Models On Celebrity Endorsement Strategy Cultural Studies Essay

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

The selection of celebrity endorsers is not an easy task; many scholars have tried to create models in order to help for the right selecting of celebrity endorsers. Hovland et al (1953) conceptually contributed one of the earliest models, which is Source Credibility Model. Afterwords, the Source Attractiveness Model (McGuire, 1985), the Product Match-Up Hypothesis (Forkan, 1980; Kamins, 1989, 1990), and the Meaning Transfer Model (McCracken, 1989) was presented through empirically researchers in turn.

The Source Credibility Model and Source Attractiveness Model are categorized under the generic name of Source Models since these two models basically show and reflect research of the Social Influence Theory/Source Effect Theory, which argues that various characteristics of a perceived communication source may have a beneficial effect on message receptivity (Kelman, 1961; Meenaghan, 1995).

The source credibility model is based on research in social psychology (Hovland and Weiss, 1951-1952; Hovland, Jani, and Kelley, 1953). The Hovland version of model present that a message depends for its effectiveness on the 'expertness' and 'trustworthiness' of the source (Hovland et al., 1953, p.20; Dholakia and Sternthal, 1977; Sternthal, Dholakia, and Leavitt, 1978), which means that information from a credible source (e.g.celebrity) can influence beliefs, opinions, attitudes, and/or behavior via a process called internalization, which occurs when receivers accept a source influence in terms of their personal attitude and value structures (Erdogan, 1999).

Expertness is defined as the extent to which a communicator is perceived to be a source of valid assertions and refers to the knowledge, experience or skills possessed by an endorser. Hovland et al (1953) and Ohanian (1991) believed that it does not really matter whether an endorser is an expert, but all that matters is how the target audience perceives the endorser. However, Aaker and Myers (1987) advocated a source/ celebrity that is more expert to be more persuasive and to generate more intentions to buy the brand (Ohanian, 1991). Hence, expert sources influence perceptions of the product's quality (Erdogan, 1999).Meanwhile; Speck et al (1988) found that expert celebrities produced higher recall of product information than non-expert celebrities, even though the difference was not statistically significant. Moreover, celebrities' professional accomplishments and expertise may serve as a logical connection with the products, and consequently make the endorsement more believable to consumers (Till and Brusler, 2000).

Trustworthiness refers to the honesty, integrity and believability of an endorser depending on target audience perceptions (Erdogan, 1999). Advertisers capitalize on the value of trustworthiness by selecting endorsers, who are widely regarded as honest, believable, and dependable (Shimp, 1997). Smith (1973) argues that consumers view untrustworthy celebrity endorsers as questionable message sources regardless of their qualities. Friedman, et al (1978) addressed that trustworthiness is the major determinant of source credibility and then tried to discover that likability was the most important attribute of trust. Thus, they recommended advertisers to select personalities who are well liked when a trustworthy celebrity is desired to endorse brands. However, Ohanian (1991) found that trustworthiness of a celebrity was not significant related to customers' intentions to buy an endorsed ethnic status could affect endorser trustworthiness and brand attitudes, because people trust endorsers who are similar to them. Their findings implied that when targeting particular ethnic groups such as Africans and Asians, ethnic background should be carefully evaluated.

Measuring source credibility in selecting celebrity

It is quite reasonable to make sense that a source's credibility is totally subjective, but research shows that in spite of individual preferences, a high degree of agreement exists among individuals (Berscheid et al, 1971). Patzer (1983) developed the Truth-of-Consensus method to assess a source's credibility and attractiveness. The method is based on the foundation that individual's judgments of attractiveness and credibility are naturally subjective, but these judgments are shaped through Gestalt principles of person perception rather than single characteristics. Notably, on the basis of extensive literature review and statistical tests, Ohanian (1990) constructed a tri- component celebrity endorser credibility scale, (see figure) which assumes that credibility and effectiveness of celebrity endorsers is associated with given characteristic dimensions, even though McCracken (1989) argued that the celebrity world consists of much more just attractive and credible individuals.

Table 2: Source Credibility Scale






Expert-Not Expert

Classy-Not Classy









Sexy-Not Sexy



Source: Ohanian, R (1990) 'Construction and validation of a scale to measure celebrity endorser's perceived expertise, trustworthiness and attractiveness', Journal of Advertising, p39-52

The Source Attractiveness Model

Advertisers have chosen celebrity on the basis of their attractiveness to gain from dual effects of celebrity status and physical appeal (Singer, 1983). Meanwhile, research showed that physically attractive endorsers are more successful at changing beliefs (Baker and Chrurchill, 1977; Chaiken, 1979; Debevec and Kernan, 1984) and generating purchase intentions (Friedman et al, 1976; Petroshius and Schuman, 1989; Petty and Cacioppo, 1980) than those unattractive individuals.

Hence, McGuire (1985) conducted an empirical research to contend that the effectiveness of a message depends on similarity, familiarity and liking for an endorser. The McGuire (1985) model holds that sources that are known to, liked by, and/or similar to the consumer are attractive and, persuasive. The source attractiveness model also rests on social psychological research (McCracken, 1989). Meanwhile, Cohen and Golden (1972) suggested that 'physical attractiveness' of an endorser determines the effectiveness of persuasive communication through a process called identification, which is assumed to occur when information from an attractive source is accepted as a result of desire to identify with such endorsers.

Petty and Cacioppo (1980) conducted attractiveness of endorsers in terms of a shampoo advertisement to comprehend effectiveness of advertising message types. In 1983, Petty et al replicated the earlier study in 1980. Their findings emphasize the interaction between involvement level and endorser type. Under low-involvement conditions, the endorser type had a significant impact on attitudes towards the product even though no impact was found on behavioral intentions. With respect to recall and recognition measures, findings indicated that exposure to celebrity endorsers increased recall of the product category only under low-involvement conditions. Besides, the endorser type manipulation revealed that celebrities had marginally significant impact on brand name recall over typical citizens.

Patzer (1985 : p30) stated that 'physical attractiveness is an information cue; involves effects that are subtle, pervasive, and inescapable; produces a definite pattern of verifiable differences; and transcends culture in its effects'. Patzer argues that people usually inflate their own attractiveness so that attractive endorsers should be more effective than average looking endorsers.

Kahle and Homer (1985) operated celebrity physical attractiveness and likability and measured attitude and purchase intentions on the same product: Edge razors. Findings indicated that participants exposed to an attractive celebrity liked the product more than participants exposed to an unattractive celebrity. Recall for the brand was greater both in attractive and likeable celebrity conditions. However, unlikeable celebrities unexpectedly performed better on recognition measures than likeable and attractive celebrities. Meanwhile, findings proved that an attractive celebrity created more purchase intentions than unattractive celebrity, but conversely an unlikeable celebrity produced more intentions to buy the product than a likeable celebrity.

Quite significantly, studies by Cabalero (1989) and Till and Brusler (1998) demonstrate that positive feelings towards advertising and products do not necessarily translate into actual behavior or purchase intentions. A possible reason for the lack of celebrity endorsers' effect on intentions to purchase is that celebrity endorsement seems to work on the cognitive and affective components of attitudes rather than the behavioral components (Baker and Churchill, 1977; Fireworker and Friedman, 1977).

In terms of gender impact between endorsers and target audience, Debevec and Kernan (1984) found that attractive female model generated more enhanced attitudes than attractive male models across both genders and particularly among males. Conversely, Cabalero et al (1989) found that males showed greater intentions to buy from male endorsers and females hold greater intentions to purchase from female endorsers. Furthermore, Baker and Churchill (1977) found a rather unexpected interaction among female models, product type and intentions to purchase products among male subjects. For instance, when the endorsed product was coffee, an unattractive female model created more intentions to buy the product than her attractive counterpart among males, whereas when it was perfume or aftershave, male reacted more positively to an attractive female model. However, Petroshius and Schulman (1989) found that endorsement gender had no impact on attitudes towards advertisements and no major impact on intentions to buy products. Consequently, based on above disparate and controversial arguments, there is no consistent and coherent direction in terms of gender interactions between endorsers and target audiences to aid practitioners.

In brief, it is apparent that attractive celebrity endorsers enhance attitudes and recall towards advertising and brands than unattractive celebrity endorsers, however there is no consistent agreement in relation to creating purchase intentions, even though a few studies found that celebrities can create purchase intentions.

Multiple celebrity endorsement

Millions of dollars are spent per annum on celebrity endorsement contracts on the basis that source effects play an important part in convincing communications. Although traditional advertising knowledge suggests the meaning of an elite product contract with the celebrity, uniqueness comes with a high price label. As a result, it is becoming familiar for companies to 'share stars' (Elliott, 1991; Sloan and Freeman, 1988). For example, former Chicago Bulls star Michael Jordan has endorsed products for 14 companies (Lipman and Hinge, 1991), and golfer Lee Trevino has had endorsement contracts with Cadillac, Motorola Cellular Phone, Spalding Top-Flight, and La Victoria Salsa (Shatel, 1991).

Multiple product endorsements set up new questions relating to our understanding of how consumers react to celebrity endorsements. If as McCracken (1989, p.311) suggests, the celebrity endorser takes on meanings that carry from ad to ad, does endorsing multiple products affect those assigned meanings such that the consumer perceives the celebrity to be less credible and less likable (Kaikati, 1987)? Do consumers have less positive approaches toward ads and brands if multiple product endorsements are involved? What consequence do multiple product endorsements have on consumers' buying aims? Does the number of products endorsed restrain the effect of frequent publicity to the celebrity endorser (Tripp, 1994)?

Practical proof concerning how consumers react to multiple product endorsement is restricted, leaving unanswered issues in an important research ground. It is known the act of multiple product endorsements guides to certain impressions about celebrity. Early studies (i.e. Mowen and Brown 1981; Mowen, Brown, and Schulman 1979) suggest that simply knowing that a celebrity endorses multiple products is satisfactory to decay consumers' insights of endorser honesty, as well as a brand and ad evaluations.

Given a limited knowledge of how the endorsement process works (McCracken, 1989), these are clearly issues with theoretical value. The current study independently manipulated the number of exposures to a celebrity in a way different from previous studies in order to investigate the effects of continual exposure to the multiple product endorsers on consumer responses. The apply of actual stimulus is important since exposure to multiple product endorsers (vs. knowledge only) may result in effects different from multiple product endorsement effects. For example, attribution theory (Kelley, 1973) suggests that assumptions may result in consumers' evaluating multiple product endorsers less favorable than single product endorsers. According to Kelley, observers identify an actor's action to be characteristic when it happens in the presence of a unit and does not occur to its absence.

In the case of endorsements, single product endorsements (even if viewed multiple times) compose characteristic actions since spokesperson endorses one brand and not other brands or products. In contrast, multiple product endorsements compose non distinctive actions because the endorsements take a broad view across products with the celebrity constant. Limited of the number of exposures to the endorser, this non distinctiveness may result in consumer's concluding that the nature of the spokesperson was the reason for the endorsement, not the nature of product. Although multiple product endorsements (i.e. non distinctive actions) influence perceptions of the spokesperson's credibility (i.e, internal attributions), the spotlight of external attributions for single product endorsements (i.e.. distinctive actions) is not obvious. Witnesses of a spokesperson who endorses only a single product may or may not trait the endorsement to the product itself (e.g. product quality). The product represents only one cause for the endorsement. Other potential causes for the endorsement exist (e.g., popularity of the endorser; endorser's ties to the product, company, or advertising agency; money paid to the endorser) (Tripp, 1994). In this respect, multiple product endorsements may lead to attribute suggestions about nature of the spokesperson (e.g. traits such as greediness) and, in turn, pressure such manifestations of affect as credibility and likability (Weiner, 1985). Moreover, affect may lead to comparative preferences or be short of of preferences toward associated stimulus (Bara and Ray, 1985) such as the ad or brand.

Attribution theory may be used to make clear consumers' assumptions about the reasons for a product endorser's support (Folkes, 1988). Commonly, when exposed to a single endorsement, consumers attribute the support to an external cause (e.g. a financial reason). However, the frequency of an action and the actions with which it co varies form the basis of many attributions (Folkes, 1988). Since a multiple product endorser is seen repeatedly and in different contexts, examination of the relationship between the number of exposures to the celebrity, endorser likability, and credibility is a critical consideration. Berlyne's (1970) two-factor paradigm suggests that increased exposure to a stimulus results in a more favorable response initially due to a learning factor. At some higher number of exposures, however, a negative response (possibly due to tedium) begins to predominate. Taken together, these two theoretical ideas suggests that the number of products endorsed and the number of exposures to the endorser may interact such that the number of products endorsed moderates the effect of number of exposures. Thus, inclusion of actual exposure allows for the first tests of any multiple product endorsement effects on both dimensions of credibility and likability beyond that due to repeated exposure to the endorser (Tripp, 1994).

The product Match-Up hypothesis

Forkan (1980) and Kamins (1990) conducted empirical experiment to test the Product Match-up Hypothesis, which contends that messages conveyed by celebrity image and the product message should be congruent for effective advertising. The determinant pf the match between celebrity and brand depends on the degree of perceived 'fit' between brand such as brand name and attributes and celebrity image (Misra, 1990). Advertising a product via a celebrity who has a relatively high product congruent image leads to greater advertiser and celebrity believability (Levy, 1959; Kamins and Gupta, 1994; Kotler, 1997).

Importance of proper match-up between celebrities and products has been emphasized. From practitioner's perspective, a senior vice president of a leading beverage company states that celebrities are an unnecessary risk unless they are very logically related to products (Watkins, 1989). Another practitioner quoted by Bertrand and Todd (1992) argued that if there is a combination of an appropriate tie-in between the company's product and the celebrity's persona, reputation or the line of work that the celebrity is in, advertisers can get both the fame and the tie-in working for them. Meanwhile, many studies report that consumers also expect congruity between celebrity endorsers' perceived images and their endorsed products (Callcoat and Phillips, 1996; Ohanian, 1991; O'Mahony and Meenaghan, 1997).

Otherwise, Evans (1988) argued that if celebrities do not have a distinct and specific relationship to the product they endorse, the use of celebrities could produce the 'Vampire effect' which happens when the audience remembers the celebrity, but not the product or service. Meanwhile, the absence of connection between celebrity endorsers and products endorsed may lead consumers to the belief that the celebrity has been bought to endorse the product/service (Erdogan, 1999).

Significantly, the proper match-up between a celebrity and a product has been based on celebrity physical attractiveness, and the match-up hypothesis predicts that attractive celebrities are more effective when endorsing products used to enhance one's attractiveness (Kahle and Homer, 1985; Kamins, 1990). Research also identifies that characteristics of a celebrity interact positively with the nature of the product endorsed (Friedman and Friedman, 1979; Kamins, 1990; Lynch and Schuler, 1994). Choi and Nora (2005), who used a cognitive approach that focuses on consumers' attributions of celebrity endorsement motives, emphasized that the level of celebrity and product congruence will influence celebrity endorsement effectiveness through the process of consumer attributions of the celebrity's motive for associating him or herself with the particular product and the subsequent effect of these attributions on the consumer evaluations of the endorser, the advertising, and the brand involved in the endorsement. Surprisingly, Kamins and Gupta (1994) found that the match-up between a celebrity endorser and the endorsed brand also enhances the celebrity endorser's believability and favorable attitudes (Till et al, 2006).

Friedman and Friedman (1978) found that celebrity endorsers are more appropriate where product purchases involve high social and psychological risk. Meanwhile, Kamins (1989) and Kamins, et al (1989) found that celebrity endorsers were able to generate desired effects on high financial and performance risk products/services such as management consultation and computers. Conversely, Callcoat and Phillips (1996) reported that consumers are generally influenced by endorsers if products are inexpensive, low-involving and few differences are perceived among available brands. As a result, these contradictory arguments lead to the conclusion that advertising is a powerful mechanism of meaning transfer that virtually any product can be made to take any meaning (McCracken's 1987; O'Mahony and Meenaghan, 1997).

The almost studies in terms of evaluating celebrities endorser and endorsed products/brands are using consumer samples. Only one study by Miciak and Shanklin (1994) investigation considered advertising practitioners when choosing celebrity endorsers based on a small sample including 21 agencies and 22 company practitioner. Remarkably, more recently, Erdogan et al (2001) investigated a larger sample that is the 300 largest British advertising agencies (Campaigh, 1997) to consider important celebrity characteristics from the practitioner's perspective when selecting an endorser by conducting exploratory interviews and a mail survey. Their findings provide implications for both theory and practice. At the theoretical level, the research firstly shows that managers do not see celebrities as undimensional individuals such as attractive and credible when selecting celebrity endorsers, because celebrities are different unknown endorsers as they represent a variety of meanings that are drawn from the roles they assume in television, film, politics, and so on (McCracken, 1989). Secondly, managers have implicitly incorporated the findings of product match-up hypothesis research in their decision-making. On the other hand, for practitioners, as none of the advertising agencies had any written documentation regarding celebrity endorsement strategy, Erdogan et al (2001) set the criteria through providing a possible 'check list' of factors in Table 3 below, when practitioners select celebrity endorsers.

Table 3 Check-List of Factors for Selecting Celebrity Endorsers


Std. Dev.

Celebrity target audience match



Celebrity product/brand match



Overall image of the celebrity



Cost of acquiring the celebrity



Celebrity trustworthiness



The likelihood of acquiring the celebrity



Celebrity controversy risk



Celebrity familiarity



Celebrity prior endorsements



Celebrity likability



Risk of celebrity overshadowing brands



The stage of celebrity life cycle



Celebrity expertise



Celebrity profession



Celebrity physical attractiveness



Celebrity equity membership status



Whether celebrity is a brand user



Scores are obtained from a scale in which 5= Very Important and 1= Very Unimportant

Source: Erdogan et al (2001) 'Selecting celebrity endorser: the practitioner's perspective', Journal of Advertising Research, Vol 41, No 3 p44

However, DeSarbo and Harshman (1985) argue that neither the source credibility and attractiveness nor the match-up research is adequate in providing a heuristic for appropriate celebrity endorser selection, although the Match-Up Hypothesis extends beyond attractiveness and credibility towards a consideration and matching of the entire image of the celebrity with the endorsed brand and the target audience.

The Meaning Transfer Model

McCracken (1989) organized an empirical research evaluating effectiveness of the endorser depends upon the meaning the person brings to the endorsement process in part. McCracken (1989) and Brierley (1995) pointed out that the number and variety of the meanings contained in celebrities are very large, which includes status, class, gender, and age and personality and lifestyles types, more importantly, the cultural meanings existing in a celebrity go beyond the person and are passed on to the products.

Fortini-Campbell (1992) argues that products just like people have personalities, and claims that people consume brands with personality characteristics like their own or ones they aspire to possess in celebrities. Similarly, according to Fowles (1996), advertisers' rationale for hiring celebrities to endorse products is that people consume images of celebrities, and advertisers hope that people will also consume products associated with celebrities. Celebrity endorsement actually is a special instance of a more general process of meaning transfer (McCracken, 1989). This process is a conventional path for the movement of cultural meaning in consumer societies through formation of celebrity image, transfer of meaning from celebrity to product, and from product to consumers. McCracken (1988) defined that meaning begins as something resident in the culturally constituted world, in the physical and social world constituted by the categories and principles of the prevailing culture. Furthermore, McCracken (1989) found that several instruments facilitate this transfer. Firstly, the movement of meanings from the culturally constituted world to consumer goods is accomplished by advertising and the fashion system. Then, the movement of meanings from consumer goods to the individual consumer is accomplished through the efforts of the consumers. Hence, meaning circulates in the consumer society.

Besides, McCracken (1986) argued that advertising is one of the instruments to move meanings from culture, to consumers, to goods; this movement is accomplished by the efforts of promotional agencies. Similarly, Domzal and Kerman (1992) claimed that advertising is an integral part of social systems, whose function is to communicate the culturally constructed meaning of products to consumers.

As the figure 1 shows, the meaning that begins in the dramatic role of the celebrity resides in the celebrity themselves in stage 1. In stage 2, this meaning is transferred when the celebrity enters into an advertisement with a product, and some of the meanings of the celebrity are now the meanings of the product. In the final stage, the meaning moves from the product to the consumer. Notably, celebrity endorsement makes a very particular contribution to each of these three stages in meaning transfer process.

Figure 1: Meaning transfer in the endorsement process

McCracken, G (1989) 'Who is the celebrity endorser? Cultural foundations of the endorsement process,' Journal of Consumer Research, 16 December, p315

In sum, as McCracken (1989) suggested, the meaning transfer model presented is intended to demonstrate that the secret of the celebrity endorsement is largely cultural in nature, and that the study of the celebrity endorsement is improved by a cultural perspective. Consequently, advertisers should assess the culture that encompasses a celebrity to determine whether these meanings are feasible for brands/products in order to achieve effectiveness of the endorser.

Definition of celebrities

Celebrities are people who enjoy public recognition by a big share of a certain group of people. Whereas characteristic like attractiveness, amazing lifestyle or special skills are just examples and specific common characteristics cannot be observed, it can be said that within a analogous social group celebrities generally vary from the social standard and enjoy a high degree of public awareness. This is factual for classic forms of celebrities, like actors (e.g. Meg Ryan, Pierce Brosnan), models (e.g. Naomi Campbell, Gisele Buendchen), sports athletes (e.g. Anna Kournikova, Michael Schumacher), entertainers (e.g. Oprah Winfrey, Conan O'Brien) and pop stars (e.g. Madonna, David Bowie) - but also for less obvious groups like businessmen (e.g. Donald Trump, Bill Gates) or politicians (e.g. Rudy Giuliani, Lee Kuan Yew). Celebrities appear in public in different ways. First, they appear in public when satisfying their profession, e.g. Pete Sampras, who plays tennis in front of an audience in Wimbledon. Furthermore, celebrities appear in public by attending special celebrity events, e.g. the Academy Awards, or world premieres of movies. In addition, they are present in news, fashion magazines, and tabloids, which offer second source information on events and the 'private life' of celebrities through mass-media channels (e.g. Fox 5 news covering Winona Ryder's trial on shoplifting, InStyle). Last but not least, celebrities work as spokespersons in advertising to endorse products and services (Kambitsis et al. 2002, Tom et al. 1992).

Advantages and disadvantages of celebrity endorsement strategy

Firstly, Erdogan (1999) briefly illustrated potential advantages and disadvantages of celebrity endorsement strategy and some preventive tactics in Table 5 below.

Table 5: Advantages and Disadvantages of Celebrity endorsement strategy

Potential Advantages

Potential Disadvantages

Preventive Tactics

Assisting product

marketing and

increased attention

Overshadow the brand

Pre-testing and careful planning

Image polishing

Public controversy

Buying insurance and putting

provision clauses in contracts

Brand introduction

Image change and


Explaining what is their role and

putting clause to restrict

endorsements for other brands

Brand repositioning

Image change and loss of

public recognition

Examining what life-cycle stage

the celebrity is in and how long

this stage is likely to continue

Underpin global



Selecting celebrity who are

appropriate for global target

audience, not because they are

'hot' in all market audience

Source: Erdogan, B.Z (1999) 'Celebrity endorsement: A literature review', Journal of Marketing Management, Vol 15, p295

Mathur et al (1997) state a variety of reasons that firms use celebrity endorsers including that firms may feel that the life experiences of endorsers fit the advertising message, that the endorser has high appeal with the firm's target consumer group, or that the endorser's universal appeal makes the advertising universal. Celebrity endorsement can bring out several positive effects. They are that advertisements become believable (Kamins et al, 1989), message recall is enhanced (Friedman and Friedman, 1979), recognition and perception of brand names is improved and attitudes about products with low purchase involvement are affected (Petty et al, 1983; Till et al, 2006), positive attitudes about brands results (Kamins et al, 1989), and distinct personalities and appeals for products and brands are created (McCracken, 1989; Dickenson, 1996). Moreover, celebrity endorsements are believed to generate a greater likelihood of customers' choosing the endorsed brand (Heath et al, 1994; Kahle and Homer, 1985; Ohanian, 1991). As a result the use of celebrity endorsement is an advertising strategy that should enhance the marginal value of advertisement expenditures and create brand equity by means of the 'secondary association' of a celebrity with a brand (Keller, 1993). Meanwhile, research indicates that celebrity endorsements can result in more favorable advertisement ratings and product evaluations (Dean and Biswas, 2001). Some of the most difficult aspects of global marketing to gasp are host countries' cultural 'roadblocks' such as time, space, language, relationships, power, risk, masculinity and femininity (Mooij, 1994; Hosfsted, 1984). Under this situation, celebrity endorsements are a powerful device by which to enter foreign markets; and celebrities with world-wide popularity can help companies break through many such roadblocks (Erdogan, 1999).

On the other hand, there are also many potential disadvantages and hazards in utilizing celebrities as endorsers as a part of marketing communication strategy. Firstly, benefits of using celebrities can reverse markedly if they for example, suddenly change image, fall popularity, get into a situation of moral turpitude, lose credibility by over-endorsing or overshadow endorsed products (Cooper, 1984; Kaikati, 1987). Secondly, negative information about a celebrity endorser not only influences consumers' perception of the celebrity, but also the endorsed product (Klebba and Unger, 1982; Till and Shimp, 1995). Thirdly, another common concern is that consumers will focus their attention on the celebrity and fail to notice the brand being promoted (Rossiter and Percy, 1987). Fourthly, celebrities who are blamed for negative events such as accidents can have detrimental influence on the products they endorse (Louie and Obermiller, 2002).

Besides, Mowen and Brown (1981) argue that if a celebrity's image ties in with many brands/ products, impact and indentify with each product may reduce since the relationship between the celebrity and a particular brand is not distinctive. This can not only compromise the value of the celebrity in the eyes of star's fans (Graham, 1989), but also can make consumers to think the real nature of endorsement that has less to do with the brand/product attributes, and more to do with generous compensation for the celebrity, leading consumers to overt doubt about their motives, so as to cause the negative influences on consumer attitudes and purchase intentions among the multiple products endorsed by celebrities (Cooper, 1984; Tripp et al, 1994).

The factors predicting the effectiveness of celebrity endorsement advertisements

Langmeyer and Walker (1991a, 1991b) demonstrated McCracken's model that is seemed as a merely theoretical concept through replicating to real life. Langmeyer and Walker (1991a) used a celebrity endorser, Cher, a celebrity endorsed product, Scandinavian Health Spas, and a non-endorsed product, bath towels, as stimuli through a response elicitation format, which means telling us what you think, to indentify the meaning in the celebrity endorsed brand and the endorser and to demonstrate the transfer meaning from endorser to product. Their results proved that Cher possessed symbolic meanings such as sexy, attractive, healthy, and independent, and indicated that Cher's meanings were transferred to Scandinavian Health Spas. Language was used to describe that association between Cher and Scandinavian Health Spas was consistent. Furthermore, Langmeyer and Walker (1991b) explored meanings communicated by celebrity endorsers, Madonna and Christie Brinkley, and by products, bath towels, VCR's, and blue jeans. The study results revealed that Madonna and Christie Brinkley are perceived differently by subjects, which are 101 undergraduate students, and when celebrities combined with products, these perceived differences affected meanings perceived in products. Meanwhile, Langmeyer and Walker (1991b) distinguished that before endorsed by celebrities, products had sole product category images; and when endorsed by celebrities, they took on the images of each celebrity. In brief, Langmeyer and Walker's findings empirically support the argument originally put forward by McCracken (1989) that celebrity embody various meanings and these meanings are passed on to products with endorsements.

I am not sure:The advertising company has to select the celebrity who contains the appropriate set of characteristics, and 'who will best be to produce the most favorable response from consumers' (Martin, 1996 p.28). L'Oreal decided to promote its lip color brand 'Shine Délice' as "sheer, sumptuous, and sensual"with"juicy shades...for luscious lips." (L'Oreal USA 2002) French model and actress Laetitia Casta with her fresh and sexy look best matched the cosmetic brand's properties and was therefore selected to portray this brand line of L'Oreal (view figure 4). Pairing the model and the beauty product in an advertisement allows the transfer of Casta's meanings to the consumer good, thus her meanings (e.g. youthful, fresh, appealing) become associated with the L'Oreal brand "Shine Délice" in the mind of the female consumer. Using a different character, for example the actress Andie McDowell, to endorse 'Shine Délice', would affect the meaning of the brand in the minds of consumers (Walker et al. 1992). The meanings associated with her, like sophisticated, moral, mature woman, and family-type, are completely different to those of Casta, thus, when transferred resulting in different associations with L'Oreal's lip color brand, for example, less youthful, more introverted, and safe.