Two questions were recently asked on the LinkedIn “Earned Value Management” forum: “Q1. What are the WEAKNESSES and LIMITATIONS of EVM? Q2. Can EVM schedule forecasting methods provide us RELIABLE EARLY WARNING SIGNAL?” Wayne Abba, a noted Earned Value expert, wrote in response:

“EVM is not a scheduling technique and therefore has no inherent ‘schedule forecasting methods’ – it measures the volume of work that has been performed relative to planned volume, with no regard to critical path or even when the work was scheduled to be performed. Nevertheless it provides one of the best possible early warning signals – a straightforward indication that work is not being performed. The larger and earlier the unfavorable schedule warning, the redder the flag… bear in mind that EVM was invented to deal not with schedule, but with cost variance. So-called ‘schedule variance’ (a better term would be accomplishment variance) is a useful bonus but limited by the mathematics. At completion, no matter how late, the variance is zero by definition. The true value of schedule variance thus is early in performance.

“Those advocating using EV to forecast schedule have built a pseudo-science on EV data that are derived from the schedule (and thus must correlate with it) but most assuredly is not the schedule. ANY schedule analysis using EVM must be compared with the ‘real’ schedule. Expect the GAO to delve into this further as the Cost Estimating and Assessment Guide evolves.”

One implication of what Wayne has written is that there need be no conflict between Earned Value Management and schedule techniques like Critical Chain. While we have known this for years, the increasing popularity of “schedule margin” (a buffer-like concept) in government contracting circles has raised the urgency for answering questions about how to make the combination work. With the help of Charlene and Chuck Budd, EV experts in their own right, I recently wrote a paper on the use of schedule margin (or buffers!) with Earned Value. See our new Resource Center to download a copy.