Accentuate the Positive

New research shows that a set of simple, psychologically realistic assumptions about consumer memory can shed light on several unusual features of advertising behavior.

Why do familiar brands such as Coca-Cola and McDonald's advertise so heavily? Traditionally, economists have seen advertising as a way of providing information about product characteristics, prices, and availability.

"With the average American drinking 10 gallons of Coca- Cola each year, it's hard to believe there is much left for most consumers to learn about what's inside the can," notes Jesse Shapiro, postdoctoral Becker Fellow for the Initiative on Chicago Price Theory at the University of Chicago Graduate School of Business. "Even if there were, it seems unlikely that images of thirsty polar bears would be the best way to get the information across."

Examples such as Coca-Cola suggest that advertising is about more than just conveying facts, yet economists have been short on theories to account for such "noninformative" advertising. New research by Shapiro aims to fill this gap by taking lessons from the psychology of human memory.

"If you start with the view that advertising is only about information, many features of advertising will seem puzzling," says Shapiro. "The picture becomes clearer if you recognize that advertising might also influence the way consumers encode and recall their consumption experiences."

In the study "A 'Memory-Jamming' Theory of Advertising," Shapiro constructs a model of how advertising affects consumer behavior, based on the idea that advertising can affect the way a consumer remembers his experiences with a product.

The Psychology of Memory

"People don't have incredibly detailed diaries of each time they've eaten a candy bar, how much they liked it, and what the circumstances were," says Shapiro. "In that sense, a consumer's memory is blurrier than their actual experience."

Shapiro's theory rests on a body of evidence about what psychologists refer to as "source memory." Psychological studies have shown that people can quite easily forget the origin of a memory for a fact or sensation. A classic example is the feeling of familiarity when an individual recognizes someone they have met before but whose face they cannot place: though the stranger's face is familiar, the individual cannot remember why. Research in the psychology of memory shows that such failures are common, and when people don't directly recall the source of a memory, they use what they know to fill in the gaps.

Experimental research further supports the view that advertising can affect remembered experiences. For example, in a prior marketing study, a researcher gave participants orange juice spiked with salt and vinegar. Results showed that people who watched advertisements for the juice after the taste test remembered the juice as tasting good, even though what they actually consumed was designed to taste terrible. A simple advertising intervention had the power to change the participants' recollection of the sensory experience of tasting the juice, even in the very short-term.

Shapiro's model transports insights from the psychology laboratory to real-world firm decisions about advertising levels. In his model, advertising can affect consumer memory in one of two ways. First, advertising may enhance the way a consumer remembers a given experience. For example, a commercial showing two people enjoying a cold soda on a hot beach may "implant" in the consumer's memory the impression of having enjoyed the soda more than the consumer actually did.

Second, an advertisement may distort the way a consumer recalls information from memory. For example, a grocery store display for a particular brand might be designed to call to mind only the most positive associations with the brand, or to induce the consumer to remember his or her best moments with the product.

Shapiro uses his model to analyze two important features of advertising behavior in mature markets. First, more familiar brands often advertise more, not less. Second, brand advertising frequently oscillates from low to high, suggesting that the response to advertising displays "increasing returns."

The first prediction of the model is that some types of advertising actually will increase as consumers become more experienced with the product. The reason is that more experienced consumers put more faith in their recollection of past product experiences, since they have more information to work with. The return to manipulating such consumers' memories is correspondingly higher, because a small change in what they remember can have a large effect on what they purchase.

The second prediction of the model is that the return to advertising will display "increasing returns," in the sense that exposing a consumer to one advertisement may actually increase his or her responsiveness to a second ad.

Traditional information-based theories of advertising tend to predict diminishing returns in the consumer's response to advertising, because once a consumer becomes informed about a product he or she is less likely to respond to additional information. However, this prediction is at odds with the frequently observed phenomenon of "pulsing," in which firms' advertising oscillates over time between low and high frequency.

The tendency for firms to prefer fluctuating advertising levels to a constant level suggests that there may be "threshold" effects in advertising: the chance to reach a consumer several times or not at all may be better than a guarantee of reaching a consumer only once.

In Shapiro's model, these "threshold" effects can arise because even one bad memory can convince the consumer not to buy the product.

"If you remember a terrible experience with a product, it must be real, because it can't be the result of advertising," says Shapiro. "For that reason, negative memories carry much more weight in this context than positive memories."

Because of this, the firm will want to "saturate" consumers with good memories, and thus ensure that consumers don't recall any negative experiences when they are deciding whether or not to purchase a product. Strategies that involve "pulsing" or oscillating advertising levels let the firm achieve this saturation at a reasonable cost. Shapiro's model predicts that pulsing strategies will be most pronounced in mature markets, where consumers are likely to have had multiple experiences with the product.

Taste Test

Shapiro tested the model's implications for consumers' response to advertising using a database provided by the James M. Kilts Center for Marketing at the University of Chicago Graduate School of Business. The database links purchases for a panel of consumers to information about in-store and in-home ad exposure.

Shapiro focused on the decision of whether or not to buy Heinz ketchup. He tested whether consumers who had already been exposed to advertising were more likely to change their behavior in response to being exposed to advertising a second time. The data supported Shapiro's prediction: television advertising was more effective for households who had already seen an ad than for those who had not.

Additionally, the frequency of television commercials for Heinz followed a "pulsing" type of pattern, oscillating between reaching a very high and very low proportion of households. This evidence, taken together with the existing research on pulsing and increasing returns in advertising response suggests that firms believe "saturating" the consumer with multiple advertising exposures is important for achieving a significant advertising response.

Shapiro also compared the results for two types of advertising that resemble the types distinguished in the model. TV advertising, which the consumer sees far from the point of purchase, is more closely related to the type of advertising that operates by enhancing the way consumers remember a product experience. In-store displays, which occur close to the point of purchase, seem more likely to distort the way a consumer accesses stored memories at the time of making a decision.

Shapiro found that television advertising was less effective than display advertising for experienced consumers, because for these consumers each memory of an experience is but one of many. Therefore transforming any given memory will count for less in the consumer's overall impression of the product than it would for a consumer with relatively few experiences.

The effect of display advertising was greater for experienced consumers, which supports the argument that in-store displays distort the recall of past experiences and have a greater effect on experienced households, who are more likely to rely on their recollections to guide their purchase decisions.

The Science of Advertising

In regards to current debates over advertising regulation, Shapiro notes that a critical issue yet to be addressed is whether the hype generated by advertising actually makes consumers enjoy their consumption experiences more. He suggests that addressing that question is at the forefront of what is necessary to have a more complete science of advertising.

"There are major public policy debates that require an understanding of what advertising does and how it works," says Shapiro. "Good economic analysis is needed to clear a path to answering those questions."

Shapiro suggests that research studying how ads affect the recall of real experiences is a critical direction for marketing and advertising research in order to make advertising more effective.