Digital Marketing: Watching Sunday’s Big Game through HP ExploreCloud

Martin Cuéllar, Director of Product Marketing for HP Software, is a Subject Matter Expert within HP Autonomy Emerging Technologies.

This post originally appeared on HP Autonomy Blog.

Our office in Plano, TX recently stood up an HP ExploreCloud multi-media insight kiosk; the same kind we put up at a Tottenham Hotspurs match, but with content about HP Autonomy, Big Data, and other “office” topics. It turns out that those “other” office topics also included football.

With the Big Game this past weekend and folks in the office talking football; I wanted to quickly change those screens to look at the impending matchup. I watched Sunday’s game through HP ExploreCloud, and here is what it had to say.

Setup and Configuration

This was a quick and dirty project—the configuration was not as elaborate as it was with Sundance—but it followed the usual basic steps. We identified where to listen, what to listen for, how to present the data, and then let HP ExploreCloud and IDOL do the rest. This is how I configured the criteria for our football championship command center:

  1. Where to listen: Everywhere except Wikipedia (Sorry Wikipedia, but page edits are not what I was looking for this time either).
  2. What to listen for: Anything mentioning the event name and hashtags. I throttled the content this time; with 24.9 million Tweets during the broadcast just 1 percent of mentions gives a great representation. Again, not limited to twitter,—listening to all social, news, blogs, images, etc. — 1 percent of all of that is a ton of content. For customer projects, we listen to all of it, but that’s not necessary to entertain the folks in the office.
  3. How to present:  The same accessible, touchable, walk-up setup that we used successfully elsewhere: 7 large touch screens with nothing between the audience and the data (You can see some pictures of it here). I also gave that audience a few Topics to select from. Topics are simple conceptual shortcuts that quickly focus in on a subset of the data. I initially chose each of the opposing teams, and commercials (commercial watching is almost as big an event as the game itself).

On Sunday, just before the game, I made one change. I turned on geographic listening to capture all content within a 1 km radius of MetLife Stadium. This captured everything publicly said in social media from people at the game, whether they mentioned our key phases or not.

What does the data say?

ExploreCloud lets the data speak for itself. This requires adjustment for customers that have used other tools. With those tools, users are used to building taxonomies, filling in dictionaries, and constructing elaborate queries to look for specific things, but that approach is flawed. It yields results only for what you are looking for (the known knowns), and it ends up showing you what you expect. As the old saying goes, “If you torture the data long enough, it will confess”.

IDOL (Intelligent Data Operating Layer) lets us ignore all of that confusion, in order to simply see what is being said. Not only does it show us those known knowns, but also the “unknown unknowns”. Each one of these projects that I’ve worked on has marvelously surfaced the unexpected; sometimes so unexpected that at first you’d think the software was broken, but then the reason reveals itself. 

In this instance, an unknown unknown immediately jumped from the data—puppies! I had forgotten about Puppy Bowl X, but people talking about the game sure hadn’t. It showed up as an IDOL Concept straightaway, and I made a quick topic so everyone in the office could share the cute. Before and during the game, the puppy content – including lots and lots of pictures – never stopped coming in.

There is a wealth of information in the data, but here are a few interesting things that jumped out at me:

  • Sentiment for Bruno Mars was overwhelmingly positive from the people at the game. No matter the viewer’s age, gender, or whether the person knew him before half time or not, they had mostly good things to say.
  • Sentiment on the west side of the stadium tended to be more negative than that on the east side.
  • Once the game was over, I started to see complaints about New Jersey transit.
  • The only talk about commercials from the people in the stands was disappointment about not being able to see them. People watching at home had a ton to say about the ads. (Also, nobody at the stadium talked about either the Puppy Bowl or the Kitty Half-time show.)
  • A constant stream of selfies flowed from the seats, and Instagram was the preferred selfie-sharing method.
  • The 3rd most used language was French, but not at the stadium. Spanish kept its 2nd place spot, but there were more Italians and Germans in the stands than French.
  • 71 percent of the people sharing in the stands were male, but looking at the audience watching at home, only 52 percent were male.

I didn’t look at a Puppy Bowl species breakdown, but with Meep the Bird tweeting; I’m sure it would be interesting.