Why does information sharing break down as companies grow?

There's a specific moment every growing company hits, and it has nothing to do with revenue or funding. It's the moment someone in leadership realizes they no longer know most of the people on the payroll. Not their names, not what they're working on, not whether they got the memo about the reorg. A year earlier, that same person could have named everyone in the building. Something structural changed in between, and it wasn't a choice anyone made on purpose.

There's actually a number for this, and it's older than most of the tools we now blame for the problem. Anthropologist Robin Dunbar's research on primate brains put a rough ceiling on how many stable social relationships a person can realistically maintain — somewhere around 150. Below that number, companies run on relationships. People know who to ask, who owns what, who to loop in before a decision goes out the door. Above it, none of that holds. The org chart keeps growing, but the mental map everyone's been using quietly stops working, and nobody sends an announcement to say so.

The math nobody accounts for

Fred Brooks wrote about this in software engineering terms back in 1975, in a book about why adding people to a late project makes it later. The insight holds well beyond software: as a team grows, the number of communication pathways between people grows far faster than the headcount does. Ten people have 45 possible one-to-one connections. Fifty people have over 1,200. It's not a linear problem, which is exactly why it sneaks up on companies that are used to solving problems linearly — hire more people, ship more work, repeat.

What actually happens at scale isn't that people stop talking. It's that they can't talk to everyone who needs to hear something, so they default to talking to whoever's nearby. Information moves through the org unevenly — dense in some pockets, completely absent in others — and nobody notices until two teams show up to the same meeting with contradictory assumptions about what was decided last month.

Proximity was never a strategy

Here's the part that's uncomfortable to admit: most early-stage companies don't have a communication strategy. They have proximity. Everyone sits near everyone else, decisions get made out loud, and anyone can catch the tail end of a conversation and stay roughly current. It works, and it works well enough that nobody thinks to build anything more durable on top of it.

Then the company doubles. Then it doubles again. A second office opens, or half the team goes remote, and proximity — the thing that was quietly doing all the work of "communication" — disappears almost overnight. What's left in its place is usually whatever tool happened to be lying around: email threads that fork into six replies, Slack channels that multiply faster than anyone can mute them, a wiki that three people update and forty people have never opened.

This is where a lot of internal comms research actually catches up with the anecdote. Staffbase's 2024 Employee Communication Impact study, run with the USC Annenberg School, found that just over half of employees don't feel especially familiar with their own company's goals and vision. Two-thirds say they're not well informed about how the business is actually performing financially. These aren't small companies with no resources for communication — some of the organizations in that study were Fortune 500 names with entire comms departments. Which tells you something important: this isn't a resourcing problem you solve by throwing more headcount at it. It's a structural one.

Leaders think it's working. Employees don't agree.

Maybe the clearest evidence of the breakdown isn't in what employees say — it's in the gap between what employees say and what leadership believes. Axios HQ's internal communications research asked both groups the same set of questions, and the answers don't line up. Seventy-three percent of leaders believe their teams can quickly find important goals, strategies, or directives when they need them. Only 49% of employees agree. That's not a rounding error. That's two groups of people standing in the same building with two completely different pictures of how well information actually flows through it.

The instinct, when you notice a gap like that, is to send more updates. Write a longer newsletter. Add another all-hands. But volume was never the constraint — reach was, and increasing volume in a system that already can't reach everyone just adds more noise to filter through. It's the equivalent of a captain who can't get the engine room to answer the radio, so he shouts louder into a receiver nobody's holding.

Where the actual gaps open up

A few specific things tend to break at growth stage, and they're worth naming individually because they don't all have the same fix.

Ownership gets fuzzy first. When ten people work on something, everyone roughly knows who's responsible for what, even without documentation. At a hundred people, that same knowledge lives in maybe three heads, and if one of them leaves, the answer to "who owns this?" disappears with them.

Context stops traveling with decisions. A choice gets made in a meeting with the right people in the room, and everyone in that room understands the reasoning behind it. Three weeks later, someone outside the room encounters the outcome with none of the reasoning attached, and it looks arbitrary — even when it wasn't.

And documentation quietly falls out of date without anyone deciding to let it.

Nobody wakes up and decides to stop maintaining the wiki. It just happens gradually, page by page, as the people who wrote it move on to other things and nobody's tasked with keeping it current. Eventually, the gap between what the documentation says and what's actually true becomes wide enough that people stop trusting it altogether, which is arguably worse than having no documentation at all — at least with nothing, people know to ask.

What doesn't fix it

It's tempting to treat this as a tooling problem, and to be fair, tooling is part of it. But swapping Slack for Teams, or bolting on another wiki nobody will maintain any better than the last one, doesn't address why the breakdown happens in the first place. The company outgrew the informal systems that used to carry information, and nothing structural replaced them. A new interface on top of the same absence of ownership just moves the same problem to a different screen.

What actually helps is treating information flow the way you'd treat any other function that stops scaling informally — with intention, ownership, and a defined system rather than an accumulation of habits. That means deciding, on purpose, where company-wide information lives permanently rather than passing through. It means assigning actual ownership to that system rather than leaving it to whoever has time this quarter. And it means accepting that at a certain size, closing the loop on communication has to be someone's job, not an occasional additional task tacked onto someone else's.

Where Happeo fits into this

None of this gets solved by picking a tool first and figuring out the strategy after — that's the same mistake as swapping Slack for Teams. But at the point where a company has accepted that proximity isn't carrying the load anymore, the shape of the fix is fairly consistent: one place where company-wide information actually lives, is actually owned, and is actually findable months after it was published.

That's the specific gap Happeo is built to sit in. It gives growing companies a structured home for the things that keep falling through the cracks mentioned in this article — a searchable knowledge base tied to the tools people already work in, clear ownership assigned to specific pages and spaces rather than left ambiguous, and a way to see when content's gone stale instead of finding out the hard way when someone acts on outdated information. Because it connects directly to Google Workspace, it doesn't ask employees to go somewhere new to find things, it meets them inside the tools they're already using every day, which matters more than it sounds like it should once a company's past the size where everyone just happens to know where to look.

The point isn't that a platform replaces the harder organizational work of assigning ownership and deciding what counts as a company-wide update. It's that once a company decides to do that work, it needs somewhere durable to put it — and that's a much easier problem to solve with the right foundation already in place.

The size where it matters most

Ironically, the companies best equipped to handle this are often the ones who need it least yet. Very large enterprises usually have entire teams dedicated to internal communications and knowledge management, built up over years of feeling the pain directly. Very small ones don't need the infrastructure yet because proximity is still doing the job. It's the companies in between — past the point where everyone fits in one room, not yet at the point where they've built a formal function to compensate — that feel this most acutely, and for the longest stretch of time.

That gap doesn't close on its own. Left alone, it tends to widen, because every new hire is one more person operating on incomplete context, and every quarter that passes is more decisions made without a clear record of why. The companies that get ahead of it don't necessarily do anything exotic. They just stop assuming that what used to work by accident will keep working on purpose — and they build something deliberate to replace it before the gap gets too wide to close.

Frequently asked questions

Why does information sharing break down even without anyone making a bad decision?
Because it's mostly a math problem, not a decisions problem. As headcount grows, the number of connections between people grows far faster than the headcount itself, which is why a system that worked fine at fifty people can quietly stop working at two hundred without anyone doing anything wrong along the way.

Is this mainly a communication tools problem?
Not on its own. Swapping one tool for another doesn't address why the breakdown happened in the first place, since the underlying issue is usually that proximity stopped doing the work informally and nothing structured replaced it. A new tool on top of that same gap tends to reproduce the same problem somewhere else.

Why do leaders often think communication is working better than employees say it is?
Leaders are usually closer to the decisions being made, so they overestimate how much context has actually traveled outward with those decisions. The research cited above shows a consistent, large gap between what leaders believe employees understand and what employees actually report, which suggests the disconnect is structural rather than a matter of leaders being out of touch on purpose.

What's usually the first sign that a company has outgrown its informal information systems?
Ownership getting fuzzy is often the earliest sign. When ten people work on something, everyone roughly knows who's responsible for what without it being written down anywhere. Once that same knowledge only lives in a couple of people's heads, and disappears if they leave, the informal system has already stopped scaling even if nobody's noticed yet.

Does this problem ever fix itself as a company keeps growing?
Not typically. Left alone, it tends to widen, since every new hire adds someone else operating on incomplete context, and every quarter that passes means more decisions made without a clear, findable record of why. It usually takes a deliberate decision about where information lives and who owns it to actually close the gap.