The Coordination Ceiling

When a portfolio starts missing dates, the response is almost always the same. Add a weekly alignment call. Stand up a programme office. Bring in another project manager to hold the moving parts together. Add a governance gate so nothing slips through unseen. Each of these is a form of coordination, and each is added in good faith, because more visibility and tighter control feel like the responsible answer to work going wrong.

For a while it helps. Then it stops helping. Then, past a point that almost nobody names, it starts to make delivery slower. That point is the Coordination Ceiling: the level of coordination overhead beyond which every additional meeting, report or oversight layer costs the organisation more capacity than it saves. Here is why the ceiling exists, why it is so easy to keep pushing through it, and what to do instead.

The reflex that feels like control

Coordination is the work of keeping separate pieces of work aligned: syncing plans, resolving dependencies, chasing status, escalating conflicts. Some of it is necessary. In a portfolio where projects share people and compete for the same scarce skills, someone has to decide what gives way when two priorities collide.

The problem is that coordination is treated as free. When a project is late, adding a coordination mechanism looks like pure upside: more oversight, more alignment, more chance of catching the problem early. The cost is real but invisible, because it appears on no project's budget. It shows up instead as time drained out of the people doing the actual work, and as a system spending an ever larger share of its energy managing itself rather than delivering.

Why coordination grows faster than the work

The first reason the ceiling exists is arithmetic. Coordination is about connections between things, and connections grow far faster than the things themselves.

Put five parties in a room who all need to stay aligned - five projects, five teams, five decision-makers - and there are ten pairs of connections to keep in sync. Double that to ten parties and the number of connections does not double to twenty. It jumps to forty-five. The work grew in a straight line, and the coordination needed to hold it together grew as a curve.

This is the same effect behind the old observation that adding people to a late software project tends to make it later. The new people create new links that everyone else has to maintain, and for a while the cost of forming those links outweighs the extra hands. A portfolio behaves the same way. Every project you run at once, and every coordinator you add to manage them, multiplies the number of relationships the organisation has to keep coherent. Beyond a certain load, the coordination required to keep the plates spinning grows faster than the capacity you have to spin them.

Coordination is a tax on the constraint

Arithmetic alone would be manageable if the cost fell on spare capacity. It does not, and that is where Flow Economics parts company with the general worry about "too many meetings."

In any portfolio there is a constraint: the one resource, skill or decision point that actually governs how fast value gets delivered. It might be a lead architect, a regulatory approver, or the executive who signs off every material trade-off. The pace of the whole portfolio is set by how fast work moves through that constraint.

Now look at who coordination actually consumes. The alignment calls, the steering boards, the escalations, the status reviews, the "quick" clarifications: they pull disproportionately on senior specialists and decision-makers, because those are the people whose judgement the coordination exists to gather. In other words, coordination overhead lands hardest on exactly the resource the portfolio can least afford to spend. Every hour your constraint spends being coordinated is an hour it is not delivering, and because everything routes through the constraint, that lost hour delays every project waiting behind it.

This has a name in Flow Economics thinking. Time lost on the constraint that pushes out delivery dates carries a drag cost: the measurable economic loss from work finishing later than it could have. Coordination that taxes the constraint is not neutral overhead. It is drag cost dressed up as diligence.

The ceiling is a work problem, not a coordination problem

Here is why organisations keep climbing past the ceiling without noticing. Each project on the dashboard still looks governed. It has its meeting, its report, its owner, its status, so the coordination looks like good management. This is a face of the Project Illusion: the belief that if every individual project is well controlled, the portfolio must be performing well. The loss is not visible on any single project. It is visible only at the level of the whole system, where the constraint is being drained by conversations about work instead of the work itself.

So leaders reach the ceiling, see delivery still slipping, and conclude they need yet more coordination. They add another layer, the constraint loses more time, delivery slips further, and that reads as proof that tighter control is required. The mechanism causing the problem is mistaken for its cure.

The way out is to see the ceiling for what it is. Past that point, the bottleneck is not a shortage of coordination. It is the sheer amount of concurrent work that has to be coordinated at all. Coordination overhead is a symptom of running too many things at once. You cannot govern your way out of it, because governance is the thing consuming the capacity.

What to do instead

The lever is not better coordination. It is less work in the system, so that far less coordination is needed in the first place.

Start by cutting concurrent work in progress. Every project in flight generates dependencies, handoffs and contention that someone has to manage. Because coordination grows as a curve, taking work out of the system reduces the coordination load faster than it reduces the delivery, not slower. Holding a firm limit on how much is in flight at once is the single most effective way to bring an organisation back below its ceiling.

Then protect the constraint's time deliberately. Treat hours on your scarcest resource as the expensive thing they are, and defend them from coordination that does not need them in the room. Most status reporting exists to reassure people who are not doing the work; very little of it changes a decision. When a slot on the constraint does open, let value per constrained resource hour, the value a piece of work returns for each hour it uses of your scarcest resource, decide what earns it, rather than whichever meeting shouted loudest.

Finally, coordinate by exception. A portfolio that has capped its work in progress and knows its constraint does not need everyone aligned on everything. It needs sharp coordination focused on the constraint and the handful of dependencies that genuinely gate delivery. Coordinate those hard, and let the rest run.

The shift

The intuition that a struggling portfolio needs more control is one of the most expensive instincts in management, because it fails so gracefully. Nothing looks wrong. Every project is covered, every meeting is well intentioned. And the whole time, the coordination is eating the constraint and slowing the very delivery it was added to protect.

The counter-intuitive move is to run less at once, spend the constraint on work rather than on being managed, and coordinate only where coordination actually changes the outcome. Do that and the paradox resolves: the organisation that coordinates less, delivers more.