Blog And Of Interest

 

Predictive Analysis

The world of business intelligence and data visualization continues to evolve from the days of “pretty” reports, to interactive but fairly dumb static visualizations. Nowadays, interactive and predictive intelligence are the game. Business Intelligence investments should show a tangible return in different spheres, and there should definitely be a payback. Sadly, the ROI and hopes invested in BI by most organizations seldom materializes as planned. Why is this, when on paper BI projects look like a slam dunk? There are several answers from beginning with a lack of skills both in-house and by vendors, collusion with by and between vendors and client staff, and competence. Business intelligence too often get’s to be guided by database vendors who have an interest to focus on the money around the databases rather than the BI where some quick wins are possible. The current vogue of “predictive analytics” should have a decent chance of success, but will most likely fail because the large database vendors do not have an interest to relinquish their drive coaxing clients to focus their budgets to get their databases perfect before serious BI can be tackled. Of course this perfection will never be achieved because there is always the diversion to the next better model of database.

Predictive analytics is not a trivial endeavor; it is often complex and those involved sometimes don’t know what they don’t know. If one were to look only at numbers without the perspective of knowledge through experience, one might be precisely wrong in understanding of a situation and in recommending strategies. It is as much art as it is science. The complexity of the predictive BI task increases almost factorially as the needs of perspectives of various departments e.g. marketing, production, finance, and HR are added into the equation.