Persuasive Advertising - J. Scott Armstrong [110]
Celebrities seem to be common in ads because their use is concentrated on high-circulation media. Of the 56 ads in the 1994 Super Bowl TV broadcast, 77 percent used celebrities (Kanner 2004, p. 48). In contrast, our analysis of 480 print ads from WAPB found that only 6 percent used a celebrity.
Evidence on the effects of using of celebrities to gain attention
Lab experiments show that even monkeys will “pay” to see images of high-status monkeys. In one study, they traded juice for views of other monkeys (Deaner, Khera, and Platt 2005).
In a lab experiment, 283 subjects viewed B&W print ads featuring a picture of Pierre Varnay, a fictional French Olympic cyclist, and Avenix, a new brand of bicycle. Subjects then received three magazine articles, one of which said that Varnay had engaged in steroid use and that he had multiple drunk-driving infractions. This information led to lower ratings for Avenix bicycles (Till and Shimp 1998).
Our analysis of quasi-experimental data supports the use of celebrities as a way of gaining attention:
Print ads with celebrity endorsements had better recall. Our WAPB analysis found ten pairs of print ads in which one ad featured a celebrity, while the other ad did not. Recall for ads with celebrities was 1.65 times better than for the other ads.
On the issue of negative information about celebrities, an analysis was conducted of 42 news events in which a celebrity spokesperson was blamed for doing something wrong, such as wife-beating. The value of the firms’ stock prices typically decreased in the days following the publicity (Louie, Kulik, and Jacobson 2000).
6.6.3. When using celebrities, make sure they are relevant and credible
Celebrities should be relevant both to the product and the target market. This principle seems obvious.
A famous actress, Lillie Langtry, endorsed Pears Soap in England in the late 1800s, saying, “I have much pleasure in stating that I have used your soap for some time and prefer it to any other.” Langtry was seen as a relevant endorser for soap because of her good-looking skin. In another example, in 1914, the Italian tenor, Enrico Caruso said, “Tuxedo [pipe tobacco] does not irritate my throat.”
In a survey of 80 advertising agencies in the United Kingdom, the 131 respondents rated the importance of factors involved in the selection of celebrities. The most important factors were the celebrity’s match with the target market and with the product. Here is their recommended list of factors to use for selecting celebrities along with their importance ratings (Erdogan et al. 2001):
Criteria for selection of a celebrity
Rating by experts*
Celebrity–product/brand match
4.6
Celebrity–target audience match
4.6
Overall image of the celebrity
4.6
Cost
4.3
Trustworthiness
4.3
Likelihood of acquiring the celebrity
4.2
Risk of controversy
4.1
Celebrity familiarity
4.1
Prior endorsements
4.1
Likeability
4.0
Risk of celebrity overshadowing brands
3.9
Stage of celebrity life cycle
3.6
Expertise
3.3
Physical attractiveness
3.1
* 5 = very important; 1 = least important.
Avoid celebrities who endorse other products. Some celebrities are so popular that many advertisers want to use them. For example, in the 1985 Super Bowl broadcast, the football player Joe Montana endorsed soft drinks, cars, sporting goods, and razors.
Unfortunately, celebrities often lack credibility. To improve trustworthiness of a celebrity, consider having them use a two-sided argument.
Evidence on the effects of a relevant match of celebrity and product
A