Culture is what people do when you’re not in the room. Not the values on the wall. Not the speech at the all-hands. The choice your engineer makes at 4pm on a Thursday while you’re on a plane. So if you want to know how to build company culture that actually holds, the question is not “what should our values be?” The question is “what am I doing, every day, that my team is copying?” Because they are copying you. Founders don’t get a vote on whether they build culture. You’re building one right now, on purpose or by accident. The only choice you get is which.
I’ve spent eighteen years coaching leaders, more than 10,000 hours of it, and the pattern holds in a 15-person startup the same way it holds in a company of thousands. Culture is never installed. It’s grown. It grows out of what the founder repeatedly does, decides and tolerates. This guide walks through the whole build: where real values come from, the fake versions that fool smart founders, how to wire values into decisions that cost something, how trust actually accumulates, and the one test that tells you whether any of it worked.
01How to build company culture: the short answer
You build company culture in three moves. Discover the values that already drive your best behaviour, instead of inventing aspirational ones. Wire those values into decisions that cost something: hiring, firing, promotion, money. Then build enough trust that people tell you the truth. You’ll know it’s working when the standard survives your absence.
That’s the map. The rest of this piece is the territory.
02You are the culture, whether you like it or not
Culture is like gravity. Everyone in the company moves inside its pull, and no single person can opt out of it, including you. Especially you. When you’re the founder, you’re the chief of the tribe. The team watches what you do far more closely than they listen to what you say, and they calibrate their own behaviour to yours with frightening speed.
A founder I worked with (details changed, as always) had “sustainable pace” printed on the careers page, the onboarding deck and the investor updates. He also answered Slack at 11pm, every night, within minutes. Within six months his leadership team was doing the same, and within a year the whole company treated an unanswered evening message as a performance issue. Nobody wrote that rule down. He wrote it with his thumbs, one late-night reply at a time.
That’s the mechanism. Your calendar is a culture document. Your Slack habits are a culture document. Who you promote, what you laugh at, which missed deadline gets a hard conversation and which gets a shrug. All of it is being read, constantly, by everyone. So before you run a values workshop, run an audit on yourself. What would a new hire conclude your real values are if they only watched your behaviour for a fortnight and never heard you speak?
Sit with that question. The gap between your stated values and your lived ones is the exact size of your culture problem.
03Step one: dig up the values you already live
Most founders build values backwards. They book a workshop, brainstorm words that sound good, vote on five, and send them to a designer. The output is a poster. Posters don’t survive contact with a bad quarter.
Real values aren’t invented. They’re discovered. Your company already has values, running silently under every decision you make. The work is excavation, not creation. That’s the whole job of the Values Discovery Protocol: dig up what’s already driving your best moments, name it in your own language, and throw away everything that’s merely aspirational.
Two digging questions do most of the work.
First: think of the moment you were proudest of the company. Not a revenue milestone. A behaviour. The support person who stayed on a call for two hours with a customer who was never going to renew. The engineer who delayed a launch because the data practice wasn’t right. Whatever made your chest swell, there’s a value underneath it. Name that.
Second: think of the moment you were angriest. Not annoyed. Properly angry, out of proportion to the event. Disproportionate anger is almost always a violated value. A founder who goes quiet and cold when someone takes credit for a teammate’s work has a fairness value, whether or not it’s on any poster.
Do this properly and you end up with three to five values that are already true. That matters more than it sounds. When values describe who you actually are on your best day, enforcing them feels like being yourself. When they describe who you wish you were, enforcing them feels like theatre, and everyone can smell it.
04Step two: watch for the near enemy of every value
Now the trap that catches the smartest founders. I call it the near enemy.
The idea comes from Buddhist psychology. Every virtue has a far enemy, its obvious opposite, and a near enemy, a counterfeit that looks almost identical and is far more dangerous precisely because it passes for the real thing. The far enemy of love is hatred. Nobody confuses the two. The near enemy of love is attachment, and people confuse them for decades.
Culture works the same way. The far enemy of a people-first culture is a performance-first culture that grinds humans up for output. Easy to spot, easy to reject. The near enemy of a people-first culture is performative HR slop. Wellness webinars, pizza on Fridays, a CEO speech that reads like ChatGPT wrote it and nobody in the room believes. It has the vocabulary of care with none of the substance, and it does more damage than honest indifference, because it teaches your team that the words mean nothing.
Every value you excavated in step one has a near enemy waiting for it. “Transparency” has a near enemy called selective transparency, where you share the good news in detail and summarise the bad. “Candour” has a near enemy called cruelty with a business case. “Ownership” has a near enemy called blame with better branding. The full framework, with the common counterfeits mapped, is in the Near Enemies Framework.
The practical move: for each of your values, write down what its convincing fake would look like in your company. Then ask your leadership team which version you’re actually running. Ask them when you’re feeling brave, because the answer is often uncomfortable, and the discomfort is the point.
05Step three: wire values into decisions that cost something
A value that has never cost you anything is not a value. It’s decoration.
This is where culture building gets real, and it’s the whole purpose of the Values Integration Protocol: take the words and wire them into the four decisions your team actually watches. Who you hire. Who you fire. Who you promote. Where the money goes.
Hiring is the easy one, and most founders half-do it. Interview questions that probe for the value in past behaviour, not agreement with it in principle. Anyone can nod at “customer obsession”. Fewer people can tell you a true story about the time it cost them.
Firing is the hard one, and it’s where culture is genuinely made. Sooner or later, every founder faces the brilliant jerk: the top performer who hits every number and quietly violates a core value while doing it. What you do next is the loudest culture decision you will ever make. Researchers call the underlying dynamic the bad apple effect. One toxic team member drags the behaviour of the whole group down, and the damage spreads faster than the output justifies. Your team already knows who it is. They’re watching to see whether the values apply to people who make you money.
I coached a founder through exactly this. Star salesperson, forty per cent of new revenue, and a trail of bruised colleagues behind him. It took her three months to make the call, and she told me afterwards the fear was almost physical. Here’s what happened when he left: revenue dipped for one quarter, two quiet people on the team turned out to be excellent once they weren’t being talked over, and her leadership team started bringing her problems eighteen months earlier than they used to. She didn’t lose forty per cent of her pipeline. She bought back the truth.
Promotion and money follow the same logic. Promote someone who hits targets by burning people, and you’ve just published the real values, whatever the poster says. Fund the initiatives your values imply, and people notice that too. Harvard Business Review’s research on corporate culture keeps landing on the same conclusion: culture and strategy live or die together, and leaders shape culture through what they systematically reward far more than through what they announce. Their guide to corporate culture is a decent map of the styles; the enforcement, though, is always yours.
If you want help pressure-testing how your values hold up in the decisions that actually cost something, that’s a conversation I have with founders every week. Book a Discovery Call.
06Step four: build trust in small deposits, not grand gestures
Values tell people what the standard is. Trust decides whether they’ll hold it when you’re not watching. And founders consistently misunderstand how trust gets built.
Trust is not built in the offsite, the retreat, or the big vulnerable speech. It’s built the way Brené Brown describes with the marble jar: in tiny moments, one marble at a time. Her daughter’s teacher kept a jar in the classroom, adding marbles when the class did small trustworthy things, and Brown realised adult trust works exactly the same way. You remember a colleague’s kid’s name. You keep the small confidence. You show up when you said you would, on the unimportant thing. Marble in. You share someone’s private frustration as a funny story in the leadership meeting. Half the jar gone.
The Micro-Trust Protocol turns this into a practice: deliberately engineering small, consistent, kept promises into how the company runs, because a hundred small kept promises beat one grand gesture every time. A few examples of what that looks like in a real week:
- Say what you’ll do in the Monday standup, then visibly do it or visibly own that you didn’t. The owning builds as much trust as the doing.
- Replace the vague open door policy with a constrained one. “Anyone can book me, any topic, between three and five.” A boundary you keep beats an offer you resent.
- Take hard one-on-ones for a walk. Side by side is a different conversation than face to face. Men’s groups have known this forever: you build a fire and sit beside each other, and people open up in a way they never do across a desk.
- When you get something wrong publicly, correct it publicly, at the same volume as the original mistake.
None of these is impressive on its own. That’s the point. Trust that arrives through grand gestures leaves the same way. Trust built in marbles is the kind that holds when the quarter goes bad.
07Step five: make it safe to tell you the truth
The last load-bearing wall is psychological safety, and it’s the one founders most reliably overestimate in their own companies.
Google spent two years studying 180 of its own teams to find out what separated the great ones from the average ones, a project they called Project Aristotle. They expected the answer to be talent density, or seniority, or some magic mix of personalities. It wasn’t any of those. The single biggest factor was psychological safety: whether people on the team felt safe to take risks, ask the dumb question and admit the mistake without being punished for it. The New York Times write-up of the study is still the best plain-language account of it, and it sits on top of two decades of Amy Edmondson’s research showing the same thing: teams that feel safe surface their errors, and teams that don’t bury them.
For a founder, this lands as one blunt question. When was the last time someone two levels down told you that you were wrong, to your face? If you can’t remember, you don’t have a truth problem in your team. You have a safety problem in your leadership.
You fix it with your own behaviour, in this order. Go first: share a real mistake before you ask anyone else to. Reward the messenger: when someone brings you bad news early, thank them in public, because the whole room is deciding whether bad news is safe here. And separate the feedback from the person’s worth, which is the entire architecture of the Effective Feedback Protocol: feedback delivered with care, on the behaviour, at the time, with the relationship intact on the other side.
Clear is kind here. The founders who dodge hard conversations to protect people aren’t protecting anyone. They’re just exporting the cost to the rest of the team, with interest.
08The real test: leave the room
Everything above builds toward one moment. Leave the room. Properly. Take the two-week holiday with the laptop at home. Skip the product review. Let the leadership team run the offsite without you.
Then watch what happens, because this is where the fragile version and the strong version of culture finally separate. A fragile culture is compliance wearing a culture costume: people hold the standard while you’re watching because you’re watching, and the moment you leave, the standard leaves with you. A strong culture holds in your absence. And the version worth actually building is one step better than strong. It’s antifragile: the culture gets tighter under stress, because the team has internalised the standard so deeply that a crisis pulls them toward the values instead of away from them. Missed quarter, key resignation, angry customer, and the response is more honesty and more ownership, not less.
You can’t get there while you’re still the enforcement mechanism. As long as every standard routes through you, you’ve built a culture with a single point of failure, and the failure point is your attention. The way out is to make the team the enforcement mechanism: people who hold the standard on each other, peer to peer, without escalating to you. That’s what I mean when I talk about building a team of CoFounders, people who act like owners of the culture rather than renters of a job, and the practical method for developing them is in The Coaching Standard: How to Turn Your Team Into CoFounders. The wider job of building a company that runs without your constant presence, decisions and all, is its own discipline, and I’ve written the full guide to it in The Founder-to-Leader Transition.
When you get back from that holiday, don’t ask “did anything break?” Ask better questions. What decisions got made, and would I have made them? Did anyone hold a standard when it was awkward to? Did bad news travel upward while I was gone, or did it wait on the tarmac for me? The answers tell you exactly which of the five steps above needs the next round of work.
09Where to start this week
Don’t try to run all five steps at once. Culture work compounds, which means the order matters less than the consistency. Start with the audit from the top of this piece: two weeks of honestly watching your own behaviour and asking what values it’s actually teaching. Then pick the step where your gap is widest. For most founders it’s step three. The values exist, the poster is fine, and the brilliant jerk is still employed.
And hold onto the standard for what you’re building. Not a culture that performs when you’re present. A culture that holds, and hardens, when you’re not. You’ll know you’re close the day you come back from leave and discover the hardest call of the month was made without you, made well, and made for the reasons you would have chosen. That’s not luck. That’s a built thing, and you can start building it this week.
If you want a partner in the build, someone who has sat with more than a thousand founders and leaders through exactly this work, Book a Discovery Call.