Celebrities in Advertising:Looking
for Congruence or Likability?
Nathalie Fleck
University of Cergy-Pontoi
Michael Korchia
BEM Bordeaux Management School
高考大省Isabelle Le Roy
CSA
ABSTRACT
The choice of a celebrity endorr for a brand is an important topic in advertising and marketing,as considerable time and effort resources are dedicated tofinding the right celebrity to reprent a given organization.Celebrities ud as endorrs in advertiments are often very popular ones. However,from a cognitive point of view(and a more academic one),congruence between brand and celebrity ems to be very important too.Bad on affective and cognitive theories to explain endorment effectiveness,congruence between brand and celebrity is shown to be as effective as celebrity likability.Moreover,congruence between brand and celebrity as well as celebrity likability have an impact on the predisposition toward the ad,which in turn influences brand beliefs and purcha intention.C 2012Wiley Periodicals,Inc.
INTRODUCTION
Brands all over the world u celebrities to adver-ti their products.As early as1890,actress Sarah Bernhardt appeared on posters for La Diaphane,a fa-mous French brand of rice powder at that time(Lehu, 1993).The u of celebrities in advertising is not, therefore,a new phenomenon,but has become increas-ingly widespread over the past20years.Brands such as Pepsi have featured stars such as Michael Jack-son,Madonna,or the Spice Girls,with varying im-pact(Erdogan,1999).Pizza Hut launched its restau-rants on the international market with icons that are almost universally recognize
d,such as Cindy Craw-ford,Linda Evangelista,and Pamela Anderson,and more recently Justin Bieber.Some stars have v-eral contracts with different brands:it is estimated that Michael Jordan,who appears in advertiments for Nike,Coke,Wheaties,Mc Donald’s,Hanes,Oak-ley,and Gatorade,has an impact worth about14 billion dollars on the American economy(Erdogan, Baker,&Tagg,2001).Nespresso has been for years the fastest growing of all Nestl´e’s products(Smale, 2009),partly becau of effective advertiments fea-turing George Clooney.Stars have become a vital com-ponent of advertising for certain categories of product,such as perfumes and cosmetics.It ems that once a celebrity becomes famous and popular,he/she immedi-ately appears in an advertiment and sometimes v-eral at the same time.For instance,in a single week, French footballer Zinedine Zidane appeared on posters for Dior perfume and Leader Price hard-discount su-permarkets,as well as TV commercials for Dannon yoghurt.
The question,for advertising agencies and brands alike,is how to choo a celebrity for advertising.What should one take into account as a priority?Are aware-ness and the popularity of the celebrity enough to en-sure advertiment effectiveness?What are the other factors that could play a role in and explain celebrity en-dorment effectiveness?Alongside this phenomenon, rearch into the role of celebrities in advertising has also incread since the1990s(Amos,Holmes,&Strut-ton,200
8;Erdogan,1999;Ohanian,1991;Pringle& Binet,2005;Seno&Lukas,2007).The works have highlighted veral explanatory factors,with congru-ence being one of the most important determinants. What are the effects of congruence on advertiment effectiveness?Is it better to choo a celebrity who is first and foremost popular or one congruent with the brand?The are the questions this rearch will more specifically investigate.
Psychology and Marketing,Vol.29(9):651–662(September2012)
View this article online /journal/mar
C 2012Wiley Periodicals,Inc.DOI:10.1002/mar.20551
651
CELEBRITY ENDORSEMENT:THEORY AND EFFECTIVENESS FACTORS
Celebrity endorment is a phenomenon where celebri-ties lend their image to brands.Several rearchers have attempted to define endorment over the last 20years(e Erdogan,1999),and for McCracken,“the celebrity endorr is defined as any individual who en-joys public recognition and who us this recognition on behalf of a consumer good by appearing with it in an advertiment”(19hope的用法
89,p.310).
Rearch has attempted to elucidate the reasons why companies invest so much money in communi-cation featuring celebrities(Erdogan,Baker,&Tagg, 2001).There are many reasons,butfirstly,it is signif-icant that celebrity endorments are a way of avoid-ing anonymity and standing out from the competition in saturated markets.For example,in the cosmetics and perfume industry,new products are being launched all the time.Cosmetics brands can no longer achieve recognition merely due to the intrinsic quality of their products,but do so,above all,by the image they project through their advertising campaigns,generally by fea-turing movie stars,models or singers.Celebrities make an impression on consumers.According to Jacques Helleu,artistic director for Chanel,“it is the only way of making a sufficiently strong impression on consumers’memories to trigger a purcha when they are in a sales outlet at a later date”(Devilliers,2004).Beyond this ob-jective of differentiation,it also gives brands an oppor-tunity to focus their positioning in consumers’minds. Sometimes,celebrity endorments also give a brand considerable exposure in the press,who are more likely to report on events concerning the brand if the celebri-ties are ,brand promotion parties;Erdo-gan,Baker,&Tagg,2001).
The impact on sales is more difficult to asss. Recent advertising campaigns for Sloggi,featuring th
e French tennis player and singer Yannick Noah, led to a47%increa in the brand’s sales for the men’s market and5%for the women’s market over the past10months,which is an excellent result in a declining market(Neumann,2006).The results of celebrity endorments are,however,highly variable as there are many factors involved,including the choice of celebrity and the quality of the advertis-ing itlf,as well as media planning,and even press relations.
This is why it is interesting to understand how en-dorment works and identify the decisive factors in its effectiveness.From this perspective,the main models devid to explain the type of relationship that devel-ops between celebrities and consumers are prented. The models are esntial to our understanding of the influence of celebrities on brand perception.Of cour endorment effectiveness depends on advertiment execution quality,media planning and public relations, etc.but above all it is linked to the choice of celebrity, as shown through the different models.Source Attractiveness and Source Credibility Models
The attractiveness model assumes that the effective-ness of the message depends on the physical character-istics of the source,which make it attractive(Chao, W¨uhrer,&Werani,2005;Till&Busler,2000).A celebrity spokesperson’s physical attractiveness has a positive impact on brand recall,attitude toward the brand and purchasing intent(Joph,1982;Kahle&
Homer,1985).Here,the endorr’s physical attractive-ness is assumed to have an influence on the target’s acceptance of the advertising,particularly due to the phenomenon of identification.The effectiveness of the message depends on the endorr’s similarity,likeable-ness,and familiarity.While there is no doubt that a celebrity endorr may improve attitude toward adver-tising and a brand,its impact on purchasing behavior is less clear(Erdogan,1999).The attractiveness of the source is not the only characteristic of a celebrity to take into account(Friedman&Friedman,1979)and is mainly significant for brands related to physical ap-pearance(Kamins,1990).
The source credibility model was developed in the 1950s by Hovland and Weiss(1951).According to this model,the effectiveness of a message depends on the perceived level of experti and trustworthiness of an endorr(Ohanian,1991;Spry,Pappu,&Cornwell, 2011).Experti refers to the extent to which the en-dorr is perceived as a valid source of information(Er-dogan,1999).It is a matter of determining whether the advertising target perceives the endorr as hav-ing a certain level of experience,knowledge,and know-how that makes their recommendation of a product and/or brand credible.Erdogan’s(1999)review of pre-vious literature indicated that a source perceived as highly credible is more effective and persuasive than a less credible one,as well as inducing a more pos-itive change of attitude toward the brand or product and more changes in behavior.Trustworthine
ss refers to the endorr’s honesty,credibility,and integrity,as perceived by the target audience.It is a matter of de-termining whether the consumer trusts the source for the honest,objective way the information is prented (Ohanian,1991).However,as McCracken(1989)and Erdogan(1999)noted,limiting the asssment of an individual to certain dimensions may em rather re-strictive.
The models may em attractive but they have been criticized on veral levels(Bower&Landreth, 2001;Erdogan,1999).They do not properly take into account the multifunctional aspect of certain charac-teristics of the source and there is no proof the dimen-sions examined are the correct ones.However,above all,the models em incomplete as they do not con-sider all the perceptions and significance connected to a particular celebrity,merely focusing on certain char-acteristics of the endorr,so we do not feel they are capable of capturing what a person,as a whole,is capable of contributing to an advertiment.We think
652FLECK,KORCHIA,AND LE ROY
Psychology and Marketing DOI:10.1002/mar
McCracken’s(1989)meaning transfer model,which in-cludes all the cultural connotations of celebritie
s,is more suitable,as it explains the endorment process as a whole.
McCracken’s Meaning Transfer
vo2
Model(1989)ppt录制视频
This model has the advantage of including affective and cognitive aspects.The transfer theory relies on the endorment process described by McCracken(1989). This process is bad on the idea that,in consumer soci-ety,which is organized around culture,celebrities have a shared cultural significance in the eyes of consumers (McCracken,1989).McCracken defines this transfer as the translation of the meaning of celebrity to a product or brand,for example,in advertiments with celebrity spokespersons.Endorment is effective when an indi-vidual who buys and consumes the product appropri-ates the meaning associated with the celebrity,which has been transferred to the product.
The McCracken model makes it possible to explain how celebrities transmit an extensive t of associations to the brands they endor.Therefore,before companies lect celebrities to reprent their brands,they need to ensure that the person conveys the right meaning.
支教感悟与收获
It is precily this congruence between personality and brand that facilitates meaning and affect transfer from one to the other(Kamins&Gupta,1994;Lynch &Schuler,1994;Misra&Beatty,1990).It is thus uful to define this concept of congruence in greater detail.
THE ROLE OF CONGRUENCE BETWEEN BRAND AND CELEBRITY
The Role of Congruence
The term“congruence”is ud in veral rearch ar-eas,particularly brand extension,co-branding,spon-soring,and endorment.In all the cas,the aim is to asss thefit between a brand and another entity (a new product category,another brand,an event,or an individual;Fleck&Quester,2007;Maille&Fleck, 2011).A variety of terms have been ud(congruence,fit,link,“match up effect,”etc.)but the general concept is the same.
In thefield of celebrity spokespersons,congruence was not really defined as such until Misra and Beatty’s work(1990,p.161).They deduced that it consisted of the fact“that the highly relevant characteristics of the spokesperson are consistent with the highly relevant attributes of the brand.”A number of authors have studied specific aspects of celebrities,such as gender and skin color(Huston,d’Ouville,&Willis,2003)or physical attractiveness(Kamins,1990),but few have analyzed
congruence in the broader n.The Two Dimensions of Congruence
Heckler and Childers(1992)analyzed the congruence between visual and verbal elements in advertiments, adapting concepts ud in social cognition and,particu-larly,the memorization of social characteristics,to ad-vertising.Thus,they started from the concept of theme, which originates in verbal discour and prents the general content of the story told by the adverti-ment.Congruence is then defined in two dimensions: relevancy and expectancy.Relevancy reflects the ex-tent to which the information contained in the stim-ulus contributes to or prevents a clear identification of the main theme or message being communicated. Expectancy refers to the degree to which an item or piece of informationfits into a predetermined pattern or structure evoked by this theme.This interesting ap-proach suggests the concept of congruence may be two-dimensional.The two dimensions are also mentioned in the area of brand extensions(Lane,2000;Tauber, 1988),but under different names.In the area of brand endorment,a celebrity can be considered as relevant if there is a clear meaning why he/she endors the brand or product.For instance,a NASCAR or Formula 1driver who endors a car brand is meaningful and this endorment induces a clear message,becau of the experti of this celebrity in the brand category.On the other hand,a celebrity is expected if he/she corre-sponds to a pattern evoked by the
message of the brand. For instance,a model would be expected to testimony on the quality of cosmetics.The two dimensions are independent as,for instance,a celebrity can be very relevant for a brand but not expected at all.For in-stance,in2007Vuitton prented a range of luxury luggage featuring Mikhail Gorbathev,Keith Richards, or Catherine Deneuve.All the celebrities are rele-vant becau they travel a lot and they are very likely to own top-of-the-range brands of luggage,or at least they have the means to buy some.However,someone such as Mikhail Gorbachev was absolutely unexpected as he almost never appeared in any advertiment,as this is totally unusual for a former statesman.
The Impact on the Brand of Congruence between a Celebrity and the Brand Endord
Atfirst glance,it may em logical that congruence between the celebrity and the brand endord should have a positive impact on the brand’s image and that, the stronger the link,the more impact the associa-tion should have on the brand.Furthermore,the better suited,more relevant,or congruent the celebrity/brand pair is perceived to be,the greater the positive respon to advertising in terms of attitude or even purchasing intent(Batra&Homer,2004;Erdogan&Baker,2000; Kahle&Homer,1985;Kamins,1989,1990;Lynch& Schuler,1994;Misra&Beatty,1990;Till&Busler, 2000).
CELEBRITIES IN ADVERTISING653 Psychology and Marketing DOI:10.1002/mar
Figure1.Rearch model.
Note:Variables on the left of the vertical line were measured at period1,variables on the right at period2.
卢雨菲However,it is possible to imagine a more complex re-lationship and examine an alternative hypothesis,that is,that a certain level of incongruence may have a pos-itive impact on respon to advertising,particularly in terms of brand image(Fleck&Maille,2010).Indeed, a moderate level of inco
ngruence between an expec-tation and an object may be beneficial,provided it is perceived as interesting and positive(Meyers-Levy& Tybout,1989).In that ca,a relatively poorfit between brand and celebrity may be stimulating and encourage individuals to process the information more intensively and elaborate more(Lee,2000).In the ca of spon-soring,a slight inconsistency between the sponsor and the event should lead to enhanced recall and more fa-vorable attitudes than in the ca of total consistency (Jagre,Watson,&Watson,2001).However,if the in-congruence is so great that consumers have to change their cognitive structures to comprehend it,they have a tendency to exhibit negative reactions,leading to neg-ative cognitive elaboration and a feeling of frustration (D’Astous&Bitz,1995).
PROPOSITIONS
Finally the effectiveness of celebrity endorment ems to be explained by two types of variables:The advertiment processing could be quite holistic and bad upon affect:a positive affect toward the celebrity could lead to a halo effect and being transferred to the brand.For instance,someone who is a fan of a certain star will have a positive affect toward all the brands this star is talking about positively.This can be due to the way that person trusts this star or to a simple mimetic and projective effect.In that ca,attitude to-ward the celebrity would be relevant as a measu
re of attractiveness,whatever the reason(physical attrac-tiveness,personality,etc.).
On the other hand,this processing could be quite analytic and bad on cognitive inferences.Here,the reasoning will be bad on the reason why the celebrity endors the product.Congruence could then be evalu-ated as a function of the experti and/or the credibility of the celebrity when he/she has to reprent the brand. In that ca,brand/celebrity congruence would be rel-evant as a measure of how well both of them match together.
So thefirst question is to know which of the two variables would have the greatest effect.Another ques-tion is to know if people would be influenced by their affect to estimate brand-celebrity congruence.Would a person who really likes a celebrity alwaysfind him/her congruent,whatever the context?So,is there any cor-relation between attitude toward the celebrity and con-gruence evaluation?
MODEL CONSTRUCTION
To address the rearch question,the model depicted in Figure1has been developed.This model is inspired by Brown and Stayman’s(1992)meta-analysis,as well as recent rearch by Bergkvist and Rossiter(2008)and our earlier discussion.Brown and Stayman’s(1992)findings support the dual medi
ation model(MacKen-zie,Lutz,&Belch,1986)“which posits a direct effect of ad attitude on brand attitude as well as an indirect effect via brand cognitions”.In their study about adver-timent likability(L ad),Bergkvist and Rossiter(2008) assd the effect of L ad on brand attitude(A b),brand beliefs,and purcha intention.In their model,the three latter variables are measured at two points in time.Each of the influences its ,A b measured at t0has an influence on A b measured a few weeks later,etc.),according to the idea that the best predictor of an attitude or a belief is the prior attitude or belief.
654FLECK,KORCHIA,AND LE ROY
Psychology and Marketing DOI:10.1002/mar
广播节目>拳师狗Figure2.The design of the rearch.
This model includes brand perceptions and purcha intentions,which improves nomological validity.Exam-ination of the relationships will make it possible to com-pare the respective effects of congruence and attitude toward the celebrity on pre-attitude toward the adver-timent(PTA).It would be awkward to u the con-cept of attitude toward the advertiment,even though it is widely ud in this kind of model,becau in this study subjects do not e a real advertiment.Actu-ally,showing a real advertiment would have intro-duced some other bias linked,for instance,to its ex-ecution.It is possible to significantly change the style and therefore the image of a celebrity for a picture. Thus,t
he subjects were simply expod to the princi-ple of associating a celebrity with a brand in an ad-vertiment for that brand,and were shown a picture of the celebrity and the logo of the brand.This is why it is rather a question of pre-attitude toward an ad-vertiment combining celebrity X and brand Y than a question of attitude toward the advertiment.To as-ss concepts before and after the stimulus,the study was conducted at two points of time(A and B).Like in Bergkvist and Rossiter’s(2008)model,brand beliefs have a direct and indirect(through attitude)effect on purcha intention.
After exposure to a stimulus(brand endorment by a celebrity),it can be expected that congruence(and its two dimensions,relevancy,and expectancy)and/or atti-tude toward the celebrity influence PTA.Also,as men-tioned earlier,we expect an“affective halo”to occur:an individual who appreciates a celebrity may think that she/he is congruent with whatever brand is endord. As demonstrated by Brown and Stayman(1992),PTA has a direct and indirect influence(through brand be-liefs and attitude)on purcha intention.As explained by Lafferty,Goldsmith,and Hult(2004)and Bergkvist and Rossiter(2008),all constructs measured in A(first period)will be strongly related to their counterparts’measures in B(cond period).DESIGN OF THE RESEARCH
In order to test this model,an experiment was con-ducted with before/after measures(three weeks be-
tween)and control groups.Figure2summarizes the re-arch design.This experiment manipulates the combi-nations between brands and celebrities so that the level of congruence between the celebrity and the brand and the likability of the celebrities could vary.Thus,the experimental design was a2(brands:Dior and Yves-Saint-Laurent)×4(celebrities for each brand,varying in terms of relevancy−high,low−and expectancy −high,low)×2(time period−before,after expo-sure)mixed factorial design,with a parate control group,which was not expod to any celebrity-brand combination.This variation in brand/celebrity congru-ence and attitude toward the celebrity may enable us to explain the eventual evolution of consumer respon to the brand/celebrity combination,may it be cognitive, affective,or conative.As explained above,predisposi-tion toward the advertiment was added to the model as a mediator.
Choice of Brands and Celebrities
All the pairings werefictive,but real brands and real celebrities were ud to increa the ecological validity of this study.Two luxury brands were lected(Yves Saint-Laurent and Dior).This is relevant with the con-text of the study,becau luxury brands u celebrities extensively in their advertising.The sample was cho-n according to the usual profile of luxury brands’cus-tomers,with subjects aged between18and65,and with high incomes.We also lected four celebrities for each br
and,chon to posss various levels of congruence, likability,and image.Each respondent of the treatment group was expod to only one condition of congruence (one brand/celebrity combination),and in the control group there was no exposure.
CELEBRITIES IN ADVERTISING655 Psychology and Marketing DOI:10.1002/mar
