You Used to Decide in a Hallway

Every company that's scaled tells the same ghost story. "We used to just decide and go. Someone would raise a problem in the hallway, three of us would talk it through, and by lunch it was done." Now that same decision takes three meetings, a deck, a pre-read, and a follow-up to loop in the people who couldn't make the first meeting. Everyone feels it. And almost everyone blames the wrong thing.

They blame size. "We're a bigger company now" ... as if speed were something you inevitably outgrow, like it's a law of physics. It isn't. Size didn't kill your speed. All the structure you added while getting bigger did, and you added most of it without ever deciding to.

Here's how it actually happens, and it's reasonable at every step. Something goes wrong, so you add an approval to make sure it can't happen again. Two teams collide, so you add a coordination meeting. A call gets made that leadership didn't love, so now that kind of call comes upstairs first. A new VP wants visibility, so there's a review. Every one of those is defensible on its own. Nobody ever proposed a step called "slow the company down." But that's the cumulative effect; a hundred sensible little additions, none of them ever removed, until a decision that used to take an afternoon takes a quarter.

Underneath it is a tension leaders rarely name honestly. They want the responsiveness of the early days and the control that feels responsible now — and when those two collide, they choose control almost every time. Control is the safe-looking option. Adding an approval feels prudent; removing one feels reckless. So process only ever ratchets in one direction. It accumulates, because the cost of each addition is invisible and the cost of taking one back feels personal.

The bill comes due in ways that are easy to misread. McKinsey surveyed more than 1,200 executives and found that only 20% of them believe their organizations are good at making decisions and that, by their own estimate, 61% of the time spent making decisions is used ineffectively (McKinsey). At a large company that adds up to something absurd: hundreds of thousands of days of manager time and roughly a quarter-billion dollars in wages, every year, poured into decision-making that goes nowhere.

And here's the part that should end the "but we have to be careful now" argument: slower is not safer. The same research found that organizations which make decisions quickly are twice as likely to make high-quality ones as the slow movers (McKinsey). Speed and quality aren't a trade-off. The conditions that let you decide fast (clear ownership, the right people in the room, context close to the call) are the same conditions that let you decide well. When you piled on process to protect quality, you were often trading away the very thing that produced it.

Meanwhile, a competitor half your size is deciding in the hallway, the way you used to. They'll be wrong sometimes. But they'll find out fast and adjust three times over while your one carefully reviewed decision is still working its way up the stack. And when your growth stalls against them, the instinct is to read it as a market problem or an execution problem. Often it's neither. It's a speed problem you built yourself, one reasonable approval at a time.

Staying nimble at scale isn't something you luck into. It's a design choice you make on purpose and keep making. It starts with pushing decisions to where the context actually lives ... the person closest to the customer and the work usually has better information than the person three levels above them, so give them the authority to match, with clear guardrails instead of case-by-case sign-off. It means keeping the teams that touch customers small and whole, because a small team that owns an outcome end to end can decide and move, while a team fragmented across five functions turns every decision into a negotiation. And it means going hunting, at least once a year, for the reflex process. The approvals and meetings and reviews that exist because of something that happened three years ago and were never removed. Somebody's actual job has to be taking process out, because no one ever removes it by accident the way they add it. Above all, measure the thing you claim to care about: how long does a real decision take to go from raised to resolved? "Faster" means nothing until you can answer "faster compared to what?" So pick the decisions that matter and time them.

This only gets sharper as AI makes iteration nearly free. When trying things is cheap, the bottleneck stops being how fast you can do and becomes how fast you can decide what to do. The companies that protected their decision speed will turn cheap iteration into a real edge. The ones buried in approvals will just generate more options they can't act on.

So the honest question isn't "how did we get so slow?" You know how. You added a step, then another, each for a good reason, and never took any of them back out. The real question is whether you're willing to do the harder, less natural thing. Treat speed as something worth defending, and start removing the weight you spent years quietly adding.

You didn't lose the hallway because you got big. You lost it because you stopped defending it.