorsement.

Emotional involvement refers to a celebrity endorser‟s motivation to be part of the adver-tisement. The individual terms that compromise emotional involvement are: like and use, which measure a celebrity‟s attitude towards the product, and passion; enthusiasm; dedica-tion and being thrilled which measure a celebrity‟s attitude towards working with the prod-uct (Hunter, 2009).

“Many consumers said they like seeing celebrities whom they think really use the products – especially if the ad is funny.” (The NPD Group/Celebrity Influence Study, May 2006)

Figure 2.3: The latent emotional involvement dimension and corresponding items, Hunter (2009)


17

2.2.5 Summary of Models The models described above have been chosen as the theoretical foundation to base the investigation on.

Table 2.1 Summary of the Theoretical Models Covered

Model Focus Celebrity characte-ristics

Implications

Source Credibility

Trustworthiness and expertise

Honest, reliable, sin-cere, dependable, ex-

perienced, knowledge-able, qualified and

skilled

The celebrity should be perceived as being a credible source of truthful statements

Source Attractive-ness

Attractiveness Familiar, similar and

likeable

The celebrity should be perceived attractive

by the consumers

Meaning Transfer

Compatibility

Holds compatible identi-ty,

personality and life-style

There should be com-patibility between the brand and the celebri-

ty

Emotional Involve-ment

Motivation

Perceived to like and use the product, show passion, enthusiasm,

and dedication and be-ing thrilled about the

collaboration

Make sure that the ce-lebrity is motivated to work with the product and appear in the ad-

vertisement


18

3 The Local Endorser: Why it should Work

This chapter presents the authors own propositions and hypotheses formed.

So far, the presented previous research has established that in order for a celebrity endorser to be effective in their communication, they should be perceived as trustworthy, expert and attractive. They should also be perceived as being emotionally involved in the endorsement process. Further, it is important that there is a successful meaning transfer between the ce-lebrity and product, as well as the product and consumer.

The main hypothesis of this study is that employing local celebrity endorsement can be as effective as traditional, well-known celebrity endorsement. It is important to establish this because many firms, especially new ventures, lack the sufficient financial capital needed for investing in typically expensive celebrity endorsement. Therefore, the previous research on general celebrity endorsement that serves as the theoretical framework is now applied to the idea of using local celebrity endorsement. This part is the authors own integration and acts as the first part of developing and supporting the hypothesis before testing it empiri-cally.

In this thesis, a local celebrity is defined as someone famous within a region, niche, cate-gory, market or culture. The value of local celebrities might be understated. In today‟s soci-ety, it can appear that everything is a race to the top. The ultimate aim is to appear on top lists, employ top achievers, associate with A-listers, appear in the Fortune 500, and be the overall „Number 1‟. It is easy to simply focus on the mass market, rather than niche mar-kets. Therefore, the value of local celebrities is easily overlooked. What should be kept in mind is that local celebrities often may experience the same type of attention from the pub-lic as a well-known celebrity, although in the limits of their particular region or context. For newly started ventures, this may be especially true as recent research by Pae, Samiee and Tai (2002) shows that localized advertising messages are especially effective when brand knowledge is low. This is because these types of new and unknown brands require creative and culturally-compatible advertising for each market (Pae, et al., 2002). Further, it was found that consumers perceive local advertisements to be more interesting, less irritating, and generally preferred them to foreign commercials. Local advertisements were even pre-ferred in terms of purchase intentions (Pae, et al., 2002). The bottom line is that firms should look beyond mass marketing and explore the values of niche markets with their re-spective local celebrities.

3.1.1 Celebrities with Benefits

The authors of this thesis propose that the benefits of using a local celebrity in marketing communications could be many. The most obvious, of course, is their more reasonable price tag compared to well-known celebrities. Intuitively, the lesser known the celebrity is to the general public, the less she or he will be able to charge for the endorsement contract. Some local celebrities might even see this as a good opportunity for wanted publicity, and the deal could be closed at an even lower price than expected. Also intuitively; local, less-known celebrities are more likely to be easier to reach out to and employ. They probably have fewer items and commitments on their agenda and not as many people and opportun-ities fighting for their attention.

Perhaps most importantly, the local celebrity can be easier to relate to. Local celebrities are more likely to be seen in their everyday life and therefore be viewed as more real. A local celebrity is someone who is seen more as an average person, someone approachable by the


19

masses. An important theoretical finding in this context is that local commercials (involving a local celebrity) seem to „naturally personify‟ cultural traits that local consumers under-stand and can relate to (Pae et al., 2002).

It has also been established that choosing the „right‟ celebrity is not simply about fame lev-el. Rather, the perceived fit between the product and the celebrity may be as important to consumers (The NPD Group/Celebrity Influence Study, 2006). Arguably, it will be easier to successfully fit a local brand with a local celebrity.

Finally, assumptions could be made about the negative aspects of celebrity endorsements. One of the major concerns of using celebrity endorsement is the risk for a negative image transfer to take place between the celebrity and the product. This can happen if the celebri-ty is involved in a real or alleged scandal, or if negative information is revealed about the celebrity (Andersen, 1983). Although it cannot be said with certainty, local celebrities can very well be assumed less prone to be involved in a scandal and in turn negatively affect the company. The argument is that local celebrities are not surrounded with as much media at-tention as well-known celebrities. This decreases the risk for false negative information about the celebrity to arise.

Under the following sub-headings, the benefits of using a local celebrity endorser will be discussed briefly in terms of the following concepts used throughout the thesis.

3.1.1.1 Trustworthiness

As stated in the section above, local celebrities are thought to be easier to relate to. It also seems that this may have positive effects on the effectiveness of celebrity endorsement. Erdogan (1999) contents that in order to gain communication effectiveness, the chosen endorser should be one that people can relate to and is similar to the target group. This is because consumers more easily can trust these individuals. As it seems, a local celebrity should be perceived more trustworthy than a more famous celebrity.

3.1.1.2 Expertise

Perceived expertise is pointed out as the second important condition for being a credible communicator (Hovland, 1953). However, it is hard to guess whether or not local celebri-ties can be perceived at least as expert as better known celebrities. There might not be a distinction between local and well-known celebrities in possession of knowledge, expe-rience and skills. However, since expertise also is defined as the ability of the source to make truthful statements (Erdogan, 1999), it could be presumed that since local celebrities might be trusted more, they are also „trusted to be expert‟ if so is portrayed.

Hypothesis 1: Local celebrity endorsement can be viewed at least as credible as well-known celebrity endors-ers.

3.1.1.3 Attractiveness

The communication effectiveness of the message partly depends on the familiarity, similari-ty and likeability of the endorser (McCracken, 1989). Arguably, celebrities that are local and seemingly closer to the public may be perceived more familiar and similar. Most notable is probably the similarity characteristic. Similarity is a reoccurring factor of importance in the study of influence and persuasion. Apparently, human beings inherently rely on the people around them for how to think, feel, and act. It is also said that influence is often best when exerted horizontally rather than vertically (Cialdini, 2001). In other words, people are more likely to be influenced by an endorser who seems closer to them. However, this being said


20

about the first two characteristics making up attractiveness, it cannot be predicted whether local celebrities should be more likeable than better known celebrities or not.

Hypothesis 2: Local celebrity endorsers can be perceived equally attractive as well-known celebrity endorsers.

3.1.1.4 Emotional Involvement

Local celebrity endorsers are likely to be more emotionally involved than a well-known ce-lebrity endorser. Recall that emotional involvement refers to the celebrity endorser‟s moti-vation to appear in the advertisement (Hunter, 2009). Even if the financial compensation for a well-known celebrity will be much higher, other factors can motivate a celebrity to take part of the endorsement. For a local celebrity, an endorsement contract may be very desirable since it may provide needed visibility and attention. A local celebrity might there-fore be more motivated to stay in good terms with the company in question, and make a true effort to live up to its expectations for the endorsement. In short, an endorsement contract could be a „bigger deal‟ for a local celebrity than a well-known celebrity, and they are more likely to make sure to keep it such.

On the other side of the coin, newly started ventures cannot provide any prestige that oth-erwise might be rubbed off on celebrities who land a major contract with one of the larger brands such as Nike, Coca Cola or L‟Oréal. Therefore, employing a well-known celebrity might outright be a bad idea. Even if they receive the sufficient financial compensation, they will unlikely show any signs of emotional involvement. If the newly started venture has yet to prove its worth, the celebrity is fairly unlikely to get involved more than neces-sary in the endorsement process.

Hypothesis 3: Local celebrities can be perceived at least as emotionally involved in the endorsement process as well-known celebrities.

3.1.1.5 Meaning Transfer

As McCracken (1989) argues, it is the cultural meaning of the celebrity that is important in the endorsement process. As already stated, consumers understand and relate more easily to local cultural traits (Pae et al., 2002). Of course, a local celebrity would personify these in a much more realistic way than a well-known superstar. Therefore, local celebrities would be more successful in transferring true meaning to the endorsed brand. In addition, local celebrities are thought to be more compatible with smaller, newly started ventures. There is a more natural „fit‟. Usually, the new ventures start out very local anyway, so there is no need to try to gain unneeded attention and employ an expensive well-known celebrity.

Hypothesis 4: Local celebrities can be as capable of transferring meaning to the endorsed product as well-known celebrities.

3.2 Communication Effectiveness of Local and Well-known Celebrity Endorsement

Below is a proposed model that extends to include not only the source models, but also the meaning transfer model as well as emotional involvement to measure the communication effectiveness of local and well-known celebrity endorsers. Ohanian‟s (1990) source credibil-ity scale is used in determining the celebrity characteristics for the trustworthiness- and ex-pertise-dimensions. As for attractiveness, the characteristics are drawn from McGuire‟s (1985) research. In the context of predicting the effectiveness of local versus well-known celebrity endorsement, these characteristics are more suitable than those proposed by Oha-


21

nian (1990). The reason for this is since the focus lies on attraction based on recognition as opposed to merely physical appeal. Also, it has been important to include the concept of meaning transfer since local celebrities are believed (by the authors) to be perceived as more compatible with the brands of smaller, newly started ventures than larger celebrities would be. Emotional involvement is also important, as local celebrity endorsers are as-sumed to show higher emotional involvement in the endorsement process than larger cele-brity endorsers might. This model should be well suited to be used in the testing of local and well-known celebrity endorsement. From this, the last hypothesis can be formed, which serves to capture the purpose of the study.

Hypothesis 5: Local celebrity endorsement can be at least as effective as well-known celebrity endorsement.

Figure 3.1: Proposed model to test communication effectiveness of local versus well-known celebrity endorsement, created by the authors


22

4 Methodology This chapter presents the research process of this thesis. It elaborates the method chosen for research and mo-tivates how the survey is structured and why. In extent, the participants of the study are presented and the validity of the survey is discussed. According to Saunders, Lewis and Thornhill (2009, p.3), method is “…the techniques and procedures used to obtain and analyze research data, including for examples questionnaires, observations, interviews, and statistical and non-statistical techniques”. The method ex-plains how research will be carried out and the implications of those choices. The method is important since it guides the research (Jacobsen, 2002). The research process is according to McGrath (1981, p.179) “a series of interlocking choices, in which we try simultaneously to maximize several conflicting desiderata”. McGrath (1981) continues to argue that there is no such thing as one true method that is the methodological answer. On the contrary, there is no guarantee to success. The different approaches of methodology will be discussed in the following sections.

4.1 Inductive or Deductive Approach

The deductive approach is used when information about theories is gathered and expecta-tions are built about the theories. A hypothesis is deduced from existing theories and tested using data to see if the hypothesis is accepted or rejected (Ghauri & Grønhaug, 2005; Saunders et al., 2009). The criticism toward this approach is that the expectations inevitably direct the researcher towards information that is relevant to their expectations (Jacobsen, 2002). This leads to a tendency to support the prior expectations. By starting with concrete expectations the researcher limits the flow of information and might overlook substantial evidence. When using this approach, the research might get the stamp of being self-fulfilling since the researchers only find what they were looking for (Jacobsen, 2002). The other approach is inductive methodology, where the researchers take the opposite take. They gather information with barely any expectations, obtain the relevant facts and finally systemize their findings (Jacobsen, 2002). Without preconceptions, the researcher has a higher chance of getting relevant and correct evidence. The conclusions from this re-search approach are drawn from empirical observations and are incorporated back into ex-isting knowledge, to improve an existing theory for example (Ghauri & Grønhaug, 2005). The critique towards the inductive method is the difficulty to search for information and try to see reality without any expectations. Jacobsen (2002), claims that it is naïve and im-possible to keep a completely open mind.

4.1.1 Chosen Method: Deductive Approach

In this thesis, the deductive research approach was chosen. In the inductive method there are no pre-determined theories. In this thesis however, theories were gathered and hypotheses developed and then tested. This corresponds to the deductive approach (Ghauri & Grønhaug, 2005; Saunders et al., 2009). From the existent theories regarding celebrity en-dorsement, the authors deducted a hypothesis; local celebrity endorsers can be at least as effective as well-known celebrity endorsers. This main hypothesis is composed of four hypotheses, which have been developed according to the models measuring communication effectiveness. Da-ta was collected through questionnaires and statistical analysis was used to test these hypo-theses.


23

4.2 Qualitative versus Quantitative Research Approach

The decision of using a qualitative or quantitative approach is an early decision when de-signing the research strategy. A qualitative research method is suitable when trying to fur-ther clarify a concept or phenomenon. A qualitative study is in general to retrieve how something is interpreted and to understand a given situation (Jacobsen, 2002). A quantita-tive method is a better measure when there is no previous knowledge about the subject at hand. It is easier to describe a phenomenon‟s frequency or relationships and trends within data (Jacobsen, 2002).

4.2.1 Intensive and Extensive Data Collection

When performing a qualitative research, an intensive data collection is used where the sam-ple size is small and the information gathered is controlled by the researcher in advance (Jacobsen, 2002). When the information is collected, the author must first standardize the findings to receive an easier systemization of the data. This is because every unit cannot be treated in a unique way; they must be categorized in the already existent groupings. When applying a quantitative approach, an extensive data collection is used, where the sample size is larger and the researcher has pre-defined what information should be gathered. In quan-titative research, the categorization and the fundamental concepts are established before the empirical research. This aspect of the methodology infers that there should be empha-sis on the defined concepts. Since the categories are predetermined it is possible to stan-dardize the information into a numerical form, which creates an effective way of perform-ing a statistical analysis of a high number of data (Jacobsen, 2002).

4.2.2 Choice of Method: Quantitative Approach

The authors of this thesis have chosen to use the quantitative approach and extensive data col-lection as research techniques. This choice has been made since the research method of this study is to measure the „trends of a phenomenon‟, namely the communication effectiveness of local celebrity endorsers. This is in line with the quantitative method described by Saunders et al. (2009). A high number of participants are needed for this sort of study; hence extensive data collection was used to measure the trends in the data. The main hypo-thesis of this thesis is: local celebrity endorsers can be as effective as well-known, interna-tional celebrity endorsers. This hypothesis was tested through a technique where partici-pants were asked to answer the same set of questions in a pre-determined order (Saunders et al., 2009) - a questionnaire.

4.3 Research Strategy

The research strategy is a general plan of how the research questions are answered by the author. The different strategies are; experiment, survey, case study, action research et cetera (Saunders et al., 2009). The choice of strategy is guided by the research questions. The pur-pose of this thesis is fulfilled through performing quantitative surveys and combining the results from this with theories gained from previous research. The thesis establishes if it is advantageous for newly started ventures targeting local markets to invest in local celebrity endorsement. An explanatory research is conducted, where the relationships between dif-ferent variables are tested (Saunders et al, 2009). The explanatory study is executed by test-ing how consumers are influenced by different advertisements presented by either local or well-known celebrities. The study is structured in a way that participants were presented with a survey containing one of eight advertisements. The eight advertisements present four local celebrities and four well-known celebrities. The results from the survey indicate


24

if the communication effectiveness of the local and well-known celebrities is equivalent or how it differs.

4.3.1 Survey Design

Regarding the methods within the quantitative approach there is one dominating way of collecting data; conducting a questionnaire with given answer alternatives (Jacobsen, 2002). This is a primary data collection, where the researcher asks questions with given answer al-ternatives within the area that has already been defined. The answer alternatives which par-ticipants are presented with could be; yes; no, completely agree; do not agree at all, very positive; very negative et cetera. There are other ways of performing a quantitative study; such as a secondary data collection, where existing statistical analysis and previous research is used as foundation. The criticism against this approach is the questionable reliability and validity of the sources. When primary data is collected, it is more time consuming since there is a great amount of preparatory work before the actual questionnaire is created (Ja-cobsen, 2002). The authors have collected primary data in this research through a ques-tionnaire with given answer alternatives.

4.3.1.1 Independent and Dependent Variables

A dependent variable changes accordingly to a change in other variables (Saunders et al., 2009). In this study, there have been different dependent variables in the different tests. Trustworthiness, expertise, attractiveness, emotional involvement and meaning transfer (as presented in Ch. 2) all act as dependent variables. Continuously, these variables combined compute a new dependent variable referred to as communication effectiveness. The inde-pendent variables on the other hand, are the cause of change in the dependent variable (Saunders et al., 2009). The independent variable within this study is “Group” which represents local celebrity or well-known celebrity. The relationship between the indepen-dent and the dependent variables were measured through statistical analysis in Statistical Product and Service Solutions (SPSS). T-tests were conducted to test if there is a significant difference between local and well-known celebrity endorsements. Based on this, the hypo-theses can be either accepted or rejected.

4.3.2 Choosing the Right Celebrity

The survey was conducted to explore the communication effectiveness of two celebrities within the same field; a local celebrity compared with a well-known celebrity. The celebri-ties in the survey were chosen to make sure that both of their images are similar and that they are active within the same field.


25

Table 4.1 Selected Celebrities for the Study

Name Celebrity Group

Profession Product

Set 1

Leila Lindholm

Local Celebrity

Chef known from Swedish

television

Designed baking utensils

Rachael Ray Well-known

Celebrity

Internationally known chef

with her own show on televi-

sion.

Set 2

Niclas Adler

Local Celebrity

Dean at Jönköping In-

ternational Business School

(JIBS).

Student lamp

Jan Björklund Well-known

Celebrity

The Swedish Minister of

Education and Science.

Set 3

Daniel Karlsson

Local Celebrity

The Jönköping based singer who partici-

pated in Swe-dish Idol.

Cough drops

James Blunt Well-known

Celebrity Singer known

worldwide

Set 4

Nikita Johnson Local Celebrity

(fictional)

Claimed to be Swedish cham-pion in badmin-

ton

Razors

Nikita Johnson Well-known

Celebrity (fic-tional)

Claimed to World cham-

pion in badmin-ton


26

4.3.2.1 Matching the Celebrities with Suitable Products

The match-up hypothesis (Kamins, 1990; Misra & Beatty, 1990) concerns the fit between the endorser‟s image and the product or brand. The hypothesis suggests that if the fit be-tween the endorser and product is well done, the endorsement will be more effective. A fit between the two also implies that the celebrity endorser‟s image matches the product‟s characteristics. When the authors assigned products to the different celebrities it was of importance to allocate products that fit the celebrities. The first celebrity set are chefs and is represented by Leila Lindholm (local) and Rachael Ray (well-known). The authors chose that the endorsed product should be baking utensils, in order to have a clear connection between the celebrities and the product. The second set of celebrities is Niclas Adler (local) and Jan Björklund (well-known). Since Adler is the Dean of Jönköping International Busi-ness School and Björklund is the Swedish Minister of Education and Science, the authors wished to assign a school related product to them, hence a study lamp was chosen. The set of singers, which is Daniel Karlsson (local) and James Blunt (well-known), were assigned cough drops. Even though this product is fairly neutral, there is a connection between the singers and cough drops since singers often strain their voice. The fourth set is represented by an artificial celebrity who the authors call Nikita Johnson. She is described as being a Swedish badminton player (local celebrity) in one of the advertisements and an Australian badminton player (well-known celebrity) in the other advertisement. She is supposedly en-dorsing razors, which can be justified by her picture focusing on her long legs.

4.3.3 Designing Survey Questions

Questions in a survey should evolve from the problematization of the research. However, the problem statements may be quite vague (Jacobsen, 2002). This implies that before the survey or questionnaire is constructed, the problematization must be detailed and concrete. When asking about a concept of choice, one must always consider the different interpreta-tions that can be made, especially if the concept in question is intangible. To measure in-tangible phenomena, operationalization is used to make the intangible phenomena measur-able. Since phenomena are not tangible, one must settle with concrete indications towards the phenomenon, hence use an indirect measurement. Regarding the structure of the ques-tions, simplicity is a key word. Questions should be structured clearly and usage of technic-al terms should be avoided to make it is easily understood (Likert, 1974). Directing the question in a way that may affect the answer and in turn lower the validity should be avoided (Jacobsen, 2002) and the survey and its vocabulary should be adjusted to match the participants of the survey.

4.3.3.1 Chosen Survey Questions to Measure Communication Effectiveness

The survey questions are based on the models presented in the theoretical framework. The questions intend to measure the concepts from the source models; attractiveness, trustwor-thiness, and expertise. The questions regarding the source models follow „The Source Cre-dibility Scale‟, developed by Ohanian (1990) as guidance for the word pairing in the an-swers. The other survey questions measure meaning transfer and emotional involvement. The dimensions of emotional involvement, which create the word pairings, can be found in figure 2.3.

There are very few occasions when a researcher is the only one interested in a field of study. Most commonly, other similar studies connected to the subject already exist. To ease the new research it is helpful to review the existent publications to be able to concretize a concept. Another advantage from reviewing existing facts and their strategies is that it


27

assists the researcher in structuring its survey questions (Jacobsen, 2002). The authors of this thesis have chosen to use a previous survey designed by the celebrity endorsement re-searcher, Erik Hunter, as a base for their questionnaire. The questions and background sec-tion are similar to those of Hunter‟s, with some minor changes.

Table 4.2 How Research Questions are Measured in the Survey

Theoretical Frame

Measures Survey Question Research Question

The Source Credi-bility Model

Perceived trustworthiness and expertise of the cele-

brity 1-2

Can local celebrity en-dorsement be viewed at least as credible as well-

known celebrity endorsers?

The Source Attrac-tiveness Model

Perceived attractiveness of the celebrity

3

Are local celebrity endors-ers perceived equally at-

tractive as well-known ce-lebrity endorsers?

Emotional In-volvement

Perceived level of emo-tional involvement of the

celebrity

11-15

Can local celebrities be perceived at least as emo-tionally involved in the endorsement process as well-known celebrity?

The Meaning Transfer Model

Perceived level of mean-ing transfer from the ce-

lebrity to the product 6-7

Can local celebrities be as capable of transferring

meaning to the endorsed product as well-known

The questions in the survey that are not present in table 4.2 are question four; „In general, how effective do you believe the advertisement for [the company] to be?‟, which measures the attitude toward the advertisement, and question five „What is the possibility that you will buy a product from [the company] in the coming year?‟, which measures the purchase intentions. These questions were later used to back up the final hypothesis five, concerning communication effectiveness in whole.


28

4.3.3.1.1 Pilot Study

A pilot study was conducted before the questionnaire was sent out. The results of the pilot study triggered the authors rephrase and clarify some of the questions to make the survey easier to comprehend. The original questions measuring meaning transfer were open-end questions where the participant was asked to describe the celebrity and the respective product with five words. However, the participants of the pilot study expressed difficulties coming up with five words and the authors changed those to questions with answering al-ternatives. The first question to replace one of the earlier meaning transfer questions is: It is natural to associate [the product] with [name of the celebrity]. The second question is: [The product] is a product that is compatible with [name of the celebrity]. The answering alternatives range from Do not agree at all to Agree completely on both the questions.

4.3.3.2 Likert Scale, Also Called the Summative Scale

To measure the attitudes towards the advertisements, the Likert scale was used as answer-ing alternatives. The Likert scale was originally developed in 1932 and is a straightforward method which is easy to use (Likert, 1974). The typical Likert scale consists of five answer options; strongly disagree, disagree, neither agree nor disagree, agree, and strongly agree. However to attain accuracy the authors of this thesis has chosen to use a seven point Likert scale which is frequently used in research and proven to be effective (Dawes, 2008).

4.4 Survey Procedure

Before handing out the questionnaires to the participants, they were told to read the infor-mation carefully and answer all questions individually. They were also asked whether or not they knew about the celebrity in question, and given a short description of him or her. Pre-cautions were taken to make sure that it was clearly understood if the celebrity was local or well-known. This information was also clearly stated in the survey package. Each package included instructions with an example of how to answer the survey questions, a section where to fill in background information, background to the advertisement, the advertise-ment and the questionnaire.

4.4.1 Background

After reading the instructions and filling out some background information about them-selves, the participants were presented with a background description of the supposed new-ly started venture, the celebrity endorser, and the advertisement.

The eight different descriptions were written in a similar way, in order to maintain constan-cy. The participants were informed that the newly started venture in question was Swedish (or specifically based in Jönköping in the case of the local celebrity from Jönköping), and was using a celebrity (either local or well-known) as their endorser. Further, the celebrity endorser was described with emphasis on their degree of fame being either local or well-known. The endorsed product was also described. At last, the participants were told that their help was needed in evaluating the company‟s chosen marketing strategy.

In order to increase credibility, a fake website was also mentioned, citing that the target group is men and women all ages. The website also supposedly stated that the endorsed product was intended to be released in Sweden (or Jönköping) in June, 2010.


29

4.4.2 Celebrity Advertisements

Before the questionnaire section of the survey, participants were presented with a fake ad-vertisement representing a large picture of the celebrity endorser along with the endorsed product and a textbox providing information of the product and where it could be pur-chased. In order to increase validity, caution was taken to make sure that all eight adver-tisements looked very similar. Especially for the respective products, it was made sure that the two celebrities were showed in similar poses and facial expressions. The celebrity was also portrayed with a speech bubble, where he or she gave a positive statement about the endorsed product.

“I always use Baka&Smaka‟s products!” – Leila Lindholm/Rachael Ray

“They [Cervix] make my throat smooth as silk before each concert.” – Daniel Karlsson/James Blunt

“I know what it is like to be a student. You study late into the night.” – Niclas Adler/Jan Björklund

“An unforgettable shaving experience, that doesn‟t leave a trace!” – Nikita Johnson/Nikita Johnson

The above quotes have been translated to English from Swedish and are taken from the ce-lebrity advertisements which can be found in appendices 3 to 10 of the thesis.

4.4.3 Participants

There are eight different advertisements, and 30 samples of each advertisement were sent out to a total of 240 different participants. Due to the fact that celebrity endorsement has been tested to be more effective when a younger consumer base is targeted (Ainsworth, 2007) the authors decided to use a younger range of age participants.

4.5 Validity and Reliability

Reliability and validity of the research concern taking control of what is to be measured (Jacobsen, 2002). Validity and reliability are important since these ensure that the research is relevant and can be applied to future cases or studies. Validity is significant since it in-creases trustworthiness of the research. To achieve a valid research, it must be performed in a proper way, regardless of the methodology used or what data is collected. General va-lidity can be parted into two subcomponents; internal validity and external validity (Jacob-sen, 2002).

Reliability is used to ensure that the data collection techniques or analysis process will yield coherent outcomes (Easterby-Smith, Thorpe, Jackson & Lowe, 2008). According to Eas-terby-Smith et al. (2008) there are three questions that should be asked to increase reliabili-ty. The first question regards if the measure will yield the same outcome in other occasions. The second question concerns the possibility that comparable observations will be reached by other observers. The last question regards the clarity in how the raw data was inter-preted.

The authors of this thesis have ensured to fulfill these requirements by using a sample that best represents the targeted population, which is a younger consumer group. As mentioned previously (in section 4.2.2), the research was conducted in a quantitative approach. The questions in the questionnaire were carefully formulated with the aim of measuring differ-ent variables among a young population. As mentioned in section 4.3.3.1 the questions in the questionnaires were based on already existing questionnaires and theories to increase the validity and the reliability. The collected data was interpreted and analyzed in a clear


30

structure, which should be easy to follow by future researchers. The authors feel confident in their method selection and believe that, given the circumstances of financial limitations and short time span, it was the most suitable method in order to answer the research ques-tions of this thesis. The authors believe that if the research will be performed by other ob-servers who use the same theoretical framework and methods as used in this thesis, they will reach similar observations.

Robson (2002) discusses four threats to reliability: subject or participant error, subject or participant bias, observer error and observer bias. Subject or participant error concerns the unusual responses that participants may provide if they are studied in a situation not con-sistent with their normal behavior patterns (Saunders, et al., 2009). The subject or partici-pant bias regards the inaccurate responses participants may provide to alter the results of a study. Observer bias may take place when the observers give false answers to distort the re-sults of a research. Observer error is systematic errors that are mistakes made by the ob-server, for example due to confusion. To avoid subject or participant error the authors of this thesis distributed the questionnaires to the participants in school during daytime (Saunders et al., 2009). To reduce subject or participant bias the participants were ensured anonymity. They were also told to fill in the questionnaire individually to avoid being influ-enced by other participants. To avoid observer error it was decided that only one of the au-thors would distribute all questionnaires to the participants. This decreased the threat to re-liability since it ensured that all participants received the same information before and after filling in the questionnaire. Since a quantitative method is used in this thesis, observer bias is not an as great threat as it would have been in a qualitative study. However, the authors of this thesis compiled the results of the questionnaires together to avoid any observer bias.

4.5.1 Increasing Internal Validity

Internal validity regards control of the research and focuses on concretizing the abstract phenomenon at hand to be able to measure it properly (Jacobsen, 2002).

4.5.1.1 Reliability Tests

To be able to form scales out of the variables, the internal consistency must be measured. It is important that the variables are correlated with one another and measure the same thing (Bland & Altman, 1997). To measure the internal consistency, reliability tests were made on the different variables. A commonly used coefficient for evaluating this is Cron-bach‟s alpha (explained further in section 3.6), with the base of α≥0.7, which means that factors with an alpha coefficient equal or greater than 0.7 are accepted (Bland & Altman, 1997).

4.5.2 Increasing External Validity

External validity regards the transferability and concerns how the results of a research can be generalized and be transferred into a new context (Jacobsen, 2002).

4.5.2.1 Multiple Experiments

To add transferability to other cases, and add external validity to the research, the authors have chosen to use multiple sets of advertisements and celebrities in the study. It was cho-sen to use three sets of celebrities to minimize the effects of personal feelings directed to-wards the celebrities, which otherwise might bias the results. If only one local and one well-known celebrity were portrayed in the advertisement, already existent negative or positive emotions related to the particular celebrity might influence the results of the research.


31

External validity is further increased by letting the fourth advertisement set be represented by an artificial celebrity, presented as a local celebrity in one case, and a well-known celebri-ty the other.

To test the validity of the research, the authors used SPSS to measure the difference be-tween the celebrities and products. The results showed that there were no significant dif-ferences between either the celebrities chosen for the advertisements or the products; indi-cating that the personal emotions of the participants did not have an impact on the results.


32

5 Findings and Analysis

This section presents and analyzes the empirical data collected from the statistical results of the survey. The outcomes are further applied to the theoretical framework and the authors‟ own propositions.

5.1 Data Analysis

The analytical tool the authors chose to evaluate the data in is SPSS, which is the most popular tool for statistical analysis (Corston & Colman, 2003).

To be able to measure the concepts in this thesis, operationalization was needed (Bryman & Cramer, 1997). The answers were already transferrable since the survey questions are based on validated models (see section 4.3.3.1). The survey questions measure the different parts of communication effectiveness (trustworthiness, expertise, attractiveness, emotional involvement and meaning transfer). The variables of interest were measured by various questions in the questionnaire, some which were divided into sub-questions. These parts were retrieved from the source credibility scale (see appendix 1). For example, to measure trustworthiness, sub-questions tested how dependable, honest, reliable, sincere, and trust-worthy the endorser was perceived by the participant. All answers were numerically pre-sented from 1 to 7 according to the Likert scale. When the 240 questionnaires were ga-thered, the authors inserted the answers manually into a spreadsheet in Excel to ensure that all the questionnaires could be used (West, 1999). Since the answers were numerical, the data transfer from Excel into SPSS was simplified. The variables that were not based on the theoretical models, such as celebrity type and gender, were converted into numerical values. Local celebrities are defined as 1 and well-known celebrities as 2. Female participants are represented with a 0 and male participants with 1.

The first step in SPSS was to perform reliability tests. These tests were conducted on the respective variables to ensure that the results are valid (Bryman & Cramer, 1997). To guar-antee reliability, the popular measure Cronbach‟s alpha (Bryman & Cramer, 1997) was used, which calculates the internal consistency. This refers to how close related a set of items are as a group. The rule of thumb is that if Cronbach‟s alpha is equal or above 0.7 it is acceptable to use the combined data (Bland & Altman, 1997). When performing the re-liability test it was shown that the Cronbach‟s alpha ranged from 0.74 and higher for attrac-tiveness, emotional involvement, expertise and trustworthiness. However, the separate va-riables measuring meaning transfer and credibility scored beneath 0.7 when scaled together, so the authors decided it was better to handle these variables separately. All variables were merged into a new variable called „communication effectiveness‟, which had a Cronbach‟s alpha of 0.775.

To simplify the analysis, the authors chose to compute the related variables into scales, hence new variables. To measure trustworthiness for example; dependability, honesty, re-liability, sincerity, and trustworthiness were after the reliability test computed into a new va-riable named „Trustworthiness Scale‟. Meaning transfer and credibility were the only scales that did not prove to be reliable. There are instead two questions measuring meaning trans-fer. Credibility is measured by trustworthiness and expertise separately. After the scales were computed an additional variable was created; the communication effectiveness scale. This scale consists of the trustworthiness scale, expertise scale, attractiveness scale, emo-tional involvement scale, and meaning transfer.


33

The sequential step conducted in SPSS was to add all the values together in each scale and divide it by the number of cases. This is also referred to as calculating the mean value (Saunders et al., 2009). Calculating the mean is a common measure and a building block for many statistical tests used to investigate relationships between variables (Saunders et al., 2009).

To test if the hypotheses are accepted or not, independent-sample t-tests were conducted. The t-test establishes the significance level of the hypothesis and was conducted with a 95 percent confidence interval. The first table provides descriptive statistics of the data where the number of cases, mean, the standard deviation, and standard error of the mean is pre-sented for independent variables compared with the respective dependent variables. The standard deviation is used to show how much the values vary compared to the mean, where a low standard deviation is a sign that the values are in fact very close to the actual mean value (Saunders et al., 2009).

The second table presents the t-test results, which show if there is equal variance between local and well-known celebrities. The Levene‟s test can assure the statistical significance of a result. Since a 95 percent confidence interval is used, the p-value must be lower than 0.05 to make the result significantly different. The p-value refers to the probability of attaining a test statistic that is at least as extreme as the one observed (Bryman & Cramer, 1997). If it is over 0.05 there is no significant difference. The second table also shows the upper and lower bounds of the confidence interval (Corston & Colman, 2003).

5.2 Overview of Findings Below, an overview of the empirical findings is given. This is followed by more detailed in-formation regarding the different variables individually as well as the outcomes from the hypothesis testing. The results are presented according to celebrity group; local or well-known. The participants range between 16 to 45 years old. The average age is 21.3. As seen in the diagram below, the older participants only constitute seven percent. The participants are ei-ther attending the university in Jönköping or Per Brahe upper-secondary school. E-mails, containing information and purpose of the thesis, were sent out to teachers at the different schools, who assigned school classes to participate in the study. To avoid biased answers the purpose of the research was not available to the participants until they completed the questionnaire.

Figure 5.1: Diagram showing the survey participants‟ age distribution

42%

51%

5% 2%

Age

16-19 20-29 30-39 40-45


34

An uneven gender distribution has occurred within the two groups. In the „Local group‟ 77 percent were represented by women, while 23 percent were men. In the „Well-known group‟ 64 percent were women and 36 percent were men. Overall, the study was partici-pated by 70 percent women and 30 percent men. As seen below, the gender distribution between the two groups is relatively dissimilar. To account for this difference, certain tests have also been presented according to gender.

Figures 5.2 and 5.3: Gender distribution according to celebrity group

In general, it was found that there is no significant difference between the values given lo-cal and well-known celebrities. The table shows that in fact, there were some cases where the local celebrities were given higher values than the well-known celebrities. The results in-dicate that the participants of the study found the local celebrities at least as effective en-dorsers as the well-known celebrities.

5.2.1 Comparison of Mean Values

Table 5.1 Mean Values for Local and Well-known Celebrity Endorser

When comparing mean values, a confidence interval of 95 percent was used, meaning that variables having a p-value equal or below 0.05 indicated a significant difference in mean values. Variables having a p-value over 0.05 had no significant difference in mean values. The test showed results under the critical value of 0.05 for the variables trustworthiness, emotional involvement and communication effectiveness. In fact, the mean values in all

Men23%

Women77%

Gender distributionLocal celebrities

Men36%

Women64%

Gender distributionWell-known celebritites

MEAN VALUES

Variables Local celebrity Well-known celebrity

Trustworthiness 4.325 3.8583

Expertise 4.1817 3.9717

Emotional involvement 3.5625 3.0153

Attractiveness 3.8583 3.5833

Celebrity product association 3.46 3.37

Image transfer 4.02 3.9

Communication effectiveness 3.933 3.6124

Attitude towards the advertisement 3.1 2.93

Purchase intentions 2.08 2.14


35

Hypothesis 1: Local celebrity endorsement can be viewed at least as credible as well-known celebrity endorsers.

these cases were higher for the „Local group‟. For the other variables, meaning transfer, ex-pertise, attractiveness, purchase intentions and attitudes towards the advertisement, the sig-nificance level was over 0.05, implying that there was a difference in the size of the inter-vals between the groups. Hence, there was no difference in mean value.

5.2.2 Hypothesis 1: Credibility

Hypothesis 1 aims to answer research question one which regards credibility. According to the source model, credibility consists of trustworthiness and expertise. In the statistical test, these two variables have been tested individually. Overall, results show that there is no relevant difference between the groups. When reviewing the result from the two groups, the standard deviation is 1.21698 for local celebrity endorsers and 1.22039 for well-known celebrity endorsers in trustworthiness. In expertise the standard deviation for local celebri-ties is 1.46075 and well-known celebrities 1.29415. Since the standard deviation is low in all cases it is possible to infer that the results are close to the respective mean values. The mean values for trustworthiness are 4.3250 for local celebrities and a slightly lower number for well-known celebrities of 3.8583. In the test for expertise the mean values are 4.1817 for local celebrities and 3.9717 for well-known celebrities.

The sigma for attractiveness is greater than 0.05, which determines that equal variance is assumed. However, regarding expertise the sigma is below 0.05 (0.042), which implies that equal variance is not assumed. The p-value in the t-test is 0.003 for trustworthiness and 0.240 for expertise. The p-value for trustworthiness indicates that there is a difference in mean values between local celebrity and well-known celebrity endorsers. The mean value for local celebrity endorsers is slightly higher than for well-known celebrities, which sug-gests that there is a positive difference between the mean values. In fact, this implies that local celebrity endorsers are perceived more trustworthy than well-known celebrities by the survey participants. For expertise, Levene‟s test resulted in a p-value higher than 0.05, which is interpreted that there is no significant difference in the mean values between the two groups. These results imply that the first hypothesis can be accepted since the data suggests that local celebrity endorsement is at least as credible as well-known celebrities.


36

Hypothesis 2: Local celebrity endorsers can be perceived as equally attractive as well-known celebrity endorsers.

Table 5.2 Statistical Results for Trustworthiness

Trustworthiness

Number of cases

Mean value

Standard Deviation

Standard Error of Mean

Sigma P-value

(2-tailed)

Lower bound of Confident Interval

Upper bound of Confident Interval

Local 120 4.3250 1.21698 0.11109

0.552 0.003 0.15673 0.77661 Well-known

120 3.8583 1.22039 0.11141

Table 5.3 Statistical Results for Expertise

Expertise

Number of cases

Mean value

Standard Deviation


Sigma P-value

(2-tailed)



Local 120 4.1817 1.46075 0.13335

0.042 0.240 -0.14098 0.56098 Well-known

120 3.9717 1.29415 0.11814

Hypothesis 1 is accepted

5.2.3 Hypothesis 2: Attractiveness

Hypothesis 2 has been tested in order to answer research question two, regarding attrac-tiveness. Attractiveness was tested by measuring perceived familiarity, similarity and likea-bility between the celebrity groups. The results for attractiveness also show no relevant dif-ference between the groups. The standard deviation in this test was low; 1.33939 for local celebrities and 1.37420 for well-known celebrities. This indicates that the mean value, of 3.8583 for local celebrity endorsers and 3.5833 for well-known celebrities, is a correct re-presentation of the average numerical answer.


37

Hypothesis 3: Local celebrities can be perceived at least as emotionally involved in the endorsement process as well-known celebrities.

In the t-test regarding attractiveness the sigma is 0.511, which determines that equal va-riance is assumed. The p-value is higher than 0.05 (0.118), which indicates that there is no significant difference between the mean values. The results indicate that the hypothesis can be accepted and that local celebrity endorser can be perceived equally attractive as well-known celebrities.

Table 5.4 Statistical Results for Attractiveness

Attractiveness

Number of cases

Mean value

Standard Deviation


Sigma

P-value

(2-tailed)



Local 120 3.8583 1.33939 0.12227

0.511 0.118 -0.07009 0.62009 Well-known

120 3.5833 1.37420 0.12545


5.2.4 Hypothesis 3: Emotional Involvement

Hypothesis 3 corresponds to research question three and tests the perceived emotional in-volvement of the celebrity group. The standard deviation is in this test low, 1.33527 for lo-cal celebrities and 1.31854 for well-known celebrities, which suggests that the average value is correctly showed in the mean values. The local celebrity endorser group has a mean value of 3.5625 whilst the mean value for well-known celebrities is 3.0153. The sigma for emo-tional involvement is 0.916 which indicates that equal variance is assumed since it is greater than 0.05. The p-value in the t-test is 0.002, which signals that there is a significant differ-ence between the mean value of local and well-known celebrity endorsers.

The results for testing emotional involvement showed statistically significant higher results for the „Local group‟ and the observation suggests that there is a positive difference be-tween the mean values. This signifies that local celebrity endorsers are perceived more emotionally involved than well-known celebrity endorser, by the survey participants.


38

Hypothesis 4: Local celebrities can be as capable of transferring meaning to the endorsed product as well-known celebrities.

Table 5.5 Statistical Results for Emotional Involvement

Emotional Involvement

Number of cases

Mean value

Standard Deviation


Sigma P-value

of T-test



Local 120 3.5625 1.33527 0.12189

0.916 0.002 0.20975 0.88469 Well-known

120 3.0153 1.31854 0.12037


5.2.5 Hypothesis 4: Meaning Transfer

Hypothesis 4 regards meaning transfer and is tested in order to answer research question four. Meaning transfer was tested by measuring two variables which the authors refer to as perceived celebrity-product association, and perceived image transfer. In the celebrity-product association test, the standard deviation was fairly low (see table 5.6), which signals that the mean value is a correct value to represent the average number. The mean values in the celebrity-product association test are 3.46 for local celebrities and a slightly lower num-ber of 3.37 for well-known celebrity endorsers.

The sigma value for celebrity-product association is 0.017, which states that equal variance is not assumed. The p-value of this t-test (0.703) is over 0.05 and indicates that there is no significant difference in the celebrity-product association between the two groups of local and well-known celebrity endorsers. The other test for meaning transfer is the perceived image transfer from the celebrity to the product. Also in this test the standard deviations are low (1.918 for local celebrities and 1.880 for well-known celebrities) indicating correctly represented mean values. The mean values are 4.02 for local celebrity endorsers and a somewhat smaller value of 3.90 for well-known celebrity endorsers. The sigma value for image transfer is 0.841, which indicates that equal variance is assumed. The p-value (2-tailed) in this test is 0.635, which indicates a similar outcome to the celebrity-product asso-ciation test. This proposes that both of these variables show no significant difference be-tween the celebrity groups. Therefore, local celebrities should be perceived equally effective in transferring meaning to the endorsed product as well-known celebrities.


39

Hypothesis5: Local celebrity endorsement can be at least as effective as well-known celebrity endorsement.

Table 5.6 Statistical Results for Celebrity-Product Association

Table 5.7 Statistical Results for Image Transfer


5.2.6 Hypothesis 5: Communication Effectiveness

Hypothesis 5 concludes the investigation, and aims to test whether the main assumption of the thesis can be accepted or rejected, namely if local celebrities can be at least as effective in the endorsement process as well-known celebrities. The test was executed by testing the combined variable consisting of credibility, attractiveness, emotional involvement and meaning transfer, to explore if the fifth hypothesis will be rejected or accepted.

The standard deviations values are 1.07487 and 0.90158 for the local and well-known ce-lebrity endorsers. This signals that the mean values of 3.9330 for local celebrities and 3.6124 for well-known celebrity endorsers are good measures.

Meaning Transfer- Celebrity-Product Association

Number of cases

Mean value

Standard Deviation


Sigma P-value

(2-tailed)



Local 120 3.46 1.978 0.181

0.017 0.703 - 0.381 0.564 Well-known

120 3.37 1.729 0.158

Meaning Transfer- Image Transfer

Number of cases

Mean value

Standard Deviation


Sigma

P-value (2-

tailed)



Local 120 4.02 1.918 0.175

0.841 0.635 -0.366 0.600 Well-known

120 3.90 1.880 0.172


40

Regarding the t-test, the sigma value for communication effectiveness is 0.024. Since this value is below 0.05, equal variances are not assumed. The p-value (2-tailed) was 0,013 for communication effectiveness. This indicates a difference in mean values of local and well-known celebrity endorsement. Since the mean value for local celebrity endorsers (3.9330) is higher than for well-known celebrity endorsers (3.6124) the data suggested that local cele-brity endorsers are more effective communicators than well-known celebrities.

Table 5.8 Statistical Results for Communication Effectiveness

A test was also conducted to see if there was a difference between gender in communica-tion effectiveness of local and well-known celebrities. When the women variable was tested, the p-value was 0.206. Since this is higher than 0.05, it indicates that women perce-ive no significant difference in communication effectiveness between local and well-known celebrities. However, when the variable for men was tested the p-value was 0.012 which signals that there is a perceived difference in communication effectiveness of local and well-known celebrity endorsement. As seen in figure 5.4 the men‟s mean values are 4.0551 for local celebrities and 3.466 for well-known celebrities, which imply that men show a clearly stronger response to the communication from a local celebrity endorser than a well-known celebrity.

Figure 5.4 Diagram presenting the mean values of communication effectiveness divided to gender and celebrity group.

4,0551

3,4663,8976

3,6941

Local Well-known

Communication effectiveness mean

Men Women

Communication Effectiveness

Number of cases

Mean value

Standard Deviation


Sigma P-value

(2-tailed)



Local 120 3.9330 1.07486 0.09812

0.024 0.013 0.06828 0.57294 Well-

known 120 3.6124 0.90158 0.08230


41

To strengthen this fifth hypothesis, two additional variables were tested as well, namely at-titudes toward the advertisement and purchase intention. For both of these tests the sigma value is above 0.05 (see tables 5.9 and 5.10) which states that equal variances are assumed. The p-values in the t-tests are 0.730 for the purchase intentions test, and 0.310 in the test regarding attitude toward the advertisement test. This implies that there is no relevant dif-ference in the mean values between the local celebrity and well-known celebrity endorsers. This indicates no difference in purchase intentions or attitudes towards the advertisement between the local celebrity, and the well-known celebrity endorsement.

Table 5.9 Statistical Results for Purchase Intentions

Purchase Intentions

Number of cases

Mean value

Standard Deviation


Sigma P-value

(2-tailed)



Local 120 2.08 1.351 0.123

0.849 0.730 -0.390 0.274 Well-

known 120 2.14 1.259 0.115

Table 5.10 Statistical Results for Attitude towards the Advertisement

Attitude towards the Advertisement

Number of cases

Mean value

Standard Deviation


Sigma P-value

(2-tailed)



Local 120 3.10 1.324 0.121

0.915 0.310 -0.165 0.516 Well-

known 120 2.93 1.348 0.123



42

Hypothesis 1: Local celebrity endorsement can be viewed at least as credible as well-known celebrity endorsers.

ACCEPTED

Hypothesis 2: Local celebrity endorsers can be perceived equally attractive as well-known celebrity endorsers.

ACCEPTED Hypothesis 3: Local celebrities can be perceived at least as emotionally involved in the endorsement process as well-known celebrities.

ACCEPTED Hypothesis 4: Local celebrities can be as capable of transferring meaning to the endorsed product as well-known cele-brities.

ACCEPTED Hypothesis5: Local celebrity endorsement can be at least as effective as well-known celebrity endorsement.

ACCEPTED

All in all, well-known celebrities do not show to be superior in any of the tests. Therefore, all the hypotheses are accepted.

5.3 Analysis in Accordance with Theoretical Framework

The main hypothesis „Local celebrity endorsement can be at least as effective as well-known celebrity en-dorsement‟ cannot be tested measuring a single variable. Therefore, this hypothesis is split in-to four. This way, communication effectiveness is evaluated by measuring perceived trust-worthiness, expertise, attractiveness, emotional involvement and meaning transfer. Since there is no previous research on the effectiveness of local celebrity endorsers in particular, the findings are analyzed to the general theory of celebrity endorsement and the proposi-tions made by the authors in chapter three.

5.3.1 Credibility

According to the source model (see section 2.2.1.1), credibility is evaluated by measuring perceived trustworthiness and expertise. The variable perceived trustworthiness actually shows a significant preference for local celebrities, where the mean value is 4.3250 com-pared to 3.8583 for well-known celebrities. This is in line with the propositions stated in section 3.1.1.1. As for the variable perceived expertise, there is no significant difference in mean value for the two groups. The interpretation of this is that local celebrities can be perceived as more trustworthy than well-known celebrities and that their level of expertise is perceived equally.

As it seems, the speculations previously stated in the thesis can now be confirmed. Local celebrities can probably be easier to relate to since they obviously are more similar to the target group in question. This makes them more trustworthy, and in turn, more persuasive. Earlier, the authors expressed difficulty predicting whether there is a distinction in the pos-session of knowledge, skills and experience between local and well-known celebrities. When reviewing the results, it is clear that there is no difference in perceived expertise.


43

However, this may also be explained by the celebrities‟ ability to make truthful statements, which was established in the test regarding the trustworthiness variable. Since local celebri-ties are found to be viewed more trustworthy, they can be said to be recognized as expert as well, making them credible communicators.

5.3.2 Attractiveness

The test for the attractiveness variable reveals no significant difference between local and well-known celebrities. Companies traditionally try to employ the A-list celebrities in hope that their attractiveness will rub off on their product. Celebrity worship is a fact and many companies wish to exploit this phenomenon. However, previous researchers have argued that the attractiveness of celebrities not only depend on their physical attributes (see section 2.2.1.2). Attractiveness is also said to be made up of familiarity, similarity and likeability which in this case seem to be attributed indifferently to the two celebrity groups.

Companies should take advantage of the fact that they can employ a celebrity that is per-ceived as attractive by the target group, without necessarily being „super famous‟ and ex-tremely expensive. The predictions made in 3.1.1.3 are confirmed by the statistical analysis made on attractiveness, where the test shows no significant difference in attractiveness be-tween local and well-known celebrities.

5.3.3 Emotional Involvement

The variable emotional involvement (see section 2.2.4) was tested and reveals a result sig-nificantly in favor of local celebrity endorsers. The mean for well-known celebrities is 3.0153 while the mean for local celebrities is impressively higher and has a value of 3.5625. This indicates that the prior speculations made (see section 3.1.1.4) can be confirmed once again. Apparently, local celebrity endorsers are perceived to be more emotionally involved than well-known celebrities in the endorsement process. Even though one might never know if local celebrities are truly more motivated in working with the product, it is clear that they are perceived as more likely to use and like the product and be enthusiastic about the endorsement contract.

However, the authors believe that local celebrities are more easily motivated than well-known celebrities. The company can entice the celebrity with promised visibility and atten-tion, which the authors assume are very desirable states for a less-known celebrity. Also, since local celebrities do not enjoy the same wealth as most well-known celebrities do, the financial compensation they receive for the endorsement can trigger the celebrity into being motivated to appear in the advertisement and keep good relations with the company.

5.3.4 Meaning Transfer

Meaning transfer (see section 2.2.3) was measured by testing two other variables, by the au-thors referred to as celebrity-product association, and perceived image transfer. The test reveals no significant difference between local and well-known celebrities, who therefore seem to be equally effective in transferring meaning to the endorsed product. In order for the endorsement to be communicated effectively, it is important that the cultural meaning of the celebrity is conveyed in the endorsement process. The finding that local celebrities are equally capable of this is important since it leaves the company with an exponentially larger selection of celebrities to choose from. Now, as long as the celebrity holds the right cultural meaning which the company wishes to be associated with the endorsed product, it does not matter how famous the celebrity is. Since media has been growing more rapidly


44

than ever, there are also more celebrities than ever. Today‟s environment is drastically dif-ferent from the time where royalty were considered the only specimen worth to employ as „celebrity‟ endorsers. Today there exist celebrities famous within thousands of niches, and these all come with their own cultural meaning. This implies that the company can find a celebrity that fits the product a thousand times better than before.

5.3.5 Communication Effectiveness

The goal of using celebrity endorsement in marketing communication is to make use of the celebrity capital possessed by the celebrity endorser. The celebrity capital, consisting of the celebrity‟s favourable awareness and associations, can transfer to the brand or product and is meant to translate into reputational capital for the company. Reputational capital is needed in order to stabilize consumers‟ demand for the product. However, in order for this process to occur, it is crucial that the communication effectiveness of the endorsement message is high. Communication effectiveness can be described as the degree to which the desired communication outcome is achieved (see section 1.6). Effective communication can prompt action, which in this case refers to increased purchase intentions. Companies who spend millions of dollars on celebrity endorsement aim for their invested celebrity dol-lars to translate into even higher revenue dollars. However, newly started ventures that typ-ically experience lack of legitimacy and liability of newness and therefore find themselves in the need to create fast reputational capital often lack the financial capital needed to achieve this by utilizing celebrity endorsement.

The research outcome of this thesis has detected a major opportunity that definitely should be exploited by newly started ventures who find themselves in the described situation. It has been proven that local, and therefore less expensive, celebrity endorsement can be at least as effective as well-known celebrity endorsement. For all variables constructing com-munication effectiveness: trustworthiness, expertise, attractiveness, meaning transfer and emotional involvement, well-known celebrities do not once gain higher values than local celebrities. Actually, for communication effectiveness as a whole, local celebrity endorsers have a mean value of 3.933 and proved to be more effective than well-known celebrity en-dorsers, who score a mean value of 3.6124. The implication of this is very important: it seems that newly started ventures evidently should choose to employ local celebrities rather than well-known in their marketing campaigns.

An additional finding that was gained in the process of statistical analysis is the fact that men showed a significantly higher tendency to be influenced by local celebrity endorsers as opposed to well-known celebrity endorsers. The difference in mean values for communica-tion effectiveness is surprisingly large: 3.466 for well-known celebrity endorsers and 4.0551 for local celebrity endorsers. This observation is interesting since it implies that it can be even more effective for firms targeting men, to use local celebrity endorsers in their mar-keting communication. These advertisement campaigns can be focused by appearing in marketing mediums targeting men, such as sports magazines.

The results gained from the statistical analysis portray a local celebrity as being more trust-worthy than a well-known celebrity. This result can be explained by the fact that they are seemingly easier to relate to, encompass local cultural meanings and are seen as people of the masses. These qualities also relate to their perceived level of expertise, which does not differ from that of the well-known celebrities. Local celebrities are as capable of transfer-ring meaning to the endorsed product as well-known celebrities. They are actually per-ceived to be more emotionally involved than well-known celebrities. Overall, a local celeb-


45

rity is more effective in communicating the endorsement message and all hypotheses can be accepted. The next section describes what this important finding implies for the use of celebrity endorsement today.

5.4 Analysis of Celebrity Endorsement Today

Celebrity endorsement is a subject that has been researched greatly from the beginning of the 1970s. It is increasingly used in advertising, and still seems to be on the up rise. The re-search themes of celebrity endorsement are many. To name a few, the difference between celebrity and non-celebrity endorsers, the effects on endorsement when negative informa-tion about the celebrity is revealed, the effects of multiple product endorsements and dif-ferent models how to chose the right celebrity have been studied. In the literature review, it was found that there was a gap in the research: there was no existent theory on the signific-ance of fame level. Consequently, a collection of primary data was made in order to test whether local celebrities could be as effective communicators of the endorsement message, as well-known celebrities.

As confirmed from the results of the survey, the level of fame does not have an impact on communication effectiveness. It seems obvious that there are still themes of celebrity en-dorsement yet to be researched. The evolution of celebrity endorsement can therefore con-tinue to expand. Since today‟s society is characterized by a major overflow of information and media, traditional advertising does not stand out. Something extra is needed, and since celebrities have proven to be a good cause of advertisement recall, celebrity endorsement can be a good strategy.

The finding of this thesis is important since it gives the companies, especially newly started ventures, who lack the sufficient financial capital, the opportunity to look beyond tradition-al celebrity endorsement.

In general, it seems that the importance of mass-markets is decreasing, and niche markets are gaining more and more significance. In a way, this signalizes that the traditional utiliza-tion of celebrity endorsement should adapt to follow this trend. This means that the major-ity of future advertisements portraying a celebrity endorser might not even be recognized by the masses. The celebrity is not known to the masses, only in its certain niche: where it is perceived to be very credible, attractive, emotionally involved and conveys the „right‟ cul-tural meaning.


46

6 Conclusion

This section concludes the output from the analysis of the research questions and fulfills the purpose of the thesis.

The purpose of this thesis was to compare the differences of local and well-known celebri-ty endorsement on communication effectiveness. The purpose is fulfilled by answering the following research questions:

RQ 1: Can local celebrity endorsers be viewed at least as credible as well-known celebrity

endorsers? Local celebrities are perceived to be as credible as well-known celebrities. Expertise is per-ceived equally, and trustworthiness is perceived to be higher among the local celebrity en-dorsers.

RQ 2: Are local celebrity endorsers perceived equally attractive as well-known celebrity endorsers? Local celebrities are perceived to be as attractive as well-known celebrities.

RQ 3: Can local celebrities be perceived at least as emotionally involved in the endorsement process as well-known celebrities?

Local celebrity endorsers are perceived more emotionally involved than well-known celeb-rities in the endorsement process.

RQ 4: Can local celebrities be as capable of transferring meaning to the endorsed product as well-known celebrities?

Local celebrities are found to be equally capable of transferring meaning to the endorsed product as well-known celebrities. Both image transfer and celebrity-product association are equally successful between the celebrity groups. This study finds that overall, local celebrity endorsers are more effective than well-known celebrities in communicating the endorsement message. When combining this finding with previous research, the authors conclude that the cause of this is that local celebrities are perceived more similar to the target group. Therefore, they are perceived more trustworthy, familiar and easier to relate to. They also encompass local cultural traits that can be trans-ferred to the endorsed product. It should be noted that this study has been conducted to test the communication effective-ness of celebrities endorsing the products of newly started ventures. However, the authors as-sume that the conclusions drawn can be applied to other small enterprises that suffer from lack of financial capital and act on local markets.


47

7 Limitations, Suggestions for Future Research and Final Thoughts

This chapter presents the critique of the study and method as well as further research suggestions by the au-thors. To finish the chapter, additional findings that were learned during the process of writing the thesis but not directly connected to the purpose of this research, are presented.

7.1 Limitations

Validity issues could be raised as critique towards this study. Even though the authors used four sets of advertisement and eight different celebrities the fact still stands that partici-pants could have been affected (positively or negatively) by the celebrity‟s face or name, which may have influenced their answers in the questionnaire. As mentioned in 1.1, this thesis was produced during a short time span which limited the authors to only use four sets of advertisements. The validity could have been increased by testing more advertise-ment sets with celebrities and products. With a longer time span, the authors would also have had the possibility to reach out to more participants, which would have increased the validity.

Due to financial limitations, the advertisements were not made by a graphic designer. In-stead they were made by the authors themselves. This fact may have decreased the trust-worthiness of the advertisements since they were not as professionally designed as real ad-vertisements.

Another limitation of this thesis is that the author used a laboratory setting instead of a typ-ical advertisement setting when distributing the advertisements and the questionnaires. This could also have decreased the validity of the advertisement and this study, since the partici-pants may be suspicious when receiving an advertisement this way.

Also, the fact that the survey had an uneven gender distribution can be seen as a limitation. This may have affected the results, especially since it was found that gender perceived communication effectiveness in whole differently. If the gender distribution had been equal, the results may have been different for other variables as well.

Despite the limitations of the study, it does have certain strengths that should be brought forward. The main strength is the sample size which can be considered large under the cir-cumstances. This compensates for some of the negative effects the limitations might have caused on the results. An additional strength of the study is that it results in a theoretical contribution, since the selected research area has not been explored before.

7.2 Suggestions for Future Research

Since there are no previous studies within the field of local celebrity endorsement the au-thors have some suggestions what could be researched in the future.

The first suggestion is to replicate and confirm the study conducted in this thesis. Howev-er, different celebrities and products could be used to investigate if new sets yield the same result.

Since the geographical perspective of this study is Sweden, an additional research sugges-tion is presented. The proposal is once again to replicate this study, but by using a different


48

geographical area as base. This can be interesting because consumers in different countries can perceive their local celebrities differently.

The next idea for future research was grazed upon in this thesis; namely to investigate if the effectiveness of local celebrity endorsement differs between genders. The authors showed in the results that the male participants of this study demonstrated a more positive attitude toward local celebrity endorsement than well-known celebrity endorsement. The female participants showed no difference in attitude toward local and well-known celebrity en-dorsers. If a more thorough investigation of this is conducted, companies could reach their target group more easily.

The next suggestion is to further investigate the prior study by Almquist & Robert (2000), who state that celebrity endorsement is more effective on younger individuals. The sugges-tion is to research if local celebrity endorsement is more effective on a younger target group.

The final suggestion for future research is to explore if well established companies can benefit from using local celebrity endorsers in marketing campaigns for their local markets. This thesis argues that local celebrity endorsement is effective for newly started ventures. However, whether local celebrity endorsers are as effective for larger, established compa-nies is not part of this thesis, and would be an interesting purpose of a future study.

7.3 Final Thoughts

Celebrity endorsement is a widely researched phenomenon. However, there are still oppor-tunities for further research. For instance, the significance of fame level is a theme that has been ignored up until now.

This study finds that for newly started ventures, local celebrity endorsers actually are more effective than well-known celebrity endorsers. The results show that a local celebrity is per-ceived as expert, attractive, and capable of transferring meaning to the endorsed product, as a well-known celebrity. Further, local celebrities are found to be more trustworthy, emo-tionally involved, and overall show higher communication effectiveness.

Beyond the conclusions drawn, there are other factors which make the local celebrity ex-tremely attractive to use in marketing communications. Insights that have been made along the way include the following:

Local celebrities will be less prone to be involved in scandals, which otherwise may impose a risk for the endorsed brand. Especially since they are surrounded with less media attention, the risk for an alleged scandal is fairly low.

Local celebrities are less “hyped”, and therefore less likely to over shine the prod-uct, otherwise known as the vampire effect. Consumers will most likely not simply re-call the local celebrity and not the product or brand.

Local celebrities will be easier to employ due to their lower demand and their wil-lingness of visibility and attention.

Last but not least we remind readers that local celebrity endorsers carry a much lower price tag, and conclude with the following:

There are celebrities who will get out of bed for less than $10,000 a day…


49

References

Agrawal, J., & Kamakura, W. A. (1995). The economic worth of celebrity endorsers: An event study analysis. Journal of Marketing, 59(3), 56-62.

Ainsworth, A. B. (2007). Public Information and Consumer Skepticism Effects on Celebri-ty Endorsements: Studies among Young Consumers. Journal of Marketing Communications, 13(2), 85-107.

Almquist, E, & Roberts, J. K. (2000). A “mindshare” manifesto: Common misconceptions squander the power of the modern brand. Mercer Management Journal.

Andersen, R. J. (1983). A Spreading Activation Theory of Memory. Journal of Verbal Learn-ing and Verbal Behavior, 22(3), 261- 295.

Ashforth, B. E., & Gibbs, B. W. (1990). The double-edge of organizational legitimization. Organization Science, 1(2), 177-194.

Atkin, C., & Block, M. (1983). Effectiveness of Celebrity Endorsers. Journal of Advertising Research, 23(1), 57-61.

Bland, J. M., & Altman, D. G. (1997, February 22). Statistics notes: Cronbach's alpha. Retrieved 2010-05-09, from http://www.bmj.com/cgi/content/full/314/7080/572

Bryman, A., & Cramer, D. (1997). Quantitative data analysis with SPSS for windows; a guide for so-cial scientists. London: Routledge.

Cialdini, R.B. (2001). Harnessing the Science of Persuasion. Harvard Business Review, 79(9), 72-79. Cooper, C. A., & Gimeno-Gascon, F. J. (1994). Initial human and financial capital as pre-dictors of new venture performance. Journal of Business Venturing, 9(5), 371-395.

Corston, R., & Colman, A. (2003). A Crash Course in SPSS (2nd ed.). Oxford: Blackwell Publishing.

Dawes, J. (2008). Do Data Characteristics Change According to the number of scale points used? An experiment using 5-point, 7-point and 10-point scales. International Journal of Market Research, 50 (1), 61–77.

Delmar, F., & Shane, S. (2004). Legitimating first: organizing activities and the survival of new ventures. Journal of Business Venturing, 19(3), 385-410.

Dowling, J., & Pfeffer, J. (1975). Organizational Legitimacy: Social Values and Organiza-tional Behaviour. The Pacific Sociological Review, 18(1), 122-136.

Drogan, J. (2006). Another Look at Communication Effectiveness. Retrieved 2010-05-23 from http://jmsdrgn.squarespace.com/storage/another%20look%20at%20communication%20effectiveness.pdf

Easterby-Smith, M., Thorpe, R., Jackson, P. & Lowe, A. (2008). Management Research (3rd ed.). London: Sage Publication Ltd.


50

Erdogan, B. (1999). Celebrity Endorsement: A Literature Review. Journal of Marketing Man-agement, 15, 291-314.

Evans, B. R. (1988). Production and Creativity in Advertising. London: Pitman Publishing

Farrel, A. & Van Riper, T. (2008). Tiger‟s next trophy: Billionair. Retrieved 2010-04-17, from http://www.forbes.com/2008/07/08/tiger-woods-billionaires-biz-sports-cx_af_tvr_0709tigerbillie.html

Farrell, K.A., Karels G.V., Montfort, K.W., & McClatchey, C.A. (2000). Celebrity Perfor-mance and Endorsement Value: The Case of Tiger Woods. Manegerial Finance, 26(7), 1-15.

Forehand, R., & Perkins, M. (2005). Implicit Assimilation and Explicit Contrast: A Set/Reset Model of Response to Celebrity Voice-Overs. Journal of Consumer Research, 32.

Friedman, H., & Friedman L. (1979). Endorser Effectiveness by Product Type. Journal of Advertising Research, 19(5), 63-71.

Ghauri, P., Grønhaug, K. (2005). Research Methods in Business Studies: A practical Guide (3rd ed.), United Kingdom, Prentice Hall/Financial Times

Glynn, M. S., Motion, J., & Brodie, R. J. (2007). Sources of brand benefits in manufacturer-reseller B2B relationships. Journal of Business & Industrial Marketing, 22(6), 400-409.

Gregory, S. (2009, September 16). Tiger Woods' Sponsors: Will Any Stick by Him? Retrieved 2010-04-17, from http://www.time.com/time/nation/article/0,8599,1948181,00.html

Hovland, C., Janis, I., & Kelly, H. (1953). Communication and persuasion: psychological studies of opinion change. New Haven, CT: Yale University Press.

Hovland, C., & Weiss, W. (1951). The Influence of Source Credibility on Communication Effectiveness. Public Opinion Quarterly, 15 (4), 635-650.

Hunter, E. (2009). Celebrity entrepreneurship and celebrity endorsement: Similarities, differences and the effect of deeper engagement. Jönköping: ARK tryckaren AB.

Hunter, E., Burger, H., & Davidsson, P. (2008). Celebrity capital as a strategic asset; Impli-cations for New Venture Strategies. In G. T. Lumpkin & J. Katz (Eds.), Advances in Entre-preneurship, Firm Emergence, and Growth, 11, 137–160, Emerald group publishing limited.

Jackson, K.T. (2004). Building reputational capital: Strategies for integrity and fair play that improve the bottom line. New York: Oxford university press.

Jacobsen, D.I. (2002.) “Vad, hur varför? Om metodval i företagsekonomi och andra samhällsveten-skapliga ämnen”. Lund: Studentlitteratur AB.

Johnson, L. A. (2005, July 18). New celebrity drug ads have subtle twist. Retrieved 2010-04-13, from http://www.msnbc.msn.com/id/8563655/ Kahle, L., R., & Homer, P., M. (1985). Physical Attractiveness of Celebrity Endorser: A So-cial Adaption Perspective. The Journal of Consumer Research, 11 (4), 954-961.


51

Kamen, J.M., Azhari, A.C., & Kragh, J. R. (1975). What a spokesman does for a sponsor. Journal of Advertising Research, 15 (2), 17-24.

Kamins, M.A., (1990) An investigation into the “match-up” hypothesis in context. Journal of Advertisement Research, June/July, 34-42

Knott, C., & James, M. (2004). An alternate approach to developing a total celebrity en-dorser rating model using the analytic hierarchy process. International Transactions in Opera-tional Research, 11(1), 87-95.

Lane, R. (1996). Nice guys finish first. Forbes, 158(14), 236-243.

Likert, R. (1974). The method of constructing an attitude scale. In G., M., Maranell (Ed.), Scaling, a Sourcebook for Behavioral Scientists, (p. 233-235). Chicago: Aldine Pub. Co.

Louie T. A., Kulik R. L., & Jacobson R. (2001). When Bad Things Happen to the Endors-ers of Good Products. Marketing Letters, 12(1), 13-23.

Maltby, J., Day, L., McCutcheon, L. E., Martin, M. M., & Cayanus, J. L. (2004). Celebrity worship, cognitive flexibility, and social complexity, Personality and Individual Differences. Personality and Individual Differences, 37(7), 1475-1482. Marchand, R. (1985). Advertising the American dream: Making way for modernity, 1920–1940. Berkeley, CA: University of California Press. McCracken, G. (1989). Who is the Celebrity Endorser? Cultural Foundations of the En-dorsement Process. The Journal of Consumer Research, 16(3), 310-321. McGrath, E. J., (1981). Dilemmatics: The Study of Research Choices and Dilemmas. The American Behavioral Scientist, 25(2), 179-210.

McGuire, W. (1985). Attitudes and Attitude Change. In G. Lindzey & E. Aronson (Eds.), Handbook of Social Psychology, 2, 233-346. New York: Random House. Mehta, A. (1994). How Advertising Response Modeling (ARM) can Increase Ad Effective-ness. Journal of Advertising Research, 34(3), 62-74.

Merriam-Webster Online Dictionary (2010). Merriam-Webster Online. Retrieved 2010-05-06, from http://www.merriam-webster.com/dictionary/endorser

Merriam-Webster Online Dictionary (2010). Merriam-Webster Online. Retrieved 2010-02-24, from http://www.merriam-webster.com/dictionary/endorsement Misra, S. & Beatty, S.E. (1990). Celebrity spokesperson and brand congruence: An assess-ment of recall and affect. Journal of Business Research, 21, 159-173

Ohanian, R. (1990). Construction and Validation of a Scale to Measure Celebrity Endors-ers' Perceived Expertise, Trustworthiness, and Attractiveness. Journal of Advertisement, 19 (3), 39-52.

O Mahony, S., & Meenaghan, T. (1998). The impact of celebrity endorsement on consum-ers. Irish marketing Review, 10(2), 15, 10.


52

NPD Group/Celebrity Influence Survey (2006). Retrieved 2010-05-03, from http://www.npdinsights.com/archives/may2006/cover.story.html Oxford Pocket Dictionary of Current English (2009). Encyclopedia.com. Retrieved on 2010-02-24, from http://www.encyclopedia.com/doc/1O999-celebrity.html

Pae, J.H., Samiee, S., & Tai,S. (2002). Global advertising strategy: The moderating role of brand familiarity and execution style. International Marketing Review, 19(2), 176-189.

Petrick, J. A., Scherer, R. F., Brodzinski, J. D., Quinn, J. F., & Ainina, M. F. (1999). Global leadership skills and reputational capital: Intangible resources for sustainable competitive advantage. Academy of Management Executive, 13(1), 58-69.

Rao, H. (1994). The social construction of reputation: Certification contests, legitimization, and the survival of organizations in the American automobile industry: 1895-1912. Strategic Management Journal, 15, 29-44.

Robson, C. (2002). Real world research (2nd ed.). United Kingdom: Blackwell Publishing.

Saunders, M., Lewis, P., & Thornhill, A. (2009). Research Methods for Business Students (5th ed.). Harlow: Pearson Education Limited.

Schaefer, P. (2006). The Seven Pitfalls of Business Failure. Retrieved 2010-05-06, from http://www.businessknowhow.com/startup/business-failure.htm

Shepherd, D. A. (1999). Venture capitalists' assessment of new venture survival. Management Science, 45(5), 621-632.

Silvera, D.H., Austad, B. (2004). Factors predicting the effectiveness of celebrity endorse-ment advertisements. European Journal of Marketing, 38(11/12), 1509-1526.

Solomon, M. (2009). Consumer Behavior: Buying, Having, and Being (8th ed.). Upper Saddle Riv-er: Pearson Education, Inc. /Prentice Hall.

Stinchombe A. L. (1965). Social Structure and Organizations, Advances in Strategic Manage-ment, 17, 229-259. Stuart, T. E., Hoang, H., & Hybels, R. C. (1999). Interorganizational endorsements and the performance of entrepreneurial ventures. Administrative Science Quarterly, 44(2), 315-349.

Suh, T., & Amine, L. (2007). Defining and managing reputational capital in global markets. Journal of Marketing Theory and Practice, 15(3), 205-217.

Supermodelguide.com – Super model Linda Evangelista. Retrieved 2010-04-17, from http://supermodelguide.com/supermodels/linda.html

Supermodelguide.com – Super model Christy Turlington. Retrieved 2010-04-17, from http://supermodelguide.com/supermodels/christy.html

Székely, I., & Sabot, R. (1997). Development strategy and management of the market economy. Ox-ford: Claredon press Oxford.


53

Talmazan, Y. (2009, December 10). Who Will Dump Tiger Woods? List of Endorsements, Lost and Kept. Retrieved 2010-04-17, from http://www.nowpublic.com/sports/who-will-dump-tiger-woods-list-endorsements-lost-and-kept-2535740.html

Teece, D. (1986). Profiting from technological innovation: implications for integration, collaboration, li-censing, and public policy. North-Holland: Elsevier Science Publishers B.V.

Till, B. D., & Busler, M. (1998). Matching products with endorsers: attractiveness versus expertise. Journal of Consumer Marketing, 15(6), 576-586.

Till, B., & Shimp, T. A. (1998). Endorsers in advertising: The case of negative celebrity in-formation. Journal of Advertising. 27(1), 67-82.

Tom, G., Clark, R., Elmer, L., Grech, E., Masetti, J. Jr., & Sandhar, H. (1992). The Use of Created Versus Celebrity Spokespersons in Advertisements. The Journal of Consumer Market-ing, 9(4), 45-51.

Towle. A. P., (2003, November 18). Celebrity Branding. Retrieved 2010-05-06, from http://www.hollywoodreporter.com/hr/search/article_display.jsp?vnu_content_id=2030984 Trout, J. (2009, July 13). Celebs Who Un-Sell Products. Retrieved 2010-04-17, from http://www.forbes.com/2007/09/12/jack-trout-marketing-celebs-oped-cx_jt_0913trout.html West, C. (1999). Marketing research. Hampshire: Macmillan Press

Yiu, D. W., & Lau, C.-M. (2008). Corporate entrepreneurship as resource capital configura-tion in emerging market firms. Entrepreneurship Theory and Practice, 32(1), 37-57.

Zimmerman M, A & Zeitz, G, J (2002). Beyond Survival: Achieving New Venture Growth by Building Legitimacy. The Academy of Management Review, 27(3), 414-431.


54

Appendix 1: Source Credibility Scale (Ohanian, 1990)


55

Appendix 2: Survey Introduction and Questionnaire (Swedish)

Internationella Handelshögskolan i Jönköping Avdelningen för EMM

Marknadsföring och Kommunikation Frågeformulär

På följande sidor ber vi dig ange lite bakgrundsinformation om dig själv och besvara frågor som gäller din attityd till, och åsikt om reklamen som du snart kommer att få se. Var god läs all information i det här frågeformuläret. Efter att ha läst varje punkt noggrant, besvara frågan genom att ringa in den siffra du anser passar bäst in på dig. Till exempel, om ett på-stående lyder: ”Jag tycker om jordgubbar”, men att du är i princip likgiltig inför smaken av jordgubbar, besvarar du frågan genom att rita en cirkel runt siffran 4. Instämmer 1 2 3 4 5 6 7 Instämmer inte alls helt

Även om några av frågorna verkar vara nästan identiska är alla lika viktiga, därför ber vi dig besvara alla frågor noggrant.

Tack för din medverkan!


56

Frågeformulär

1. Med utgångspunkt från annonsen tycker du att [namn på kändis] är (rita en cirkel

runt lämplig siffra för varje ordpar):

Opålitlig 1 2 3 4 5 6 7 Pålitlig

Oärlig 1 2 3 4 5 6 7 Ärlig

Ej tillförlitlig 1 2 3 4 5 6 7 Tillförlitlig

Ej uppriktig 1 2 3 4 5 6 7 Uppriktig

Icke trovärdig 1 2 3 4 5 6 7 Trovärdig

2. I relation till produkten anser du att [namn på kändis] är:

Icke expert 1 2 3 4 5 6 7 Expert

Oerfaren 1 2 3 4 5 6 7 Erfaren

Okunnig 1 2 3 4 5 6 7 Kunnig

Okvalificerad 1 2 3 4 5 6 7 Kvalificerad

Oskicklig 1 2 3 4 5 6 7 Skicklig

3. Enligt dig är [namn på kändis]:

Osympatisk 1 2 3 4 5 6 7 Sympatisk

Obekant 1 2 3 4 5 6 7 Bekant

Olik dig 1 2 3 4 5 6 7 Lik dig

4. Generellt sett, hur effektiv anser du att annonsen för [namn på företaget] är?

Extremt 1 2 3 4 5 6 7 Extremt ineffektiv effektiv

5. Vad är sannolikheten att du kommer att köpa [produkten] från [namn på företaget]

till dig själv under det kommande året?

Kommer 1 2 3 4 5 6 7 Kommer definitivt definitivt inte att köpa att köpa


57

6. Anser du att det känns naturligt att sammankoppla/associera [produkten] med

[namn på kändis]?

Instämmer inte 1 2 3 4 5 6 7 Instämmer alls helt

7. [Namn på kändis] image påverkar din syn på produkten:

Instämmer 1 2 3 4 5 6 7 Instämmer inte alls helt

8. Enligt dig, är [namn på kändis] entusiastisk till att representera [namn företaget]s

[produkt]?


9. Det verkar som att [namn på kändis] är passionerad om att representera [namn på

företaget och produkt]:


10. [Namn på kändis] verkar vara engagerad i sin representation av [namn på företaget

och produkt]:


11. Det verkar som att [namn på kändis] ofta använder [namn på företaget och pro-

dukt]:


12. [Namn på kändis] gillar [namn på företaget och produkt]:



58

13. [Namn på kändis] känns exalterad över att representera [namn på företaget och

produkt]:


14. Vad tror du syftet är med det här experimentet?

□ Jag vet inte Om du tror att du vet syftet, skriv ned ditt svar här:


59

Appendix 3: Survey Introduction and Questionnaire (English)

Jönköping International Business School Department for EMM

Marketing and Communications Questionnaire

In the following pages, we ask you to fill out some background information about yourself. We ask you to answer questions regarding your attitude and opinion regarding the adver-tisement you are about to see. Please read all the information in this questionnaire. After reading each item carefully, answer the question by circling the number you believe suits you best. For example, if a statement reads:”I like strawberries”, but you are basically indif-ferent to the taste of strawberries, you answer the question by circling the number 4. Completely 1 2 3 4 5 6 7 Completely do not agree agree

Even if some of the questions seem almost identical, we ask you to answer all questions carefully.

Thank you for your participation!


60

Questionnaire

1. With the advertisement in mind, do you consider that [name of celebrity] is (circle

the number that suits you best):

Not dependable 1 2 3 4 5 6 7 Dependable

Dishonest 1 2 3 4 5 6 7 Honest

Unreliable 1 2 3 4 5 6 7 Reliable

Insincere 1 2 3 4 5 6 7 Sincere

Untrustworthy 1 2 3 4 5 6 7 Trustworthy

2. In relation to the product, do you consider [name of celebrity] to be:

Not expert 1 2 3 4 5 6 7 Expert

Inexperienced 1 2 3 4 5 6 7 Experienced

Unknowledgeable 1 2 3 4 5 6 7 Knowledgeable

Unqualified 1 2 3 4 5 6 7 Qualified

Unskilled 1 2 3 4 5 6 7 Skilled

3. According to you, [name of celebrity] is:

Unlikeable 1 2 3 4 5 6 7 Likeable

Unfamiliar 1 2 3 4 5 6 7 Familiar

Dissimilar 1 2 3 4 5 6 7 Similar

4. Generally, how effective do you consider the advertisement for [name of company]

to be?

Extremely 1 2 3 4 5 6 7 Extremely ineffective effective

5. What is the possibility that you will purchase [the product] from [name of compa-

ny] for yourself, in the upcoming year?

Will 1 2 3 4 5 6 7 Will definitively definitively not purchase purchase


61

6. Do you consider it natural to connect/associate [the product] with [name of celebr-

ity]?

Completely 1 2 3 4 5 6 7 Completely do not agree agree

7. The image of [name of celebrity] influences your view of the product:


8. According to you, is [name of celebrity] enthusiastic about [name of company]‟s

[product]?


9. It seems like [name of celebrity] is passionate about endorsing [name of company

and product]:


10. [Name of celebrity] seems to show dedication in endorsing [name of company and

product]:


11. It seems like [name of celebrity] often uses [name of celebrity and product]:


12. [Name of celebrity] seems to like [name of company and product]:



62

13. [Name of celebrity] seems to be thrilled about endorsing [name of celebrity and

product]:


14. What do you think is the purpose of this experiment?

□ I do not know If you think you know what the purpose is, write your answer here:


63

Appendix 4: Advertisement Leila Lindholm (Including English example)

Bakgrundsinformation

Kön (kryssa i en ruta): □ Man □ Kvinna Ålder: ____

Är Svenska ditt modersmål? □ Ja eller □ Nej

Bakgrund

Baka&Smaka är ett nystartat svenskt företag som använder sig av TV-kocken Leila Lindholm som sitt ansikte utåt. Leila är känd från TV, där hon medverkat i program såsom Leila Bakar, Leilas Mat, Leila på Landet och Leilas Jul. Nu uppträder hon alltså snart även i tryckt reklam, TV-reklam, radioreklam och webbreklam för företaget Baka&Smaka.

Reklamen som ni snart ska få se handlar om företaget Baka&Smaka. De har specialiserat sig på försäljning av designade bakredskap som Sveriges alla matlagare kan använda hemma i köket. I sortimentet finns bland annat vispar, slickepottar, bakpenslar och måttverktyg i diverse färger och snygga mönster. Bilden i annonsen föreställer kändisen Leila Lindholm. Syftet med det här experimentet är att utvärdera hur effektiv Baka&Smakas eftersträvade marknadsföringsstrategin är.

Enligt hemsidan www.bakaochsmaka.com, är målgruppen män och kvinnor i alla åldrar. Sortimentet är tänkt att lanseras i Sverige i juni 2010.

Vänligen, ta en stund på dig och studera annonsen noggrant.

Background information

Sex (check one box) □ Male □ Female Age: ____

Is Swedish your mother tongue? □ Yes or □ No

Bakgrund

Baka&Smaka is a newly started Swedish company that uses the TV-chef Leila Lindholm as their celebrity endorser. Leila is known from TV, where she has participated in shows such as Leila Bakar, Leilas Mat, Leila på Landet and Leilas Jul. Now she will soon appear in print ads, TV-ads, radio ads and Internet ads for the company Baka&Smaka.

The advertisement that you are about to see concerns the company Baka&Smaka. They have specialized in selling designed baking utensils, that all of Swedish home cookers can use in their own kitchens. The assortment includes whisks, spatulas and measuring cups among others, in lively patterns and colors. The picture in the advertisement features the celebrity Leila Lindholm. The purpose of this experiment is to evaluate how effective Ba-ka&Smaka‟s marketing strategy is.

According to the webisite www.bakaochsmaka.com, the target group is men and women in all ages. The assortment is meant to be released in Sweden in June, 2010.

Please, take a moment and study the advertisement carefully.


64

3

3 Picture: Leila Lindholm, retrieved 2010-03-16, from http://marianova.com/mat/mat-leila-jag-slapper-inte-in-ovantat-besok


65

Appendix 5: Advertisement Rachael Ray




Bakgrund

Baka&Smaka är ett nystartat svenskt företag som använder sig av den amerikanska TV-kocken Rachael Ray som sitt ansikte utåt. Rachael Ray är känd från amerikansk TV, där hon medverkat i program såsom Tasty Travels, $40 A Day, Inside Dish och 30-Minute Meals. Nu uppträder hon alltså snart även i tryckt reklam, TV-reklam, radioreklam och webbre-klam för företaget Baka&Smaka.

Reklamen som ni snart ska få se handlar om företaget Baka&Smaka. De har specialiserat sig på försäljning av designade bakredskap som Sveriges alla matlagare kan använda hemma i köket. I sortimentet finns bland annat vispar, slickepottar, bakpenslar och måttverktyg i diverse färger och snygga mönster. Bilden i annonsen föreställer kändisen Rachael Ray. Syf-tet med den här studien är att utvärdera hur effektiv den eftersträvade marknadsföringsstra-tegin är.

Enligt hemsidan www.bakaochsmaka.com, är målgruppen män och kvinnor i alla åldrar. Sortimentet är tänkt att lanseras i Sverige i juni 2010.



66

4

4 Picture: Rachael Ray, retrieved 2010-0318, from http://mycrains.crainsnewyork.com/100women/view/78


67

Appendix 6: Advertisement Niclas Adler




Bakgrund

Uni-Lumos är ett nystartat studentföretag från Internationella Handels Högskolan i Jönkö-ping som använder sig av IHH‟s VD Niclas Adler som sitt ansikte utåt. Niclas Adler är allt-så främst känd som skolans VD, men han har även ett förflutet som framgångsrik entre-prenör. Nu uppträder han alltså snart i reklam för företaget Uni-Lumos på Jibsnet och www.ihh.hj.se, i Junited och på affischer runt om skolan.

Reklamen som ni snart ska få se handlar om studentföretaget Uni-Lumos. De har låtit pro-ducera smarta läslampor i diverse snygga färger som snart finns för försäljning. Bilden i an-nonsen föreställer kändisen Niclas Adler. Syftet med den här studien är att utvärdera hur effektiv den eftersträvade marknadsföringsstrategin är.

Enligt hemsidan www.uni-lumos.com, är målgruppen studenter i alla åldrar. Sortimentet är tänkt att lanseras på högskolan i september 2010.



68

5

5 Picture: Niclas Adler, retrieved 2010-03-24, from http://www.hj.se/doc/8794


69

Appendix 7: Advertisement Jan Björklund




Bakgrund

Uni-Lumos är ett nystartat studentföretag från Internationella Handels Högskolan i Jönkö-ping som använder sig av Sveriges utbildningsminister Jan Björklund som sitt ansikte utåt. Jan Björklund har varit utbildningsminister sedan 2007 men är också känd sedan tidigare som kommunalpolitiker och skolminister. Nu är han även partiordförande. Snart uppträder han alltså i reklam för företaget Uni-Lumos på Jibsnet och www.ihh.hj.se, i Junited och på affischer runt om skolan.

Reklamen som ni snart ska få se handlar om studentföretaget Uni-Lumos. De har låtit pro-ducera smarta läslampor i diverse snygga färger som snart finns för försäljning. Bilden i an-nonsen föreställer kändisen Jan Björklund. Syftet med den här studien är att utvärdera hur effektiv den eftersträvade marknadsföringsstrategin är.

Enligt hemsidan www.uni-lumos.com, är målgruppen studenter i alla åldrar. Sortimentet är tänkt att lanseras på högskolan i september 2010.



70

6

6 Picture: Jan Björklund, retrieved 2010-03-16, from http://www.google.com/imgres?imgurl=http://sp.marksgymnasieskola.se/inriktningar/samhaellsvetens-kap/Samhall/edce44ce25e270c5.jpg&imgrefurl=http://sp.marksgymnasieskola.se/inriktningar/samhaellsvetenskap/Samhall/halloffame.htm&usg=__RbxExyKiHrvIU9JZxyoXK7TwCdc=&h=400&w=317&sz=23&hl=en&start=5&um=1&itbs=1&tbnid=qbueRewzah2w0M:&tbnh=124&tbnw=98&prev=/images%3Fq%3DJan%2BBj%25C3%25B6rklund%26um%3D1%26hl%3Den%26sa%3DN%26tbs%3Disch:1


71

Appendix 8: Advertisement Daniel Karlsson




Bakgrund

Cérvix är ett nystartat svenskt företag som använder sig av idolstjärnan Daniel Karlsson från Jönköping som sitt ansikte utåt. Daniel Karlsson är känd från Idol 2007, där han kom på fjärde plats i finalen. I oktober 2008 släppte Daniel Karlsson sin första singel Would you believe? som också gjordes till musikvideo. Nu uppträder han alltså snart i tryckt reklam, TV-reklam, radioreklam och webbreklam för företaget Cérvix.

Reklamen som ni snart ska få se handlar om företaget Cérvix. De har specialiserat sig på tillverkning av välsmakande halstabletter som är extra skonsamma för halsen. Bilden i an-nonsen föreställer kändisen Daniel Karlsson. Syftet med den här studien är att utvärdera hur effektiv den eftersträvade marknadsföringsstrategin är.

Enligt hemsidan www.cervix.com, är målgruppen män och kvinnor i alla åldrar. Sortimen-tet är först och främst tänkt att lanseras i Jönköping i juni 2010.



72

7

7 Picture: Daniel Karlsson, retrieved 2010-03-17, from http://www.salongk.se/artiklar/import/2007/10/4/2244/index.xml


73

Appendix 9: Advertisement James Blunt




Bakgrund

Cérvix är ett nystartat svenskt företag som använder sig av världsstjärnan James Blunt från England som sitt ansikte utåt. James Blunt slog igenom 2004 med sitt debutalbum Back to Bedlam och speciellt låten You're Beautiful. Nu uppträder han alltså snart i tryckt reklam, TV-reklam, radioreklam och webbreklam för företaget Cérvix.

Reklamen som ni snart ska få se handlar om företaget Cérvix. De har specialiserat sig på tillverkning av välsmakande halstabletter som är extra skonsamma för halsen. Bilden i an-nonsen föreställer kändisen James Blunt. Syftet med den här studien är att utvärdera hur ef-fektiv den eftersträvade marknadsföringsstrategin är.

Enligt hemsidan www.cervix.com, är målgruppen män och kvinnor i alla åldrar. Sortimen-tet är först och främst tänkt att lanseras i Jönköping i juni 2010.



74

8

8 Picture: James Blunt, retrieved 2010-0316, from http://becauseican.co.za/2008/05/28/


75

Appendix 10: Advertisement Nikita Johnson (Local)




Bakgrund

Touché är ett nystartat svenskt företag som använder sig av den kända badmintonspelaren Nikita Johnson, ursprungligen från Helsingborg, som sitt ansikte utåt. Nikita har spelat badminton sedan barnsben och är regerande svensk mästare sedan fem år tillbaka. Nu upp-träder hon alltså snart även i tryckt reklam, TV-reklam, radioreklam och webbreklam för företaget Touché.

Reklamen som ni snart ska få se handlar om företaget Touché. De har specialiserat sig på tillverkning av rakhyvlar för extra skonsam rakning. De gör hyvlar för både män och kvin-nor som vill slippa skärsår och hudirritationer. Bilden i annonsen föreställer kändisen Niki-ta Johnson. Syftet med den här studien är att utvärdera hur effektiv Touchés eftersträvade marknadsföringsstrategin är.

Enligt hemsidan www.touche.com, är målgruppen män och kvinnor i alla åldrar. Sortimen-tet är tänkt att lanseras i Sverige i juni 2010.



76

9

9 Picture: Nikita Johnson, retrieved 2010-03-16, from http://pictures.com/s?q=long+legs#pg=3


77

Appendix 11: Advertisement Nikita Johnson (Well-known)




Bakgrund

Touché är ett nystartat svenskt företag som använder sig av den kända badmintonspelaren Nikita Johnson, från Australien, som sitt ansikte utåt. Nikita har spelat badminton sedan barnsben och är regerande världsmästare sedan fem år tillbaka. Nu uppträder hon alltså snart även i tryckt reklam, TV-reklam, radioreklam och webbreklam för företaget Touché.

Reklamen som ni snart ska få se handlar om företaget Touché. De har specialiserat sig på tillverkning av rakhyvlar för extra skonsam rakning. De gör hyvlar för både män och kvin-nor som vill slippa skärsår och hudirritationer. Bilden i annonsen föreställer kändisen Niki-ta Johnson. Syftet med den här studien är att utvärdera hur effektiv Touchés eftersträvade marknadsföringsstrategin är.

Enligt hemsidan www.touche.com, är målgruppen män och kvinnor i alla åldrar. Sortimen-tet är tänkt att lanseras i Sverige i juni 2010.



78

Local Celebrity Endorsement337633/FULLTEXT01.pdf · celebrity endorsement but lack sufficient financial capital, which conducts the purpose. Further, the research questions are discussed

Documents