Between roughly 50 and 400 people, an operation crosses a line: the way work actually runs — policy, onboarding, decision rights, exceptions, escalation — stops fitting in anyone's head, but nobody has written it down yet. The company keeps operating on memory and proximity. It works until it doesn't: an audit, a departure, a new location, a volume spike.
The common responses buy capacity, not infrastructure. A senior title inherits the fog instead of clearing it. Contractors cover the gap while the knowledge stays temporary. A strategy deck names the work and leaves the team to execute it. All three leave the operating layer where it was: undocumented, unowned, and priced by the next emergency.
The alternative is to treat operations the way engineering treats systems: assess the risk, build the standard, write it down, and hand it over — so the operation survives the people who built it, including us.