Can the Presidential Election be Predicted by Media Attention?

Like most people in the US right now, we’re watching the 2016 Presidential Election pretty closely. Starting in the primaries, we noticed a lot of discussion around the candidates’ media attention —who was getting the most and who wasn’t getting enough.

One morning before the daily meeting, we were discussing media’s role in the election. Someone brought up Nate Silver’s piece over at FiveThirtyEight about a 1:1 correlation between media attention and polling data during the primaries. We looked around the table and quickly realized we had all the skills to build a site that had machine learning and could track the candidates’ media attention during the generals. We got curious. We got to work.

We don’t have a lot of spare time around here so we worked fast to get the Media Election website up. We launched it right on time, hours before the first debate. Designed and developed by our team, Media Election tracks Trump and Clinton’s media attention using artificial intelligence to scan over 150,000 publishers. We built pages to track the amount of articles over time and the trending topics. On the results page we pulled in the Huffington Post polling API to see the numbers side by side.

We don’t know what the results will be, this election is far from typical. We’ve been checking it obsessively and so have a few thousand other people. We bet you can’t guess who’s ahead today, have a look! Check it out here.



FOUR32C is a design firm, New York. We’re curious digital natives who bring creativity and intelligence to the world’s most beloved brands. We create thoughtful, compelling interactive experiences by starting every project with questions. Our process leads to evocative digital products designed with a purpose for every pixel.