Super Bowl advertisers are shortchanging their brands by relying so heavily on celebrities, according to new research from System1 Group. System1 tested consumers’ emotional reactions to Super Bowl ads of the past four years (2020-2023) and found that commercials featuring brand characters consistently outscored celebrity spots for appeal, brand recognition and commercial impact.
Ads with brand characters and branded situations (e.g., Snickers “you’re not you when you’re hungry” setups) – what System1 calls fluent devices – averaged 3.8 Stars, yet only 10% of ads used them.
By contrast, 39% of ads featured celebrities and averaged only 2.7 Stars. System1 scores ads for brand impact on a scale of 0 to 5.9 stars. What’s more, brand characters outscored all other ads by a similar margin. And brand character ads registered greater immediate sales potential than celebrity, music, sports and non-celebrity ads. For example, brand characters averaged a 1.38 Spike Rating, compared to 1.24 for celebrity ads, demonstrating a significantly higher propensity to spike sales over 10 days.
Despite the big budgets and creative attention dedicated to Super Bowl commercials, System1 found their brand Fluency (i.e., strength of brand recognition) is lower than that of typical US ads. Since 2020, the average brand recognition score for Super Bowl ads has dropped from 85 to 83, while the score for general US ads is 85. The issue? Advertisers aren’t introducing their brands early enough or making them central enough to the ads.
Last year’s M&M’s Super Bowl ads bear this out. Early in the game, when M&M’s pretended to change their candy mascots by replacing the hard shell chocolate characters with Maya Rudolph, a popular actress and former “Saturday Night Live” star, the ad received a low 1.0 Star Rating. Emotional response trended negative. Later in the game, when M&M’s aired an ad with their candy characters front and center, they received a 4.8 Star Rating, the second highest of any of the 2023 Super Bowl ads.
“Almost 20% of viewers leave Super Bowl ads not being able to recall what brand the ad was for. This is causing serious wastage,” said Jon Evans, chief customer officer at System1. “Can you explain your creative idea without mentioning your brand, service or product? That’s a red flag.”
System1 tested the 308 Super Bowl ads airing between 2020 and 2023 with 46,200 respondents in the United States, then consulted its extensive database of over 150,000+ ads for comparisons.
When the System1 Star Rating is combined with ESOV (Excess Share of Voice), it is a much more accurate predictor of market share growth than ESOV alone. When looking at projected growth based on ESOV alone, correlation stands at about 25 percent. However, when you account for System1’s Star Rating and ESOV, there is an 85 percent correlation between projected growth and actual growth.