Culture at 50 people vs culture at 500: what actually breaks

Nobody loses their culture in one big moment. There's no meeting where it happens, no quarter you can point to. It slips quietly, somewhere between 50 and 500 people, and by the time it shows up in engagement scores, regretted attrition or a glassy-eyed all-hands, it's expensive to put right.

We've spent 25 years working with companies through exactly this stage of growth, from Rightmove when they were fewer than 15 people through their growth into a £4.5bn business, to scaleups across food delivery, fintech and beyond. The pattern is remarkably consistent. Three things break, at three predictable points, and almost nobody plans for any of them.

At 50 people, culture needs no management

‍Here's the uncomfortable truth about the culture you're so proud of at 50 people: you didn't build it deliberately, and you're not maintaining it deliberately either.

At 50, everyone can see the founders. Information travels by proximity. Standards are enforced by the simple fact that everyone knows everything. When someone behaves brilliantly, the whole company sees it. When someone behaves badly, they feel 49 pairs of eyes. Culture transmits by osmosis, and osmosis is free.

That's why so many founders believe culture is easy. It has been easy, for them, so far. The systems that will carry culture at 500 people don't exist yet because nothing has required them to exist.

Then growth removes the conditions one by one.

Break one: the founder can no longer be in every room

The first thing to go is proximity to the founder.

In the early days, every hire is interviewed by a founder, onboarded within earshot of a founder, and corrected or celebrated by a founder. People don't learn the culture from a document. They learn it by watching how the founder handles a difficult client, a missed deadline, a hard call on a colleague.

‍Somewhere around 50 people, that stops. Not because the founder changes, but because arithmetic does. There are now rooms the founder isn't in, decisions they never see, and new joiners who will go months without a meaningful interaction with them. What the founder rewards and what the founder tolerates, the two signals that actually teach a culture, are no longer being broadcast to everyone.

The dangerous part is the lag. The founder still feels present. The old guard still quote them. But the newest 30 people are learning the culture from whoever manages them, and nobody has checked what that person is teaching.

Break two: values stop transmitting by osmosis

‍The second break is subtler. The values that once travelled through the air now have to travel through structure, and most companies haven't built any.

‍Robin Dunbar's research on social group sizes suggests humans can maintain genuinely stable relationships with roughly 150 people. Below that number, an organisation can run on informal networks: everyone knows who to ask, who to trust, and how things are done here. Above it, the informal network fails. People know of each other rather than knowing each other, and "how we do things" starts arriving second-hand, third-hand, or not at all.

This is the point where most leadership teams reach for a values exercise. An away day, a workshop, five words, a poster in reception. And this is where the laminated values problem begins: values that exist on the wall but not in the room. The words are fine. The problem is that words on a wall do nothing until they're translated into a small set of concrete, observable behaviours that a manager can actually coach and an interviewer can actually test.

‍You can coach "we challenge each other directly and kindly." You cannot coach "integrity." One is a behaviour someone can practise on Tuesday. The other is a poster.

Break three: the first "us and them"

‍The third break is the one leaders take most personally: the day they discover an old guard and a new guard.

It usually surfaces as nostalgia. "It's not like it used to be." "We've gone corporate." "The new people don't get it." The early employees experience growth as dilution. The new employees experience the old guard as a club they can't join, full of in-jokes, shared history and unwritten rules nobody will write down for them.

‍Left alone, this hardens into two cultures under one logo. The old guard hold the founding story hostage. The new hires build their own norms, because they have to build something. And managers get caught between the two, enforcing rules for a culture they were never inducted into themselves.

‍The fix isn't to freeze the old culture and demand the new people absorb it. The companies that get this right don't preserve their culture. They translate it: they take what actually made the place work at 50 and rebuild it in a form that works at 500.

What that translation looks like in practice

When Rightmove set out to protect their culture through growth, the work had four stages, and they're the same four stages we'd point any scaling leadership team towards.

‍Discover. Find out what the culture actually is, from the CEO to the front line. The real culture, how it feels to work here on a wet Tuesday, is usually different from the espoused one on the careers page, and you can't scale something you haven't honestly described.

‍Design. Translate the values into a small set of concrete behaviours, co-created with the people who'll live them. At Rightmove that became six behaviours, a shared language for how the business worked that anyone could recognise, recruit for and coach.

Deliver. Induct everyone. At Rightmove that meant a five-day cultural induction with no one exempt: new joiners, long-timers, the executive team, the chairman. The moment anyone is too senior for the culture, you've announced that the culture is optional.

Embed. Build the capability internally so the culture no longer depends on any one person, founder included. That's the whole point. At 500 people, your culture is carried by your managers, your hiring and your rituals, or it isn't carried at all.

The results are the kind you can measure: a 9.8 out of 10 programme rating, 91% of participants reporting improved confidence in performance conversations, and 100% clarity on behavioural expectations.

The manager multiplier

If you take one idea from this piece, take this one: at 500 people, your culture is exactly as good as your most average manager.

Growth forces promotion. Your best scientist, engineer or dealmaker gets a team, usually overnight, and usually with no preparation for the fact that leading people is a different job from being brilliant at the previous one. Every one of those managers is now a broadcast tower for your culture. What they reward, what they tolerate, and how they behave under pressure is the culture, for every person who reports to them.

Which means the highest-leverage culture investment in a scaling company is rarely another values workshop. It's developing the people you've promoted into management, so that the culture you built at 50 has somewhere to live at 500.

The Sunday night test

Strip away the frameworks and culture comes down to something simple: how your people feel on a Sunday night. Whether they're dreading Monday or quietly looking forward to it. Everything above, the behaviours, the induction, the managers, exists to protect that feeling as the headcount climbs.

Pressure doesn't change people. It reveals them. And scaling is sustained pressure, applied to your entire organisation at once. The companies that come through it with their culture intact are the ones that stopped relying on osmosis before osmosis stopped working.

The Russell Partnership has spent 25 years helping leadership teams scale their culture deliberately, from Rightmove at fewer than 15 people to organisations of thousands. If any of this hit a nerve, our culture transformation work is where to start, or get in touch and compare notes with us.