August 9, 2021

Smart government leaders find ways to validate projects before heavily resourcing them. They spend small dollars to learn, then large dollars to execute. 


It’s the reverse of the process used in most government efforts. The traditional approach is based on extensive planning and documentation. In the typical way of doing things, project leaders report upfront on all possible conditions, permutations, combinations, and contingencies. The confident presentation of an executive summary squashes all doubt, and the executive champion expects reality to unfold more or less as described.


The problem with this approach is that it assumes we can predict the future by knowing isolated facts. Project leaders deconstruct a massively complex system into subsystems, which are further reduced ad nauseum until each unit can be analyzed on its own. Once assessed, units are then “recomposed” into ever-greater subsystems until we arrive again at the whole. 


The process is time-consuming, wastes funds and other resources, and affects our ability to stay competitive. According to a 2020 GAO report: major defense acquisition programs “have accumulated over $628 billion (or 54 percent) in total cost growth since program start, most of which is unrelated to the increase in quantities purchased. Additionally, over the same time period, time required to deliver initial capabilities has increased by 30 percent, resulting in an average delay of more than 2 years.” 


In other words, large projects are 54% more expensive and take 30% more time than planned. These are dramatic changes of budgets and timelines that were extensively researched by highly trained, motivated, patriotic Americans. The problem isn’t with them. The problem is with a system that reduces complexity into tiny chunks, then makes decisions based on an awkward reassembly of those chunks into a crude whole. 


Why doesn’t this approach work? Because complex systems don’t behave predictably. Combining the functions of all parts does not add up to any useful sum. Aristotle said as much over 2,300 years ago: “the whole is something besides the sum of its parts.” 


Government leaders should instead focus on funding experiments to validate basic system performance. In our work with government innovation, this phase is often overlooked, but slowly we’re seeing this change, with positive results. Leaders who embrace this phase enjoy much greater results than their peers using outdated development models. 


One customer did several months of work that cost under $150,000 to develop a simplified model of a system. It showed the art of the possible. The team then used those conceptual models to source funds for $5,000,000 worth of follow-on prototypes over the course of a year. The most promising prototype led to the establishing of a funded program office within two years. 


This is Gall’s Law in action, which states that complex systems evolve from simple systems that do one thing well. Complex systems do not get designed and then built. The key, then, is building a simple system as quickly and cheaply as possible. Program Managers who take time to learn about the essential functions of a stripped-down system  can effectively plan for development, delivery, and sustainment. 


21st century government leaders know they are building complex systems in a dynamic environment. They can more effectively and efficiently deliver results by embracing an upfront phase of building simple systems to learn, rather than making critical assumptions that lead to delays, cost overruns, and poor performance.

Photo credit: Shutterstock photo