The rollout of political advertising transparency tools by Google, Facebook, and Twitter, together with the impressive work done by others in the Democratic digital space to visualize this information in meaningful ways, has ushered in an era of unprecedented insight into what campaigns are doing online.
These changes are a welcome step, both for the digital advertising industry and for our democracy. But for the casual political observer, they have also bred a swath of misconceptions that don’t hold up upon more careful analysis.
Misconception #1: All digital spend is intended to reach voters.
Axios recently took a look at the age of Facebook users being targeted by 2020 presidential candidates, noting that the Trump campaign is disproportionately communicating with individuals aged 65-plus and that the most common message being communicated was on immigration. Other publications took this data and ran with it, drawing conclusions like Vanity Fair’s that Trump’s 2020 strategy is “targeting grandpa with anti-immigrant ads.”
This thinking is an artifact of the broadcast age. A clear advantage of digital is the ability to narrow-cast specific messages to specific audiences. Eighteen months out from a general election, and in a world where the Democratic National Committee has set grassroots donation standards for debate participation, sophisticated campaigns are spending the majority of their resources communicating with the small slice of their population who they believe will be their grassroots donors—not necessarily persuadable voters.
Take a closer look at the creative that the Trump campaign is running. The majority of ads that are running in the Facebook ad archive have a donation as their primary ask. Immigration is certainly being discussed. But that’s because Trump’s team is seeing data demonstrating that his base is motivated to give by this issue—not because it is a message they have determined is persuasive to the general public.
This isn’t to say that there aren’t any persuasion ads being run currently. There are, albeit in much smaller numbers, and there are interesting strategic insights to be gleaned from the data we have from these ads. The Trump campaign is currently running a series of Facebook ads on criminal justice reform in a tone quite different from those on immigration. A current video series opens with an emphasis on bipartisan support for criminal justice reform and goes on to prominently feature a news clip from CNN.
Where are those ads being targeted? Not nationally, but to the fourteen states most likely to decide the presidency: Nevada, Arizona, Colorado, Minnesota, Wisconsin, Michigan, Ohio, Pennsylvania, Maine, New Hampshire, Virginia, North Carolina, Georgia, and Florida. Unsurprisingly, the age demographics on this series of ads look significantly different from the aggregate data that was reported on, with the majority of spend going to individuals 25-44 and only a sliver going to individuals 65-plus.
The broader takeaway is clear: Campaigns are using digital advertising for a variety of objectives and each requires communicating with different audiences. Looking at aggregate data ignores that a significant portion of what campaigns are doing online, especially at this point in the cycle, is not geared toward persuadable voters.
Misconception #2: The transparency tools cover all that a campaign is doing online.
To date, Facebook, Google, and Twitter have debuted the most comprehensive transparency tools available. While Facebook and Google should be significant components of any political media plan, they’re merely a piece of it. Analyzing only these three platforms ignores other social platforms (Snapchat, Pinterest, Tumblr, LinkedIn), other search engines (Bing and Yahoo) streaming audio (Pandora and Spotify), direct buying with local newspapers (The Charlotte Observer), non-Google programmatic ad networks, native advertising platforms, and video inventory you watch from your cable provider on non-television devices—all of which a digital media buyer may purchase to reach voters and none of which have transparency portals. For fundraising purposes, there’s an entirely different constellation of content creators and vendors who also don’t fall under current transparency platforms.
Tech for Campaigns recently released a well-covered report looking at campaign investment in digital. Their report is well-researched and presents a number of findings that Democratic campaigns should take seriously. In order to estimate how much each campaign spent on digital, they assumed that spend reflected in Facebook and Google represented 60 percent of that campaign’s total digital investment, consistent with those provider’s market share nationally.
This is likely to be inaccurate, though, because the forces that drive digital political advertising in midterm statewide elections aren’t the same that drive national advertising. First, in competitive statewide elections, a number of advertisers are competing to buy the same inventory, meaning that even if an advertiser wants to max out on Facebook or Google, they may not be able to due to inventory constraints.
Second, the age demographics of regular midterm voters aren’t the same as those of the audience most advertisers are attempting to reach. As a result, persuasion media plans skew toward those platforms where older voters are most represented, which isn’t always the platforms that have implemented spend transparency.
Misconception #3: The number of ads an advertiser is running is a sign of sophistication.
This one is easier than the others. Publications have from time to time have reported on the number of ads an advertiser is running. A CNN headline from last May read, “Trump's campaign has run 4,400 ads on Facebook so far this month.” The New York Times reported last July that the Trump campaign “has run dozens of ads on Facebook recently that seek to rally support to confirm Judge Brett M. Kavanaugh.”
The number of ads being run is meaningless. It’s incredibly easy to duplicate an ad any number of times into any number of targeting groups to artificially increase the number reflected in the Facebook archive. There could be actual strategic value to doing so in terms of audience testing, but the targeting information revealed in the transparency platform is insufficient to draw those conclusions.
Now, digital advertising doesn’t stop at the transparency reports. I’ll be the first to tell you that being able to cover digital advertising with the level of accuracy that campaigns currently track television would be incredibly beneficial, but we are far from achieving that. Until then, it’s critical that your digital strategy (and analysis) be based on reality, not the limited snapshot provided by a few big tech players.
As vice president of client strategy at Rising Tide Interactive, Jake Sticka helps clients meet their goals online. He can be followed on Twitter @Sticka.