Facebook’s come in for plenty of criticism since Election Day, in part for its role in helping fake election news stories and conspiracy theories spread online in the months before Trump’s victory. But did the vast social network help put Trump in the White House through other means?
Check out this little nugget, buried in a terrific Bloomberg story about the Trump campaign in the days before the election:
In San Antonio, a young staffer showed off a South Park-style animation heâ€™d created of Clinton delivering the â€œsuper predatorâ€ line (using audio from her original 1996 sound bite), as cartoon text popped up around her: â€œHillary Thinks African Americans are Super Predators.â€ The animation will be delivered to certain African American voters through Facebook â€œdark postsâ€â€”nonpublic posts whose viewership the campaign controls so that, as Parscale puts it, â€œonly the people we want to see it, see it.â€ The aim is to depress Clintonâ€™s vote total. â€œWe know because weâ€™ve modeled this,â€ says the official. â€œIt will dramatically affect her ability to turn these people out.â€
“Dark posts”? Scary! Until you realize that we’re talking about the kind of targeted, promoted Facebook stories that advertisers use all the time (I’m running several for clients as we speak). The “dark” part is that the posts don’t appear on a campaign or company’s main Facebook page, instead showing up only in news feeds of people who meet the criteria set by the advertisers (example: women between 25 and 50 in Travis County, Texas).
Ironically, I suggested earlier this year that DEMOCRATS should think about using targeted online ads to eat into Trump’s support among the more marginal members of his coalition. Such a “suppression” campaign wouldn’t necessarily persuade people to vote for Hillary Clinton rather than The Donald; its main goal would be to get stories in front of potentially wavering Trump supporters who might just decide to stay home.
Could Trump’s work to take the wind out of Clinton’s sails via highly targeted Facebook content have made the difference? It’s certainly possible: higher African-American turnout in cities like Milwaukee and Philadelphia might have put Clinton over the top in the Electoral College, and we know that the black vote lagged behind 2008 and 2012.
But of course, we can’t parse out these Facebook posts as the single variable: too many OTHER factors were at play, including the fact that Barack Obama was not on the ballot and that Clinton’s campaign didn’t pour resources into those communities until too late. And, we’d need to find proof that the people who stayed home actually saw the “dark posts” to even prove correlation, since those particular voters might not have been active on Facebook at all.
We DO know that Trump’s campaign focused on the digital more than any presidential campaign before it, thinking about online advertising and peer-to-peer outreach as a true substitute for TV. As Issie Lapowsky describes in Wired, to facilitate their social campaign, they built a data infrastructure to support multi-variant testing of Facebook content on a massive scale:
On any given day…the campaign was running 40,000 to 50,000 variants of its ads, testing how they performed in different formats, with subtitles and without, and static versus video, among other small differences. On the day of the third presidential debate in October, the team ran 175,000 variations.
Trump digital director Brad Parscale credits the campaign’s Facebook outreach (along with Trump’s own use of Twitter) for their victory, particularly because the social network was their primary channel into the hearts of prospective supporters and donors (more than two million of whom gave the campaign money). Another grain of salt: of course Parscale’s going to credit the Facebook program with victory, since it was his baby. But win Trump did, relying on social media in ways we haven’t seen presidential campaigns do before. Watch for other organizations to learn from his achievement…and for Parscale’s services to be in high demand on the Right.