I’ve been eagerly listening to some of my colleague’s tales from Nate Silver’s presentation at Hubspot’s Inbound 2013 conference. A marketing and media superstar in the making, Silver deserves all accolades and more for shifting part of the news paradigm where high quality data is prominent in the telling of great stories. Highly quantitative but still accessible to everyone, Silver’s style is to take other’s people data (much of it in the public domain) and methodically interpret results to where he makes highly relevant and accurate predictions, often, in the stupidly partisan world of Federal elections.
For instance, the success of his approach across the two most recent Presidential elections has stamped him prominently as a credible political news analyst. It has also jettisoned his career from baseball stats geek, to New York Times columnist, to multi-faceted guru for ESPN who reportedly has big and diverse plans for Silver to help recoup the large coin that lured Silver to its network.
Always a statistician and social scientist first, Silver has shown that being both bright and statistically fluent can be an effective force to add to the American news dialogue and dramatically lessen the waves of ambiguity that frame discourse in the worlds of marketing, elections and everywhere in between. For some this is threatening. For instance to the ubiquitous election consultant who sees his profession as a combination of media cat-and-mouse, alchemist and high stakes street game of Three Card Monte, Silver is an anathema—forever dousing BS and spin with a style of lucid fact presentation that is remarkable in both its simplicity and sophistication.
Silver has achieved this coincidentally at a time when certain parts of Big Media and overall Big Commerce have become zealot proponents for the “it is great for business” concept of Big Data. In doing so some have adopted Silver as the father of Big Data. They are as wrong as wrong can be.
Silver and those like him employ the same techniques of inferential statistics and probability sampling that has been around since Galileo and Pascal. Such approaches are scientific, based in the rhythm of methodical inquiry, and driven by theoretical hunch or hypothesis. It’s what I might call “Wise Data” and has abundant applications in today’s marketing world.
The early advocates of Big Data are endorsing approaches that have none of these qualities. As we have written here before (and detailed in our forthcoming white-paper), Big Data is a vastly different animal. It presumes that having more data is somehow always preferred even though the canons of inferential statistics prove otherwise. For instance the fundamental statistical concept of sampling is ignored in Big Data’s aim to link all data toward finding “hidden” associations. Not only is such an approach indefensible methodologically, it invites findings that are potentially spurious, silly or both.
If you believe (as I do) that anything “worth doing is worth measuring”, then Big Data wants you to believe that any measurement is worthy of continued analysis until its significance can be discovered. That’s bunk and not evident in the approaches of Silver or reputable everyday research and data folks who apply their skills across many commercial marketing and communication disciplines. Big Data advocates are too frequently on a mission that supports the commercial aims of Big Data—more hard technology, more technology consulting, more endless data mining that may or may not yield returns practical for scale of investments made. Silver (to date) and those like him have no such aims. Perhaps the twain will meet—who knows? But while Big Data is pimping its interests on the coattails of those like Silver, it would do well to realize its fortunes benefit from not finding a “father” to symbolize its hype, but rather, to gain a patron of statistical common-sense.
_____
John Durante is marketing services director for WordWrite Communications.
