I haven’t criticized a PM icon this blog for a while, so it’s about time I did. My targets? None other than EIA-748, Earned Value Management Systems, and documents associated with it from the National Defense Industrial Association (NDIA), specifically the NDIA/IPMD Surveillance Guide and Intent Guide. Where I believe these documents are creating problems in the PM world is in their macro-effect of influencing the whole PM codex away from actual Project performance, and towards process policing, not giving enough attention to its real-world consequences.
First let me lay down some basics, so intuitive that I’m genuinely surprised that they have not already been universally accepted in the PM universe. The point of all Project Management Information Systems (PMISs) is to put into the hands of decision-makers that which they need to make, well, informed decisions on their Projects, on a day-to-day basis. This information needs to meet three criteria:
·It must be timely,
·It must be accurate, and
·It must be relevant.
Based on my reading of these documents, I think they went light on that last bullet. Instead, they spend thousands of words describing process: how baselines must be created, must be changed, must have a documented audit trail to ensure no informal alterations have been made, so on and so forth. According to EIA and the NDIA, what are signs that your EVMS is suspect, or even invalid? These include the absence of:
·Documented processes and procedures,
·Precise alignment between the Work Breakdown Structure (WBS), Organizational Breakdown Structure (OBS), Schedule and Cost Baselines,
·Work Authorization documents, properly laying out the Scope, and with the appropriate signatures, and
·Change Control documents, laying out the nature of the change to the Scope, Cost, and Schedule Baselines, also with the appropriate signatures,
…among others. Don’t get me wrong – all of these certainly point to a mature and robust PMO. But there’s a problem with them: they are almost exclusively process oriented, and, to my thinking, successful Project Management is not about scrupulously obeying someone’s else’s idea of correct process. Success in PM is about, well, success (again, something that should be so intuitively obvious that I’m shocked it hasn’t been widely adapted already). It’s analogous to having a team of the best engineers in the land working in your automobile factory, evaluating every part and bolt-tightening process, but then not arranging for anyone to actually start the car at the end of the assembly line.
“Okay, Michael” I can hear the more process-aligned members of GTIM Nation say, “what do you think should replace these criteria as audit standards for Earned Value Management Systems?” Simple: do they work? And by “do they work?”, I mean do they actually help PMs bring their projects in on-time, on-budget? Here’s how to make that determination.
First off, does the organization’s project portfolio have a preponderance of Projects that come in on-time, on-budget? The proof of the pudding, as they say, is in the eating. If a strong majority of the Projects in the portfolio do come in on-time, on-budget, then I’d say that’s evidence of that organization’s PMIS functioning rather well. Either that, or they are staffed with PMs who are so exceptional that they don’t need timely and accurate cost and schedule performance information in order to succeed.
Of course, most organizations can’t point to an overwhelming Project success rate to provide ipso facto evidence of PMIS efficacy, but that doesn’t necessarily mean that such systems aren’t reliable. After all, some published estimates put the percentage of Projects coming in on-time, on-budget at 30 – 50%, depending on industry.[i] I believe that the litmus test for EVMS reliability should be: did the EVMS accurately predict those Projects that did overrun or come in late, by the near- actual amount of the overrun or delay? And how far in advance did the EVMS make that prediction? Because if the baselines were acceptably derived, percent complete data reasonably assessed, and actual costs correctly collected, even basic EV analysis will reliably predict the overrun/delay to within ten points (typically), and do so months in advance for anything but short-duration, low-budget work.
I think that part (most?) of the reason that there’s some level of truth in the axiom that you can put fifty PMs into a room and they won’t agree on the color of an orange is due to this movement of the determiner of PM Information System validity in general, and Earned Value Management Systems in particular, away from actual success rates, and towards process and function policing. And that, dear GTIM Nation, is the problem with PM audit standards.
[i] Retrieved from Chat GPT 6 Luna on October 27, 2026, 20:03 MDT.
First let me lay down some basics, so intuitive that I’m genuinely surprised that they have not already been universally accepted in the PM universe. The point of all Project Management Information Systems (PMISs) is to put into the hands of decision-makers that which they need to make, well, informed decisions on their Projects, on a day-to-day basis. This information needs to meet three criteria:
·It must be timely,
·It must be accurate, and
·It must be relevant.
Based on my reading of these documents, I think they went light on that last bullet. Instead, they spend thousands of words describing process: how baselines must be created, must be changed, must have a documented audit trail to ensure no informal alterations have been made, so on and so forth. According to EIA and the NDIA, what are signs that your EVMS is suspect, or even invalid? These include the absence of:
·Documented processes and procedures,
·Precise alignment between the Work Breakdown Structure (WBS), Organizational Breakdown Structure (OBS), Schedule and Cost Baselines,
·Work Authorization documents, properly laying out the Scope, and with the appropriate signatures, and
·Change Control documents, laying out the nature of the change to the Scope, Cost, and Schedule Baselines, also with the appropriate signatures,
…among others. Don’t get me wrong – all of these certainly point to a mature and robust PMO. But there’s a problem with them: they are almost exclusively process oriented, and, to my thinking, successful Project Management is not about scrupulously obeying someone’s else’s idea of correct process. Success in PM is about, well, success (again, something that should be so intuitively obvious that I’m shocked it hasn’t been widely adapted already). It’s analogous to having a team of the best engineers in the land working in your automobile factory, evaluating every part and bolt-tightening process, but then not arranging for anyone to actually start the car at the end of the assembly line.
“Okay, Michael” I can hear the more process-aligned members of GTIM Nation say, “what do you think should replace these criteria as audit standards for Earned Value Management Systems?” Simple: do they work? And by “do they work?”, I mean do they actually help PMs bring their projects in on-time, on-budget? Here’s how to make that determination.
First off, does the organization’s project portfolio have a preponderance of Projects that come in on-time, on-budget? The proof of the pudding, as they say, is in the eating. If a strong majority of the Projects in the portfolio do come in on-time, on-budget, then I’d say that’s evidence of that organization’s PMIS functioning rather well. Either that, or they are staffed with PMs who are so exceptional that they don’t need timely and accurate cost and schedule performance information in order to succeed.
Of course, most organizations can’t point to an overwhelming Project success rate to provide ipso facto evidence of PMIS efficacy, but that doesn’t necessarily mean that such systems aren’t reliable. After all, some published estimates put the percentage of Projects coming in on-time, on-budget at 30 – 50%, depending on industry.[i] I believe that the litmus test for EVMS reliability should be: did the EVMS accurately predict those Projects that did overrun or come in late, by the near- actual amount of the overrun or delay? And how far in advance did the EVMS make that prediction? Because if the baselines were acceptably derived, percent complete data reasonably assessed, and actual costs correctly collected, even basic EV analysis will reliably predict the overrun/delay to within ten points (typically), and do so months in advance for anything but short-duration, low-budget work.
I think that part (most?) of the reason that there’s some level of truth in the axiom that you can put fifty PMs into a room and they won’t agree on the color of an orange is due to this movement of the determiner of PM Information System validity in general, and Earned Value Management Systems in particular, away from actual success rates, and towards process and function policing. And that, dear GTIM Nation, is the problem with PM audit standards.
[i] Retrieved from Chat GPT 6 Luna on October 27, 2026, 20:03 MDT.
Posted on: October 31, 2026 10:54 AM |
Permalink
