Coordination tax is the cost a company pays to keep people agreeing on what’s being done and why. At ten people it’s barely visible. At fifty it’s a few hours of meetings and shared docs every week. At two hundred it’s a real fraction of working time. At five hundred it becomes a meaningful share of senior attention.
The tax gets paid whether or not the coordination it buys actually works. A weekly meeting of six people for an hour costs roughly three hundred hours a year, not counting prep or the work that follows. A review doc costs whatever it costs to write and read, whether or not anyone changes course because of it. Companies treat this as work because the people doing it are working. Whether the work is buying real agreement is a separate question, and it gets asked less often than it should.
Run the tax like any other cost center. Audit it periodically. Ask which meetings and shared docs are producing the agreement they claim to produce, and retire the ones that have drifted into overhead for its own sake. In most companies, some meaningful share of coordination survives after the reason for it disappeared. Cutting it is usually cheaper than people fear, because much of what gets cut is no longer carrying much load.
Any new coordination should be priced before it’s added. A standing weekly meeting, a new review pack, a new approval step, a new dashboard to watch: each one is a real spend on attention. If it isn’t producing more value than it consumes, it’s overpriced, and the fact that nobody had to sign off to create it doesn’t change what it costs.
One of the differences between companies that compound past two hundred people and companies that stall is what they do with the coordination tax. Everyone pays it. Few people own it. The ones who do buy back attention that everybody else is quietly spending away.