Composite, drawn from patterns across engagements, not one specific client.
At five people, every project decision ran through one weekly meeting, everyone in the room, decisions made fast, informally, and everyone left aligned. It worked well, genuinely well, for years. At twenty people, the exact same meeting was actively slowing the business down, and nobody had changed the process on purpose, it just stopped scaling.
Why the same process broke
At five people, "everyone in the room" meant five people who each had full context on every project. At twenty, "everyone in the room" meant fifteen people sitting through decisions on projects they had no context on and no stake in, while the two or three people who actually needed to decide waited for their five-minute slot buried in an hour-long meeting. The process hadn't changed. The math underneath it had completely changed.
What the actual cost looked like
Roughly ten person-hours a week, spent by people listening to decisions that didn't involve them, just to reach the same handful of actual decisions that used to take twenty minutes total when the room was smaller. Nobody had measured this, it had just quietly become normal, the way inefficiency does when it arrives gradually instead of all at once.
What replaced it
Decisions moved to the smallest group that actually had context and authority to make them, async by default, with a short written summary posted for visibility instead of requiring live attendance. The full-team meeting dropped to twice a month, for things that genuinely needed everyone in the room, real cross-team decisions, not routine ones.
The lesson, generalized
A process built for one headcount doesn't fail because it was badly designed. It fails because it wasn't designed to scale, and few processes are, because nobody designs for a headcount they don't have yet. The fix isn't blaming the original process, it's building a habit of periodically asking whether last year's process still fits this year's size.