Earned Value Lies and Truths

There was a great quote from Benjamin Disraeli in David Hillson's letter to the editor in the latest PM Network Magazine.

Disraeli allegedly* said, "There are three kinds of lies: lies, damn lies, and statistics."

* As an aside, there's apparently some debate over the actual origin of this phrase.

In any case, Hillson's interesting letter was cautioning those who frequently misapply statistics, and offered some clarification the terminology----specifically, the mean (average), mode (most frequently occuring item), and median (the middle item if all were lined up in order).

I find that many misuse Earned Value statistics the same way. The intent of EVM is to be an early indicator of a potential cost or schedule overrun (and I personally feel that it's better at predicting cost than schedule). However, much like the Ghost of Christmas Future, it's not set in stone. There are many things a project manager can do to get things back in order. More importantly, sometimes there are reasons for the apparent variance that indicate that the variance is explainable and not a concern at all.

The key with EVM (much like any metric) is to not take the statistics at face value, and to use them as a trigger to do further subjective examination. It's a tool, and organizations often overuse such tools (much like they do with Six Sigma). If all you have is a hammer, everything looks like a nail.