In the Star Trek original series episode Friday’s Child, Captain Kirk, Commander Spock, Doctor McCoy, and Lieutenant Grant beam down to planet Capella IV in order to negotiate mineral rights with the planet’s inhabitants, a tribal people with an aggressive, warrior culture. The Enterprise contingent is quickly reduced by one, as Lieutenant Grant instinctively draws his phaser weapon upon seeing a Klingon among the Capellans, prompting one of their warriors to throw a kleegat at the unfortunate officer (and, yes, he was wearing a red shirt). This kleegat weapon bears a striking resemblance to a table saw blade, and the Capellans throw them much like we humans would throw a frisbee, except overhanded. As it turns out, the unfortunate red-shirt’s instincts were correct: the Klingon is present to help stage a coup d’état on behalf of Maab, who seeks to control the planet’s Ten Tribes in place of the current Teer, Akaar.
Meanwhile, Back In The Project Management World…
We PM-types have lots and lots of tools at our disposal, so many, it seems, as to obfuscate why we’ve been hired by our organizations in the first place (cough, risk register, cough). Rather than mock the use of the particularly superfluous ones, I’d rather focus on the over-engineering of one of the more fundamental ones, the Earned Value Management System (EVMS). In many business environs the EVMS actually bears the title of “Cost Processor,” though that’s not really its function. Costs are collected by the General Ledger, which leads me to the first of the invalid but commonly accepted business practices, that of attempting to understand cost performance as a function of spending. The way (not how much, mind you, but how) that an organization – including Project Teams – expends its resources is important to Asset Managers, since spending behavior often determines tax rates. It has nothing to do with Project cost performance, as the following mental exercise clearly demonstrates.
Imagine you are the PM of a $100,000 (USD) project, and in your winning proposal’s Basis of Estimate (BOE) you have bid $75K in labor, and $25K in equipment. At project completion you have spent $25K in labor, and $75K in equipment so, by Project Management standards, you’re fine, right? I mean, you came in on-time, on-budget. Obviously, you had to make several major changes in management strategy in order to get to an acceptable PM outcome, but that’s what Project Management is all about. If it could be done successfully using a template or robot, there wouldn’t really be a need for even the Project Management Institute®, amirite?
Don’t believe it for a second. Those “analysts” who pore over the spending reports are completely freaking out at this turn of events. You’ve overspent on equipment! And underspent on labor! This is a disaster, in their eyes. The insistence that project spending needs to mirror the original BOE or else the project is a failure has to be one of the most intellectually vacuous notions permeating the business world, and yet it will not go away. I think it stems from the concept that all management information dealing with costs simply must come from the General Ledger, where deviations from plan represent the only way of assessing cost performance from that particular source.
The other dopey notion I want to address has to do with the belief that EVMSs simply must have such features as an automated link to the General Ledger, Critical Path Scheduling package, or risk management (no initial caps) documents, or a dozen other aspects that really have nothing to do with its core function, that of reliably quantifying actual project performance. As I’ve mentioned many times in this blog, probably the most valuable pieces of PM information – at this rate of performance, how much will my project cost at completion, and when will that happen? – can be accurately (within ten points) derived from three data points. How much will it cost? Divide cumulative percent complete into cumulative actual costs, and there’s your answer. How long will it take? Take the same cumulative percent complete, and divide it into cumulative duration. Oh, sure, you can have an extremely sophisticated EV package, that can do all sorts of gee-whiz analyses, many of which I’m convinced are unnecessary, just as a wood working shop can have a variety of table saws with different blades installed for specific effects.
Or, you can just take the table saw blade, call it a kleegat, and throw it at anyone who’s threatening your life. It really doesn’t need to be installed on a powered-up machine to be planetary-government-changingly effective, just as EV information doesn’t need to have all the extra stuff attached to it to be operant. It just needs someone who knows how to
throw use it.