Johnson railed against group decision-making in design. He argued that while committees might avoid truly stupid outcomes, they reliably prevent the bold, singular vision required for breakthrough advances. Brilliance, he believed, requires individual authority and conviction.

To maintain small, elite teams, Johnson decoupled compensation from headcount. This removed the bureaucratic incentive for managers to grow their teams for status or pay, ensuring the organization stayed lean and focused on individual contribution and impact.

Many leaders fight bureaucracy like an external threat. The real cause is the organization's design: too many layers, functional silos, and distant decision-making. To fix bureaucracy, you must fundamentally change the organizational structure, not just treat symptoms.

Johnson's model demanded lean interfaces with the military and intelligence customers, sometimes limiting their team to just six people. This ensured quick decisions and minimal correspondence, making the entire project ecosystem faster, not just his internal team.

An engineering team's velocity is often bogged down by non-engineering work, which can consume a significant portion of their time. A leader's primary role in accelerating projects is to identify and systematically remove these obstacles, freeing engineers to focus on creative problem-solving and core design tasks.

Transformative defense systems like the Skunk Works' U-2 were championed by accountable, visionary founders like Kelly Johnson. Modern programs, often built by committee with components spread across congressional districts for political reasons, lack the focused leadership required for true breakthroughs.

Beyond hiding projects from adversaries, secrecy served a critical internal function: it insulated the team from corporate bureaucracy and distractions. This allowed a compact, focused group to maintain high velocity without interference from the larger organization.

The primary bottleneck to organizational speed isn't how fast individuals work; it's decision latency—the time it takes for decisions to be made and flow through the organization. This stems from unclear decision rights, poor communication, or lack of empowerment. Reducing this latency is the key to accelerating engineering and overall business velocity.

The idea that you need a massive framework to scale agility is a lie. Agility doesn't scale; bureaucracy does. To increase speed and responsiveness, you must relentlessly de-scale the organization by breaking down silos into smaller, cross-functional, autonomous units.