Too much process or too little structure: the two ways growing companies fail

Growing companies break in one of two directions, and each one is usually a panicked reaction to the other.

Failure mode one: process until nobody can move

Something goes wrong, so a rule is added. A hire misfires, so another interview round appears. A launch slips, so a weekly status meeting is born, then a pre-meeting to prepare for it. None of these steps is crazy on its own. Each one was a reasonable response to a real incident. But the steps never get removed, only added, and after two years of growth the company is wearing every scar as permanent armor.

The tell: elapsed time explodes while touch time stays flat. Work is not harder, it just waits more. People spend their week in coordination and call it collaboration. Decisions are technically distributed but practically require five sign-offs, so everything important escalates anyway.

Failure mode two: flexibility until nobody knows who decides

The opposite reflex, common in companies proud of their culture: keep it flat, keep it informal, structure is for corporates. It works beautifully at twelve people, because twelve people can coordinate in a hallway. Somewhere past twenty or thirty, the hallway stops working. Informal coordination has a ceiling, and it arrives earlier than almost anyone expects.

The tell: everything is everyone's and nothing is anyone's. Two people quietly duplicate work; a third thing falls between them entirely. New people take months to become useful, not from lack of ability but because the map of who owns what exists only in veterans' heads. And in the vacuum, decisions default upward: the founders become the routing layer for a company that swears it has no hierarchy.

Why do companies swing between the two?

Because each failure mode looks like the cure for the other. The chaotic company adds process and overshoots into bureaucracy. The bureaucratic company declares a culture reset, strips the rules, and rediscovers chaos. We have watched companies complete this full pendulum swing in under three years, at enormous cost, without ever passing through the point they actually needed.

What is the point they need?

Minimum deliberate design. Not more structure, not less: the smallest amount, chosen on purpose. In practice that is three things done explicitly and almost nothing else. Clear goals: what the company must be able to do next year, stated so a new joiner could repeat it. Clear ownership: every significant domain has one name on it, and overlaps are resolved rather than tolerated. Clear decision rights: the recurring decisions each role owns, written down, so authority stops defaulting to the loudest voice or the founders' calendar.

The best small teams in the world already run this way: total clarity about who does what and who decides, standards instead of supervision, and very little else. Most companies never get there, not because it is hard, but because nobody's job was ever to design the company itself. Everyone was too busy working inside it.