Persuasive Advertising - J. Scott Armstrong [143]
However, the applications often overlook key aspects of the principle. Our WAPB analysis found that of the 480 tested full-page print ads by leading U.S. firms, only 22 percent used a novel metaphor, and only 32 percent used one related to a selling point. Overall, only 16 percent used a metaphor that was both novel and related to a benefit.
Evidence on the effects of metaphors
Sopory and Dillard (2002) found 29 empirical studies with 38 comparisons of literal statements versus metaphors. The researchers used diverse methods in these studies, which involved about 4,000 subjects. Metaphors were persuasive (a correlation of 0.07) with respect to attitude change. The authors provided correlations for each of the relationships. (Correlations provide a poor way to communicate the strength of relationships, so think of them as a measure of relative importance of the relationships in the discussion that follows.) This meta-analysis showed that:
• Metaphors enhanced the credibility of low-credibility sources; as a result, they were more persuasive (a correlation of 0.12). Highly credible sources had little need for metaphors (a correlation of 0.02).
• Novel metaphors were much more persuasive than familiar ones (a correlation of 0.12 for the novel ones versus 0.01 for the familiar). Novel example: “Now the great cloud cat, darting out his lightning tongue, licks the creamy moonlight from the saucepan of the sky.” Familiar example: “She has a heart of gold.”
• Extended discussions of a metaphor were more persuasive than brief mentions (a correlation of 0.09 versus 0.05).
• Oral metaphors were more persuasive than written ones (an average correlation of 0.09 for audio experiments versus 0.06 written).
• Persuasiveness dropped off as the messages went from including one metaphor (a correlation of 0.08), to two through eight metaphors (a correlation of 0.06), to nine or more metaphors (a correlation of 0.02).
Sopory and Dillard also found that under ideal use—a single extended novel metaphor introduced early in the message—metaphors were especially persuasive.
In a lab experiment, 13 actual full-page magazine ads were shown to 103 subjects. The ads had metaphors with either concrete or abstract words. The concrete metaphors related the product to other objects through the five senses, while abstract metaphors used an intangible trait or concept. The subjects reported that the concrete metaphors were more understandable. In addition, the three ads that presented visual support for metaphors led to higher comprehension than the three that did not (Morgan and Reichert 1999).
In another lab experiment, 242 subjects viewed ads for eight products in a test magazine. There were two versions of each ad, one with metaphors, either though words or pictures, and the other ad without. For example, a flashlight ad said, “The gift idea that leaves everyone beaming,” while another ad substituted “happy” for “beaming.” Recall was twice as high for the ads with metaphors (McQuarrie and Mick 2003).
Our analysis of data on quasi-experimental data on print ads supported the value of concrete words in metaphors:
Print ads using metaphors with concrete words had better recall. Our WAPB analysis found 37 pairs of print ads in which one ad used concrete metaphors, while the other did not. For example, an ad for Panasonic that said, “Panasonic folding cordless. It’s an open and shut case,” had a much better recall score than that of a Panasonic ad that used no metaphor. Recall for ads with metaphors was 1.34 times better than for the other ads.
We also analyzed non-experimental data in WAPB. In comparison with the industry norms for each print ad, the average persuasion score for the 42 ads with concrete metaphors was 8 percent higher than the comparable score for the 180 other ads.
7.9. Simplicity
Simplicity was a revolutionary advertising concept in the early 1900s when John E. Powers, advertising manager for John Wanamaker’s department stores, began to use simple,