Too Big To Fail? Maybe Not.
One of the dubious advantages of life in beautiful downtown Ohio is its abundance of drivers who have no earthly idea of where they are going, and do not care a bit how long it gtakes to get there. On the plus side, it does generate plenty of time for contemplative examination of many of life's mysteries. One of those is the latter-day popularity of the phrase "too big to fail."
The magic words reached public consciousness in the wake of panic embracing financial institutions of all sorts in the meltdown that jump-started the Great recession. But, there are plenty of other examples.
One that comes to mind is the case of the late Andre the Giant, a staple of the WWE (nee WWF) for decades. A one-time singles and tag team champion, reportedly never being pinned (even in losses) in that peculiar theater of the absurd, Andre weighed in at something over 500 pounds, and was thought by some to be approaching a light-on-his-feet 700. He exemplified the cliche immovable object, and was credited by a reliable source with downing over 125 beers in a single sitting.
But the disease that made him also brought him low, and he left us prematurely, which rendered the sport one character short of an asylum.
Comes now the not-so-shocking news that the much-ballyhooed Big Data phenomenon may not be the silver bullet that some early adopters had hoped. A recent study by one of the mega-consultancies raises questions about the payback on investments to date, with several survey respondents expressing disappointment in results, and a small percentage branding their commitment an utter failure.
I've written before about the importance of Big Analysis to give Big Data a shot at being genuinely useful. We, as a professional community, really need to get our heads around the undertanding that a pile of Big Data is not, all by itself, going to automatically make our lives easier and our decisions wiser.
Actually, there is no amount of data that will answer all of our questions and clarify our uncertainties. All it can really do, is give us increasing stacks of ammunition to ask better and better questions. In all cases, with either a little or a lot of data, the next steps are up to us, the hard work of finding out what is behind the messages the data seems ro be sending.
Its a sortation exercise, understanding what are the reflections of special cause variation, which are the random outbursts of lumpy demand, how much is universal systemic change, and which elements are systemic but local. No army of analysts, with only mathemetics in their arsenals can hope to effectively categorize these nuggets plucked from the data pile without the cultural, behavioral, and business model grounding against which they are playing out.