Forget about the contest between US Republican Presidential candidate Donald Trump and presumptive Democratic nominee Hillary Clinton. 

There's another competition taking place during the 2016 elections.

Analytics vendors large and small are eager to show us what they can do with the firehoses of data the campaigns and activities around them are generating.

Some, like Facebook and Microsoft via Bing, want to convince us they've created analytics tools and/or apps to help both the media and individual citizens glean insights. 

Others like Alteryx, Qlik, and Expert System are using the election as an opportunity to show off their analytical prowess.

Regardless of the motive or your personal politics, they've all come up with interesting approaches that marketers, especially, should find interesting.

If All Press is Good Press

Leave it to Trump to spin a story in a tweet. This week he took to twitter to suggest that the uproar around his wife Melania's speech — which sounded oddly like Michelle Obama's 2008 speech in certain parts — might have been a good thing.

Not only did #FamousMelaniaTrumpQuotes become a top national trend overnight, but broadcasters collectively couldn't stop talking about Trump. Gartner MQ leader Qlik created a dashboard that tracks broadcast media election mentions by leveraging data from the television news archive. Their application looks specifically at information from Aljazeera America, Bloomberg, CNBC, CNN, Comedy Central, FOX Business, FOX News, LinkTV and CNBC via the Internet's Television Archive.

While the insights might not seem all that surprising in aggregate, it's interesting to note that CNBC spoke less about Trump the day after Melania's speech than the day before. Trump mentions on Comedy Central, on the other hand, skyrocketed from five on Monday to 127 on Tuesday, after the start of the RNC.

Qlik's dashboard will be available from now until a new president is elected. Note that it's always more than 24 hours behind because it syncs with the Internet's Television Archive, which runs on a slight delay. To get the most recent data, you'll have to manually change the last date in the URL address.

Understanding the Context of Comments

While media mentions are one way for campaigns — and marketers, for that matter — to measure their effectiveness, context is of extreme importance, according to Daniel Mayer, CEO of Expert System.

Leveraging Cogito, the company's semantic analysis technology, Expert Systems took the complete texts of Michelle Obama’s “One Nation” speech at the 2008 Democratic National Convention and Melania Trump's speech Monday night and found that when they performed a quantitative analysis on the extracted data and then asked, "From a linguistic standpoint, did Melania and Michelle say basically the same things?"

The answer was "no." They did note, however that this may not be the case had they looked only at specific passages. You can see the report here.

What we found especially interesting in the report is the verbs that each woman used most frequently.

In 2008, action words like "work, do, know, can, see, hope, love get, bring, should, come, drive, go, believe," in that order of cumulative frequency, filled Mrs. Obama's address. Mrs. Trump used "can, love, work, do, want, do, see, fight, thank, help, keep, need, come change" most often.

The report noted that while Obama's speech focused on "hope," Trump's focused on "fight."

But forget the politics. What can marketers learn from this kind of analysis?

"Analyzing what people say is essential for marketers to properly understand how their products are perceived, and, going forward, how to properly position them. Politics has been doing this for a very long time. What is totally new is that today's technology enables this on a larger scale than ever, making it possible to gain deep contextual intelligence in quasi real-time," Mayer told CMSWire.

An Unprecedented Divide?

While political banter might be popular in places like Washington D.C., in other parts of the country the mere mention of the election might clear a room.

Learning Opportunities

That is the polarizing effect that the candidates, the political parties and their views have on some of us. And while that may feel uncomfortable, it's likely to change.

"Relax. We (as a nation) have been here before," Dan Putler, chief data scientist at Alteryx told CMSWire.

Putler, who looked at partisanship trends over time, concluded that the level of polarization (data is current to 2012) is similar to what it was in the 1950's. "When you look at political party preferences at a county-level across the US since World War II, we sometimes look bi-polar," he said.

Today's map, using the Cook Partisan Voting Index (PVI) Trend (taken from Tableau public), illustrates exactly that.

Play with it interactively, and you’ll see that over time, it usually doesn’t stay that way. One item of particular interest is that we tend to vote similarly to our neighbors, according to Putler.

He calls the phenomena “the great sort.” It has come into play since 1952. Putler also pointed out that data scientists can use cartographic analysis to figure out where swing voters are most likely to live and that that’s where campaigns might be wise to spend their resources.

But from a marketing perspective there are other things that can be learned.

Both campaigns and marketers can leverage data science insights for micro-targeting, for example. Behavior matters, too. RFM — Recency, Frequency, Monetary — might be an even better predictor of outcomes, according to Putler.

Marry RFM to demographic and cartographic data and you're even more likely to hit the bulls-eye.

Of course, that's not easy to do which is why data scientists are so important. But most of us don't have data butlers at our beck and call, so self-service tools are becoming increasingly important.

If you're just fooling around with data gleaned from the elections, Microsoft's Bing has an Election Experience that might be fun for you, your kids or anyone who might benefit from a broader vantage point.

Title image by Asa Aarons Smith