ImageSo how well is the world doing?  Are we in a state of chaos or emerging from it?  These are all valid and vitally important pieces of information because our feelings and impressions have been conditioned to take in lots and lots of information.  At one time the news could be viewed as a reliable source.  Based upon facts and not opinions.  Credible sources were the order of the day, the mandate of management in order to provide credibility to viewers, listeners and readers.  Today I’m not so sure.  Obviously I suffer from the opinion that anything and everything you get exposed to must be filtered, studied, measured and analyzed against what you know, feel or have knowledge about.  For that reason one cannot being accepting but it also means that you are responsible none the less to use this information in a proper way.

Predictions are the last thing we need in business.  While it might give us promise that we are on track (possibly because we are the track to which the predictions are being made) we have to consider it in the much large category of ‘information’.  To the extreme this might be used to promote and elevate our agenda, but certainly it should never be used to justify the commitment.  We need definitive and concrete information that can then examined and evaluate for the possibilities that it contains.  Even then we adopt an element of risk because we are conjecturing about what might happen.  If we are able to leverage the ‘happen’ element then we are more apt to have information that can be used to lay strategy groundwork.   Be cautious however, the best laid plans do not always materialize size your range of control is limited to what you do and not that of the consuming public.

A fundamental rule is the order of control.  Control over analytics should follow the Prevention, Detection and Correction paradigm.  I would that only after these steps have been followed Predictions can occur.  We must,

  • Prevent predictive analytics being used beyond their level of authentication.  Will useful in making public announcements and sparking curiosity that can be examined further they simply too unreliable to bet your entire business on.  If you think I’m blowing smoke on this take any if not all of the analytic company predictions and see for yourself what the reliability factor has been based on actual historical outcomes.   then predict with improved reliability.
  • Detect the level of use and the need for added study or possibly collaborative comparisons.  Blind acceptance is again not an option for rational things.  You are risking allot to possible gain allot but don’t be fool hearty.  Don’t bet on a high pair in your hand if you haven’t seen what cards get shown on the table.
  • Correction is probably the most time consuming aspect.  Correction is about making change to the predictions but adapting the conclusions drawn into a strategic approach to follow in a ‘guarded’ way.  Never go all in when there remains some significant unknowns that make all of your ambitions a waste.
  • The preceding three points can then lead you to a point where predictive predictions can occur.  Taking what you have, what you know, what actions might be taken and melding them all together you arrive at actions plans and adaptive predictions that serve to guide your business.  These provide the necessary bullet proof protection necessary to make ‘right’ movements in ‘right’ directions with ‘right’ measures of control.

Predictions are guesses.  Like the news they were to be based on facts.  But as we have seen the complexity and the number of factors grow we have seen a similar response in limiting the sources for analytic purposes.  Some would ask why do this in the age of technology?  Couldn’t we just amalgamate this all and come up with more accurate results?  The answer is yes… we could, however we face the economics of analytics.  I am tempted to call it as a lean approach but would prefer to call it brand leverage lean.  Companies rely upon their name to care the information, absent of sufficient analytic depth but opportunistic enough to sell for a hearty sum.   I’m not really certain how long the charade will last.  The economic reason for it retains popularity is that its sold at the top of organizations whereby the analytics can be used for the purpose supporting a public campaign they cannot not, should not and must not be relied upon at face value.  Reckless abandonment puts business at risk.  You might have nothing to loose professionally, maybe the golden parachute will do you just fine but my gosh give consideration to the role that you have and the impact of your decisions on others.  When I look at real leaders their righteous role is defined by the consideration of the entire community that surrounds them.  They often place themselves well down on who gets pleased,  As leaders I plead with you to be responsible (some of you may blow this off or quickly discount this request because you feel  you are… but please-please-please reflect for just a moment as to whether you or not).  Behavioral redemption is possible even at the late hour in your career.

Predictive power is driven by the means/methods, data, source, age/timing and motives.  We  have all heard of creative statistics the permit just the right pieces to shine through to achieve a purpose.  Real truth based analytics are near real-time (event based and is current as possible), defined by appropriate methods including both mathematical and population selection, and must articulate the limitations.  If these steps are not being followed then the numbers are just the numbers.  They are unreliable and essentially unusable in any form.   What would you say that 99 out of 100 CEOs wore socks would you agree?  Would  you ask what was the 1 that didn’t, what country or region was this information taken from?   But more importantly would  you go about making socks for CEOs without understanding that the climate in which they live or even travel to.  This will and should impact the way you use analytics.  On the predictive side of the house, if I was to say that 3 million out of 25 million graduates will not have a job upon graduation in 2017 would this compel you to create jobs or would you see this as an opportunity to serve this unemployed population with career entry preparation before reaching graduation (to drive the 3 million jobless down)?   These are examples that help to illustrate the power of numbers but also the risk when blindly using them.

Their precision is grounded upon the means/methods, data, sources, age and motives.  We have all heard of creative statistics.  These can and often create suggestive answers that pander to goals but fail to sufficiently reflect complete truth.  Applied to strategic and tactical missions they run a deplorable high risk of failure.