In 2008, when Obama cruised past his better connected, better known rival Hillary Clinton, and defeated John McCain to take the presidency, he made history by using social media and the internet to directly interact with millions of potential voters. Pre-2008, the internet was used mainly for soliciting political donations. But Obama did a whole lot more. The Obama campaign gathered data on millions of volunteers using their website. They fostered organic online activity, and spread popular content, including YouTube videos. Obama didn’t outfox his rivals online. He was the only player. By all metrics, Obama dwarfed his competitors in online engagement. And he spent less than $8 million, much less than his TV spending.

In 2012, the Obama campaign again innovated online, amassing data on millions of potential voters. The campaign targeted ads and encouraged volunteers to reach out to specific friends that the data suggested. It was the first time a sitting president compiled a massive, centralized hub of data gathered from the Internet to target potential voters and refine the messaging of the campaign. 

In 2016 a candidate again innovated in the use of technology. If Obama’s use of social media was grassroots and innovative, Trump’s use was an all-out blitz. Trump spent millions on Facebook ads, using, in part, data acquired against Facebook’s wishes from Cambridge Analytica. The campaign ran tens of thousands of ads in a day, with subtle differences in the political messaging, the imagery, and the background music, and then let the successful ads proliferate. Trump’s data allowed targeting of voters in specific counties, and in a few counties the Trump campaign advertised tailored messages to traditional democratic voters in an attempt to suppress voting. For instance, the campaign targeted Haitians in Miami with an ad that highlighted the Clinton foundation’s controversial activities in Haiti. In a public Facebook post, Facebook executive Andrew Bosworth claims that Trump was elected because he ran the best digital ad campaign Bosworth had ever seen. 

Regardless of the impact Trump’s social media strategy had on the outcome of the election, one clear result was a loss of trust in internet companies, and a general mistrust of election fairness. After the election, Clinton supporters pointed to many things, including targeted advertising, attempted voter suppression, and Russian intervention through social media, as undermining the fairness of the election. Sanders and Trump supporters also joined in, complaining of DNC conspiracy or undocumented residents voting, respectively. After the 2016 election, the idea that shady behavior on social media could influence the outcome of the election, by overwhelming people in a flood of newsfeed propaganda, or subtly influencing a person to not vote, became mainstream. And a Pew poll this year found that only 1 in 4 Americans are confident that tech companies will prevent the misuse of their platforms in the 2020 election. 

But the same strategies used in 2016 are back this election season, fueled by even more money. Each presidential candidate has already spent hundreds of thousands on targeted Facebook ads this year. And candidates are scrambling to amass as much personal information–cell phone numbers, addresses, names, and email addresses–as they can. While in 2008, Obama acquired voter information from his website and app, now candidates fish for information online in numerous ways. In 2016, Trump’s methods included an online store and requests for donations. Perhaps the oddest fishing strategy are the “official polls,” which might ask for an opinion on socialism, then for your phone number. One thing that’s new this time around are peer-to-peer texting apps, which allow individual volunteers to text hundreds of people. The apps autofill numbers and messages, volunteers press send. Part of campaigning is now the race to collect as many phone numbers as possible, as texting is perhaps the most effective avenue of messaging–well over 50% of political text messages are read. 

Though the same strategies are at play, the rules have changed since 2016. Social media companies are working to combat divisive media created and spread by foreign actors from countries like Russia and Iran. They’re removing or labelling false news, and trying to label media organizations that are sponsored by governments, like Xinhua news agency. Further, Facebook has reduced the ability for campaigns and companies to target advertisements based on user demographics. 

As Facebook sets rules for political advertising, and changes how their platform can be used, politicians have become major customers. Facebook.com/politics, was, in 2008, a hub for Facebook’s involvement in politics, for debate streams, discussion forums, and user polling. Now it’s a storefront for politicians or activists seeking to get their messaging across, complete with guides on how to use the platform effectively and examples of good political messaging. Since May 2018, politicians, activist groups, and companies have spent $1.1 trillion on political ads on Facebook, the number displayed on the frontpage of the new Facebook ad library. The public database includes all political advertisements purchased on Facebook in the last couple years, and tracks the ad spending of political candidates and interest groups.

Politicians have gone from generators of activity on social media to major customers. But if the last few elections are any indication, 2020 will be defined by innovation, the leading candidates will use social media in original ways. Because of the backlash from the 2016 election, social media companies are on the defensive, struggling to deal with new strategies as they arrive. Take, for example, Michael Bloomberg’s meme program, in which he paid popular meme pages on Instagram to post cheeky, self-aware ads. Facebook recommends that posters recognize sponsorship, but the policy is not always followed. Facebook employees hunted down the content to monitor it. The flexibility of social media means that candidates are bound to invent new strategies between now and election day.

On election day, expect a text from one (or both) candidates asking for your vote. Expect to see political ads everywhere you go online, some encouraging you to vote, others perhaps designed by a candidate to decrease your willingness to vote for their rival. But also expect that in the days after the election, the losing side will be trying to figure out what the other side’s innovation was, how they got the edge in online engagement. 

Tech has always been in politics, but the last few election cycles have shown that the possibilities for online engagement are too endless for each candidate to stick to the same, trusted strategy. The memes, text messages, and advertisements of 2016 may well become personalized phone calls and VR meet and greets by next election. Presidential campaigns are now like startups, experimental and creative. And the prize of online innovation is effective, instantaneous outreach to millions of potential voters.