Why Clever Ads Can Backfire

SMITH BRAIN TRUST — When it comes to display advertising — especially online — simpler can be better. That's the implication of new research from the Robert H. Smith School of Business, at the University of Maryland.

One theory of advertising holds that display ads need a degree of nuance or visual complexity in order to capture the viewer's attention. But that fails to take into account the increasingly cluttered and hectic context in which ads are viewed today, according to Michel Wedel, distinguished university professor and PepsiCo Chair in Consumer Science at the Smith School.

"A lot of advertising is being tested over fairly long exposures — several seconds, or even 10 to 20 seconds," says Wedel. "The problem is that ads that do well in that scenario may not do well in short exposures." Complexity especially does not pay off online, Wedel says, where eye tracking research shows that people actively try to avoid ads. But billboards and even many print ads are often taken in with a glimpse, too.

A new paper by Wedel, Millie Elsen and Rik Pieters (of Tilburg University) tests reactions to ads over periods as short as 100 milliseconds, which is less than a full "eye fixation," or a full glance. The authors broke ads into three categories. "Upfront" ads, those that present a product in a straightforward, expected, typical way (a photo of a bottle of orange soda to sell orange soda, for example) are grasped and received positively by viewers in those 100 milliseconds, the authors found. They continue to be viewed positively over 5, 10 or even 30 seconds.

"Mystery ads," whose visual complexity require work on the part of the viewer to decode, are viewed less positively than upfront ads in the initial glimpse, but they gain in approval over time, reaching similar levels. One example in the study showed a ninja severing a rope holding a refrigerator, which was about to crush apples to create juice.

A third kind of ad uses a clear image of one thing to sell something different, as when a headshot of a blonde woman is used to sell wheat beer. The authors called that a "false front" ad. Such ads are initially appealing, because they appear comprehensible, but are liked less once viewers reorient themselves to the right interpretation. "We find very little justification for false-front ads," Wedel says. "People don't like to be duped." Sponsored content, ads that take the form of news articles, would fall into this category.

The pleasure derived from ads was closely connected with whether viewers believed they grasped their gist, surveys of test subjects demonstrated.  Positive reviews had little connection to visual appeal, visual complexity, or the ratio of text to image.

The authors tested the three types of ads in three experiments, both in labs and online, involving a total of 1,360 test subjects and 50 advertisements. The advertisements were a mixture of real-life ads and ads modified by the researchers to fit into one of the categories. The first experiment looked at reactions at 100 milliseconds, or 2, 5 or 30 seconds. The second looked at 100 milliseconds, 500 milliseconds, 2 seconds and 10 seconds. In the final experiment, test subjects could look at ads as long as they wanted.

In some cases, the researchers controlled for the recognizability of brands. In others, they looked at the effect of brand recognition on the ad ratings. Test subjects tended to rate brands they recognized higher, but brand recognition did not change the order of preference for the ad types.

In a demonstration that upfront ads are underrated, the authors analyzed a set of 150 recipients of a major advertising prize. 75 percent were mystery ads, 15 percent were upfront ads, and 10 percent were false-front ads.

"We aren't saying that ad agencies shouldn't be creative anymore," Wedel says. There are some contexts when you can be sure an ad will be viewed with great attentiveness, like the Super Bowl. But for online banner ads, for example, advertisers should realize that they'll have only a tenth of a second of a viewer's attention, if that. And so they should stick to the basics: What's the product? And what's the brand?

"Thin Slice Impressions: How Advertising Evaluation Depends on Exposure Duration," by Millie Elsen, Rik Pieters and Michel Wedel, has been accepted at the Journal of Marketing Research.