You get bigger, you add structure, things get more complicated — that seemed like the natural order of things. The longer I've worked inside growing organizations, the less convinced I am that's the whole story.
Some structure is necessary. A company of two hundred can't operate like a company of ten, and pretending otherwise is its own kind of failure. But somewhere along the way, structure stops helping the organization move and starts slowing it down. The shift is quiet, and by the time you notice it, the complexity feels permanent — part of what the company is rather than something it chose.
Complexity rarely arrives as an obvious mistake. It shows up as a reasonable response to something real. A team adds a review step because quality slipped once and nobody wants it to happen again. A leader adds a planning layer because coordination has gotten harder. A company adds a process because it wants the same outcome every time instead of depending on whoever happens to be in the room. I've made versions of all these calls myself, and most were right at the time. The cost shows up later, usually somewhere other than where the decision was made.
The first signs are mundane. Meetings multiply. Decisions that used to take an afternoon take a couple of weeks. Work passes through more hands before it ships. Planning gets more elaborate without execution getting any better. None of it looks like a problem on any given day, which is exactly why it's so easy to live with. The people are still talented, and the strategy is still sound, but the distance between deciding something and doing it has filled with translation, negotiation, waiting, and rework.
I watched a version of this play out at one company. The leadership team added a weekly status meeting because they wanted more visibility into why things kept slipping — reasonable enough. Within a couple of months a pre-review had appeared so the status meeting itself would go smoothly, and teams were spending the second half of the week preparing updates instead of fixing what the updates were about. The meeting was excellent at surfacing friction and useless at removing it. Nobody set out to build a machine for reporting problems instead of solving them. It assembled itself, one reasonable addition at a time.
That's the mechanism, more or less. Someone adds a process to solve a specific, local problem, and it works. The original problem fades, but the process stays, because nothing forces it to leave. People start optimizing for the process itself — getting through the review, satisfying the template, preparing for the meeting. Decisions slow as more of them require coordination, and eventually the organization spends more energy moving work between people than doing it. No single step is unreasonable. The problem is that the chain rarely reverses on its own.
It's worth being honest about why this happens, because it isn't carelessness. Companies accumulate functions, systems, controls, and rituals precisely because they're solving real problems, and most of the structure that gets added is a sincere attempt to manage genuine complexity. The issue was never that structure exists. The issue is that structure almost never removes itself. A process will tell you when it's needed — something breaks, someone escalates — but it will never tell you when it has stopped earning its keep. That signal doesn't exist unless someone goes looking, and most of the time nobody does. I've seen companies add process for good reasons and then forget to ask whether it was still worth what it cost.
There's also a quieter reason the structure stays, one that has nothing to do with whether it works. Removing complexity is politically harder than adding it. Adding a process looks responsible — you saw a risk and did something about it. Removing one takes judgment. You're deciding the thing it guarded against is now worth risking, and if something goes wrong afterward, it goes wrong on your call. Adding a control spreads responsibility across a system; removing one concentrates it on a person. So the safe move, almost always, is to leave it in place and route around it — which is how organizations end up permanently carrying the cost of avoiding a risk that may not even be there anymore.
The mistake I see most often is what leaders reach for when execution slows down. The instinct is to add visibility — another status meeting, another report, another planning artifact — on the theory that if things are moving slowly, the answer is to see more of it. Sometimes that helps. But in an organization that's already heavy, visibility just becomes more work. I've watched teams spend more time explaining why work was slow than removing what made it slow. The reporting gets very good. The work doesn't.
The more useful question is what the new process is actually compensating for. Usually it's something underneath — ownership that was never clear, or a roadmap with too many things on it that all claim to be the priority. That's harder to look at than a missed deadline, which is part of why the meeting is always more popular than the fix. But that's where the opportunity lives.
None of this is an argument for tearing structure out. The job isn't to make the company simple; past a certain size it can't be, and forcing it there breaks things that were quietly holding it together. The job is to keep structure in service of outcomes, which mostly means doing the unglamorous work of asking, on some regular basis, which parts of the operating system still create value and which parts only create motion. Both feel like activity. Only one of them is.
That's harder than it sounds, because complexity often feels like control. It gives you more places to look — more dashboards, more checkpoints, more assurance that nothing is happening without someone watching. But seeing more of the work is not the same as improving how the work moves. A lot of what passes for control is really just visibility into complexity you've decided to live with.
I don't think the goal is a simple organization. The goal is one that stays light enough to move. Structure is meant to help a company move; when it mostly helps the company explain why moving is hard, it has quietly become something else.