Case Study: How Facebook Ads Defeated a Florida Ballot Initiative

Here’s something to chew on while you’re waiting for those promised Netroots Nation after-action reports: the results of an attempt to use Facebook advertising as the primary paid outreach tool for a ballot-initiative campaign. And though the results as published do not completely rule out the idea that other factors were at play, the ad campaign itself seems to have been a deciding factor in defeating a Florida initiative that would have led to larger class sizes in the state’s public schools. And, it was the product of Chong & Koster, which is cool because Josh Koster is a long-time reader and hence an all-around smart guy.

Key Takeaways

The article contains plenty of details that you should definitely check out, but here are a few key takeaways:

  • The campaign used Facebook Ads for two key purposes, first to research the effectiveness of different message points to various demographics, and second to “saturate Facebook users in Florida with targeted messages in the month prior to the election.”

  • The campaign used geo-targeting to place its ads in front of residents of the two most populous counties in the state.
  • It also used polling data to identify the initial demographics to research, which were then hit with different messages and imagery to measure how well they worked with each distinct audience.
  • Besides breaking people down demographically, the campaign also tried interest-targeting, for instance reaching folks who listed “math teacher” or “I love my son” on their Facebook profiles.
  • The results of this testing helped to drive the campaign’s other online advertising, influencing the content and targeting of display (banner) ads and video ads.
  • Though the Facebook element of the outreach effort ran for two months, the bulk of the impressions landed in the four weeks prior to the election — the early ads provided the research, while the later ads provided the saturation.
  • Website visitor tracking (I’m presuming via cookie) let the campaign target banner ads at people who had been to the site, reinforcing its messaging through many different online channels (and no doubt making a few people think that this group was stalking them).

Read the full article for the results (hint: they served some 75 million ad impressions, and they won), but here’s an excerpt:

“Where the Facebook Ads appeared, we did almost 20 percentage points better than where they didn’t,” says Josh. “Within that area, the people who saw the ads were 17 percent more likely to vote our way than the people who didn’t. Within that group, the people who voted the way we wanted them to, when asked why, often cited the messages they learned from the Facebook Ads.”

Of course, our academic brethren will decry the absence of a true control group to validate these tactics’ effectiveness, but the numbers sound pretty good from here. Good work!

Update: One more point to note is that an all-online approach like this is perhaps particularly well suited to a low-profile campaign like a typical ballot initiative, since success in this kind of campaign often depends on getting a core group of supporters to the polls. In a campaign to get someone elected, or one involving a higher-profile ballot initiative, Facebook and other online messaging will be competing with TV and radio ads, media coverage, etc., and may not work quite as well. Still, the basic approach seems valid for MANY different issue and electoral campaigns. Political advertisers, ignore it at your peril!


Written by
Colin Delany
View all articles