Moral Support
Looking for other way to run programs
The conventional practices in program management are extremely procedural and process driven. Newcomers are asked to accept as obvious that all one needs to do is create a plan and drive a team to follow it. It seems obvious that plans will not go as expected, or that not all risks can be accounted for - but more often than not to speak the obvious results in looks of confusion or disappointment. It has been nice to find a trail of papers and studies that show that the battle against the proceduralists has been going on for some time - though admittedly the proceduralists continue to win. Still - good to find others that think about this and that are better at summarizing their positions in writing. Below is a short summary of the best that I found over the last couple of months. I think they are more than enough for anyone who is inclined to follow down the path a little more.
Programmed to Fail
Everything started for me with the amazing work by Eric Lofgren in Programmed to Fail. Since WWII, the US military has been the largest single customer for large, complex technology programs.1 The US government has had a pervasive - and perhaps pernicious - influence on how programs are managed as a result. Programmed to Fail walks through the history of the US governments influence on Program Management from WWII through the 70’s. The work is detailed, but thoughtful and interesting throughout. Two areas that I found particularly interesting where Lofgren’s deep dive into cost account management/PERT and his great explanation of complexity science and John Boyd’s contribution in the defense world.
Lost Roots
The references in Programmed to Fail put me on the trail to other great work. While Lost Roots does cover quite a bit of history like Programmed to Fail, it also spends time proposing alternatives to the practices that have taken over program management. The authors argue that trial-and-error and parallel paths have a home in program management. This line in the paper made my heart flutter:
Companies do apply trial-and-error and parallel approaches in their novel projects because they have no choice, but in doing so they go against their professional PM training rather than being supported by it.
Programs that I lead of any complexity will have trial-and error and parallel paths. But I often have to keep it quiet. As I have become a veteran, I have to really work to convince young engineers that a program going down alternate paths is not failure when all the teaching they get from the PMI certified people in the Program Management Office (who almost always have never managed a complex program of any consequence2) tell them differently.
Armen Alchian
Lost Roots spends some time talking about an iconoclast economist at the RAND corporation named Armen Alchian. In Uncertainty, Evolution and Economic Theory, Alchian argues that real progress is made by an evolutionary process rather than through plans. Success that looks planned likely only looks that way when looking backwards in time. In reality, the environment selects the correct actions needed at the time and lets others die on the vine just like evolution in nature. Alchian’s work seems like a predecessor to some of my other favorites like Kahneman and Taleb.
People like to find reason in random events after-the-fact and are ill-suited probabilistic thinking. I think this line of thinking goes a long way to explaining why there is such a strong institutional drive towards procedure in program management. People like certainty and are uncomfortable with the idea that it is impossible to root out randomness from their work. When programs fail, people want reasons. Much more often than not, they latch on to some step in the process that was not followed as reason for failure - that’s an easy explanation that can be presented to management as a corrective-action. Instead, we should be taking the advice of Alchian, Kahneman and Taleb and looking for ways to best live in a random world.
Disclaimer: I don’t have the numbers at hand to back this claim up. I am fairly confident that over any reasonable time period - it is true. Apollo, Manhattan, JSF, ISS, Stealth, beat that. It does seem that the tech private sector R&D has been the most serious threat to US government dominance.
Nope, no disclaimer needed.