The system is finally working, so you move to make it explicit.
Write the playbook. Set the targets. Formalise the operating model. Roll the winning approach out across the organisation. This is what maturing looks like – turning what works into something repeatable.
But there is a question underneath all of it that decides whether any of it helps: has the system actually shown it can decide well under pressure, consistently, on its own? Not once, in a good quarter, carried by a few strong people. Consistently – when conditions shift, when the senior person is not in the room, when the case is one no playbook anticipated.
If it has not, then making it explicit does not capture judgement.
It captures the absence of it.
Because each of these moves – articulation, scaling, optimisation – assumes there is settled judgement underneath to encode. A playbook written before the system reliably decides well does not record how the organisation thinks. It records how it happened to act, in a handful of cases, before anyone knew which of those actions would hold. You write it down, and it looks like capability captured. It is a snapshot of behaviour that had not yet proven itself, made permanent.
And the order is not interchangeable. There is a sequence to building a system that can carry its own intent, and articulation sits near the end of it, not the start. Decisions have to land in the right place first. Shared judgement has to form through real cases, not be declared. Structure has to settle around what actually works. Only then – once the system is genuinely deciding well, and silence about it would start to cause drift – is it time to name it. Articulation is the act of describing judgement that has already proven durable. Done in that order, it stabilises. Done early, it substitutes – the written version quietly replaces the reasoning it was supposed to capture, and people start following the playbook instead of thinking.
The tell is what the artefact gets used for. When intent is named at the right time, people use it to say no – to decline the deal that does not fit, because the logic underneath is real and shared. When it is named too early, they use it to negotiate – to argue the exception, because the words are there but the judgement they describe never actually formed.
Same document. Opposite function. One makes the system more decisive. The other makes it more defended.
This is the hard part, because the pressure to make it explicit is heaviest exactly when you should resist it. The system looks like it is working. The board wants the model. It feels like the responsible, mature next step. And naming it then is precisely what arrests the judgement that was still forming – you freeze the system at “almost”, and teach everyone to perform the words rather than keep developing the reasoning underneath them.
So the discipline is not to articulate less. It is to articulate later – to let judgement prove itself under real pressure first, and only then write down what has earned the right to be written down. What you lose by waiting is the comfort of looking finished.
What you keep is a system that is still thinking. Where in your organisation are you trying to make intent explicit, scalable, or optimisable before the system has demonstrated consistent judgement under pressure?
This question closes a chapter of The Architecture of Commercial Performance, where the thinking behind it is developed in full.
