Persuasive Advertising - J. Scott Armstrong [111]
In two lab experiments, subjects who received a print ad where a relevant celebrity, a track athlete, endorsed a new brand of athletic shoe had higher purchase intentions than did subjects who received the same ad with an irrelevant endorser (Goldsmith, Lafferty, and Newell 2000).
On the issue of avoiding celebrities who advertise other products, 20 print ads were shown to 461 subjects in a lab experiment. Actors Dustin Hoffman and Matthew Broderick were used as endorsers for Visa credit cards, Kodak film, Colgate toothpaste, and Certs mints. Using an experimental design, variations were made in the number of products endorsed by each celebrity. As the number of products endorsed increased from one to two to four to eight, the credibility of the endorser went down—as did purchase intentions (Tripp, Jensen, and Carlson 1994).
6.7. Expert endorsements
In the 1870s, Henry Parson Crowell, founder of the Quaker Oats Company, was one of the first to use expert endorsements in ads.
6.7.1. Consider support from an expert unless the target market already agrees with the message.
This principle is more effective for high-involvement products because people will be more likely to consider experts’ opinions. Expert endorsements are expected to be most effective when the target market lacks awareness of the brand or firm.
When the advertiser’s intent is to support current attitudes and behavior, expertise is unnecessary. Interestingly, in this situation the use of expertise is likely to be harmful to persuasion, perhaps because it leads the people to rethink their belief. When the intent is to reinforce existing opinions, use other approaches, such as customer testimonials.
When using credible experts, introduce them early in the ad. That helps to establish the credibility of the claims that follow. Conversely, if the source lacks credibility, put it at the end, so that people focus on the message, not the source.
The effect of expertise drops over time because people forget about the source. Thus, high expertise is more important when the advertiser expects a quick response.
The expert source in an ad should be someone that the target market recognizes. In 1955, P&G introduced Crest toothpaste with stannous fluoride. On August 1, 1960, the American Dental Association (ADA), a well-respected source, endorsed Crest’s claim that the fluoride in its toothpaste prevented tooth decay. Crest had been advertising this benefit prior to the ADA announcement, but with only modest success. Following the endorsement, Crest’s market share rose sharply in a little over two years.
The expert should be independent of the advertiser and easy to locate, as the Crest example illustrates. However, this is not always the case. For example, in 1935, Fleishmann’s Yeast showed pictures of medical doctors in various countries and provided their statements on the health benefits of Fleischmann’s. It was never determined whether these were real doctors, or whether they even existed.
The mass media have reported that endorsements by experts have substantial effects on sales. For example, in 1989, sales of Buick’s LeSabre automobile allegedly rose 62 percent after it advertised that J.D. Power, a consumer satisfaction survey company, had rated it as the most trouble-free American car.
Evidence on the effects of expert endorsements
A meta-analysis of 114 field and lab experiments found that expertise had a much stronger effect on persuasion than any of the other four elements that were examined: credibility, trustworthiness, similarity, and physical attractiveness (Wilson and Sherell 1993). However, in five studies with ten experimental comparisons, the use of credible sources harmed persuasion when people already had a strong favorable opinion (O’Keefe 1987).
In lab experiments,