We are in the final stretch of the US presidential election and the different campaigns are ramping up their social media effort (including Twitter ads). During this election cycle, we have run multiple analysis using Twitter data, today we are taking a closer look to get clues on who follows more than one candidate, a large percentage of those shared followers might be undecided voters, or voters who might be easily persuaded to cross over.
Before doing so, let’s take a step back to look at the followers composition for each candidate. From the location perspective, only 50% of Clinton’s followers are from the US, compared with 59% for Trump, 83% for Stein and 89% for Johnson. This comes as no surprise since both Clinton and Trump were well known figures internationally before this election campaign, whereas both Stein and Johnson are new to the international arena. It is worth noting however that only 1% of Johnson’s followers are from New Mexico despite being the former governor of that state.


In terms of demographics, Clinton has the highest percentage of female audience (at 41%), with Trump and Stein not far behind (38, 37% respectively). It is worth noting that Johnson is far behind having only 29% female followers. Clinton also has the highest percentage of non-English followers at 21% (with 6% of them being Spanish), Trump follows with 13%, with only 4% and 2% for Johnson and Stein respectively.
Trump, Johnson and Clinton have a large number of inactive followers (28%, 29% and 31% respectively), interestingly however, only 14% of Stein’s followers are inactive. Stein also commands the highest percentage of most active followers with 70% of her followers with Clinton and Trump having 33% and 32% respectively.

Let’s zoom onto the Undecided:
Drilling into the common followers between the different candidates we found that Clinton and Trump share 2.8M followers (making 33% of Clinton’s over all followers and 24% of Trump’s). What is interesting however is that the vast majority of Stein and Johnson’s followers follow other candidates, meaning that they stand to lose the most as the race tightens. Johnson and Stein also happen to share the least percentage of followers between them (15% for Stein and 11% of Johnson). The biggest percentage of shared followers however goes to Johnson (34% of his followers also follow Trump), the Trump campaign can benefit here by targeting that demographic of followers. Similarly, Clinton can target 28% of her followers that also follow Stein.

twitter-undecided-voters
Looking at the top 20 followers shared between the different candidates, the majority of them are in fact media! This contrasts for example with Clinton’s top 20 followers who are mostly famous artists. It is also worth noting that Wikileaks follows all of them 🙂

top followers Clinton-Trump

In terms of demographics, 37% of those that follow both Clinton and Stein are female while only 21% of those following Trump and Johnson are female.

Election undecided female
Generally, 30% of common followers are under 24 with the exception of those who follow Trump and Johnson, only 25% are under 24 years old.

Looking at geography, there appears to be a pattern of shared followers, the top 4 states are California, Texas and New York and the battleground state of Florida, with Pennsylvania and Illinois competing for the 5th spot, with the exception of Clinton-Johnson which happens to be Washington DC!

So it appears that the different campaigns can make significant gains by targeting the shared followers with other candidates especially with their GOTV effort (Get Out the Vote), both Clinton and Trump stand to gain the most since they happen to share millions of followers.

The presidential race has tightened up significantly in the last few weeks, and Twitter remains to be an effective tool to reach the right audience as demonstrated by Trump, it remains to be seen if other candidates can use it as efficiently as Trump does.

Posted on: November 7, 2016 | Author: Samir

Leave a Reply

Your email address will not be published. Required fields are marked *