The work of dividing a plan gets a meeting; the work of reintegrating it rarely gets an owner.
Why it matters when the plan changes
Dividing work feels like progress because it produces visible artefacts: a structure, a set of owners, a list of workstreams. Integration produces nothing until it fails. That asymmetry in visibility is enough to explain why the second half of the job is routinely under-resourced, without anyone deciding to under-resource it. The mathematics compounds this: Brooks's law shows communication channels rise sharply as more people are added, while the effort assigned to managing them does not rise to match.
The tension is that the bias is not irrational at the individual level. A leader is rewarded for what their own part delivers, and integration is by definition somebody else's boundary as well as theirs. Fixing it needs someone whose remit is the interface itself, which most structures do not create. Harvard Business Review reported in 2015 that only 9% of managers say they can rely on colleagues in other functions and units all the time, the shape neglected integration produces.
In practice
A transformation is broken into five workstreams with five named leads, each with a clear remit and a reporting line. No one is given the interfaces between them. Each workstream delivers, and the combined result does not work, because the assumptions each made about the others were never compared.
Evidence
Coordination cost rises with every party added, while the effort spent on integration does not rise with it.
Brooks's law, after Frederick P. Brooks, The Mythical Man-Month (1975)Cross-unit reliance is reported as far weaker than reliance within a line, which is what neglected integration produces.
Donald Sull, Rebecca Homkes and Charles Sull, Why Strategy Execution Unravels and What to Do About It, Harvard Business Review (2015)
What it cannot tell you
Coordination neglect explains why integration is under-resourced; it does not identify which interface will fail first or how much rework will follow. It is a tendency, not a diagnostic for a specific project, and it says nothing about whether a given workstream division was the right one to begin with.
Questions
Frederick P. Brooks noted in The Mythical Man-Month (1975) that communication channels multiply rapidly as more people join a project, a mechanism now called Brooks's law. Later research on how groups divide and reintegrate work found the same asymmetry: partitioning is done carefully, integration is not.
Poor communication is a shortfall in how well people exchange information. Coordination neglect is a shortfall in how much effort is allocated to integration in the first place. Better communication inside each workstream does not close it, because the gap sits between them.
Rework, sequencing failures and delay that appears late and resists attribution to any single workstream. Harvard Business Review reported in 2015 that only 9% of managers can rely on colleagues in other functions all the time, which is roughly the state of the interfaces where this cost is paid.
Give the interfaces owners at the same moment the work is divided, not afterwards. The decision that creates a boundary is the cheapest place to name who holds it, and almost every structure fails to do that because the boundary has no natural home.
Yes, on both sides. Larger organisations create more interfaces, and the distance between the parties on each interface grows, which makes the informal repairs that hold small organisations together less available. The neglect stays constant while the thing being neglected grows.