Flat Organizations Are Accumulating Coordination Debt. Almost None of Them Know It.
Flat Organizations Are Accumulating Coordination Debt. Almost None of Them Know It.
Every flat org announcement contains the same claim: less bloat, faster decisions, more agility. The board deck says the layer was unnecessary. Work continues to ship. For a while, sometimes a long while, the metrics confirm the story. There is a cost the dashboard does not capture. It has a name: Coordination Debt.
Technical debt is the accumulated cost of decisions made without quality architecture. Code continues to function after you skip the design work, with increasing friction, increasing brittleness, and increasing cost per change, until the debt must be paid. The payment is always more expensive than the original investment would have been. Coordination Debt works by the same mechanism. When a management layer is removed without redesigning the coordination architecture, the coordination work does not disappear. It redistributes and compounds.
The redistribution is predictable. The work goes downward, to the most senior people at the level below the removed layer. They absorb it without a title, without a budget, without a mandate. Work continues to ship. The dashboard looks fine. Upper management is the last to know, because what gets measured is output, and output is still arriving. What is not visible is what the output is costing: the informal meetings that never appear on any calendar, the decisions that now require five people in a room instead of one, the senior engineers carrying coordination work that is not in their job description and is slowly consuming the capacity they were hired to apply to something else.
What those engineers are doing has a structural name: Signal Staging. The continuous judgment about when information from one team has matured enough to transmit to another, without generating rework, misaligned expectations, or consumed capacity that cannot be recovered. The management layer was performing this function at every team boundary it held. Remove the layer without redesigning the boundary, and Signal Staging does not stop. It migrates to whoever will absorb it. That person is now performing a critical infrastructure function as an informal obligation layered on top of their actual job.
The conventional case for removing the layer treats it as headcount. Fewer managers means less overhead, less coordination tax, more speed. This framing is wrong in a specific way: it assumes the layer’s value was in managing, not in coordinating. When the actual function was carrying context across team boundaries, translating decisions across functions, and holding the interface between systems that answer to different governing realities, removing it without a replacement design does not reduce the coordination tax. It reassigns it. The tax becomes invisible because it is now distributed across job descriptions where it does not appear.
The leaders who made this call are not operating without information. They are operating without an inventory of what they are deciding about. That is a different problem, and a correctable one. The decision to remove a layer is not the failure. The failure is making that decision before completing the function inventory: what was this role actually doing, not what should it have been doing? The delta between normal operations and the week that manager was out sick is the coordination surface that must now find a structural home. Some of it should: a defined process, a different role, a mechanism that carries the function without the headcount. Some of it was overhead without function and does not survive the inventory. The inventory must precede the announcement, not follow the compounding.
The diagnostic is direct: can you name, right now, every coordination function the removed role was performing, not what the job description said, but what actually happened the last time that person was out sick for a week? If the list is difficult to produce, the organization is carrying coordination debt it has not yet measured. The payment is coming. The only question is whether it arrives as a design decision or as a departure.
I write about structural leadership for technical leaders in high-stakes operating environments. If this way of thinking resonates, it runs deeper in The Edge Case: http://TheEdgeCaseBook.com
I write about structural leadership for technical leaders in high-stakes operating environments. If this way of thinking resonates, it runs deeper in The Edge Case: http://TheEdgeCaseBook.com
