October 27, 2016 | Written by: Heather Green
Categorized: Data | New Thinking
Share this post:
We all know the role social media has played in this presidential election. Twitter, Facebook, Snapchat, and other social media tools have been aggressively deployed by the two campaigns and their supporters. But in the final stretch, it may be mastery of Big Data that will determine the next president of the United States. And here, observers say the two campaigns are taking very different approaches.
Trump, the ultimate celebrity, harnessed social media early to a remarkable advantage. An estimate this fall pegged the amount of free publicity he’d garnered at $3 billion. But for all his social media prowess, he did not see the value in collecting and analyzing election data, choosing instead to rely on his personal appeal.
“I’ve always felt it was overrated,” Trump said in May in an interview with the Associated Press. “Obama got the votes much more so than his data-processing machine. And I think the same is true with me.”
Trump’s rejection of data analysis gave Hillary Clinton the opportunity to build even further on the data-driven lead established by the Obama’s campaigns in 2008 and 2012. As the party’s nominee, Clinton inherited a centralized database that combines data from traditional outreach, including calls and polls, what voters said or did on social media, and commercial consumer data.
Big data is key during the entire campaign, but it’s critically important in the last few weeks of getting out the vote, as volunteers call voters and knock on doors, and the campaign crafts last-minute personalized email and social media messages and ads. In an electorate where voters are increasingly calling themselves independents — 39% say they’re independent in 2014 versus 30% in 2004, according to Pew Research — analytically understanding who these voters typically swing for and smartly targeting them is more crucial than ever.
“Technology will be a real important piece of this election for a couple of reasons,” explained Mikah Sellers, Chief Digital Officer for D.C. branding agency Grafik, in a Washington Post video in August. “This is going to be one of the most expensive elections in history, and with fewer people watching TV, digital is going to be a medium where, from a marketer’s perspective, you’re going to be able to achieve the reach and the repetition that is required to either hold onto a constituent or to convert a constituent.”
During the final stretch of the election, the AP reported that Trump’s campaign switched gears, funneling millions of dollars into digital data and services to drum up donations and woo voters. Neither campaign is discussing in detail its digital operations, so how much ground Trump can make up in using data remains to be seen. The test, in the end, will be Election Day.
The opportunity to exploit big data was bigger this campaign because it is the first in which Republicans and Democrats have access to data on a national scale, says Sophie-Charlotte Moatti, general manager of Products That Count, a think tank advising public companies, investment firms and startups. That’s the result of two trends: mobility adoption and innovations in database systems focused on analyzing exceptions within data rather than averages.
This database system innovation was driven in large part by the demands of government agencies after September 11 to improve national security by collecting and processing huge amounts of data, says Moatti. “The combination of the technology supporting that, plus the data available, means you have something that can be really disruptive,” she says. “These political parties now have access to a lot more data than they used to.”
Looking forward to 2020, Moatti says that artificial intelligence, machine learning, and chatbots could play an outsized role. “The first party to do that will have a great competitive advantage,” says Moatti, because that will give campaigns the ability to target undecided voters and Americans who usually don’t vote, getting them to the polls by understanding what would change their votes and their behavior. That could help break through some of the rising polarization that increasingly defines elections.