The moment you notice is never the moment it happened
We have written elsewhere about what actually breaks as a company scales, and about the three points at which it tends to happen. This piece is about something that sits underneath all of that, and which is the reason the whole problem is so much harder to act on than it looks: by the time anyone can see it, it has been happening for years.
If you ask a leadership team when their culture changed, you will almost always get a number, and it will almost always be delivered with a slight shrug. We were fine until about eighty. It went wrong somewhere past two hundred. What you will not get is a week, or a meeting, or an incident, because as far as anyone can recall there wasn't one. The company simply got bigger, and at some point it stopped feeling the way it used to.
That vagueness is usually treated as a failure of memory, or as evidence that culture is too woolly a thing to manage properly. It is neither. It is the signature of a problem whose cause and whose symptoms are separated by a very long gap.
The arithmetic nobody runs
Think about who in your organisation actually learned the culture from the source, rather than from a description of it.
In a company of fifty, that is everybody. They have all watched the founder handle a week that went badly. They have all heard what got said about a client who let the team down, and noticed whether anyone was thrown under a bus to protect a relationship. Nobody sat them down and explained the standards, because the standards were visible from wherever they happened to be sitting, all day, every day.
Then the company doubles, and half of it has never had that experience. Those hundred people are not learning the culture from the founders; they are learning it from the fifty who did, which is a reasonable approximation but a second-hand one. Double again and three quarters of your people are working from a copy. At four hundred it is seven eighths, and the handful who were there at the beginning are now a rounding error in most rooms, however senior they have become.
Nothing dramatic happens at any of those points, which is precisely the difficulty. Each doubling simply halves the proportion of the company who learned it properly. The people who did are still there and still behaving well; there are just proportionally fewer of them every time you grow, and nobody has ever put that number on a slide.
Why it feels sudden when it finally shows
A culture does not run on its average. It runs on what looks normal, and what looks normal is set by whoever a person sees most of.
For as long as the well-inducted group makes up a decent share of the company, they set the tone more or less everywhere, because they are simply the majority of the people any new joiner encounters. The culture reproduces itself without anyone having to manage it, which is why this stage feels so easy and why so many leadership teams conclude that culture is something they are naturally good at.
At some point that group stops being the majority in most rooms. It happens gradually, and it happens unevenly, and it happens last in the leadership team, which is the single most important thing to understand about it. The executive floor is still full of people who learned the culture first hand, so the view from there remains genuinely reassuring. Down where the work actually gets done, new joiners are increasingly learning from other new joiners, and the culture that results is a copy of a copy, each generation a little further from the original and none of them aware that they are drifting.
So what gets experienced as the culture dying at two hundred is usually the culture having stopped transmitting properly at eighty, with the last well-inducted cohort finally becoming a minority. The thing you notice is not the cause. It is a consequence that has taken a couple of years to surface.
What the delay actually costs
The first cost is that you are always repairing old damage. When the problem becomes visible, you are not addressing something that happened this quarter; you are addressing two doublings of accumulated drift, in an organisation that is now considerably larger and slower to turn than it was when the drift began. Anything you do will take a year to show, which is roughly a year after everyone started asking why it hasn't.
The second cost is that your instincts are calibrated to a company that no longer exists. When a leadership team says the culture feels fine, they are not managing upwards or kidding themselves. They are reporting honestly on the one part of the business where it is still true, and they have no particular reason to suspect that their sample is unrepresentative, because it never used to be.
The third cost is that the measures most companies rely on will not warn them in time. Engagement scores move late, and they move for all sorts of reasons that have nothing to do with culture. Regretted attrition moves later still, and by the time it moves the person has already gone. Both of them describe damage that has already happened, which was never the part you needed help noticing.
Something that moves earlier
If headcount is the wrong trigger and the usual metrics are too slow, you need a signal that shifts while there is still time to do something about it. There is one, and it costs an afternoon.
Ask your twenty most recent joiners what the standards are here. Then ask your twenty longest-serving people the same question, and compare the two lists.
The wording matters. Do not ask what the values are, because anyone can read those off a wall or a careers page. Ask what actually gets rewarded, what gets quietly tolerated, and what would earn someone a difficult conversation. Ask what they would do if they found a mistake that nobody else had spotted yet, and whether they would expect that to go well for them.
If the two lists broadly agree, transmission is still working and you have time to build something deliberate before you need it, whether that is translating your values into behaviours you can actually interview for or an induction that inducts rather than processes. A systematic review of the onboarding research found that structured, supported on-the-job training had the strongest evidence behind it of any approach, and that leaving new starters to informal mentoring alone was consistently the weaker option, which is worth knowing if informal is what you are currently relying on. If they do not, the mechanism that used to carry your culture has already retired, and the reason nobody told you is that the mechanism which would have told you is the one that stopped working. Run the comparison at every doubling and you will find out years before an engagement survey would have got round to mentioning it.
Why this gets harder rather than easier
There is a reason all of this becomes more urgent as you grow rather than less.
Once informal transmission has gone, culture is carried by managers, and not by the leadership team but by the individual each employee actually reports to. Which means that the quality of your culture is set by your average manager rather than your best one, and most organisations have never really looked at that distinction. Gallup, working across 27 million employees and more than two and a half million work units, put the manager's share of the variance in engagement between teams at a minimum of 70%, which is another way of saying that the person someone reports to matters more than almost everything else you could change.
The average manager, meanwhile, has been given almost nothing to work with. The Chartered Management Institute, working with YouGov across more than four and a half thousand UK workers and managers, found that 82% of people who enter management have had no formal management or leadership training at all. The same research found that one in three people have left a job because of a negative workplace culture.
Put those two things side by side and the delay stops being an abstract problem about measurement. The point at which your culture starts depending entirely on managers is the same point at which it starts depending on a group of people who were handed the responsibility and very little else, and if you only discover that two doublings late, what began as a culture question has quietly turned into a management development one.
Where to start
You do not need a programme to begin, and you certainly do not need to wait for a number. Headcount tells you how big you are, which is not the same as telling you what is happening to the thing you care about.
Run the two-list comparison this month, whatever size you currently are, so that you have a baseline to argue from. Then run it again every time the company gets meaningfully bigger, and treat any widening gap as the early warning it is.
And if the answers already differ, you have not discovered a new problem. You have discovered an old one, which is rather the point.
If you want to go further than the two-list comparison, our guide Culture at 50 vs. Culture at 500 sets out where culture predictably breaks as a company grows, and includes a ten-question self-assessment for working out where yours is actually living at the moment.
If the gap has already opened, culture transformation is the work of closing it deliberately, and management effectiveness is what stops it reopening at the next doubling.
We have spent 25 years helping leadership teams keep their culture intact while their headcount climbs, from Rightmove at fewer than 15 people through to global businesses. If any of this landed, we are always happy to compare notes.
