Process that speeds people up, not slows them down.
The team had grown from eight to twenty-five, and everything that used to take a hallway conversation now took a week. So they did what growing companies do: they added process. Approval steps, status meetings, sign-off chains. Six months later they had order, and they had also lost the speed that made them worth joining. They had cured the chaos and killed the patient.
Every operator hits this wall. Too little process and the company descends into confusion as it grows. Too much and it calcifies into the slow, cautious place its founders left a big company to escape. The art is building process that speeds people up rather than slowing them down, and knowing the difference.
Process should answer a question, not perform control
The test for any process is simple: does it help people make a good decision faster, or does it mostly exist to make someone feel in control? A checklist that stops a recurring mistake is good process. An approval step that exists because a manager wants visibility is usually ceremony. Build the first kind relentlessly. Resist the second, because it multiplies as the company grows.
Add process in response to real pain, not imagined risk
The healthiest way to introduce process is to wait for a genuine, recurring problem and then solve exactly that problem, no more. When the same mistake happens three times, build the lightest possible system to prevent the fourth. Process added pre-emptively, to guard against things that have not actually happened, is where bureaucracy comes from. Let the pain earn the process.
The three-times rule
Do not systematise something the first time it happens, or even the second. On the third occurrence, you have a pattern worth a process. This stops you from building rules for one-off events that never recur.
Default to the lightest version that works
When you do build something, reach for the simplest tool that solves it. A shared document often beats a new piece of software. A clear owner often beats a committee. A single written rule often beats a meeting. The goal is the outcome, not the apparatus. Heavy process feels thorough and serious, but weight is a cost, not a virtue.
Give every process an owner and an expiry
Process rots. A rule that made sense at twenty people can be dead weight at eighty, but nobody removes it because removing things feels risky. Assign each process an owner whose job is to keep it useful or kill it, and revisit your rules regularly. A company that only ever adds process and never removes it is on a one-way road to bureaucracy. Pruning is as much an operator's job as planting.
⚖️ Process that speeds people up
- Ask whether a process helps people decide faster or just performs control.
- Add process in response to real, recurring pain, not imagined risk.
- Use the three-times rule before systematising anything.
- Default to the lightest tool that solves the problem.
- Give every process an owner and revisit it, so dead rules get pruned.
The companies that scale well are not the ones with the most process or the least. They are the ones whose operators treat process as a tool with a cost, add it deliberately, and remove it without sentiment. Keep the friction low and the focus high, and the company can grow without losing the speed that made it special.