When a capable company keeps missing by a margin it cannot explain, the reason usually does not sit in the strategy and does not sit in the talent. It sits between them. There is a layer underneath the visible work that turns a decision into a set of handoffs, a priority into a cadence, a team’s output into something another team can pick up without losing half of it on the way. That layer has a shape. It is rarely written down in one place. Even where parts of it sit with operations, program management, or a chief of staff, almost nobody owns it as a system. And it is almost always the quietest reason the company is slower than it should be.
The public version of this problem is almost always told in the language of strategy or talent. The strategy is unclear, or the talent is thin, or the culture is off. Those stories are comfortable because they are the ones with obvious owners. A CEO can reset strategy. A head of people can raise the talent bar. A leadership team can work on culture. The coordination layer, by contrast, rarely has a named owner with end-to-end authority. It is produced as a by-product of everything else, and it degrades the same way.
What sits in this layer is mundane once you start listing it. The meetings where cross-functional trade-offs get resolved. The decision rights that determine where a call gets made. The review cadences that surface problems early enough to matter. The handoffs and escalation paths by which work moves. The reporting relationships through which context travels, not just status. The shared records and vocabularies that stop the same confusion being solved again and again.
None of this is glamorous. All of it compounds. A company that runs these things well can execute a mediocre strategy far better than a company with a brilliant strategy and a threadbare coordination layer. The second company will look busy. It will look smart in the room. It will deliver less than it should, and the reasons will be hard to name, because the things that are failing are the things that never get named.
The coordination layer usually accretes rather than being built. Someone set up a weekly meeting two years ago for a reason that made sense at the time. The reason changed. The meeting stayed. Someone added a step to the handoff because a deal went wrong. The deal was solved. The step stayed. Someone routed a decision through a VP because the previous leader could not be trusted with it. The previous leader is long gone. The route stayed. Over a few years you end up with something that still works, expensively, and that nobody remembers designing. Most executives inherit a version of this and treat it as the terrain.
The concrete shape is easiest to see in a cross-functional launch. Product has a roadmap. Engineering has a plan. Marketing has a calendar. Sales has a quota. Finance has a forecast. Each function is internally coherent. None of those five plans slot together without a coordinating forum that sits above any one of them. In companies where the coordination layer is healthy, that forum happens on a known cadence, with a named decision-maker and a short memory of what was decided last time. In companies where it is thin, the launch becomes a series of surprises, and whoever is loudest in the final week tends to get their way.
Another version shows up in slipping initiatives. A program is behind. Two leaders are nominally responsible. Neither has the authority to make the call that would unstick it, and nobody above them has made it clear who does. Weeks pass in which both leaders work hard, run meetings, produce plans, and the actual call does not get made. When the program finally lands late, the story in the board deck is usually about execution difficulty. The real story is about a decision right that was never assigned.
Meetings are the fossil record of the coordination layer. Walk through the recurring calendar of a mid-sized company and you will see, in that pattern, the company’s accumulated answer to the question of how decisions get made. Some of the meetings are doing serious work. A surprising number are solving a problem that no longer exists, or coordinating between teams that no longer need coordinating, or maintaining a relationship that has since been rerouted. A good operator reads a calendar for what each meeting is holding in place, whether it is still needed, and where the first leak would show up.
The mistake most leaders make is to treat coordination as a soft skill problem. If everyone would just communicate better, meet less, be more aligned, the friction would go away. That framing misses the point. Coordination is infrastructure. It has capacity, failure modes, and bottlenecks of its own. It can be over-specified and become rigid. It can be under-specified and become fragile. The problem is usually not the people. It is the layer they are running on.
There is a particular failure mode that is worth naming. Companies with a weak coordination layer develop a pattern where individual teams do excellent work that does not compound. The analytics team ships a good model that another team was too busy to integrate. The engineering team builds a platform that product never reorients onto. The sales team closes deals with a customer profile the operations team has no way to serve. Each team looks good on its own. The company looks sluggish. That gap between team-level quality and company-level output is the clearest sign that the coordination layer is not carrying the load.
An operator who learns to read this layer has an edge over one who does not, because the fastest hidden gains in most companies live here rather than in the visible work. A week spent pruning three dead meetings and clarifying two decision rights can shift a quarter’s output more than a round of reorganisation. This is unfashionable, because it does not look like leadership in the heroic sense. It looks like maintenance. It is maintenance. Running a company well is largely the patient maintenance of things that look boring until they stop working.
If you are a senior leader and you want to find the coordination layer in your own company, look for the ambient signs. Where are decisions being made twice. Where do initiatives keep landing late with everybody surprised. Where does a meeting exist whose charter nobody can cleanly state. Which handoffs generate the same kind of rework every quarter. Which escalation paths take a week longer than they should. These are the places where the layer is thin, or accreted, or pointed at a problem that moved. They are where the slowness is coming from.
The work is quiet. The payoff is not. A capable company running on a well-maintained coordination layer feels, from the outside, like it has better people or a sharper strategy than its peers. From the inside, it has the same people and a roughly similar strategy. What it has is a coordination layer running underneath both of them that rarely draws attention and almost nobody is paid to look after. Maintained, the coordination layer is invisible. Neglected, it becomes the ceiling on what the company can actually do.