Former Secretary of State Hillary Clinton can take heart from a real-time big data analysis of viewer sentiment during the Oct. 13 Democratic debate that showed the public liked her better after the debate was over than they did before it started.
There was a substantial movement in overall positive sentiment that far exceeded that of her opponent Senator Bernie Sanders (I-VT). However, Sanders also exhibited a gain in positive sentiment. Perhaps more important over the course of the Presidential campaign to come is that Sanders has shown far more engagement with users of social media than did Clinton.
But most worrisome for both Democratic candidates is that Republican candidate Donald Trump’s engagement on social media far exceeded all of the Democrats combined.
We worked with ICG Solutions to analyze Twitter and Facebook posts as a way to gauge voter interest in each of the candidates and in each of the major issues in the Presidential campaign. The process was similar to the analysis we performed on the Republican debates in September, and as was the case with the earlier analysis, we worked at the National Press Club.
There were some differences in how we approached this latest analysis of viewer sentiment. Most notable was that the LUX2016 analysis tool has been refined by ICG so that it’s now able to determine gender in the social media traffic. In addition, we also included sentiment numbers for Republican candidate Donald Trump because he is leading that party’s nomination race in the polls.
The results in the Democratic primary debates were unlike those in the GOP version in a couple of ways. First there were two clear front-runners followed by three candidates with small chances of success.
Second, there was a much greater emphasis on issues important to voters, which is something that didn’t appear in the 11-way race for relevance on the Republican side.
Readers can see a graphical representation here of the LUX2016 findings. Just scroll down until the page until you see the OCT-13-2016 date next to the Democratic mule symbol.
LUX2016 displayed the measurements of voter sentiment and engagement in a series of bar graphs that broke down the number and sentiment of social media mentions during very short periods of time ranging from five to 15 seconds. We only counted a mention by a unique sender during those periods as a way to keep from counting mass tweets created by automated systems.
What we found is that Bernie Sanders had a substantially higher number of social media followers active during the debates, frequently more than twice as many as Hillary Clinton.
Analysis Shows Trump’s Social Media Following Beat Democrats in Debate
But it should also be noted that during the entire debate, Republican candidate Donald Trump was getting a continuous level of social media support that was higher than all of the Democratic candidate support combined. Trump made a point of live-tweeting the debates, which certainly stirred up some of his supporters. But in some cases, Trump was ahead of the Democratic line-up by an order of magnitude.
Of course we were measuring social media activity, not conducting a scientific poll, but in the previous test we found that our real-time social media results tracked very closely with the results from polling reported a few days later. For example, we saw Carly Fiorina’s surge as it happened and correctly predicted that Scott Walker would withdraw from the primary, which he did a few days later.
In the Democratic contest our results also tracked well with preliminary polling. This would suggest that while Hillary Clinton remains ahead in the polls, the support for Bernie Sanders, especially when considering his lead in social media, may well help the Vermont senator gain additional voter and fundraising clout.
Perhaps more remarkable is Trump’s surge in social media support, which continues at a much higher level than any of the Democrats. If the Trump organization can leverage that sentiment and turn it into votes, then the Clinton camp, especially, has reason to worry.
As interesting as it was to use LUX2016 to predict how the debates turn out before anybody gets to see polling results, what’s more interesting is its use for instant viewer sentiment analysis, which is only a minor example of what this big data analysis engine is capable of producing. While I was working with ICG Solutions CEO David Waldrop in preparing for this analysis, I got a glimpse of far greater potential than the debates.
Imagine for example that you could track every ship on every ocean in the world in real time and then using machine learning predict where that ship would be at some time in the future. This would be nice for a shipping company, obviously, but imagine using it to determine when a freighter in the Indian Ocean decides to make an unscheduled stop to deliver weapons to terrorists. You could be alerted instantly, which would be handy if you were fighting terrorism.
The fact that LUX2016 can process a nearly unlimited amount of data in real-time and analyze it according to pre-built and easily configurable rules means that this is a tool that can analyze data in ways that exceed the imagination. This goes far beyond just tracking voter sentiment and response, but even there it can make a difference.