Creating the Perfect Super Bowl Ad

Super Bowl Sunday is sure to bring the usual barrage of clever soda, beer, and car ads featuring everyone from Vegas showgirls selling Coca-Cola to pop celebrity Psy charming us to buy pistachios Gangnam-style.

But at a whopping $3.7 million per 30-second spot, perhaps it's time for advertisers to question whether ads, including the over-the-top Super Bowl spots, have become too entertaining, says Thales S. Teixeira, an assistant professor in the Marketing unit at Harvard Business School.

In upcoming research, Why, When and How Much to Entertain Consumers in Advertisements? A Web-based Facial Tracking Study, Teixeira and colleagues question whether advertisers are striking the right balance between entertaining and promoting their brands. Could companies entertain less and get consumers to buy more?

Teixeira's research offers insight to modern-day Don Drapers, perhaps showing advertisers ways to spend ad dollars more shrewdly.

"They need to use entertainment judiciously," he says. "It's very expensive. They have to ask: Why do we need to entertain so much?"

During an interview in his office, Teixeira, who has analyzed why ads go viral, reviews a quick history of the use of entertainment in advertising. In 1963, British advertising pioneer David Ogilvy, in his book Confessions of an Advertising Man, argued that ads should be all business, selling products by informing consumers rather than entertaining them. A few Ogilvy classic lines that reflected the information-only mantra included: "At 60 miles an hour the loudest noise in the new Rolls-Royce comes from the electric clock," and the straightforward "man in the Hathaway shirt."

By 1985, Ogilvy changed his tune, agreeing that a little fun is fine, Teixeira says. (Proof of that attitude is clear in Ogilvy's award-winning Sprite Shower campaign, which features men, women, and children splashing under Sprite-logo showers on a Brazilian beach.)

Teixeira does not argue that entertainment—fun content, creative stories, upbeat music—isn't used effectively in ads. But many companies don't use entertainment to their best advantage. To find out what works, Teixeira went directly to consumers, partnering with the MIT Media Lab and Waltham, MA-based Affectiva, which has developed the first online facial tracking system for testing advertising.

The research team worked with 275 randomly selected people, showing them a sample of 82 ads for 30 brands in their homes or workplaces.

The research relied on webcams that track visible changes on a person's face as they watch an ad, responding to funny, or amusing content. The software detects the slightest deviations of facial movement in the eyes, cheeks, and mouth, tracking whether a person smiles, grins, or breaks out in laughter. Continuously measuring smile responses and facial cues allowed the team to determine a person's interest in viewing the ad and their intent to purchase the brand.

The ad watchers were monitored remotely, a setup intended to be unobtrusive and provide a more natural out-of-lab environment, Teixeira says. Also, participants were allowed to skip through commercials they didn't want to watch. They viewed ads in three product categories: confectionary (Skittles, Snickers, M&Ms, and others), beverage (including Coke, Pepsi, Mountain Dew), and alcohol (Budweiser, Dos Equis, Captain Morgan, etc.). The team built a model relating the amount of entertainment a participant experienced with both their viewing interest and their intent to purchase a brand.

Entertaining Only So Far

Results determined that the level of entertainment increases a viewer's interest in watching an ad. After a certain amount, however, that persuasiveness starts to decrease and advertisers lose the link to the brand and the power to influence buying. In other words, too much entertainment can actually reduce viewers' desire to buy.

These observations can help companies strategize. Ideally, an advertiser that seeks acquiring customers should entertain initially to capture attention, but also introduce the brand early on, and follow up with a moderate amount of entertainment, Teixeira says.

"If you want to persuade consumers as do Pepsi, Skittles, and Coke, it's about getting people to associate your brand with fun," he says. "But you have to show the brand and then entertain. That's when the conditioning occurs. The other way around—entertain then brand—doesn't work as well."

Three Pepsi MAX ads from Super Bowl 2011 viewed in Teixeira's study show that Pepsi does not always maximize its entertainment/purchase intent balance, he says.

Pepsi MAX's Torpedo Cooler ad (also known as "Catch"), for example, featuring a beer can-shooting ice chest, scored low (13^th percentile) on both entertaining and purchase intent, according to Teixeira's facial tracking study. Pepsi Max's First Date, an ad portraying gender differences of a couple on their first date, fared a bit better. Viewers found it extremely entertaining—as she judges him as possible marriage material, he is thinking of the bedroom rather than the altar—and 81 percent viewed it until the end . Yet it fell short for purchasing intent in the low 22nd percentile score. "First Date has too much entertainment," Teixeira says. "People view but are not very persuaded by this ad."

The most successful Pepsi MAX effort, Love Hurts, features a woman trying to get her husband to eat better by snatching a burger from his mouth and kicking him under the dinner table when he orders fries. Sitting on a park bench, the couple agree that drinking Pepsi MAX is OK—until a pretty jogger waves at the husband. The wife throws her can at her hubby, missing him and hitting the jogger instead. The viewers found the ad moderately entertaining yet it generated a high purchasing intent in the 77th percentile, "striking a balance between entertaining viewers to grab their attention and maximizing the brand's persuasiveness," Teixeira says.

Importantly, he adds, all three ads had exactly the same amount of information content on Pepsi MAX.

Advertisers should pay attention to that balance, taking into account not only how soon into the ad that the brand is presented, but also how the branding aligns with the entertainment. Many advertisers choose to show the brand name at the end, when it's too late. Others entertain too much, failing to connect the entertainment parts effectively to the brand.

If the balance is off, Teixeira says, entertainment competes with brand awareness and reduces purchase intent. But when entertainment follows the brand introduction a cooperating effect is created, which increases both interest in the brand and intent to buy. The idea is aligned with Pavlov's theory of classical conditioning, he says. If you show somebody a stimulus (the brand) and then you show qualities (fun) right after, people better associate those two things.

Finally, advertisers should consider the stage of the consumer purchasing cycle they aim to impact: do they want to create awareness, interest, or spark a purchase? If the goal is to simply create a buzz, then entertain as much as possible—as many advertisers will do during the Super Bowl. If the intent is to convert consumer interest to purchases it might be time to reduce the volume on talking babies and dancing monkeys and up the attention on brand.

Teixeira is now talking with companies that are trying to connect television ad campaigns with online viral advertising. "What companies tell me is that they want to relate this research to viral advertising—to get people's attention and then to get them sharing. They want to understand how to use these media collectively, such as by spending money on the Super Bowl to get attention and then saving money by using viral ads to get engagement," he says.

In whatever medium, the trick remains the same: strike a balance that holds consumer attention optimally in both media. "If you entertain too much on TV, you're wasting an expensive opportunity to sell."