The most expensive GCC mistake is keeping a great delivery leader too long

Standing up a GCC rewards a particular kind of leader: hands-on, improvisational, willing to do 3 roles at once because there’s no one else to do them yet. That leader is often exceptional. They set the culture, made the first 10 hires personally, and carried the parent’s trust when the centre was still a bet rather than an org chart. A great builder can carry a centre a remarkably long way: through set-up, through the first scale, through the phase where hands-on excellence is exactly what the moment needs.

But there’s a point where the centre needs something the builder isn’t. Someone who leads leaders, who owns the charter, who runs through structure, and who manages the parent relationship as a strategic peer. The founding profile and the scaling profile rarely live in the same person. Building got the centre here. It won’t take it where the centre now needs to go.

The mistake is keeping them in the top seat past the moment the seat changed, because they’re good, because they built the place, and because no one wants to move a person who has done nothing wrong.

This mistake is expensive precisely because it doesn’t look like one. The builder keeps delivering. The numbers hold. Everyone points to the person who stood the place up. What’s missing is what isn’t happening: the charter that has stopped expanding, the leadership bench that isn’t deepening below the top layer, the parent relationship that stays transactional when it should have turned strategic two phases ago. The centre stops climbing, and because there’s no failure to point at, the cause goes unnamed for a long time. In our experience the plateau can run a year or more before anyone connects it to the seat.

The signs are quiet, which is why they get missed. The charter hasn’t grown in the last 2 planning cycles. The best senior people have started to leave, or have quietly stopped putting their hands up, because the layer above them isn’t opening. Decisions that should sit 3 levels down still funnel to the top of the house. The parent has begun routing its most ambitious work to another centre without ever saying why. None of these reads as a crisis on its own. Together they describe a centre that has outgrown the way it’s being led.

By the time it’s obvious, you have paid twice: once in the growth the centre didn’t capture while it plateaued, and again in the harder, later transition you now have to force under pressure instead of managing early by choice.

Here’s where good centres get it wrong. The parent has no vocabulary for “great leader, wrong stage,” so when the plateau finally registers, the conversation defaults to the one frame everyone knows: performance. Reviews cool. Scope quietly narrows. A new layer gets hired “to help” without anyone naming what it means. The builder feels the shift, reads it as a verdict on their work, and starts looking. When they go, they take the institutional memory of how the centre was actually built: the relationships, the context, the reasons behind 100 decisions that were never written down.

So the centre loses a leader and gains a gap, and a mismatch that could have been managed becomes a rupture that has to be survived. None of that was necessary. The builder was doing everything right by the standard that got them there. The standard changed underneath them, and no one said so out loud. The problem was the absence of an honest, early conversation about what the seat had become.

You can catch this while it’s still a fit question if you’re watching the seat rather than the person. The tell is that a set of things have stopped happening. Ask a few questions on a planning cadence, not a review cadence. Has the charter expanded in the last 2 cycles, or has it held flat while the headcount grew? Is there a credible layer of leaders below the top seat who could each run a piece of this, or does everything still route to one desk? Are your strongest senior people staying and stretching, or getting quieter?

Then look at the parent relationship, because it’s the clearest signal and the easiest to miss. When the parent trusts a centre’s leadership with genuine ambiguity, it hands over more ambitious, less-defined mandates over time. When it has quietly decided the centre executes but doesn’t shape, the work that arrives stays well-specified and the interesting problems go elsewhere. That drift is the market telling you the seat has moved before your internal reviews will. Read it as data about the stage.

The whole cost of this mistake hinges on when you look. Catch the shift while it’s still a question of fit and you have room to move. Leave it until it has hardened into a question of performance and you have already narrowed your options to the expensive ones.

The centres that navigate this well do two things, and they do them early. First, they name the shift for what it is: the centre has moved into a phase that calls for a different mode of leadership, and that’s a statement about the stage, not the person. Said plainly and in time, it lands as respect, because it’s true and the builder usually already senses it.

Second, they design a role that keeps the builder’s real value inside the building. There’s almost always a version that fits the founding strength precisely: a charter-expansion mandate to open a new capability the centre doesn’t have yet; a hard technical or domain area they are uniquely placed to seed and own; a hand in choosing, hiring and coaching the scaling leader who takes the top seat, and then mentoring the layer beneath. Meanwhile you bring in the leadership the centre now needs for scale, and you bring it in alongside the builder.

Handled this way, the move looks like what it is: the reassignment of a scarce asset to the work that person does best. Get the timing right and you keep the person, the institutional memory, and the momentum, and you get the scale leadership on top of all three. Get it wrong, leave it until the plateau is undeniable and the frame has already curdled into performance, and you pay for each of those three twice, once in the losing and once in the rebuilding.