Somewhere between 50 and 500 people, a company stops being able to absorb new joiners by osmosis.
At 50, a new starter learns the culture by sitting near it. They watch how the founder handles a bad week. They hear how a difficult client gets spoken about when the client is not in the room. Nobody has to explain the standards, because the standards are visible from most desks.
At 300, that same new starter learns the culture from whoever interviewed them, whoever onboarded them, and above all from whoever they report to. If those three carry it well, the culture holds. If they do not, it drifts, one hire at a time, and nobody notices until the drift has a name and a leaving date attached.
Culture at scale is, in the end, a manager problem. It is held or lost in a hundred small moments between a manager and their team, most of which nobody else sees. But hiring is where you decide who is in those moments in the first place, and it is the one part of the system you can fix with process rather than luck. Get the front door right and every manager after you is playing an easier game.
This piece is about the front door.
Why growth makes hiring worse
Recruiting well takes three to six months from the moment you decide you need someone. Whenever you hire, you are predicting what the business will need in another six to twelve months. There is always guesswork involved.
Meanwhile the team is carrying the gap. Deadlines do not move because you are a person short. Everyone is tired. And that is exactly the moment when a candidate who is competent, available and only slightly wrong starts to look like the sensible option.
Urgency creates pressure, and pressure is where standards quietly go. Every leader we have worked with who has hired badly at scale describes some version of the same moment: they knew, and they did it anyway, because the alternative was another two months short-handed.
Which is why the decision needs to be made before you are under pressure, in the form of a process you have agreed to follow when you are tired.
The problem with hiring for culture fit
Most companies say they hire for culture fit. Very few can say what they mean by it.
In practice, "culture fit" gets decided in the last five minutes of an interview, by feel, using some version of "would I want to be stuck in an airport with this person?" That produces two outcomes, both bad.
The first is that it selects for similarity. People rate candidates who remind them of themselves as a stronger fit, which over a hundred hires produces a company that is demographically and cognitively narrower than the one you started with. You end up with a culture that agrees with itself and cannot see its own blind spots.
The second is that it is unfalsifiable. If a hiring manager says a candidate "wasn't quite right culturally" and cannot point at anything specific, there is no way to challenge it, no way to coach the manager, and no way to defend it if the decision is ever questioned.
The fix is small and it changes everything. You cannot interview for a value. You can interview for a behaviour.
The translation
A value is a word on a wall. A behaviour is something a person can observably do, or fail to do, in a specific situation. Once you have the behaviour, the question writes itself, and so does the scoring.
That gives you a chain with four links:
Value → Behaviour → Question about the past → What strong and weak actually sound like
The chain matters more than any particular question, because your values are not ours. A company built on speed and a company built on precision should end up with different questions, and both should be able to defend them. What follows is the method, with worked examples deliberately drawn from cultures that pull in different directions.
Worked example one: a company that values speed
VALUE: We move fast
BEHAVIOUR: We ship the imperfect version and improve it in the open
QUESTION: Tell me about something you released before you were comfortable with it. What was still wrong, and what happened next?
STRONG ANSWER: A specific thing shipped, a clear-eyed account of what was unfinished, and evidence they improved it after rather than defending it.
WEAK ANSWER: An example where nothing was actually unfinished, or a story that quietly becomes about how they persuaded everyone else to wait.
Worked example two: a company that values precision
VALUE: We get it right
BEHAVIOUR: We stop the line when something looks wrong, even when it is expensive
QUESTION: Tell me about the last time you held something up because you were not sure. What did it cost, and how did you handle the pressure to let it go?
STRONG ANSWER: A real cost named, and an account of who pushed back and how they held.
WEAK ANSWER: They cannot think of one, or the only example is a decision that turned out to be obviously correct.
Those two companies have opposite instincts. Both questions are fair, evidenced and coachable. Neither would work in the other building.
Worked example three: a value nearly everyone claims
VALUE: We are one team
BEHAVIOUR: We share bad news within 24 hours, upwards first
QUESTION: Tell me about a time you had to tell someone senior something they did not want to hear. What did you say, and when did you say it relative to when you knew?
STRONG ANSWER: A specific example, a short gap between knowing and telling, and honesty about the discomfort.
WEAK ANSWER: A hypothetical, or an example where the news was already public by the time they raised it.
Note what the behaviour column does here. "We are one team" is unrecruitable. "Within 24 hours, upwards first" can be asked about, evidenced, scored, and then managed once the person is in the building — which is the point. The interview and the management conversation eighteen months later are now about the same thing.
Building your own set
Start from the values you actually have, not the ones you published. The test is whether the people who have been there longest recognise them. A set handed down from a leadership away day arrives already dead.
Translate each one into a behaviour someone could be filmed doing. If you cannot picture the footage, it is still a value. Keep going.
Keep the set small. Five or six behaviours at most. Then pick four to interview for. Ten questions is an interrogation, not an interview.
Write the strong and weak columns before you meet anyone. Do it with the panel, in the same room, in advance. It takes forty minutes once and removes most of the arguing later.
Ask about the past, not the hypothetical. "What would you do if a colleague missed a deadline?" tells you what a candidate believes a good answer sounds like. "Tell me about the last time a colleague let you down" tells you what they did. Everyone can describe the ideal. Far fewer can describe having done it.
Ask everyone the same questions. Structured interviews, where every candidate faces the same questions in the same order and is scored against agreed criteria, predict job performance considerably better than unstructured conversations. A major 2022 re-analysis of the selection research put structured interviews at the top of the table, ahead of job knowledge tests, work samples and cognitive ability. It remains one of the most ignored findings in hiring.
One note on scoring. Rate the answer against the behaviour, not against how much you enjoyed the conversation. The candidate who interviews smoothly and gives you nothing concrete should score lower than the one who is nervous and gives you three specific examples.
What it looks like when it holds
When Rightmove worked with us, they were fewer than 15 people. As they scaled towards IPO, their values were translated into six concrete behaviours, which they called The Rightmover Way. Those behaviours became the spine of both the interview process and the induction.
They also inducted everyone into them. A five-day cultural induction across the entire organisation, from new joiners to the chairman, nobody exempt. The moment you exempt a senior hire from the culture programme, you have told the whole business that the culture is for junior people.
The company grew into a £4.5bn business with the culture intact. The behaviours are the reason it travelled.
Where the front door stops being enough
Every scaling company eventually gets the same problem, and no interview process prevents it. Someone delivers brilliantly and behaves badly. They hit every number and leave a trail. The leadership team has a quiet conversation about how they are difficult but they are worth it.
Everybody watches what happens next. If nothing happens, you have just published your real values, and they override whatever is on the wall.
MIT Sloan Management Review analysed why people were leaving and found that a toxic corporate culture was 10.4 times more powerful than compensation in predicting a company's attrition rate against its industry. The cost of tolerating one brilliant, corrosive performer is rarely just that one person.
This is the point where hiring hands over to management. You can select for the behaviour at the front door, but whether it survives contact with a bad quarter depends on managers who will name it in the moment. That is a different problem, and a harder one.
Where to start
If you are hiring into a company between 50 and 500 people, three things are worth doing in the next month.
Write down what your values look like as behaviour, while the people who lived them from the start are still in the building to check it against.
Build four questions off those behaviours, agree the strong and weak columns with your hiring managers, and use the same four for every role.
Look at your onboarding and ask whether a new joiner could name three behaviours you want them to copy by the end of week one.
None of this requires a programme, a platform or a budget line. It requires deciding what you want people to do, and then asking about it consistently.
If you want the wider picture this sits inside, our guide Culture at 50 vs. Culture at 500 covers where culture predictably breaks as you grow, and includes a ten-question self-assessment for working out where yours lives right now.
The harder half of this — what happens once they are through the door and it is down to their manager — is what Terry Russell is covering in a free hour on Thursday 24th September: Culture at 50 vs. Culture at 500. Save a place here.
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.
