Most of my regular readers know I like to take accountants to task for pretending to be able to deliver cost performance or estimate-at-completion information to decision-makers based on generally accepted accounting principles. But that door swings both ways: Earned value practitioners are also guilty of trying to further their technical agenda using the resource managers' arguments and analysis, which, in my opinion, is profoundly flawed.
The most prominent of these tactics is to try to justify the cost--or even the existence--of the project management office by running some sort of ROI analysis. This is simply illogical if for no other reason than the ROI calculation pertains to assets, not capabilities.
Less notorious but every bit as pernicious is the tendency of earned value practitioners and accountants to compare the time-phased budget's basis of estimate document with its associated actual costs at the line-item level.
In the earned value world, comparing budgets to actuals is worse than useless: It's actually misleading.
And yet, some practitioners seem to think that if such an analysis were simply done at a very detailed level, it would suddenly become relevant. It doesn't.
Oh, they may try to make some lame argument about the need to benchmark the estimators' work, but this assertion lacks validity that can be demonstrated in the following scenario:
A US$100,000 task is estimated to require US$25,000 in heavy equipment and US$75,000 in labor. At task end, US$74,000 was spent in heavy equipment and US$25,000 was spent in labor. An earned value management system correctly--would not raise the red flag for cost performance, but the system that compares budgets to actuals would erroneously report a severe problem--never mind that the task came in under budget.
Any management information system that reports a phantom cost performance problem isn't good for very much.
Next up: The absurdity of maintaining milestone lists in lieu of real schedules.