Every transformation deck has the same four words on a slide somewhere: Vision, Architecture, Structure, Governance. Draw the boxes, connect the arrows, and it looks complete. It isn’t.

Vision is where you want to go — what you want to become, or the difference you want to make. Architecture is the blueprint for getting there: the plan that supports the vision and the goals sitting under it. Structure is the people, processes, and tools that coordinate the work — the thing that turns a room of individuals into a working unit. And Governance is supposed to hold the other three in place: keep them aligned, protect them, make sure they stay complementary, and make sure the right decisions get made.

On paper, that’s a complete system. Vision sets the direction. Architecture translates it into a plan. Structure executes the plan. Governance protects the whole thing from drifting apart.

Here’s the problem. That model describes a system in equilibrium. It says nothing about what happens when the equilibrium breaks — and in every program I’ve sat inside for the last twenty years, it always breaks. Not dramatically. Quietly.

Structure is the part that erodes first, and it erodes in a specific way: the people who understood why the structure was built that way move on, get promoted, or get tired. The process that was designed to force a hard conversation becomes the process everyone routes around. The tool that was supposed to create a single source of truth becomes one more system nobody trusts, so people keep the real numbers in a spreadsheet on the side. None of this shows up as a failure. It shows up as a status report that still says green.

That’s the piece missing from the four-box version: a mechanism for noticing the structure has quietly stopped matching the architecture it was built to serve. Governance is defined as the function that protects alignment — but protection isn’t a state, it’s a habit. If Governance is just a charter and a steering committee that meets monthly to hear good news, it isn’t protecting anything. It’s a formality that lets drift travel further before anyone names it.

And this is where the framework needs a fifth element it doesn’t have: an independent check. Everyone inside Structure has a reason, usually a good one, to report that the structure is working — their job, their sponsor’s confidence, their own belief that they’d know if something were wrong. That’s not dishonesty. It’s what happens when the people executing the plan are also the only ones checking whether the plan still holds. You need someone whose job is only to look at the gap between the architecture on paper and the structure in the room, and who has no stake in saying it’s fine.

The notes I started with also treat Leadership as the closing chapter — the list of virtues that make the first four boxes work: making hard choices, walking the talk, sacrificing for the cause, leading with humility and courage. All true. But written that way, leadership reads as a personality trait some people have and others don’t. It’s more useful, and more honest, as a structural question: who, specifically, has the standing to say “this has drifted” when saying so is inconvenient — and what happens to that person when they say it. If the answer is nobody, or if the answer is someone whose next pay review depends on the report staying green, you don’t have a leadership gap. You have a governance failure wearing a leadership costume.

I think about this in three stages, not four: Direction, before the decision gets made. Oversight, while the work is in flight. Stewardship, after go-live, when the architecture is no longer new and nobody’s watching as closely. Vision and Architecture live inside Direction. Structure and Governance are supposed to carry through Oversight and Stewardship both — but most organisations build them once, at the Direction stage, and never revisit them once the program is declared a success.

The framework isn’t wrong. It’s incomplete in the one place that matters: it describes how to build the system, not how to notice when the system has quietly stopped being the one you built. Your project looks fine. That’s usually the risk.