Big Data just got much bigger at AOL. At its second annual “Programmatic Upfront” event in New York on Monday, AOL announced the launch of a data management platform (DMP) with multi-touch
attribution using technology from Convertro, the consumer tracking platform and attribution tech firm AOL acquired in May.
A fascinating piece in Techcrunch this week, titled Big Data Analytics vs The Gut Check raises the most important question in today’s data-heavy world: can we really turn a stunning amount of statistics into something useful?
For a successful digital video provider, the sheer quantity of data can be literally dizzying. For every play’ selected by a viewer, there’s the basic information, like when it was loaded, when an ad ran, when the video itself started and what quality it was running at. But then there’s the moment-to-moment tracking, which helps us see which elements of a particular asset were captivating, and which got scanned right past. One play can deliver dozens of data points; multiply that by a few hundred (or a few thousand, or a few million) plays, and pretty soon it can get hard to see the forest for the trees.
This is why, when it comes to Big Data, the genius is not really in getting answers to questions – it’s in formulating the questions in the first place. In the Techcrunch piece, they identify a situation where a real, live person saw a pattern among some bank customers which led her to ask: are these really individuals, or are they small businesses who aren’t getting the best possible service because they’re using the wrong products?
When we’re looking at video, there are often way more variables than we can work out by just looking at reports. Sure, we can see which videos are playing most frequently, and which elements within them go viral. But to build a differentiated business, with a sustainable business model, there is no substitute for the human brain and the creativity it hauls along with it. “OK,” we say, “this video seems light on viewers, but we feel like we get a lot of positive anecdotal feedback – how do those things reconcile?”
With so many potential perspectives, the winners are going to be those who take a pile of data and ask the right questions. If an anecdotally-popular show doesn’t seem to have the view, let’s ask: is it more of a niche, where we can identify that, say, iPhone viewers in the UK really love it? If so, it’s way more valuable than we thought: although it won’t have the raw volume of views, its demographic is well-defined and can command a higher CPM from advertisers.
For Big Data and human intuition to come together, we’re going to need the tools that allow us to combine all the perspectives and build tangible answers to complex, multi-dimensional questions. It’s the meeting of left brain, right brain, and silicon brain, and it’s the future.