Advertising Principles - Evidence-based principles

Should Super Bowl Ads Emphasize their Brands—and follow other evidence-based principles?

Should Super Bowl Ads Emphasize their Brands—and follow other evidence-based principles?

To gain attention, advertisers on the Super Bowl ignore many persuasion principles. Here are two examples:

Mystery Ads: Many companies used “mystery ads” in the Super Bowl. That is, they decided to conceal the name of their brand. About 90% of print ads by major brands regard their brand as a way to communicate useful information. Experimental evidence to date shows that by prominently mentioning the brand name at the beginning of an advertisement, persuasion would increase by 44%. Assuming a $500,000 ad that broke even, the firm would have earned an additional $220,000. (See Principle 5.5.1 in Persuasive Advertising.)

Double Branding: Procter and Gamble ran a Super Bowl ad in 2017 that ignored the principle of “double-branding.” The idea is that companies with well-respected reputations should use the company’s brand name as well as the product’s brand. So P&G could have advertised “Procter & Gamble’s Tide.” Experimental research showed that this would enhance persuasion by 70%. Given that the ad cost about $500,000 and assuming that it did no better than breakeven, they could have earned $350,000. (See Principle 5.5.2 in Persuasive Advertising.)

Why do companies overlook such opportunities? Designing an ad is a complex process. There are 195 evidence-based principles and about 1/3 of these are relevant for a typical TV commercial, so this cannot be solved in one’s head—even if a person was familiar with all of the principles. However, you can hire non-experts at about $10 per hour (for example by using Amazon’s Mechanical Turk), and with the help of a one-hour training self-training course and freeware, they can rate an ad in less than 30 minutes. They would see when principles, such as the two above, were ignored. If you doubt that evidence-based principles are useful for advertising, read Predictive Validity of Evidence-Based Persuasion Principles.

Those creating a 2018 Super bowl ad might want to spend a few hundred dollars to test their commercials to ensure they use evidence-based persuasion principles. They are freely available at


5.5.1. Make the brand or company name prominent if it conveys a good image

In the successful Bose Wave Radio print ad, “Why you should pay $349 for this radio,” Bose was mentioned ten times.

        A grapefruit ad presented a series of statements, including, “You can tell a melon’s perfect by squeezing it, you can tell a pear’s perfect by sniffing it, and you can tell a banana’s perfect by peeling it.” Then, under a photo of a grapefruit with “Jaffa” stamped on it, it stated, “You can tell a grapefruit’s perfect just by reading it.”

        A focus on brand name is especially important when the brand enjoys a well-known and favorable image among the target market. Firms expend enormous sums to enhance their brand and company names to elicit favorable feelings in their target market—it’s like seeing an old friend.

The brand name should typically be placed at the beginning and end of an ad, in logical places in the text, such as in illustrations and captions. Space the repetitions throughout the ad so that it is subtle. When feasible, ads should include clear visual and audio identification of the brand.

To show that a company will be there for customers in the future, emphasize that the company is large and successful, has been in business for many years, and is growing. This is especially important for a Glossary Link high-involvement product. For example, Mayflower, a household moving company, emphasizes that it was founded in 1927.

If the firm is small and not well known, provide a street address, rather than a post-office box number, to show that it can be located.

        The condition “if it conveys a good image” is important. “Mystery ads,” those that conceal the brand and the seller until very late in an ad—or not at all, should only be used where the advertiser does not have a good reputation.

        If a strong argument exists and the brand and company image is not favorable to the target market, mystery ads are appropriate. A mystery ad with a strong argument might draw customers into the ad, whereas the company or brand image might lead people to skip the ad. For example, an ad from a Republican candidate might immediately turn off a potential Democrat voter even if the message would be useful to that voter.

        Nokia, a well-known and respected company, used mystery ads in a 2004-2008 Glossary Link campaign. Its large color ads, which appeared in the Wall Street Journal and other media, never mentioned the brand, product, or company. One contained the picture of a man jumping into a swimming pool with the Glossary Link caption, “splashed at 2:14.” I asked people what they thought was being advertised and hardly anyone knew. Finally, I found a version of the ad with text describing Nokia cell phones. Later ads also ran with no clues.

In the late 1990s, several well-known companies used mystery ads. Foot Locker, Oldsmobile, Reebok, Sony’s promotion of a movie (Godzilla), and Lee Jeans “leaders” were called leaders in this movement.[1]  Given that they are all well respected companies, it seems odd that they would use mystery ads. And, in retrospect, these examples do not provide good endorsements for mystery ads: Foot Locker’s stock price dropped sharply right after this campaign, Oldsmobile went defunct in 2004, Reebok was acquired by Adidas in 2005, Sony’s Godzilla was a judged an economic failure, and Lee Jeans suffered a sharp stock decline soon after the ads were run.

This principle is widely used. On average, brand names were repeated almost four times in a sample of mostly 30-second TV commercials by leading advertisers. In addition, 95% of the commercials used visual sign-offs and 71% used auditory sign-offs (Stewart & Furse 1986).

        Our WAPB analysis found that, of the 480 print ads (almost all from well-known brands), 93% repeated brand names. In addition, 89% of them mentioned the brand name near the beginning of the ads.

Evidence on the effects of prominent mention of brand names

                Our analysis of quasi-experimental data supports this principle:

Print ads that displayed brand names prominently had better Glossary Link recall. Our analysis of WAPB found 18 pairs of print ads in which one ad displayed the brand name in prominent manner (with at least three repetitions) while it was not prominent in the other ad. The ad with the prominent display was recalled 1.44 times better than for ads that did not display the brand name as prominently.

        In a small-scale experiment, 21 TV commercials for high-involvement products, such as cars, computers, and automobile tires, were shown to 228 subjects. As part of this experiment, the researchers tested recall for four unknown products. Brand recall was better for each of the unknown brands when the brands were mentioned only at the end of the ads, thus supporting the use of mystery-ads for unknown brands (Fazio et al. 1992)

Our analyses suggest that mystery ads performed poorly with print ads for well-known brands, even when readers could jump to the end to see the brand:

Ads with only late brand identification had lower recall and persuasion. Our WAPB analysis found 22 pairs of ads for well-known brands in which one mentioned the brand early and the other mentioned it only at the end. Recall for the “mystery ads” was only .69 times thatof the other ads.


      Analyses of non-experimental data on TV commercials are relevant as this principle is not highly dependent on the conditions.

       Recall for an analysis 30-second TV commercials was 6% below-average when there was no mention of the brand, 4% below for one oral mention, 1% below for two, and 4% above for three or more mentions. Persuasion was 7% below-average for no oral mentions and 2% above for three or more. A brand audio sign-off improved both recall and persuasion by 10% (Walker 2008).

        An analysis of TV commercials (90% of which were 30-seconds in length) found that the number of times the brand name was mentioned and the number of times the brand name and Glossary Link logo were shown on the screen were positively related to recall (Stewart & Furse 1986). Stanton and Burke (1998) added support, especially for short (15-second) commercials. In an analysis of 832 radio commercials covering 31 product categories, those with more brand name repetitions had better recall (Sewell & Sarel 1986). TV commercials with a brand auditory sign-off (mentioning the brand in the last three seconds) were rated as more persuasive than those that did not include such a sign-off (Stewart & Koslow 1989).

         For the non-experimental data in WAPB, in comparison with the industry norms for each ad, the average persuasion score for the 192 ads that led with the brand was 8% higher than the comparable score for the 14 mystery ads.

 A survey of website users added support: The most important in a list of 30 possible factors that lead to high credibility was, “The site represents an organization that you respect. . .” (Fogg 2003, p.154).

5.5.2.      Include brand and company names (double-branding)


W. K. Kellogg’s invention of corn flakes in the early 1900s was so successful that by 1911 there were 108 brands of corn flakes being sold in Kellogg’s hometown, Battle Creek, Michigan. Kellogg, who wrote his own ads, had always been careful to pair his company name, Kellogg’s, with corn flakes. When the courts ruled that “corn flakes” was merely descriptive of a product and anyone could use the term, Kellogg did not suffer much because most customers did not think “corn flakes,” they thought “Kellogg’s Corn Flakes” (Goodrum & Dalrymple 1990).

        Persuasion is enhanced when an ad includes both the brand and company name—as long as both are familiar and well liked, such as P&G’s Tide.

        Given that “double branding” is a commonly applied and inexpensive to implement principle, it is surprising that more advertisers do not use it. Only 32% of TV commercials from major firms used double branding (Stewart & Furse 1986). The situation was better in our WAPB analysis of 480 print ads by leading firms, as 82% used double branding.

Evidence on the effects of double branding

Our analysis of quasi-experimental data for double branding provided strong support:

Print ads with double branding had much better recall. Our WAPB analysis found 21 pairs of print ads in which one ad included both the brand and company name while the other did not. For example, a Tilt Wheel steering ad that stated "Tilt-Wheel Steering from Saginaw” generated much better recall than did a Tilt-Wheel ad that did not mention Saginaw. Recall for the double-branded ads was 1.71 times better than for ads that did not include double branding.

Non-experimental data provides further support. TV commercials with double branding had better comprehension and persuasion than those that did not (Stewart & Furse (1986). A related analysis found double-branded TV commercials were more persuasive (Stanton & Burke 1998.)

[1] Neuborne, Ellen (1998), “Great ad! What’s it for?” Business Week (July 20), 118-119.

© Copyright J. Scott Armstrong and Kesten C. Green. All rights are reserved.
Web Design by Zoe Communications Ltd.