La legge delle conseguenze inaspettate

Per chi si vuole avvicinare alla Business Intelligence Microsoft e nello specifico PerformancePoint, esistono sul mercato (solo in inglese per ora), diversi buoni libri.

Uno fra questi è "Pro PerformancePoint Server 2007", dove a pagina 7 viene riportato un esempio della "Law of Unintended Consequences", che ben spiega l'importanza, e quindi anche la difficoltà insita nel definire KPIs rilevanti e, soprattutto, in grado di orientare correttamente l'organizzazione verso gli obiettivi desiderati.

Riporto qui di seguito un breve estratto (ringraziando l'autore, Philo Janus):

Esempio 1:
"A company has a metric of "number of cases held over 20 days." The net result of this metric is that when your case is 18 days old, you'll see a flurry of activity, but when you hit the 3-week mark, it will suddenly go dead. Why? Because there is no metric to differentiate between a case that's 21 days old and one that's 90 days old. Once you're past the magic 20-day mark, there is no incentive to work on your case."

Esempio 2:
"One computer manufacturer implemented a metric on its support line counting "number of calls lasting more than 10 minutes." Their cost of support skyrocketed. When they dug into the background, they flound that their support technicians would work hard to help customers for 9 minutes. As the clock entered that ninth minute, they would simply offer to ship the customer a new system to get them off the phone".