It’s election season! That means millions of Americans taking to the Internet to express their opinions about who should be in charge of running our country. Yay! Free speech!

What with all of the tweeting, posting and tumbr-ing going on, I decided to pose a question – could measuring social conversations about candidates be an accurate way to predict the outcome of the presidential election? I’m talking beyond the polls, or what comes from the media. It seems like understanding the stuff real people say online (when no one asks them) is important.

While I usually care a lot more about how consumer perception relates to brands rather than public figures, it’s basically the same principle and I think it’s worth testing. Quantifying things like volume (how often people are mentioning something or someone), sentiment (whether what they say is positive or negative) and passion (how strongly they feel that way) presents an invaluable opportunity to create understanding about how people feel about things, whether that’s a politician or a box of cereal.

So, with the Iowa Caucus just around the corner, I decided to conduct a little experiment.

Side note: if you need a refresher on what the Iowa Caucus actually is, I suggest this article from Vox. I majored in political science and still wasn’t 100% sure. Good thing I didn’t use that degree. Anyway…

Getting into the data

My goal was to quantify what people say and how they feel about the leading presidential candidates. I used Santy’s social intelligence tool of choice, NetBase, to get to the data. It pulled in lots and lots (and lots) of mentions about the candidates. As it turns out, people express their opinions about politicians frequently. Who knew.

Then, I filtered all of the results to the best of my ability to achieve maximum accuracy and representation of the public. If you’re interested in the details of how this was conducted, feel free to tweet me. I just figure it’s more information than most of you care about. But, here are a couple of top line parameters to note:

  • In some cases I limited the conversations to include only people who had fewer than 10k followers (most “real” people / no major media / pundits) and also excluded retweets and the candidates themselves as authors to ensure that all thoughts are original. Data goes back 6 months unless otherwise noted.
  • In all cases I only included the United States and data sources that were inherently social (I also did not have access to Facebook data at the time of this pull. I know.)

HERE IS A LINK to all of the dashboards I generated comparing public conversations around the following presidential candidates:

  • Donald Trump
  • Ted Cruz
  • Marco Rubio
  • Jeb Bush
  • Hillary Clinton
  • Bernie Sanders

I encourage you to take a look at everything I put together. There is much more information there than is acceptable to put in this blog post.

There are some pretty clear trends

So, what are my observations so far? Take a look at the table below:


  • Bernie Sanders seems to be the guy people like the most. He’s leading in public sentiment and even beats Rubio, Cruz and Bush in number of mentions.
  • No surprise here, Donald Trump continues to overpower his peers in terms of sheer volume of mentions, but his sentiment score varies. It’s a 13% when you only include people with less than 10k followers and only original mentions. It’s a -1% when you exclude the filters listed above. So, does that mean that real people don’t hate him as much as the media does? It looks that way.

Here’s another view of that – Trump seems to have a pretty secure hold on the “hate” quadrant:


And, no matter the candidate, the top emoji used in mentions about them is this….


I’ll leave the interpretations of this up to you.

Why Cruz and Clinton Will Win in Iowa

Let’s look at the numbers in in the state of Iowa alone:


For the Republicans, Donald Trump obviously has the most people talking about him, but, with a -8% sentiment score, it doesn’t look like many Iowans want Donald in The White House. There just aren’t that many people talking about Bush, and if they are talking about him, it isn’t very positive. While Rubio’s sentiment score beats Cruz, I think Cruz’s lead in share of voice (total mentions) is going to position him as the winner in Iowa.

Between Bernie and Hillary, it’s a tighter race. Clinton leads Sanders by only 4% in sentiment score, but she has a solid lead in number of mentions. If this means more voters, then Hillary is in. Let’s see if I’m right.

As the election progresses I’ll continue to provide updates on how the candidates stack up against each other. Want to see something else? Let me know. In the meantime…USA! USA! USA!

To track your own insights around the election – or anything else – ask us how to get started with NetBase today!

Image from TumblingRun

Premier social media analytics platform

Expand your social platform with LexisNexis news media

Power of social analytics for your entire team

Media analytics and market intelligence platform

Enrich your media analytics with social data

Media coverage for historical & real-time monitoring

Data streams & custom KPIs for advanced data science

AI, Image Analytics, Reporting Tools & more

Out-of-the-box integration with other data sources