Thoughts on Efficiency
I have been thinking about organizational efficiency. There are many schools of thought on how to run an efficient business: Six Sigma, Lean, Agile, Just-In-Time, etc. Going back to the inspiration for this newsletter, I reread Kelly's 14 Rules of Management. Reading through the 14 Rules, I initially thought of Tiger Teams. The difference I see between the Tiger Team construct and the 14 Rules, is that Tiger Teams are assembled to act quickly, solve a problem, and disband. What makes Kelly's vision unique, a management philosophy designed to maintain that same level of urgency and innovation indefinitely. This led me to a parallel idea that I had read recently. In “The Defense Reformation”, Palantir CTO Shyam Sankar lays out 18 theses to reform industrial inefficiencies in the DoD. His ideas challenge traditional bureaucracy, just as Kelly’s Rules did.
To explore this, I’ll compare two of Kelly’s Rules with two of Shyam’s Theses. Starting with rule #5: “minimize reports while ensuring thorough documentation of important work.” Interestingly, this aligns with the 5th thesis: “the only requirement is winning.” To quote the thesis directly,
The most important projects don’t come from requirements. America’s cultural strengths are fundamentally creative and improvisational. The requirements process ensures we play to our weaknesses. In a fight, no one cares about the requirements document. The only requirement is winning. And winning requires engaging in the messy, overlapping, seemingly wasteful but actually efficient process of being better. Validating requirements leads to solving yesterday’s problem without today’s context. We have countless validated problems.
Both rule #5 and the 5th thesis center on removing barriers to innovation. Kelly’s rule recognizes that process is necessary, but only in its most efficient form. Similarly, Shyam’s thesis doesn’t call for eliminating process altogether; rather, it shifts the focus from rigid system requirements to stakeholder needs. This distinction mirrors traditional systems engineering considerations: verification vs. validation. Verification asks, “has the widget been built correctly?” Validation asks, “has the correct widget been built?” I am disagreeing with Shyam’s use of “validation”. I believe that within the context of the article, his use of “validation” closely aligns with my definition of “verification”. Shyam argues that traditional processes overemphasize verification; ensuring compliance with predefined requirements, and does not emphasize whether those requirements are still relevant in the first place. It’s easy for engineering firms to dismiss Shyam’s thesis as a software-centric viewpoint, but the need for a big-picture focus applies across disciplines. Kelly wasn’t leading software teams, yet he fought to reduce excessive reporting requirements. The A-12 wasn’t built by a massive, bureaucratic machine. It was the product of a lean, focused, and highly motivated team. The trade-off between rigid processes and innovation is clear: excessive structure ensures control, but it also slows progress. In an industry driven by cutting-edge technology and rapid execution, the challenge is balancing discipline with agility. Innovation requires flexibility, risk, and a willingness to fail in order to succeed. Finding that balance is what separates stagnation from breakthrough.
Now, let’s pivot to rule #13, “access to projects and personnel must be strictly controlled for security.” In contrast, the 17th thesis states, “let the people speak to the mission.” For context, I’ll quote an abridged version of the 17th thesis:
… You can’t get clearances unless you have a classified contract, but you can’t get clearance unless you are part of the existing class of cleared people. The same is true for SCIF sponsorship and access to classified networks… There are too many monopolies, and we have long since passed the point where they resembled legitimate security concerns… the legacy primes have enriched themselves because they are the only ones with access behind the SAP door. Private industry will pay its own way here for this enabler. It does not require the government to front these funds. Enable American capital to show up and purchase network, SCIF, classified cloud compute, and clearances — all governed by investor confidence that the company and team can credibly turn that investment into value.
This sets up an inherent tension, security versus accessibility. Kelly’s rule emphasizes controlled access to protect sensitive information, while Shyam’s thesis suggests that broader communication fosters alignment and mission success. As part of the national security establishment, it is no surprise that I believe that classification and compartmentalization is vital. That said, I have a confession, when abridging the 17th thesis, I strategically omitted some context. For one, I completely understand frustration with the process to obtain classified network access, especially if you've been navigating the process for decades. But more importantly, I believe the real intent behind the 17th thesis is not to eliminate security measures but to create a fair and transparent process for accessing classified networks and SCIF sponsorship. Shyam argues that by enabling new entrants, the DoD can encourage competition, diversity of thought, and ultimately, more innovation. Kelly’s philosophy, by contrast, was about identifying talent, empowering it, and then getting out of the way. His small, controlled teams allowed him to manage security concerns without letting them stifle innovation. These are categorically different philosophies. Kelly's produced the SR-71, but how much of that was a result of his management style versus Kelly's singular brilliance.