Celebrity advertising in the case of negative ...

15 downloads 64 Views 96KB Size Report
discourse analysis of weblogs. Ulun Akturan. Galatasaray University, Istanbul, Turkey. Abstract. Purpose – The purpose of this paper is to examine celebrity ...
The current issue and full text archive of this journal is available at www.emeraldinsight.com/2040-8269.htm

MRR 34,12

Celebrity advertising in the case of negative associations: discourse analysis of weblogs

1280

Ulun Akturan Galatasaray University, Istanbul, Turkey Abstract Purpose – The purpose of this paper is to examine celebrity advertising in the case of negative associations. Design/methodology/approach – In total, 451 posts were captured permanently from randomly selected weblogs, and analyzed by discourse analysis. The basic emerging themes therein have been isolated and interpreted, and a model of celebrity endorsement in the case of negative information has been developed. Findings – When a celebrity becomes involved in an undesirable event, the consumers’ perceptions of the celebrity may or may not change. This study argues that this situation is influenced by “the level of negativity”, “the level of blameworthiness”, “admiration” and the “message content”. In addition, it was found that the admiration affects the perceived image of the celebrity and it is also affected by “the level of negativity” and “the level of blameworthiness”. Furthermore, the message given is directly associated with the “negativity”, “blameworthiness”, “perception of the celebrity”, and “the perception of the brand”. Research limitations/implications – The present study examined only one case of celebrity advertising. Practical implications – The expectations and attitudes of the identified segments should be taken into consideration when firms develop marketing programs. Originality/value – This study, unlike others, analyzes the case of the continuance of the endorsement relationship between the company and the negatively publicized celebrity. Keywords Advertising, Celebrities, Celebrity advertising, Celebrity endorsement, Discourse analysis, Negative associations, Web sites Paper type Research paper

Management Research Review Vol. 34 No. 12, 2011 pp. 1280-1295 q Emerald Group Publishing Limited 2040-8269 DOI 10.1108/01409171111186405

Introduction Companies invest billions of dollars in paying celebrities for their endorsements in advertising campaigns, recognizing they are powerful mechanisms in transferring meanings to their brands. Celebrity endorsers typically have a symbolic aspired to reference group association (Assael, 1984) and act as cues in peripheral routes to persuasion (Petty and Cacioppo, 1985). Famous endorsers are exploited mainly for two reasons. First, they are viewed as dynamic, attractive and likeable. Second, their fame attracts attention to the product (Atkin and Block, 1983). From a theoretical perspective, the meaning attributed to the celebrities moves from the celebrity endorser to the product or brand. In the consumer’s mind, the associations related to the celebrity become associated with the brand. When the consumer The author thanks Galatasaray University Scientific Research Projects Commission for the support provided (Research Project No. 10.102.002).

purchases the product, and therefore acquires the meaning of the brand, the meaning transfer process is completed (McCracken, 1989). In this process, it is paramount to create positive and favorable associations in the mind of the consumer. However, over time the endorsing celebrity can be involved in an undesirable event which shifts public attention and interest in a polar direction, producing negative associations for the celebrity (Gupta, 2009). The fame of the celebrity and the event itself, tied together in something called a scandal, can be widely publicized (e.g. O.J. Simpson, Mike Tyson, Madonna, Michael Jackson, Tiger Woods, etc.). O.J. Simpson was charged with murder but acquitted. More than a decade later he was arrested and charged with numerous felonies, including armed robbery and kidnapping. Mike Tyson was convicted of sexually assaulting a woman, and served three years in prison. Michael Jackson was accused of child sexual abuse. The case was settled out of court and no formal charges were brought. Later new charges were made, and he was acquitted in a lengthy jury trial. Madonna released her music video, Like a Prayer, which is about a young girl struggling with feelings of guilt and sexuality. It contains images of burning crosses, stigmata, romantic love with a priest or saint, and thus was highly criticized in the media. Multiple researches have revealed that “people have a ‘negativity bias’ in evaluating individuals’ moral behaviors” (Gupta, 2009, p. 2). Furthermore, because there is a high exposure in the media and people weigh negative behaviors more heavily than positive ones, negative associations become more and more critical in the continuity of celebrity endorsement. Companies can overnight choose to terminate their contracts with the celebrity who engages in “bad deeds”, both to prevent the transfer of the negative associations to the brand and second to show disapproval for the behavior. Recently, there has been an extraordinary case in which a scandal rather than ending the endorsement contract, instead dramatically continued it with the airing of a new purposively created and dedicated “self-conscious” ad. The Nike and Tiger Woods reunion, following the globally damaging revelations about his private issues, became fully visible when on the eve of the Masters Nike released a brand new Woods ad which marked the golfer’s return to the sport. This study aims to examine both consumers’ reactions and attitudes in the case of the negative associations. The study differs from other studies in examining what is perceived by the consumers when the scandal, rather than ending endorsement relationship is used instead to call attention to the brand. Literature review A celebrity is an individual “whose name has attention-getting, interest-riveting and profit generating value that stems from the high level of public attention and interest” (Gupta, 2009, p. 2). The use of celebrities as spokespeople for brands dates back to the nineteenth century when Queen Victoria was associated with Cadbury’s Cocoa (Erdogan, 1999). In the literature, there are a number of studies that have examined the degree and conditions under which celebrities make appropriate endorsers for products (Atkin and Block, 1983; Freiden, 1984; Kamins, 1989; Kamins et al., 1989; Ohanian, 1991; Tripp et al., 1994; Agrawal and Kamakura, 1995). It was found that the appropriate use of celebrity endorsers can be highly effective in product promotions (Mitchell and Olson, 1981; Atkin and Block, 1983; Misra and Beatty, 1990). In marketing communication, utilization of celebrity endorsement has risen over the years (Erdogan, 1999). A company can accrue several benefits from celebrity endorsement,

Celebrity advertising

1281

MRR 34,12

1282

but at the same time incurs some costs and risks. Table I displays the potential advantages and disadvantages of the celebrity endorsement strategy, and the available preventive tactics (Erdogan, 1999). In the literature, there are numerous studies exploring the issue of the impact of negative information related to the endorser of a product or brand. Till and Shimp (1998) conducted three comparative studies of a real vs a fictitious endorser who both have negative associations. They found negative information about both resulted in lowered evaluations of the brand, but that for the real celebrity endorser the impact was weaker than it was for the fictitious celebrity endorser. They argued this difference could have resulted from the fact of the target subjects having a richer association set for the real celebrity. Gupta (2009) studied the public reaction to Michael Vick immediately after his involvement with dog fighting became known. Gupta found from his qualitative netnographic research that “respondents who found Vick blameworthy attributed his negative action mostly to causal locus closely followed by the causal controllability dimension” (Gupta, 2009, p. 7). White et al. (2009) empirically tested the impact of negative information about a celebrity spokesperson on the consumers’ perceptions of the endorsed brand. They found that when respondents are exposed to negative information about a celebrity endorser, a negative transference of effect in the endorsement relationship will occur. However, they also found that in the reverse situation the transference of effect is less severe. Louie and Obermiller (1998) found that the company image was perceived to benefit in any case, were they either to retain an endorser who had low blame for the negative event or to dismiss one who had high blame for the same negative event. Tripp et al. (1994) found that when the celebrity with a negative perception endorsed multiple products, the celebrity’s credibility and likeability decreased, which then negatively impacted attitudes towards the specific advertising. Louie et al. (2001) revealed that stock prices declined in the case of negative publicity about a celebrity who endorsed one of the company’s brands. Potential advantages

Potential hazards

Preventive tactics

Table I. Potential advantages, hazards and preventive tactics of the celebrity endorsement strategy

Increased attention Image polishing Brand introduction Brand repositioning Underpin global campaigns Overshadow the brand Public controversy Image change and overexposure Image change and loss of public recognition Expensive Pre-testing and careful planning Buying insurance and putting provision clauses in contracts Explaining what their role is and putting clauses to restrict endorsements for other brands Examining what life-cycle stage the celebrity is in and how long this stage is likely to continue Selecting celebrities who are appropriate for the global target audience, not because they are “hot” in all market audiences

Source: Erdogan (1999, p. 295)

The case of Tiger Woods The case of Tiger Woods and Nike’s dramatic reunion is often commented on in a speculative tone, but yet to be carefully analyzed. Tiger Woods is an American professional golfer who is ranked as the most successful golfer of all time. He was the highest paid professional athlete in 2008, having earned an estimated $110 million from winnings and endorsements (Badenhausen, 2009). He has also been the world’s most marketable athlete; in 1996 he began signing endorsement deals with numerous companies. During the years 1996-2009 Woods signed contracts with General Motors, Titleist, General Mills, American Express, Accenture, Buick, TAG Heuer, Electronic Arts, Gillette, Gatorade and Nike. In 2000 he signed a five-year $105 million contract extension with Nike. Nike Golf is one of the fastest growing brands in the sport, with an estimated $600 million in annual sales. Woods has been described as the “ultimate endorser” for Nike Golf, frequently seen wearing Nike gear during tournaments. He even appears in advertisements for other products wearing Nike products. Woods receives a cut from the sales of Nike Golf apparel, footwear, golf equipment, golf balls, and has a building named after him at Nike’s headquarters campus in Oregon. In 2009, Forbes identified Woods as the world’s first athlete to earn over a billion dollars (before taxes) in his career, and estimated his net worth to be $600 million (Badenhausen, 2009; Miller, 2009). Then suddenly, in November 2009, news surfaced in the media about his multiple marital infidelities. On December 11, 2009, Woods admitted the infidelity and announced he would take an indefinite leave from professional golf to focus on his marriage. Woods returned to competition for the 2010 Masters on April 8, 2010, after a break lasting 20 weeks. After the scandal relating to Woods was massively publicized, most of the companies with which he had signed endorsement deals announced that they would no longer work with him. However, Nike announced they would continue to work with Woods, and aired a commercial that stirred controversy just as the Masters golf tournament began. It was Wood’s first TV ad since the scandal. The 30-second clip revealed him looking remorseful as he was being questioned by his father Earl Woods, who had died in 2006. In the commercial the voice of Earl Woods asks his son: Tiger, I am more prone to be inquisitive, to promote discussion. I want to find out what you are thinking was. I want to find out what your feelings are. Did you learn anything?

Woods is dressed in a Nike cap and vest, and looks directly into the camera as his late father talks to him. In this study, the consumer reactions relate to the continuance of the endorsement deal and the perception of the message of the commercial. The research questions are: RQ1. What are the reactions of consumers when the company keeps an endorsement deal with a negatively publicized celebrity? RQ2. What are the attitudes of consumers in relation to a commercial which has a message content concerning the negative situation? RQ3. How is the company’s image affected by the perception of the celebrity’s negative image?

Celebrity advertising

1283

MRR 34,12

1284

Research methodology Conversation is a fundamental tool that people use to construct their reality (Berger and Luckmann, 1991). Discourse is a key concept for anyone interested in the workings of construction processes, reality-maintenance and change. Discourse is “a set of meanings, metaphors, representations, images, stories and statements which together produce a particular version of the world” (Berglund and Johnson, 2007, p. 79). A discourse is a stretch of language that is typically longer than a single sentence. The texts used in discourse analysis may consist of formal written records such as news reports, company statements and reports, academic papers; transcripts of social interactions such as conversations, focus group discussions, and individual interviews, advertisements, magazines and novels (Elliot, 1996). The field of discourse analysis varies from the study of linguistic features in texts (Van Dijk, 1981) to a focus on the dominant themes in respondents’ discourse (Elliot, 1996). In addition to semantic or pragmatic units of sentence grammars, the theory of discourse has introduced new notions – the coherence of themes (Van Dijk, 1981). A theme “is a pattern found in the information that at minimum describes and organizes the possible observations and at maximum interprets aspects of the phenomenon” (Boyatzis, 1998, p. 4). The method involves coding the phrases and sentences of discourse into the three domains of behavior: the affective domain, the cognitive domain and the sensorimotor domain (Nahl, 2007). In that sense, discourse analysis is concerned with naturally occurring contextualized speech rather than linguistic organization per se (Elliot et al., 1995). In qualitative studies, coding is analysis. It involves the differentiation and the combination of the retrieved data and the reflections that the researcher makes (Miles and Huberman, 1994). In discourse analysis, the concern is to understand how a text (written or spoken) is constructed. The key task is to understand what the text is trying to achieve – assert, justify, accuse, defend or explain. The complimentary next task is to examine how the text achieves its aims through the words, phrases and propositions (Dick, 2004). In order to determine the reactions of consumers to the Woods-Nike reunion commercial, discourse analysis was used to search for the dominant themes, or what are often labeled macrostructures, in our data. Thematic analysis is a well-established technique commonly used in handling qualitative data. It is not simply just another qualitative method, but rather a process for objectively encoding qualitative information (Boyatzis, 1998). Looking for themes in written material involves analyzing and marking the text through repetitions, unfamiliar words, metaphors and analogies, transitions, similarities/differences, linguistic connectors, missing data, theory-related materials, cutting-sorting, word lists, word co-occurrence and metacoding (Bernard and Ryan, 2010). Among these, we focused in our analysis on repetitions, transitions, similarities/differences and cutting and sorting techniques to identity the themes, all of which was brought together in an evolving codebook which included structural codes, theme codes and memos. The themes articulated in this study were generated inductively from the large but finite collection of raw data. Data collection and sampling In the study, the data were collected from the blogosphere, a term in common usage referring to the universe of blogs and their links to each other. Blogging has become

a popular way for people all around the world to publish information. Bloggers write blog posts, share their likes and dislikes, voice their opinions, provide suggestions, report news and form groups in the blogosphere (Agarwal et al., 2008). There are roughly 1.6 million posts every day; 346 million people worldwide read blogs (Social Media White Paper, 2008). The collection of such data, taken as a kind of self-reporting from the web, and marked as documentary evidence, no doubt increases “the external validity and generalisability of attributional research findings beyond the experimental setting of the self-probe method or survey research common in literature” (Gupta, 2009, p. 6). Self-reporting on the internet and the communications within the cyberspace is mostly used by the ethnographic researches in which researchers “[. . .] adapt ethnographic research techniques to study the cultures and communities that are emerging through computer-mediated communications” (Kozinets, 2002, p. 62). The data can be directly copied from the communications among the virtual communities (Kozinets, 2002; Fox and Robert, 1999; Kanayama, 2003), and can be analyzed by content analysis, constant comparative techniques and discourse analysis (O’Reilly et al., 2007; Sandlin, 2007; Nguyen et al., 2006; Sneijder and te Molder, 2005). To collect data from the blogosphere blog search engines should be used. They are similar to web search engines. They automatically gather and store large quantities of information from the web, and offer a free-fast interface allowing the public to drill-down their databases. The main difference between the web search engines and blog search engines is that blog search engines mainly index blogs and ignore the rest of the web (Thelwall and Hasler, 2007). In this study, the Google Blog Search Engine was used. The commercial we are concerned with was broadcasted on April 7, 2010. The search was done with the keywords “Tiger Woods Nike ad” between the April 7 and April 10, 2010 and detected exclusively blogs written in English. The search results displayed 10,997 blogs, of which 50 were chosen by simple random sampling. These posts were downloaded, and as a result 451 posts were obtained for the data analysis. Once the raw data had been generated, it was essential to develop a codebook centered on the posts as the units of analysis. There were five fundamental steps involved in developing a code: (1) reducing the raw information by reading each unit of analysis again and again; (2) identifying themes within sub-samples; (3) comparing themes across sub-samples; (4) creating a code; and (5) determining the reliability of the code. During the process, the discursive analysis constantly moved from general to particulars, alternating the phases of coding, writing memos and connecting the postings. As a result, numerous disparate codes emerged. These gradually gave way to new similarities and were merged in order to make progress toward “conceptualization” (Sitz, 2008). The code development procedure is given in Table II in details. In that process of reading and comparing the constructs, it was detected that there were general patterns in different postings. Thereafter, once the key constructs were identified, the relationships and linkages among the constructs were drawn and displayed visually. This preliminary tentative model, shown in Figure 1, eventually formed the basis of the proposed model.

Celebrity advertising

1285

MRR 34,12

1286

Table II. Code/dominant theme development

Steps Reducing the raw information All the written posts were copied to a table formed in Excel and consecutively numbered All of the postings were read; while reading, the repetitions and metaphors were highlighted with different colors Identifying themes within sub- The related phrases were categorized samples The general initial categories were determined. General categories: Advertising, Tiger Woods, Nike and Behavioral Action (to buy or not to buy) Sub-categories were settled as positive and negative in each of the general categories Comparing themes across sub- The similarities and differences of the statements were compared samples across sub-samples Meanings were determined Creating a code A general code/dominant theme was created and named in relation to the general meaning. Dominant themes are: level of negativity and responsibility (blameworthiness), admiration, celebrity image, brand image and message content Determining the reliability of The reliability of the code was discussed with two marketing the code academicians

Research findings In the first step, all of the 451 posts were analyzed as to whether they contained negative, positive or neutral reactions towards the Nike commercial, Woods, Nike and the purchase decision. It was found that 45.5 percent of the consumers talked negatively about the Nike commercial. In addition, 36.1 percent of the consumers had negative feelings about Woods, 23.1 percent of them talked negatively about Nike, but just 6 percent of them said they would boycott the products of the company. There were also positive posts: 27.5 percent of the consumers had positive feelings for Woods, 13.5 percent had positive attitudes towards the commercial, but only 7.5 percent of them talked positively about Nike. The negative attitudes in the commercial included the words “creepy”, “bizarre”, “strange”, “disgusting”, “not wise”, “horrible”, “uncomfortable”, “mental game”, “manipulative”, “sick”, “ill”, and “not like”. The positive attitudes included the words “brilliant”, “ingenious”, “like it”, “love it”, “nice”, “good” and “sincere”.

Interpretation of the emerging themes and model development The postings of the consumers were analyzed by discourse analysis to identify the basic patterns about the celebrity endorsement in the case of negative information. Also the emerging themes were interpreted. On the basis of that interpretation a model was developed. The proposed model is shown in Figure 2. Consumer perception of celebrity. The perception of celebrity is important in celebrity endorsement. It is likewise critical in understanding the process of how consumers form positive or negative opinions about the celebrities. In the case of negative information, consumer reactions and perceptions of celebrity may vary: they can change negatively or not change at all, depending on several dominant factors.

Celebrity advertising

Admiration as an athlete

1287

High

Negativity Low

The event

Positive Celebrity Image

High

Negative Celebrity Image

Blame

Positive Brand Image

Negative Brand Image

Low

Positive Message Content

Negative Message Content

Figure 1. The themes and the linkages

Admiration Level of Negativity Level of Blameworthiness

Message Content

Consumer Perception of Celebrity

Image Transfer

Consumer Perception of Brand

Figure 2. Proposed model

MRR 34,12

1288

Level of negativity. The level of negativity is related to how bad the event is perceived to be. In people’s minds each crime has different levels (Louie and Obermiller, 2002). If the situation is not perceived as a major transgression, the consumer tends to not change his celebrity image perception. In this case, those having had a positive image perception of Woods routinely mentioned his multiple infidelity was not a severe crime such as murder or rape. Some of the positive posts said that: He cheated not raped or molestered somebody! [. . .] [. . .] half of you people out there who hate Tiger now because you just found out he is a human are stupid. He is allowed to make mistakes, and unlike when you all cheat on your spouses and get away with it, he gets to have it put front [. . .] Tiger is still one of the worlds’ best golfers, and I know I will still be cheering for him. [. . .] 97% of married men cheat [. . .] [. . .] Tiger is human and a fool so what else is new. Politicians do it every day [. . .] Why so hard on this guy? He is a man. Give me 10 men and I will show you 8 of them if not all that have cheated on their wife. Who cares. Their business. Just thank God he is not your husband. The man had an affair, he didn’t commit murder, lol. Affairs are no ones business but the offenders and their spouses.

While some of the negative postings included: He’s not anybody’s little boy anymore. And, you don’t go around committing terrible crimes against your wife and children, and your (former) fans – those fans who are actually “honest with themselves” about life. Yes, “marketing” will bring back all of the “gullible”, but the reality of life has nothing to do with “marketing” and “image.” I will not be purchasing anything from Nike going forward. If he was in politics and not a golf star he would be unemployed. He’s a great golfer, probably the best, but I have lost a lot of respect for him.

This problem is labeled a personal trouble for him and his family and called a “mistake”. Consequently, it is not related to his professional performance. However, for some people infidelity is a low moral violation indicative of “character”, with cheaters being seen and judged as “liars” and “dishonest people”. In that case people perceive the celebrity image as low and disgusting. Hence it is proposed that as the level of negativity increases, the negative perception of the celebrity image also increases. As a result, it was proposed that: P1.

When the negativity of the event is perceived as low (high), the celebrity image is perceived as positive (negative).

Level of blameworthiness. Blame theory argues that: [. . .] questions of blameworthiness arise only when at least one of the causal elements participating in the production of the effect for which blame is to be assigned is a human action (Shaver, 1985, p. 162).

The blameworthiness depends on whether or not the person is perceived to be responsible (Louie and Obermiller, 2002). This study of the corpus of weblog evidence found that the consumers blaming Tiger Woods for his infidelity tended to have a lower image perception of him than those who were not blaming him. Consumers posting messages about him and responsibility said:

Celebrity advertising

I think it’s creepy because they’re repurposing his dead father’s words from years ago to generate sympathy for someone who was in complete control of his decisions. I’m sure Tiger is real sorry. Sorry he got caught.

1289

[. . .] he knew what he did so he is not an innocent victim here [ . . . .] The ONLY reason he is apologizing now is because he got caught. I condemn him [. . .] [. . .] his actions being none of our business and completely between him and his wife (and whoever else was directly involved) [. . .] [. . .] he’s got to take responsibility for his own actions.

And, it was proposed that: P2.

When the blameworthiness of the celebrity is perceived as low (high), the celebrity image is perceived as positive (negative).

Admiration. As the data dissected, it was found that consumers’ “admiration” was directly related to the professional performance of Woods as an athlete. People who admired his profession just wanted him to go back and play golf. The admiration of people was related to “the level of negativity”, “the level of responsibility” and “the level of blameworthiness”. This was determined because in the posts his personal life and profession were repeatedly perceived as different things. The authors of posts thought that he was not responsible to the public concerning his infidelity for basically two reasons: on the one hand they thought his private life was essentially private business, and on the other hand they judged infidelity to essentially not be a serious crime. Thus, their feelings towards Woods were a dual affinity constraining a positive bias towards him. Consequently, they literally “just” wanted him to play golf. The following postings clearly illustrate this tendency: For me, it’s kinda like comparing apples and oranges [. . .] how the man plays the game and what he does in his personal life are two VERY different things. Tiger’s personal life shouldn’t have anything to do with the sport he plays. Pathetic! Leave his private life to him and his family. Just have him play golf. He’s good at that. He is an amazing athlete, And that is what promotes him with Nike and everyone else. Not his Sex Escapades. So just let it go, And let the man do what he does best. PLAY AMAZING GOLF !!!!!!! [. . .] it’s their career we are interested in, not whose bed they are in. I liked the commercial a lot, and I like Tiger. Despite the best golfer in the world, he is a human. Just play golf, hold it together, and do your job. do it well and we’ll be back.

MRR 34,12

1290

In relation, it was proposed that: P3.

When the negativity of the event is perceived as low (high), the admiration is perceived as positive (negative).

P4.

When the blameworthiness of the celebrity is perceived as low (high), the admiration is perceived as positive (negative).

P5.

There is a positive relationship between the admiration and the consumer perception of celebrity.

Consumer perception of brand. In the study the consumer perception of the brand was identified as being highly influenced by the consumer perception of celebrity. Those who had a negative attitude towards the celebrity also conveyed a negative attitude towards the brand. The negative postings are: I do not like Mr Woods and I will not buy any Nikes. Nike has gone way to far with this. All I could think was, damn NIKE, you guys are really playing minds games with us. but Nike’s cynical behavior guarantees I will never buy another Nike product.

The positive postings are: I do like Mr Woods. For every Nike that you won’t buy, I will and so will many others. The man is just a man. I’ll buy Nike products. Tiger is just a man, with a great skill. People buy the products because of the skill. That’s all. Normal people could care less what happened in his personal life.

The “Just do it!” slogan of the brand was also associated with the event negatively. Some posting noted that it was wise (!) not to display the slogan at the end of the commercial. Therefore, it was seen that the brand slogan gained another meaning; this incurs a risk for the brand. Therefore, it was proposed that: P6.

There is a positive relationship between the consumer perception of celebrity and the consumer perception of brand.

Message content. In that case of celebrity advertising Nike released a commercial containing a message about Tiger Woods’ personal issue. The commercial attracted comments in several ways. First, some posts criticized the usage of the dead father to create sympathy. In these posts Woods, seen as a 12-year-old boy, played the victim card and did not take responsibility for his actions; rather, he hid behind his father. Second, some posts questioned the credibility of his father, since his father was stereotyped as a former “cheater”. Third, in the commercial his father asks the question “did you learn anything?” and the commercial ends without Woods answering. Consequently, posters said they therefore concluded that “he did not learn anything”. Finally, people commented that Nike and Woods were exploiting his dead father for profit.

In sum, the comments concerning the commercial were negative/positive, and this negativity (positivity) was projected onto the celebrity and the brand. Here are a few exemplary negative postings:

Celebrity advertising

[. . .]strange, uncomfortable, sad, painful, and creepy new Nike ad. Nike is protecting it’s (sic) investment so it can sell product and Tiger will kiss the golden goose [. . .] I bought new Adidas today! Def creepy and inappropriate. Nike milking his homewrecking to sell gear? Shameful.

In contrast, here are a few exemplary positive postings: I like the ad. Redemption and learning is the key. Go Tiger. I love it and I love tiger! Welcome back! Effectively Nike is saying- we stood by you because you are family.

Therefore, it was proposed that: P7.

There is a negative relationship between the consumer perception of the message and the level of negativity and blameworthiness.

P8.

There is a positive relationship between the consumer perception of the message and the perception of celebrity.

P9.

There is a positive relationship between the consumer perception of the message and the consumer perception of brand.

Conclusion This study aims to investigate the reactions of the consumers when the company keeps an endorsement deal with a negatively publicized celebrity. In the study, the case of the Woods and Nike reunion is analyzed by discourse analysis of the weblog postings. As a result a model was proposed. When there is a negative event that the celebrity involved in, the consumers’ perceptions of the celebrity may change or may not change. This “change or not change” situation is related with the perceptions of “the level of the negativity of the event” and “the level of the blameworthiness”. When the consumer perceives the event as highly unfavorable and the celebrity as responsible, his/her perception of celebrity image is affected negatively. Moreover, celebrities are seen as role models and admired. In this study admiration is directly related to the professional performance of Woods as an athlete. It was found that the admiration affected the perceived image of the celebrity and it was affected by “the level of negativity” and “the level of blameworthiness”. Finally, this study reached the same conclusion as the other studies on celebrity endorsement that the perception of the brand is related with the perception of the celebrity. The perceived image of the celebrity and the associations related to the celebrity is transferred to the brand. Besides, in the case analyzed, the firm retained the celebrity in endorsement and broadcasted a commercial about the celebrity’s personal issue. In relation to the commercial it was found that, the message content created unfavorable perceptions and influenced negatively even the people who have positive feelings. Therefore, it was concluded that

1291

MRR 34,12

1292

there is also a meaning transfer from the message content to the perceptions of the negativity/blameworthiness, and celebrity and brand image. Implications This research extends the literature on celebrity advertising and broadens our understanding of the consumers’ perspectives of negatively publicized celebrity endorsement. The main theoretical contribution is the development of a model which helps us make sense of what is essentially an anomaly, namely consumers’ perceptions of the negative event, celebrity and brand image, and the message content. This study further provides managerial implications potentially useful to firms in developing strategies on celebrity advertising and brand image. The decision for a firm to (dis)associate itself from the celebrity involved in an undesirable event hinges on consumer perceptions of blame (Louie and Obermiller, 2002) and the level of negativity. When an individual is not culpable for undesirable actions, responses to that individual include sympathy and liking. However, when an individual is blameworthy for undesirable actions, reactions to that individual include anger and reduced levels of liking (Louie et al., 2001). Besides, negative behaviors attract more attention, and are better recalled than positive behaviors (Money et al., 2006). Moreover, people tend to evaluate such negative information more strongly than the positive information (Ito et al., 1998). In the case of negative event, this study demonstrated that it is important to understand the consumers’ attributions of blameworthiness and negativity for the undesirable event, and put forward that there is a negative relationship between the level of negativity/blameworthiness and the celebrity image perception. There is a wealth of evidence that brands should be careful in deciding on giving a message. The sympathetic and empathetic reactions to TV commercials are important since they directly enhance positive attitudes to the advertisement (Escalas and Stern, 2003). However, in this study the consumers’ reactions towards the commercial were mostly negative; and it was found that the perceptions of the message are not only transferred to the celebrity and the brand image but also transferred to the perceptions related with negativity and blameworthiness. In other words, negative perception of the message increases the perception of the negativity and blameworthiness, and simultaneously negatively affects the celebrity image and brand image. This is an exploratory study based on qualitative data analysis of just one celebrity endorsement case – the Woods and Nike reunion in 2010. Thus, it would be foolhardy to try and generalize these limited results, pending further research on the reactions of consumers when the company does and does not keep its endorsement deal, along with investigations of the long-term impact of celebrity image on brand image. It is hoped that the model and nine propositions will be further tested in future research on such scandalous moments, the celebrities involved and their endorsements in contemporary media. References Agarwal, N., Liu, H., Tang, L. and Yu, P.S. (2008), “Identifying the influential bloggers in a community”, Proceedings of the International Conference on Web Search and Web Data Mining, Palo Alto, CA, USA, pp. 207-18. Agrawal, J. and Kamakura, W.A. (1995), “The economic worth of celebrity endorsers: an event study analysis”, Journal of Marketing, Vol. 59 No. 3, pp. 56-63.

Assael, H. (1984), Consumer Behavior and Market Action, Kent Publishing Company, Boston, MA. Atkin, C. and Block, M. (1983), “Effectiveness of celebrity endorsers”, Journal of Advertising Research, Vol. 23 No. 1, pp. 57-61. Badenhausen, K. (2009), “Woods is sports’ first billion-dollar man”, Forbes, Yahoo! available at: http://sports.yahoo.Com/golf/pga/news?slug¼ys-forbestiger100109& prov¼yhoo&type¼lgns (accessed October 2, 2009). Berger, P.L. and Luckmann, T. (1991), The Social Construction of Reality, Anchor Books, London. Berglund, K. and Johnson, A.W. (2007), “Constructions of entrepreneurship: a discourse analysis of academic publications”, Journal of Enterprising Communities: People and Places in the Global Economy, Vol. 1 No. 1, pp. 77-102. Bernard, H.R. and Ryan, W.G. (2010), Analyzing Qualitative Data: Systematic Approaches, Sage, Thousand Oaks, CA. Boyatzis, R.E. (1998), Transforming Qualitative Information: Thematic Analysis and Code Development, Sage, Thousand Oaks, CA. Dick, P. (2004), “Discourse analysis”, in Cassell, C. and Symon, G. (Eds), Essential Guide to Qualitative Methods in Organizational Research, Sage, London. Elliot, R. (1996), “Discourse analysis: exploring action, function and conflict in social texts”, Marketing Intelligence & Planning, Vol. 14 No. 6, pp. 65-8. Elliot, R., Jones, A., Benfield, A. and Barlow, M. (1995), “Overt sexuality in advertising: a discourse analysis of gender response”, Journal of Consumer Poicy, Vol. 18 Nos 2/3, pp. 187-217. Erdogan, B.Z. (1999), “Celebrity endorsement: a literature review”, Journal of Marketing Management, Vol. 15 No. 4, pp. 291-314. Escalas, J.E. and Stern, B.B. (2003), “Emotional responses to advertising dramas”, Journal of Consumer Research, Vol. 29 No. 4, pp. 566-78. Fox, N. and Roberts, C. (1999), “GPs in cyberspace: the sociology of a ‘virtual community’”, The Sociological Review, Vol. 47 No. 4, pp. 643-71. Freiden, J.B. (1984), “Advertising spokesperson effects: an examination of endorser type and gender on two audiences”, Journal of Advertising Research, Vol. 24 No. 5, pp. 33-41. Gupta, S. (2009), “How do consumers judge celebrities’ irresponsible behavior? An attribution theory perspective”, Journal of Applied Business and Economics, Vol. 10 No. 3, pp. 1-14. Ito, T.A., Larsen, J.T., Smith, N.K. and Cacioppo, J.T. (1998), “Negative information weighs more heavily on the brain; the negativity bias in evaluative categorizations”, Journal of Personality and Social Psychology, Vol. 75 No. 4, pp. 887-900. Kamins, M.A. (1989), “Celebrity and non-celebrity advertising in a two-sided context”, Journal of Advertising Research, Vol. 29 No. 3, pp. 34-42. Kamins, M.A., Brand, M.J., Hoeke, S.A. and Moe, J.C. (1989), “Two-sided versus one-sided celebrity endorsements: the impact on advertising effectiveness and credibility”, Journal of Advertising, Vol. 18 No. 2, pp. 4-10. Kanayama, T. (2003), “Ethnographic research on the experience of Japanese elderly people online”, New Media & Society, Vol. 5 No. 2, pp. 267-88. Kozinets, R.V. (2002), “The field behind the screen: using netnography for marketing research in online communities”, Journal of Marketing Research, Vol. 39, pp. 61-72. Louie, T.A. and Obermiller, C. (2002), “Consumer response to a firm’s endorser (dis)association decisions”, Journal of Advertising, Vol. 31 No. 4, pp. 41-54.

Celebrity advertising

1293

MRR 34,12

1294

Louie, T.A., Kulik, R.L. and Jacobson, R. (2001), “When bad things happen to the endorsers of good products”, Marketing Letters, Vol. 12 No. 1, pp. 13-23. McCracken, G. (1989), “Who is the celebrity endorser? Cultural foundations of the endorsement process”, Journal of Consumer Research, Vol. 16 No. 3, pp. 310-20. Miles, M.B. and Huberman, A.M. (1994), Qualitative Data Analysis, Sage, Thousand Oaks, CA. Miller, M. (2009), “The wealthiest black Americans”, Forbes, available at: www.forbes.com/2009/ 05/06/richest-black-americans-business-billionaires-richestblack-americans.html (accessed December 17, 2009). Misra, S. and Beatty, S.E. (1990), “Celebrity spokesperson and brand congruence: an assessment of recall and affect”, Journal of Business Research, Vol. 21 No. 2, pp. 159-73. Mitchell, A.A. and Olson, J.C. (1981), “Are product attribute beliefs the only mediator of advertising effects on brand attitude?”, Journal of Marketing Research, Vol. 18 No. 3, pp. 318-32. Money, R.B., Shimp, T.A. and Sakano, T. (2006), “Celebrity endorsements in Japan and the United States: is negative information all that harmful?”, Journal of Advertising Research, Vol. 46 No. 1, pp. 113-23. Nahl, D. (2007), “A discourse analysis technique for charting the flow of micro-information behavior”, Journal of Documentation, Vol. 63 No. 3, pp. 323-39. Nguyen, L., Torlina, L., Peszynski, K. and Corbitt, B. (2006), “Power relations in virtual communities: an ethnographic study”, Electronic Commerce Research, Vol. 6, pp. 21-37. Ohanian, R. (1991), “The impact of celebrity spokespersons’ perceived image on consumers’ intention to purchase”, Journal of Advertising Research, Vol. 31 No. 1, pp. 46-54. O’Reilly, N.J., Rahinel, R., Foster, M.K. and Patterson, M. (2007), “Connecting in megaclasses: the netnographic advantage”, Journal of Marketing Education, Vol. 29 No. 1, pp. 69-84. Petty, R.E. and Cacioppo, J.T. (1985), “Source factors and the elaboration likelihood model of persuasion”, Advances in Consumer Research, Vol. 77 No. 1, pp. 668-72. Sandlin, J.A. (2007), “Netnography as a consumer education research tool”, International Journal of Consumer Studies, Vol. 31, pp. 288-94. Shaver, K.G. (1985), The Attribution of Blame: Causality, Responsibility, and Blameworthiness, Springer, New York, NY. Sitz, L. (2008), “Beyond semiotics and hermeneutics: discourse analysis as a way to interpret consumers’ discourses and experiences”, Qualitative Market Research: An International Journal, Vol. 11 No. 2, pp. 177-91. Sneijder, P. and te Molder, H.F.M. (2005), “Disputing taste: foods pleasure as an achievement in interaction”, Appetite, Vol. 45, pp. 51-61. Social Media White Paper (2008), available at: http://technorati.com/blogging/state-of-theblogosphere/ Thelwall, M. and Hasler, L. (2007), “Blog search engines”, Online Information Review, Vol. 31 No. 4, pp. 467-79. Till, B.D. and Shimp, T.A. (1998), “Endorsers in advertising: the case of negative celebrity information”, Journal of Advertising, Vol. 27 No. 1, pp. 67-82. Tripp, C., Jensen, T.D. and Carlson, L. (1994), “The effects of multiple product endorsements by celebrities on consumers’ attitudes and intentions”, Journal of Consumer Research, Vol. 20 No. 4, pp. 535-47. Van Dijk, T.A. (1981), “Discourse studies and education”, Applied Linguistics, Vol. 2 No. 1, pp. 1-26.

White, D.W., Goddard, L. and Wilbur, N. (2009), “The effects of negative information transference in the celebrity endorsement relationship”, International Journal of Retail & Distribution Management, Vol. 37 No. 4, pp. 322-35. Further reading Kahle, L.R. and Homer, P.M. (1985), “Physical attractiveness of the celebrity endorser: a social adaptation perspective”, Journal of Consumer Research, Vol. 11 No. 4, pp. 954-61. McGuire, W.J. (1968), “The nature of attitudes and attitude change”, in Lindzey, G. and Aronson, E. (Eds), The Handbook of Social Psychology, 2nd ed., Vol. 3, Addison-Wesley, Reading, MA, pp. 136-314. McGuire, W.J. (1985), “Attitudes and attitude change”, in Lindzey, G. and Aronson, E. (Eds), Handbook of Social Psychology, 3rd ed., Vol. 2, Random House, New York, NY, pp. 223-346. Ohanian, R. (1990), “Construction and validation of a scale to measure celebrity endorsers’ perceived expertise, trustworthiness, and attractiveness”, Journal of Advertising, Vol. 19 No. 3, pp. 39-52. About the author Ulun Akturan is Assistant Professor in the Department of Marketing, Galatasaray University (Istanbul, Turkey). She has specialized in consumer behaviour, brand management and marketing research and has published national and international papers on consumer behaviour, brand management and young adults. Ulun Akturan can be contacted at: [email protected]

To purchase reprints of this article please e-mail: [email protected] Or visit our web site for further details: www.emeraldinsight.com/reprints

Celebrity advertising

1295