Most founders believe a healthy culture is something you define, build and protect. So when a healthy culture starts to break, the instinct is to look for the leak. The toxic hire. The values no longer being lived. The founder who got too busy to care. The fix often becomes a values offsite, a new set of behaviors on the wall or a renewed commitment to “people first.”

It rarely works. Why? Because diagnoses are almost always wrong.

Culture doesn’t break because founders stop caring. It breaks because the company has outgrown the way it coordinates. The trust, decisions and accountability that once moved naturally through the organization no longer match the company the founder is running or the game the company is playing. That’s the lag. And when a founder tries to close that lag with care alone, the culture keeps breaking.

Part of what makes the lag so tempting is that a one-sided culture is easier to build. High care and high performance each make sense on their own. You only have to lean on one instinct, and most founders have one they trust more. Care-led founders build warmth, loyalty and belonging, but often let performance stay soft. Performance-led founders build intensity, output and ambition, but often let care stay thin. Either can carry a company a surprisingly long way. I know. The harder and much rarer thing is building a culture that holds both at once: high care and high performance, without one paying the bill for the other.

Balancing culture and performance is not easy

What almost no founder wants to admit is that building the both/and culture is hard even when everyone agrees they want it. I’ve watched leaders endorse the need for objective performance standards, agree on what an excellent hire looks like, agree that excellence belongs in every leadership seat and still avoid the mechanisms that would make performance measurable, but not through open disagreement. More often, it looks like full agreement in the room and very little appetite for the hard work that follows, especially once the measurements need to be installed or the standard needs to be applied.

That resistance is not duplicity, and it does not mean you have a weak leader. It is one of the most important things a founder can learn to read. People can sincerely believe in a standard and still resist the mechanism that makes it real. The standard fits what they consciously value. The resistance often comes from what they fear, even if they would never say it that way: that being measured precisely means being found wanting or that explicit standards will be used against people rather than for them. You cannot argue an assumption like that out of someone. It only moves under sustained, patient pressure, sometimes over years.

There is also a structural source. Objective performance standards are a coordination mechanism. They replace “we all know who’s good” with something more explicit. Early-stage cultures run on relationships, and relationships protect a certain amount of ambiguity. Making performance visible removes that comfort. That is why the resistance often comes not from weak performers, but from the leadership layer itself. They have the most to lose from clarity and the most power to stall it. What looks like a fight about a measurement tool is usually the harder shift from relationships to agreements, showing up in the place where it is hardest to win.

Culture often breaks as your company grows

This is why culture breaks at scale. As I wrote about in Work 9.0: The Evolution of Work, every group, especially companies, coordinates through one of three systems. Early on, alignment is relationship-based. It holds because people are personally close to one another, especially the founder. The founder can sense when something is off and correct it in a hallway conversation. It works beautifully, and that is the trap. As groups grow, relationships stop scaling and alignment has to move to agreements: explicit roles, standards and expectations that let people who barely know each other still work as one. Eventually, especially in faster or more ambiguous markets, even agreements get too slow. The company needs guiding principles, meaning compressed judgment internalized widely enough to travel where no rule reaches.

The break happens in the gap. A company outgrows its coordination system long before anyone notices because the old system still feels like culture. Relationships feel warm, and warmth can feel like health. Investing real time in designing the culture a 100-person company will need, while you are still 30 people and everyone eats lunch together, can feel almost like bad karma. Like you are borrowing trouble instead of enjoying what you built. So the work gets deferred. The culture doesn’t collapse all at once. It leans, then leans further, until one day it is broken and no one can point to the moment it happened.

This is harder for first- and second-time founders, and not because they care less. Many have never worked inside a company that successfully made the journey from Day 1 to 300-plus people with a genuinely well-designed culture. They have no felt reference for the journey, or for what an agreements-based or principles-based culture is supposed to feel like. So they default to the only system they have personally experienced, relationships, and run it well past its limit.

Two forces make the lag worse. The first is the market itself. A company does not only outgrow its coordination system by adding people. The environment can move underneath it. If the market gets faster or more ambiguous, the demands on coordination jump even if headcount holds flat. Neither the founder nor the culture got worse. The game changed, and the old system cannot carry the new weight.

The second is layering. Past roughly 30 people, you no longer have one culture. You have layers: executives, managers and frontline teams, each interpreting the agreements slightly differently. The founder’s relationship-based culture may still glow at the center, but it no longer reaches the edges. Out at those edges, departments start operating by different rules, not because of bad intent, but because relationships are the only thing binding them. And relationships do not scale to that distance.

How to keep your company culture from breaking

The both/and culture is even harder to retrofit under pressure. When a high-care, low-performance culture meets a sudden rise in pace or ambiguity, the common outcome is not transformation. It is replacement. The board installs a new CEO, or the founder brings in new executives, and performance arrives through new people. The warmth is often a casualty.

A genuine both/and transformation, where performance gets built into a care-strong culture without hollowing out the care, rarely happens without some leadership change and disciplined change management. When it does happen without a leadership change, it is usually because the founder built the capacity for it before the pressure arrived. The window is before the spike, not during it.

So how do you prevent the break? Five concrete moves.

1. Name the game you are actually in. Culture has to survive a specific competitive environment, defined by two things: how fast your market moves and how much can be known in advance. A founder designing the culture they personally find comfortable, rather than the one their market demands, is already taking on risk. Name your environment honestly before you design anything. In my experience, it helps to think of this as a 3×3 matrix, from low to high pace and low to high ambiguity.
2. Match your coordination system to that game. Know which system you are running today, whether relationships, agreements or principles, and which one your next stage will require. Most scaling companies break because they are trying to hold a market with a noncompetitive coordination system. Make that shift deliberately before reality forces it.
3. Design the 100-person culture while you are still 25. Treat it as architecture, not a daydream. The instinct that says “it is too early to think about this” is not modesty. It is the actual risk. The agreements and standards a larger, high-growth company needs take years to take hold because you are changing what people unconsciously assume, not just what they say. Start before you need them.
4. Install the performance half before you are forced to. Objective standards, including a clear, measurable picture of the value each seat generates against what it costs, meet the least resistance when stakes are low and trust is high. Founders who wait until growth exposes the gap are trying to install accountability during the exact period when people have the least goodwill to absorb it. Expect resistance anyway, and read it as information rather than betrayal.
5. Watch the layers, not just the center. Your experience of the culture, from the founder’s seat, is the least representative one in the building. Spend time at the edges. Hold leaders, who carry more of the culture than you do at scale, to the same standard everywhere. The first sign of a breaking culture is rarely a dramatic event. It is two or more teams operating by different rules.

Culture, in the end, is not a feeling you protect. It is a system you build and refine every time the company or the market outgrows the last version. The founders whose cultures endure are not the ones who care the most. They are the ones who saw the lag coming, closed it on purpose and refused to let their version of the easy half of culture stand in for the whole thing.