Building Company Culture: How the Team Decides When You're Not There
Topic: Building Company Culture·4 Principles
"Culture is how a company makes decisions when you're not there." — Ben Horowitz
This week's premise: Most "culture building" is self-deception—you print the values on mugs and lobby posters, then keep rewarding the behavior that violates them. Culture is not what you proclaim; it's what you reward, what you tolerate, who you promote, and what you cut first under pressure. As a tech lead manager in a big company, you can't control company-level culture—but you can fully decide the culture of your own small team, and it's often different from what the company claims. Four steps: see culture's real layers, write it as an executable contract, preserve the founder's mentality as scale dilutes it, and make culture outlive your departure.
PRINCIPLE 01
Culture Is the Behavior You Reward and Tolerate, Not the Values on the Wall
Culture Is What You Reward and Tolerate
EssenceSchein's 3 levelsSay-do gap
The Principle in One Line
The values poster is culture's outermost—and least trustworthy—layer. Real culture lives in the underlying assumptions: the team's default sense that "here, X is safe and Y gets punished." To change culture, change what you reward and tolerate, not the poster.
In Their Words
"Your culture is how your company makes decisions when you're not there. It's the set of assumptions your employees use to resolve the problems they face every day."— Ben Horowitz, What You Do Is Who You Are (Introduction)
Framework: Schein's Three Levels of Culture
Scenario
Situation: Your team's wall says "Quality First," but last quarter's promotion went to the engineer who pulled an all-nighter to suppress a production incident—while the person who quietly eliminated that whole class of incident, so the system never broke again, went unmentioned.
✗ Self-deceiving culture work
You give the "we value quality, we value the long term" speech again at the all-hands and make a fresh values slide. What the team actually learns: firefighting makes headlines; fire prevention is invisible. Next quarter, no one wants the silent preventive work.
✓ Change the reward signal
At the promotion review you say explicitly: "The person I want to recognize is the one who made the incident never happen—their work had no demo, none of firefighting's drama, but it's the culture we actually want." Then write it into the promotion rationale doc, publicly visible. Culture is calibrated by one real reward decision, not one speech.
Checklist: To Know Your Team's Real Culture, Skip the Poster—Read These 5 Signals
What behavior did the last public recognition / promotion reward? That's the strongest culture broadcast.
What behavior gets tolerated with no one addressing it? (Tolerated = tacitly endorsed = the de facto standard.)
When pressure hits and something must be cut, what goes first? (Exposes the real priority.)
In month three, what unwritten "here's how it actually works" rules do newcomers hear from veterans?
What "legends" do people retell privately—about whom, about what? (Stories are values.)
Common Mistakes
Mistaking proclamation for building. Revising the values slide a hundred times matters less than one promotion decision.
Rewarding against your values, unaware. You preach collaboration but pay bonuses only to individual heroes—the team believes the bonus.
Touching only the surface. New logo, new slogan, more offsites—underlying assumptions untouched.
PRINCIPLE 02
Write Culture as an Executable Contract, Not Platitudes Anyone Would Sign
Write Culture as a Testable Contract
Culture DeckFalsifiableTrade-offs
The Principle in One Line
Unwritten culture can't be inherited, hired for, or calibrated. But write it as behavior that distinguishes people, has a counter-example, and requires a trade-off—"we value integrity" is something anyone signs, which means it says nothing. A good culture line should make some people uncomfortable.
In Their Words
"We're a team, not a family. ... Adequate performance gets a generous severance package."— Patty McCord & Reed Hastings, Netflix Culture Deck / Powerful
Scenario: Writing the Team's Operating Principles
Situation: You want to set a few culture principles for the team, to put in the onboarding doc and hiring bar.
✗ Platitudes anyone would sign
"We pursue excellence. We embrace change. We respect each other. We are customer-centric." —Not one of these helps you choose between two candidates, and not one makes anyone uncomfortable. It's not culture; it's decoration.
✓ Falsifiable, trade-off-bearing lines
· "Async by default; meetings are the last resort. If it can be a doc, don't meet; a meeting needs an agenda, and one without an agenda can be declined." · "Bad news flows upstream, the faster the better. Whoever reports a problem is protected; hiding it, once found, is the offense." · "We're accountable for output, not hours. No one tracks when you shipped; delivery quality is the only yardstick." Each implies a trade-off, a counter-example, and a kind of person it would screen out.
Checklist: Is a Culture Principle Sound? (5 Questions)
Does it distinguish people? Can it help you decide between two candidates? If not, it's too vague.
Does it have a counter-example? Is there a reasonable opposite choice? ("Async-first" vs "sync-first"—both are valid stances.)
Is it observable? Can you point to concrete behavior and say "this fits / this violates"?
Does it involve a trade-off? What does it give up? A principle that sacrifices nothing is a slogan.
Can you hire for it? Can you design an interview question to test fit?
Common Mistakes
Copying a big company's deck. Netflix's "freedom and responsibility" comes paired with high pay and high turnover; copy the words without the supporting mechanisms and you get an empty shell.
Writing it, then locking it in a drawer. A culture doc that doesn't enter hiring, onboarding, and performance is a dead file.
Too many lines. 15 equals none; 3–5 with real trade-offs are memorable and binding.
PRINCIPLE 03
The Founder's Mentality: Rebuilding Ownership in a Big Company's Sub-Team
The Founder's Mentality at Sub-Team Scale
Founder MentalityOwnershipFrontline obsession
The Principle in One Line
Scale creates complexity, and complexity dilutes the owner's mindset—Bain calls it "the paradox of growth." You can't stop the company from growing, but in your own small team you can rebuild the founder's mentality's three traits: an insurgent mission, frontline obsession, and an owner's mindset.
In Their Words
"The founder's mentality has three traits: an insurgent mission, an obsession with the front line, and an owner's mindset. Growth creates complexity, and complexity is the silent killer of these traits."— Chris Zook & James Allen, The Founder's Mentality (Bain, 2016)
Scenario
Situation: Your team used to be the one in the company that "took on any dirty job and bit down hard on user problems." Two years on, it's become an execution unit that "delivers by ticket, has full process, but no one truly cares about outcomes." Engineers own their tasks, not the output.
✗ Filling the ownership vacuum with process
Add more process, finer Jira fields, more status meetings, trying to buy back quality with control. Result: higher complexity, lower ownership—everyone is just clocking the work.
✓ Scripts to rebuild the three traits
Insurgent mission: "We're not the maintenance crew for system XX. We exist so that [a class of users] never has to endure [a specific pain]—that's our fight." Frontline obsession: "This sprint, everyone reads 5 raw user reports / sits in on one oncall. Cut off from the front line, we drift into PowerPoint." Owner's mindset: "As of today this metric is yours—not 'a task assigned to you' but 'your territory.' It does well, you get credit; it breaks, you decide how to fix it."
Checklist: Is the Owner's Mindset Still There? (Read Signals, Not Slogans)
Does anyone pick up ownerless work on their own, rather than "that's not in my scope"?
Does anyone dare say "we shouldn't be building this feature at all", even after it's scheduled?
How many layers separate the team from the front line (real users, real incidents)? More layers, thinner ownership.
When something breaks, is the first reflex "whose fault" or "I'll fix it"?
Is there an "insurgent mission" at all—or only "finish what's assigned from above"?
Common Mistakes
Mistaking process for culture. Process is a byproduct of complexity, not a source of ownership; heavier process often means lower ownership.
Thinking ownership can be obtained by exhortation. It comes from real delegation—you have to hand over decision rights and credit together.
Denying the cost of growth. The honest trade-off: scale genuinely brings necessary process and coordination cost. The goal isn't to return to a chaotic garage shop—it's to keep the ownership core inside necessary complexity.
PRINCIPLE 04
The Culture Handoff: Once You Leave, Does It Still Live?
Culture Handoff — Will It Outlive You?
InheritanceEmbedding mechanismsSurviving reorg
The Principle in One Line
A culture sustained by your personal charisma is a single point of failure—one transfer or reorg and it collapses. Culture must be inherited through what Schein calls embedding mechanisms: what you attend to, what you reward, how you react in a crisis, who you hire and promote. Institutionalize these, and culture lives without you.
In Their Words
"The only thing of real importance that leaders do is to create and manage culture. ... If you do not manage culture, it manages you."— Edgar Schein, Organizational Culture and Leadership
Scenario
Situation: The team you built from scratch is being handed to a successor, or folded into a larger org by reorg. You fear the hard-won culture (candor, no-blame for bad news, output orientation) will evaporate once you're gone.
✓ Make tacit culture explicit, turn it into mechanisms
1. Write down decision principles: "Here's how we've always decided"—turn the judgment criteria that live only in your head into a one-page doc for the successor. 2. Tell the founding story: behind "why we do it this way" is an origin event. Tell it to the successor and to newcomers—stories outlive rules. 3. Institutionalize rituals: make blameless postmortems and bad-news protection fixed calendar events that don't depend on your attendance. 4. Name the culture carriers: "Zhang and Li are the living embodiments of this culture. Keep them and give them a voice—worth more than any doc."
Checklist: Before the Handoff, Make Sure Culture Survives
Are the core decision principles written down, or do they live only in your gut?
Have you identified 2–3 culture carriers? Does the successor know who they are?
Do the key rituals (postmortems, 1:1 cadence, bad-news channel) run without your presence?
Are reward / promotion criteria institutionalized, rather than your personal call?
Does the successor buy into this culture—or will they tear it down on arrival? (A handoff without buy-in is a guaranteed collapse.)
Common Mistakes
Assuming culture continues automatically. Culture with no institutional carrier reverts to the company average within six months.
Handing off projects but not culture. The handoff doc is full of architecture and oncall, not a word about "how we decide, how we treat bad news."
Denying the successor has changed. A new leader has the right to reshape culture; what you can do is articulate and institutionalize the good parts—not demand they enshrine it unchanged.
Female Leader's Note
Female Leader's Note
"Culture building" contains a lot of invisible glue work—offsites, onboarding buddy, organizing rituals, maintaining psychological safety. Research (Lise Vesterlund, The No Club) shows women are disproportionately assigned this "no-one-can-decline, doesn't-count-in-reviews" office housework. Two moves: make the work explicit and rotated, so it enters the team's public ledger rather than landing on one person by default; and account for the glue work you do and claim it in reviews—"I built and maintained the team's postmortem and bad-news channels" is real leadership output, not background noise to be erased.
Going Deeper · Pushing the Edges
1. Can a small team's culture stay at odds with the company's mainstream long-term without being assimilated?
It can hold for a while, but there's a ceiling. A subculture is like an island against the current—your dike holds back the mainstream (promotion mechanics, evaluation framing, your boss's preferences). As long as you're there and your team keeps producing results to buy autonomy, it survives and can even become a model. But once a reorg, a new boss, or a dip in results arrives, the dike breaks. The pragmatic move isn't to resist head-on—it's to frame the subculture as a "playbook" that serves company goals, giving leadership a reason to protect it rather than treat it as a deviant. Cultural autonomy is, in essence, bought with performance.
2. Can Netflix's "high-freedom, high-turnover" strong culture be copied elsewhere?
Mostly not. A strong culture is a tightly self-consistent system: freedom paired with responsibility, responsibility with high pay, high pay with high turnover, turnover with a "team-not-family" contract. Remove any link and copy only the words and it deforms—freedom without accountability becomes license; turnover without high pay and decent severance becomes exploitation. Schein's insight is that culture solves adaptation problems in a specific environment; Netflix's answer fits its high-talent-density industry. Don't copy the answer—copy the method of pairing every value with a corresponding mechanism.
3. Is the founder's mentality's "insurgent mission" still fitting in a stable, mature large org?
The form must change; the core still holds. A mature org may not need a grand "disrupt the industry" narrative, but every team needs a reason to exist beyond "maintain the status quo"—even just "make this legacy system stop tormenting users." A team without a sense of mission degrades into a pure execution unit, and ownership evaporates with it. But be honest: not every role can have an epic mission, and manufacturing one rings false. Safer to land the mission at the scale of a concrete user pain—real and tangible, not a slogan.
4. How do you balance culture inheritance against the successor's autonomy?
This is a real tension with no clean solution. If the successor copies everything wholesale, the culture is yours, not theirs, and they can't steer it; if they tear it all down, what you built is lost and the team endures pointless turbulence. The healthy band: you make the "why" of the culture clear and institutionalized (so the good parts have inertia and don't vanish overnight with a personnel change), but genuinely hand over the right to modify the "how." What most needs handing off isn't the rule list—it's the judgment behind the rules; once the successor buys the judgment, they'll naturally evolve a form that fits them.
5. When culture and performance clash—a high-output person who damages culture: keep or cut?
This is the hardest real call, and culture's most consequential broadcast. Horowitz and McCord agree: keeping the "brilliant jerk" announces that your culture lines are fake. The team reads it precisely—so output can buy out the principles. But don't romanticize either: when they're irreplaceable and their departure carries major business risk, the right move is often to draw a clear line + give a deadline to change the behavior, not to remove them instantly. Count the cost honestly: the cost of tolerating is a devalued culture; the cost of cutting is short-term loss of firepower. There's no painless option—only which one you're willing to pay for.
This Week's Exercise · Your Day 40 Action
Do two concrete things this week—not reflection, not reading:
Step 1 (culture checkup): Run an honest diagnosis on your team with Card 1's 5 signals—especially ask yourself: "What behavior did my last public recognition / promotion reward? Does it match the culture I preach?" Write down the line where they don't match—that's your team's real culture gap.
Step 2 (write it down): Using Card 2's 5-question test, write 3 operating principles with real trade-offs—each must distinguish people and have a counter-example. Then test them: read them to a colleague of the opposite temperament; if they nod at all of them and say "isn't this obvious?", you're still writing decoration—go back and rewrite until at least one makes them frown.
One reflection question: If you transferred tomorrow, how many lines of your team's culture could survive six months without you? The ones you can't answer for are what you should institutionalize this quarter.