The Advertising Super Bowl is almost upon us, and we’re looking forward to pretending we know which teams are playing just as much as the next Ad Agency exec. But this year there’s a twist! Two billionaires are running for president (well technically 3 billionaires – sorry Tom Steyer) and they’re both breaking with tradition by taking out dueling $10M bets on what are sure to be the most talked about 60 second spots of the game. With some very small exceptions, never before in history has a political candidate advertised in the Super Bowl. And since this year’s Super Bowl falls just one day before the polls open in Iowa, let’s just call this game what it actually is: The Miami Caucus.
Now what’s at stake in this game is more than just Michael Bloomberg’s candidacy. Although his long-shot standing in 4th or 5th place (and roughly 8% of the polls) is certainly troubling for a candidate with no plans to attend any of the 3 February debates, and zero of the four early-voting states, we can’t neglect the other, more important underdog in play here:
The State of Linear TV Ad Efficacy
While the TV industry is modernizing with audience and performance-based buying that looks like digital and things like business outcome benchmarking help bring scale to attribution measurement what works for brands doesn’t always work for pols.
As it turns out, most of the modern literature concerning the efficacy of TV advertising is pretty bleak. Researchers at Yale, UCSD, and the University of Pittsburgh have all recently cited that most TV ads have only modest effects on people’s vote choice; that heavy campaign spending yields diminishing returns; and that any effects of political TV advertising tends to be short-lived. In one 2016 study, 34,000 people were bound, tortured, and forced to watch 29 consecutive weeks of US Presidential ads. Technically, half of them watched the campaign ads, and the control group watched a placebo ad for Nationwide Insurance featuring Peyton Manning singing in the back of his camper van. (This is still a true story btw…)
The punchline here is that seeing a pro-candidate ad (or even an anti-opponent ad) shifts a person’s perception by about 1%. The researchers are careful to point out that this effect is not zero per-se. But although the small effect has been widely reproduced, it is also not statistically significant.
But what about if you ran a lot of ads? Like $100M worth of ads….
Recent history holds a great example showing that approximately $100M worth of TV ads can buy approximately 2.5% of the Democratic primary vote (if you’re Tom Steyer) or 7% of the primary vote (if you’re Mike Bloomberg). So the open question is, can Mike Bloomberg buy an additional 7% with his $10M Super Bowl spot to push into the top tier for Super Tuesday when 40% of the Democratic delegates are awarded. If that feat is too crazy for him to accomplish, his campaign is pretty much over.
The Super Bowl is the most watched tentpole event on all of TV. Something like 90M people will watch it, which is almost 3 out of 4 US adults. The game will spawn something like 30 million social media interactions. And the ads themselves are the most reshared ads on social platforms, with reach well into the 100M range the day after the Super Bowl.
It regularly registers a Gross Ratings Point of around 40, a number that although declining, represents a very intriguing opportunity for Mike Bloomberg…
For starters, Bloomberg is a centrist and the Super Bowl is going to deliver a big centrist audience. While it’s true that fewer and fewer young people watch the Super Bowl every year—with a recent Super Bowl putting up the worst numbers for that demographic in over a decade—older, wealthier, more suburban Gen X and Baby Boomer audiences remain the core viewers for the big game. So this election for Bloomberg is a game of capturing the attention of Middle America, a large and reliable voting block, represented by the giant cluster of cities on the far right (aka the “most Middle America” cities) of the chart below…
And Bloomberg has excellent traction with these audiences.
Bloomberg’s well-documented strategy is to drive a wedge between the approximately 10-15% of Trump 2016 voters who claim to be “exhausted” with the Trump presidency. This block generally has their minds made up about the further-Left proposals of a chunk of the Democratic field, but their familiarity with Bloomberg is significantly lower. According to the latest Economist/YouGov poll, over 30% of voters are undecided or “don’t know” about Bloomberg, in comparison with 15% undecided for Joe Biden and only 5% undecided for Donald Trump. Simply stated, Bloomberg has considerable room to grow within his own party, and a fighting chance to steal votes from Trump.
Now all he has to do is deliver an amazing ad! The trick is, does he know his audience well enough to sway their hearts and minds? There’s a great argument to be made that learning from A/B tests on Digital can translate into great creative on Television. (Latticework published a piece on exactly this subject last fall) The Trump Campaign mastered this strategy throwing pots of spaghetti at the wall (over 360,000 ad versions to be specific) in order to validate what resonates. The Bloomberg campaign has kept pace, testing over 31,000 ad units since he announced in late November, a rate on par with President Trump.
Bloomberg has also made deep investments in a personal data shop called Hawkfish, and has been poaching top tier talent from the likes of Facebook, Foursquare, Tik Tok, and more and offering outsized benefits like a guaranteed salary until the general election.
With such a deep well of resources informing both Bloomberg and Trump’s creative teams, it’s a safe bet the Super Bowl ads are going to hit some serious nerves. So polarizing, in fact, that Fox is isolating the ads from brand advertising, fearing the backlash that could come from sandwiching a political message between corporate messaging. The last thing they want is another 84 Lumber situation on their hands. Remember, Super Bowl ads generated over 390 million social interactions in the days following the game, so this is definitely going to be a “buckle your seatbelt” moment, right as the 2020 election really heats up.
There is no precedent for Presidential Super Bowl ads, and major measurement vendors will be likely share an abundance of literature on the results. It will be telling in the week and month that follows to see whether this colossal TV buy has a true impact on poll numbers (and even more importantly, votes), or whether the current literature about linear TV’s efficacy will ring true.