The Friction in the Room

I sat in many meetings where people from NASA and SpaceX were at the same table. This was during the Commercial Orbital Transportation Services (‘COTS’) programme, a joint initiative that put two very different organisations in the same room and asked them to produce something together. I was in those meetings enough times to notice a consistent pattern.

Smart people in a room generate positive tension. Ideas get sharpened, assumptions get challenged, decisions get better. That is what you want. But when the people in the room are operating from genuinely different philosophical foundations, the tension becomes friction. Not conflict necessarily, but a subtle resistance – energy going into the gap between orientations rather than into the problem. It was visible if you knew what to look for. Most people in those rooms only knew one side.

I had been inside both. That made the difference apparent.

What We Usually Think Makes Elite Organisations Special

Writing about elite organisations tends to focus on the visible aspects. The selection processes, the extreme commitment stories, the bold leadership moves, the rituals, the stated values. These are real, and they matter. But they are also what gets copied when organisations try to replicate elite performance without understanding what actually produces it. The result is what you might call cargo cult leadership: organisations that have borrowed the artefacts without the underlying structure that gives them meaning.

NASA and SpaceX are the most instructive examples I have encountered because they look almost nothing alike on the surface. NASA was programmatic, structured, and rooted in science and discovery during that period. The people it attracted were thorough, careful, and oriented towards rigour. SpaceX was agile, hungry, often young, and oriented around a specific and ambitious mission: the democratisation of space travel. The surface contrast was stark.

And yet both are genuinely elite. Both have achieved things the rest of their industry said were impossible. Which means the surface contrast is not the interesting thing. What sits underneath both of them is.

How Failure Flows

The clearest structural difference I observed across every elite organisation I was part of was not about how they approached success. It was about what they did with failure.

In the Royal Marines, failure was an opportunity. The after-action review was not an exercise in blame. It was an assumption that something had gone wrong, a systematic attempt to understand what had gone wrong, and a commitment not to repeat it. Failure was disaggregated from the person and treated as information the whole team needed.

NASA’s unofficial mantra for a long time was ‘failure is not an option.’ This is sometimes misread as arrogance. It is not. It is the accumulated knowledge of an organisation that has experienced catastrophic, public, lethal failure and built a culture around preventing it. The rigour, the review processes, the safety culture – these are not bureaucracy. They are what you build when you have learned the hard way what happens without them.

SpaceX inverted the frame entirely. Failure was expected. Sometimes it was the point. It is often faster to find the solution by rapidly determining what does not work than by trying to design the right answer in advance. Test vehicles were budgeted to fail. A failed test was a data collection event, not a disaster.

In my startup, I made a conscious choice to take the best of all three. Analyse and learn from the Marines. Build specific non-negotiable standards from NASA. Use failure as a method, as SpaceX did, when the stakes allowed it. That synthesis was only possible because I had been inside all three and understood why each approach was rational in its context.

Most organisations have not made that choice consciously. They have inherited an approach to failure from whoever built the culture before them, and they have never examined whether it still fits what they are actually trying to do.

Where Authority Actually Lives

The second structural difference concerns where decisions are made.

In NASA, authority sat with the programme. Programmes owned things. Responsibility was clear and bounded. That produced accountability but also edges – places where ownership stopped and something important fell between structures.

SpaceX’s model was democratic at the operating level. Individuals were expected to own problems, challenge bad decisions, and move without waiting for permission. This was genuine, not rhetorical. But it also came with the understanding that the founder could override any decision at any moment. Empowered day to day, with a ceiling that everyone could feel, even when they could not see it.

The Royal Marines model is the cleanest version. Mission command: You are told what success looks like and why it matters, and you are trusted to determine how. The authority at the edge is real, not provisional. A junior commander making a call in the field is not acting outside their authority. They are exercising it as designed.

What all three have in common is that the authority model was explicit rather than assumed. Everyone understood where decisions sat, what they were allowed to own, and what they needed to escalate. Most organisations I have observed since have never made that explicit. People guess at the boundaries, and the guessing produces exactly the kind of friction that looks like a cultural problem but is actually a structural one.

What Organisations Are Actually Sorting For

The third structural difference is about who stays and who does not, and on what basis.

NASA operated as an elite club. Advancement tracked against known performance equations. Once inside, the sorting was visible and relatively predictable. SpaceX sorted differently. The democratic, empowered culture was real at the operational level, but there was a hard boundary: fundamental misalignment with the mission got you removed. The day-to-day empowerment was genuine, but the direction was not up for debate. The Royal Marines, at the level I operated, sorted through what people could actually do when it mattered.

Three different mechanisms. What they shared was that the mechanism was applied consistently and, crucially, people inside each organisation broadly understood what they were actually being sorted on. That sounds obvious until you look at most organisations, where people believe they are being assessed on one thing and are actually being assessed on another. That gap – between the stated sorting criteria and the real ones – is where good people leave, and mediocre ones survive.

The question for any leader is not whether their sorting mechanism is the right one. It is whether the people in their organisation actually know what it is.

What Gets Made Explicit and What Gets Left to Guesswork

The fourth structural difference is about standards – not values, which are easy to write on a wall, but specific expected behaviours that people are actually held to.

NASA had a framework it called the Business Imperative, which defined five expected behaviours: trustworthy, key player, open-minded, innovative, and accountable. At the working level I operated within, these were visible and taken seriously. They gave people a common language for what good looked like that went beyond technical performance. Whether they were applied with the same consistency at senior levels, I cannot say with confidence – I was not in those rooms. But that caveat is itself instructive. Explicit standards are only as powerful as their consistent application, and the most common failure point is precisely at the level where people assume the standards apply to everyone else.

SpaceX’s expectations were less formally documented but equally explicit in practice. You were expected to challenge bad decisions regardless of where you sat. You were expected to own problems rather than route them. Those expectations were enforced, not just stated. The Royal Marines had a leadership framework that was specific enough to be actionable rather than aspirational.

The common thread is not that these organisations had values. Every organisation has values. It is that they had made specific behaviours explicit, visible, and applied – so that people knew not just what was expected but what it looked like when it was not being met.

What Is Actually Transferable

The question worth asking is: what can a chief executive or a board member running something other than a space programme or a special forces unit actually take from this?

The answer is structural, and it is specific. You do not need extreme selection processes or existential stakes to build these patterns. But you do need to make four decisions explicitly that most organisations leave to drift.

What is your actual approach to failure, not the stated one but the real one, and does it fit what you are trying to build? Where do your people believe decisions actually sit, and does that match where you think they sit? What are people genuinely being sorted on, and do they know it? And what specific behaviours are you holding people to, visibly and consistently, rather than leaving to interpretation?

These are not cultural questions. They are design questions. And unlike culture, design can be deliberately changed once you have decided to do so.

The Choice You Are Already Making

If you are reading this and feeling the gap between what you have just read and what your organisation actually looks like, that is the right response.

Most organisations are not elite. Most of them know it. The question worth sitting with is not how to become more like SpaceX or the Marines. That path leads to the cargo cult version – borrowed artefacts, no underlying structure.

The question is where you are currently making authority ambiguous, absorbing failure quietly, sorting for the wrong things, or relying on standards nobody has made explicit. These are structural choices, even when they feel cultural.

The harder version of that question is whether you are willing to act on what you find. Making failure visible means admitting your organisation has problems you have been managing around. Distributing authority means letting people make decisions you would not make. Explicit standards mean measuring against them, which means being willing to act on what you find.

Most organisations choose not to. That is also a choice. It just usually goes unnamed.