Have political ad researchers discovered the holy grail?
Thoughts on ads, research, and social media before Election Day
(Editor’s Note: Tuesday is election day here in California, along with 15 other states. Make sure to vote if you haven’t already. While you’re at it, please share this Substack with friends or colleagues that may be interested! )
A study was released last week with the intriguing headline “Researchers discover what makes a political ad truly persuasive.” I was eager to finally learn the answer to the question I’m asked most about political spots: “do these ads even work?”
Yet the conclusion from the authors was the unsatisfying “the unifying theme of the results is inconsistency.”
In my career I've seen great ads get axed thanks to research. Recent studies show technology will play an even greater role in fine-tuning what we all see - including in political ads.
Essentially, the study found that political consultants have no idea of what makes for an effective ad, but when they use testing, especially with new, super fast, affordable data-driven analytics that are available on platforms like Swayable, who (surprise!) provided the data for the researchers, they can predict better results.
This is not exactly an “a-ha” moment for most brand marketers, who tend to test every single idea to an inch of its life. I can tell you though that in political campaigns the problem with research always boiled down to two things: time and money, which is something most campaigns don’t have nearly enough of.
The study does offer a path to more effective ads, although in the rapid response world of politics more data is likely needed on just how fast — and cheap — ad development, testing, and actual production can be done.
Among the insights:
Even in high-stakes political advertising with experienced consultants, assumptions about message effectiveness often fail.
Optimizing through testing provides a disproportionate edge to the best-funded campaigns.
The average effects of the ads tested were small but meaningful politically - between 0.8 and 2.3 percentage points depending on the election type.
“What works" changes each election as cultural sentiments shift. Adaptability matters.
Drilling down just a bit deeper, the core insight around optimization is that the more a campaign can afford to test (i.e. produce multiple ad concepts and evaluate them experimentally), the greater their odds of discovering the most effective messaging approaches for that particular election cycle. This allows for optimal ads to persuade substantially better than average or untested ones.
So in environments where testing is used, the benefits disproportionately go to resourced campaigns able to fund more trials. And the ads convert better once the spend is made.
But what if political ad makers think they know better? An additional study by the same researchers may offer lessons for both political and brand marketers.
Why Intuition Fails: Political Guru Guessing Games
Cable news is full of political pundits. A few of them are pretty smart and know their stuff. Others, not so much. But another study by these same researchers pretty much blows the doors off the idea that these insiders have more insight on what makes for effective communications than the general public.
“(T)he political practitioners who are responsible for determining how the public is persuaded have poor baseline intuitions about how to persuade the public.”
While this may seem to be just an issue for the political world, I’m pretty sure more than a few agency creatives and senior agency leaders, as well as clients 😲 have the same problem.
I know how many feel about over-testing, and I totally get it. But tech is changing, and so is advertising.
We’re already pretty deep in the analytical weeds in this issue of ctrl+alt+delete, and the study paints a pretty bleak picture of the political world’s understanding of advertising effectiveness so I won’t dive to deep.
But if you want to learn more paid subscribers can get full findings and insight from both of these studies.
Social media ads are missing the target…by a lot.
I’ve been looking at tons of ads and media buys for the course I’m teaching on Political Advertising in the 2024 Election at the USC Annenberg School. It must be punishment for something I did in an earlier life, because the majority of this stuff is a tough watch — not gonna lie.
You probably know about social media platform Snapchat. With a market cap of $18.4B, it has an audience that skews very young. In fact, the most popular age group on Snapchat is 15-25, which makes up 48% of users.
Snap is known for its unique image filters along with the ability for users to share content like images or texts that disappear almost immediately.
It’s “short attention span theater” with the best performing ads meant to be 3 to 6 seconds.
Here is an example of an effective Snap ad.
So how are political campaigns doing?
CA Senate
By far the biggest political spending on Snap is happening in the CA US Senate race. Here are two ads (each) from Adam Schiff and Katie Porter.
Thoughts: These spots are perfectly fine, but are they right for this platform? Maybe not a total swing and a miss, but def not MLB-level stuff. The 1st ad is a :30 that is running on Snap. Good luck. The mom targeted ad doesn’t quite hit the Snap audience where they live, but conceptually I get it.
Thoughts: Oh Katie. This is not good. You need young voters, You’re a single mom with three kids. You are relatable, and your whiteboard is a winner. So instead: 1) You choose to drive GOP votes in CA on a social app overwhelmingly used by 18-24 year olds? and; 2) The good news? It’s 6 seconds long, perfect for Snap. Bad news? This is not exactly an audience looking to make voting choices based on “earmarks.” As one of my students noted, most all Snap viewers won’t even know what the word “earmark” means. Neither would most voters.
Bottom line: opportunity missed with an audience that is essential to her reaching the coveted #2 slot on Tuesday.
Presidential Campaign
Thoughts: Not a lot of presidential candidate sponsored ads on Snap yet, but this spot from a union (AFSCME) supporting Biden is on message, and at least they get to the key message in under 6 seconds.
Thoughts: No, no, no, And no again. Where to start? OK, Ronna McDaniel is no longer chairwoman of the Republican Party. That happened like a week ago, but this ad is live now and scheduled to run until March 16th, with no other spots in sight. Bad, right?
Not as bad as trying to raise money from 18-24 year old Snap users. Ugh.
Other Campaigns
Thoughts: Cutting a super short web version of an already running :30 can be tricky, but the Sherrod Brown campaign does a fine enough job in getting it done here. Not sure the steel issue tests through the roof with young Ohioans, but it can’t hurt.
Thoughts: Right copy, right tone, right visuals. Simple message. This is how it’s done.
Other approaches:
Thoughts: Fundraising on Snap might be hard but wow, does AOC know how to communicate and connect. Especially with this audience. They may even watch it past 6 seconds.
What’s Next?
Election Day is Tuesday. Expect no great surprises at the national level. All eyes here will be on the CA Senate race, with the top two finishers moving on to November.
We will keep an eye out on things for you. Until then — happy advertising!
PS - Share any ads or content (good, bad, or awful) here.
Also on the lookout for contributors. If you have something to share about the ad world and Election ‘24 reach out.
Thanks for reading!