While the need to have normalized commodity classifications to drive better strategic sourcing and compliance is now widely accepted, a debate has raged on for some time in regards to spend classification approach.  There are two fundamental options - to classify in the eProcurement system at the time of requisition or to aggregate all organization spend after the transaction has been completed.  The latter is the traditional spend analysis approach that Ariba recommends as best-in-class, but since this is not always easy to do, we're often asked, why not leverage codes from eProcurement.


One of my colleagues at Ariba, Chris Rodriguez from our Solution Consulting group, works with customers and prospects often and posed the following concerns and shortcomings that he has experienced with the eProcurement option that make good food for thought:


1) If you want to include some historical spend in the activity for baseline and trend reporting, the intial data extract effort does not go away.


2) There are additional costs to applying a classification method to multiple eProcurement systems and each instance represents a separate implementation effort.  Also consider that each time an eProcurement system is upgraded, the integration must be re-reviewed.


3) The approach considers only the spend that flows through the eProcurement system.  What about other spend that needs to be classified?


4) The approach ignores the valid classifications that already come from some eProcurement systems allowing a tool or user to update to a different classification reducing consistency and often, accuracy.