The Trouble with Performance Indices

PMI defines SPI and CPI [EDIT: See here for a primer on these] as ways to compare the man hours planned (or money budgeted) for a project, the man hours (or money) actually spent and the man hours (or... you know!) earned at a certain point in the project. Generally speaking these should be pretty good indicators of how you are performing in a project. However there is one big problem in using this method: It assumes that your initial plan was the best one that you could have devised, so that any deviation from the plan will indicate the quality of your performance. Problem comes when the initial plan was not the best one it could have been, because either the planner was not experienced enough, calculation error were made, information was incomplete, wrong assumptions were made, or a host of other reasons.

The reason some projects deviate from the plan is that the plan did not take local conditions in consideration, or did not include enough resources. In such cases the CPI and/or the SPI will be below 1 as the PM tries to bring the project to face reality.

In a sense, the CPI and SPI values over the course of several projects may be used as an indicator of how good the planner is. If a planner consistently produces plans for projects that show significant deviations in performance indices then, may be, the fault lies with the planning team and not the project manager.

So variations should not be taken as absolute indications of poor management. On the hand, however, sometimes they really do so... and when they do that's when you are in trouble!