On Monday night, Ted Cruz clinched a victory in the GOP Iowa caucus, while the Iowa Democratic Party declared Hillary Clinton the winner on Tuesday after lingering in a "virtual tie" with rival Bernie Sanders overnight.
In the days leading up to caucus night, Americans discussed the first vote of the 2016 campaign across multiple social media platforms, including Facebook, Twitter, and Google. Those platforms tracked Iowa caucus conversations, collecting data on what users were talking about, what was trending, and what kind of social media feedback candidates were receiving, and released their findings before Iowa's contests got underway.
What correlation is there, if any, between social conversation trends and who wins an election? Not a whole lot, it would seem. Conversation trends on Facebook, Twitter, and Google all differed, and each platform had different leaders at the front of the pack.
Twitter conversation data from January 24 through January 31 showed that Hillary Clinton had a 51.6 percent share of the conversation in Iowa, while Sanders had a 42.9 percent share of the conversation. On the GOP side, Ted Cruz had a 31 percent share of conversation, followed by 30 percent for Trump and 14 Percent for Rubio.
According to Facebook, Sanders dominated pre-caucus discussion. Their data indicated that Sanders was the most talked-about of all candidates among Iowans on Facebook, capturing 42 percent of candidate conversation in Iowa on Monday, followed by Trump at 21.7 percent, Clinton at 13.1 percent, and Cruz at 10.7 percent. And when the conversation was filtered to just Iowa Democrats, 73 percent of conversation was about Sanders while just 25 percent was about Clinton.
And Google, which released search trend data, announced on Monday afternoon that Trump was getting the most Google searches on the Republican side while Sanders was getting more searches on the Democratic side.
Of the three websites, Twitter discussion data came closest to mirroring the actual outcome of the caucuses. Twitter was the only platform where the most discussed candidates from both parties prior to the caucus were the eventual winners -- Clinton and Cruz. Facebook and Google data both indicated higher interest in Sanders and Trump. But social media chatter can only measure one swath of the population -- those with regular access to the internet and computers or smartphones, often those who are younger digital natives, and those who are passionate enough about candidates to regularly post about them on social media. It's also important to note that the data released doesn't measure or account for sentiment (negative conversations about a candidate, for example, likely wouldn't translate to an in-person vote).
In the end, all three digital media platforms reported different trends in conversations among Iowans before the caucuses, and all three had different leaders, making it difficult to draw any conclusions about whether online conversation trends can predict who would actually win.