Wednesday, March 20, 2013

Big Data? Big Deal.

For some reason March Madness brings gambling, odds and statistics into our daily lives like no other. 

Some ridiculous percentage of Americans fill out an NCAA tournament bracket just to try their odds and see how well they can predict who the eventual national champion will be.  It's particularly intriguing since most can't remember who won two weeks after the tournament is over ... unless it was "your team" or you happened to renew your subscription to Sports Illustrated and received the commemoration pack. 

In the corporate world there is a continual push to get more data and measurement.  No matter how much data you have, you want and need more.  Why?  

Because it's important to understanding what happened in the past (with 100% accuracy) and more importantly to predict what's going to happen in the future  How is the company doing, but more importantly how is it going to do.  Companies seek quantifiable statistics about nearly every aspect of their business: revenue, profits, cost of sales, turnover, ROI, productivity, etc ... It's eloquently called BIG DATA.  Sounds pretty cool huh?  It's one of the new phrases being uttered in executive boardrooms of the very best companies.  And yet it falls short, way short.  

On a parallel track the NCAA tournament has incredible amounts of data to evaluate teams and predict outcomes.  How many different statistics do you think is viewed by the NCAA selection committee?  Do you think they consider statistics that aren't being covered by all the analysts on ESPN?  

And isn't it interesting how when it comes to decisions of magnitude that we all seem to attempt to extract those intangible, unquantifiable elements from BIG DATA?  We have to in order to validate our analysis because we can't figure out a way to include it meaningfully in the equation.  So we tend to trivialize, discount and discredit the importance of these unquantifiable elements in predicting success.  

We all do it, but it's WRONG.  Yep, we are all wrong.  Literally hundreds of millions of people are wrong in completing their brackets.  I know I am, and I'm betting yours is too.  There is no shortage of BIG DATA for all to utilize, and its readily available for anyone who wants to wade through it.  So given this, why is success so fleeting?

Because things like teamwork, coaching strategy, engagement, playing style, referee selection and even team nutrition play crucially into the formula for winning a basketball game.  And guess what, they play crucially into winning companies as well.  Not to mention luck ... which almost always gets mentioned when a game is close as the potential difference maker.

Where are the statistics on NCAA tournament teams for these elements?  Suddenly BIG DATA doesn't seem quite so big.

BIG DATA has given us nearly every imaginable quantifiable statistic.  How many times did this player make the front end of a 1-and-1 on Tuesday nights on the road over the past two seasons?  Just a minute, I've got that number.  How many times did the Irish win when facing a non-conference team that had a better record than they did and where they choose to where fluorescent green?  Got that number too.  

But all that adds up to squat.  It's a classic case of the law of diminishing returns.  So here's my pitch ...

If you want to predict the outcome of a basketball game or your company's success you need to include the emotional, intangible, human factors that can and often do make the difference.  Use your human judgment and gut in a way that can't be replicated by BIG DATA generated from a computer.  And using these intangibles in a meaningful way in your decision making requires strong intuition, experience and understanding these intangibles better than others.  

Sure, use the statistics and BIG DATA that is available to you.  But also rely on those intangible components and include them in your decision making process.  Doing so won't just ensure you aren't replaced by a computer someday but will also increase your likelihood of being RIGHT, something everyone can appreciate.

 

No comments:

Post a Comment