Priorities USA, the Democratic Party’s most influential super PAC, is making a play to try and tilt the electoral landscape with just a small portion of their campaign budget.
The group has put a $5 million investment into its analytics operation in hopes that it can more efficiently target need-to-reach voters at a fraction of the cost of traditional campaigning. Whereas in 2016, the super PAC had no one on staff doing in-house analytics, in 2018, it had a team of four. Its projected staff will end up at 13 during the 2020 cycle, aides said, with four targeting analysts, three experiments analysts, four data scientists, and a director and coordinator.
The strategy was derived from two sobering realizations made by officials at the super PAC. The first was that Republicans had leapfrogged Democrats in the online ops field. The second was that President Trump’s re-election campaign would be spending an unprecedented amount of money to define Democrats as they battled among themselves over who would be their party’s nominee. Forced to fill that gap in time and money, Priorities has recast its own operations.
“We are investing in our analytics program because Republicans are going to have more money than us, so we need to be smarter about how we’re spending it,” said Patrick McHugh, executive director at Priorities. “We can’t just run the same old campaign we’ve run before, we have to constantly be innovating and questioning assumptions to help us run the most effective campaign possible. Our hope is that this program will help us refine our creative to avoid unintended backlash while being smarter about how and where we’re advertising.”
To get to that point, Priorities is pushing four major changes to its analytics operations.
The first is to refine voter data to make it more detailed than traditional voter files. The goal, ultimately, is to put together a more predictive data set by asking individuals for, among other things, their music preferences, levels of education, and whether they have student or credit card debt. To get that information, Priorities USA said it was running online surveys in battleground states with the help of like-minded Democratic institutions. So far, it has collected roughly 2 million responses, with plans for 50 or so more panel tests this cycle.
The second component is to recast its messaging on voter persuasion and mobilization. To date, Priorities says it has shifted its thinking, looking to find voters not just equally torn between voting and not voting but those who are leaning (even heavily) toward not voting but could be persuaded based on one specific message or issue. And it is looking at everything from different types of inventory (whether high-quality video at a low frequency is more effective than medium-quality video at a high frequency) to targeting (what added lift look-a-like online models provide).
The expectation among Priorities’ team is that voter turnout in 2020 will be massive for both Democrats and Republicans, making it imperative that it find and mobilize even marginally interested voters. To do that, it plans to lean not just on the voter data that it is collecting but to rely aggressively on Facebook and Google to help find potential targets.
The third component is to dramatically rethink what a data scientist looks like. Too often, Priorities concluded, analytics teams are composed of white men only—giving their inputs and outputs the potential for bias. To that point, the person leading the Priorities team is Nick Ahamed, a Muslim mixed-raced immigrant. No one else on staff is a white male.
The fourth component is perhaps the most critical. The group is running digital testing to refine its operations in real time, with plans to run at least four field experiments testing a total of 26 distinct mobilization and persuasion programs. Such real-time experimentation, it believes, could have potentially major benefits not just for the campaigns themselves but in convincing candidates that analytics is a worthy investment.
“In 2018, we ran an experiment where we gave two entry-level staffers basic training and gave them budget to run a digital ad buy in 50 percent of precincts in two congressional districts,” Ahamed recalled. “Where we targeted voters, support was higher by 0.6 percentage points, meaning that among people we reached, support for the Democratic candidate was at least 1.3 percentage points higher.”
Officials at Priorities say they do not expect findings like these to be a panacea for Democrats as they head into 2020. Nor are they pretending that their operation is now meant to be a lean alternative to the type of efforts that super PACs traditionally undertake. Priorities will continue to run massive Get Out the Vote operations and television ad operations in a group of key battleground states.
But for an election that could, once again, be decided on the margins, every bit of advantage helps, and the investment in a more robust analytics operation could greatly matter.
“We have a year to iterate and experiment,” said one official at the group. “So by the time the Democratic convention is done and the candidate is chosen, we will be ready with our findings.”