What You Do Is Who You Are: An expert guide to building your company’s culture

by Ben Horowitz

When I was the CEO of LoudCloud, I figured that our company culture would be just a reflection of my values, behaviors, and personality. So I focused all my energy on “leading by example.” To my bewilderment and horror, that method did not scale as the company grew and diversified.

Because 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. It’s how they behave when no one is looking. If you don’t methodically set your culture, then two-thirds of it will end up being accidental, and the rest will be a mistake.

Breakthrough ideas have traditionally been difficult to manage for two reasons: 1) innovative ideas fail far more than they succeed, and 2) innovative ideas are always controversial before they succeed. If everyone could instantly understand them, they wouldn’t be innovative.

Hierarchies are good at weeding out obviously bad ideas. By the time an idea makes it all the way up the chain, it will have been compared to all the other ideas in the system, with the obviously good ideas ranked at the top. This seems like common sense. The problem is that obviously good ideas are not truly innovative, and truly innovative ideas often look like very bad ideas when they’re introduced.

Companies—just like gangs, armies, and nations—are large organizations that rise or fall because of the daily microbehaviors of the human beings that compose them.

most attempt to dissect successful companies’ cultures after the companies have succeeded. This approach confuses cause and effect. There are plenty of massively successful companies with weak, inconsistent, or even toxic cultures; a desirable product can overcome a miserable environment, at least for a while.

Creating a culture is more complex than just trying to get your people to behave the way you want them to when no one is looking. Remember that your employees are far from uniform. They come from different countries, races, genders, backgrounds, even eras. Each one brings to your organization a different cultural point of departure. To get all of them to conform to and be reasonably happy with a common set of norms is a challenging puzzle.

Critics love to attack companies for having a “broken culture” or being “morally corrupt,” but it’s actually a minor miracle if a culture isn’t dysfunctional.

As a final word of discouragement: a great culture does not get you a great company. If your product isn’t superior or the market doesn’t want it, your company will fail no matter how good its culture is. Culture is to a company as nutrition and training are to an aspiring professional athlete. If the athlete is talented enough, he’ll succeed despite relatively poor nutrition and a below-average training regimen. If he lacks talent, perfect nutrition and relentless training will not qualify him for the Olympics. But great nutrition and training make every athlete better.

If a great culture won’t ensure success, why bother? In the end, the people who work for you won’t remember the press releases or the awards. They’ll lose track of the quarterly ups and downs. They may even grow hazy about the products. But they will never forget how it felt to work there, or the kind of people they became as a result. The company’s character and ethos will be the one thing they carry with them. It will be the glue that holds them together when things go wrong. It will be their guide to the tiny, daily decisions they make that add up to a sense of genuine purpose.

Without trust, communication breaks. Here’s why: In any human interaction, the required amount of communication is inversely proportional to the level of trust.

Company cultures organize around a simple goal: build a product or service that people want. But when those companies progress beyond their initial battles they must evolve to take on new challenges.

Building a great culture means adapting it to circumstances. And that often means bringing in outside leadership from the culture you need to penetrate or master.

Louverture began building trust by being trustworthy himself.

For a culture to stick, it must reflect the leader’s actual values, not just those he thinks sound inspiring. Because a leader creates culture chiefly by his actions—by example.

Jobs narrowed the product line to ensure that the company focused on delivering great experiences to individual humans rather than an impersonal set of specs, feeds, and speeds aimed at no one in particular.

Here are the rules for writing a rule so powerful it sets the culture for many years: It must be memorable. If people forget the rule, they forget the culture. It must raise the question “Why?” Your rule should be so bizarre and shocking that everybody who hears it is compelled to ask, “Are you serious?” Its cultural impact must be straightforward. The answer to the “Why?” must clearly explain the cultural concept. People must encounter the rule almost daily. If your incredibly memorable rule applies only to situations people face once a year, it’s irrelevant.

The side effect of creating a massive set of rules that governed behavior—of optimizing for removing all errors rather than for encouraging exploration and free thinking—was to stifle creativity. Hastings vowed not to make that mistake again.

the point is that when you are a leader, even your accidental actions set the culture.

That’s the nature of culture. It’s not a single decision—it’s a code that manifests itself as a vast set of actions taken over time. No one person makes or takes all these actions. Cultural design is a way to program the actions of an organization, but, like computer programs, every culture has bugs. And cultures are significantly more difficult to debug than programs.

Bushido looks like a set of principles, but it’s a set of practices. The samurai defined culture as a code of action, a system not of values but of virtues.

A value is merely a belief, but a virtue is a belief that you actively pursue or embody. The reason so many efforts to establish “corporate values” are basically worthless is that they emphasize beliefs instead of actions. Culturally, what you believe means nearly nothing. What you do is who you are.

Meditating on your company’s downfall will enable you to build your culture the right way. Imagine you’ve gone bankrupt. Were you a great place to work? What was it like to do business with you? Did your encounters with people leave them better off or worse off? Did the quality of your products make you proud?

To build the culture, we would hold study groups once or twice a week.

Culture is weird like that. Because it’s a consequence of actions rather than beliefs, it almost never ends up exactly as you intend it. This is why it’s not a “set it and forget it” endeavor. You must constantly examine and reshape your culture or it won’t be your culture at all.

All leaders get surprised by feedback like “Our culture is really harsh” or “We’re arrogant,” but when they try to examine it directly to figure out what’s going on, they fall prey to the Heisenberg Uncertainty Principle of Management. The act of trying to measure your culture changes the result.

Go around your managers to ask new employees these questions directly after their first week. And make sure you ask them for the bad stuff, the practices or assumptions that made them wary and uncomfortable. Ask them what’s different than other places they’ve worked—not just what’s better, but what’s worse. And ask them for advice: “If you were me, how would you improve the culture based on your first week here? What would you try to enhance?”

People’s receptivity when they join, and the lasting impact of first impressions, is why the new-employee process is the most important one to get right.

If your company’s process for recruiting, interviewing, orienting, training, and integrating new employees is intentional and systematic, great. If any part of it is accidental, then so is your culture.

At Slack, “collaborative” means taking leadership from everywhere. Collaborative people know that their success is limited by uncollaborative people, so they are either going to help those people raise their game or they are going to get rid of them.

If you have a crisis situation and you need the team to execute, meet with them every day and even twice a day if necessary. That will show them this is a top priority. At the beginning of each meeting you say, “Where’s my money?” They will start making excuses like “Boo Boo was supposed to call me and didn’t,” or “The system didn’t tell me the right thing.” Those excuses are the key, because that’s the knowledge you’re missing.

When Wilderotter arrived at Frontier, she told me, “Everybody wanted to show me the org chart, to make sure I understood the pecking order. I didn’t even look at it, because I believe that work gets done through the go-to people. They may not have titles and positions, but they’re the ones who get the work done.”

This interview revealed a key to leadership: you must be yourself. Other people will always have ideas of what you should be, but if you try to integrate all those ideas in a way that’s inconsistent with your own beliefs and personality, you will lose your mojo.

When you need a large organization to work in concert and execute with precision on an enormous number of tasks, you simply don’t have time to explore every nook and cranny of every issue in every conversation. So my discursiveness—which I still like to think of as curiosity—could be extremely disadvantageous. I learned to counterprogram the culture against my inclinations in three ways: I surrounded myself with people who had the opposite personality trait. They wanted to finish the conversation as soon as possible and move on to the next thing. I made rules to help manage myself. If a meeting was called without a tightly phrased written agenda and a desired outcome, we’d cancel it. I announced to the company that we were committed to running meetings efficiently—talking the talk that I did not like walking, and forcing myself to walk it as much as I could. The company still suffered from my inefficient personal style, but we were mostly able to work around it.

Smart. It doesn’t mean high IQ (although that’s great), it means disposed toward learning. If there’s a best practice anywhere, adopt it. We want to turn as much as possible into a routine so we can focus on the few things that require human intelligence and creativity. A good interview question for this is: “Tell me about the last significant thing you learned about how to do your job better.” Or you might ask a candidate: “What’s something that you’ve automated? What’s a process you’ve had to tear down at a company?” Humble. I don’t mean meek or unambitious, I mean being humble in the way that Steph Curry is humble. If you’re humble, people want you to succeed. If you’re selfish, they want you to fail. It also gives you the capacity for self-awareness, so you can actually learn and be smart. Humility is foundational like that. It is also essential for the kind of collaboration we want at Slack. Hardworking. It does not mean long hours. You can go home and take care of your family, but when you’re here, you’re disciplined, professional, and focused. You should also be competitive, determined, resourceful, resilient, and gritty. Take this job as an opportunity to do the best work of your life. Collaborative. It’s not submissive, not deferential—in fact it’s kind of the opposite. In our culture, being collaborative means providing leadership from everywhere. I’m taking responsibility for the health of this meeting. If there’s a lack of trust, I’m going to address that. If t...

The questions employees everywhere ask themselves all the time are “Will what I do make a difference? Will it matter? Will it move the company forward? Will anybody notice?” A huge part of management’s job is to make sure the answer to all those questions is “Yes!”

If your organization can’t make decisions, can’t approve initiatives quickly, or has voids where leadership should be, it doesn’t matter how many great people you hire or how much work you spend defining your culture. Your culture will be defined by indifference, because that’s what you’re rewarding. If I work hard and my neighbor does nothing and we both have the same impact at the company, then her behavior is obviously the way to go.

The Heretic Every company needs lots of smart, super-engaged employees who can identify its particular weaknesses and help it improve them. But some employees look for faults not so they can fix them, but so they can build a case. Specifically, a case that the company is hopeless and run by a bunch of morons. The smarter the employee, the more destructive this type of behavior can be—because people are that much more likely to listen to him. He will convince engaged, productive employees to become disengaged and to rally others to do the same. They will question every management decision, break trust, and cause your culture to disintegrate. Why would a smart person try to destroy the company he works for? He is disempowered. He feels he can’t access the people in charge, so complaining is his only way to get the truth out. He is fundamentally a rebel. Sometimes these people actually make better CEOs than employees. He is immature and naive. He cannot comprehend that the people running the company do not know every minute detail of its operations. He therefore believes they are complicit in everything that’s broken. It’s very difficult to turn heretics around. Once a heretic takes a public stance, the social pressure to be consistent is enormous. If he tells fifty friends that the CEO is the stupidest person on the planet, reversing that position will cost him enormous credibility. Most people aren’t willing to take the credibility hit.

The team, frustrated that all their hard work is being thrown away, will be generally pissed. The natural thing for the manager to say is, “I hear you and, quite frankly, I agree with you, but I was overruled by the powers that be.” This is absolutely toxic to the culture. Everyone on the team will feel marginalized because they work for someone who’s powerless. This makes them one level less than powerless. They have just been demoted from the bottom of the totem pole to the ground beneath it. The strong-willed among them will make their displeasure known throughout the company, causing other employees to question the leadership team and to wonder if their own work will ultimately matter. The end result will be apathy or attrition or both. So it’s critical to a healthy culture that whatever your decision-making process, you insist on a strict rule of disagree and commit. If you are a manager, at any level, you have a fundamental responsibility to support every decision that gets made. You can disagree in the meeting, but afterward you must not only support the final decision, you must be able to compellingly articulate the reasons the decision was made.

In the speed-versus-accuracy calibration, the cultural question of empowerment plays an important role. How far down the org chart can a decision get made? Do you trust lower-level employees to decide important matters, and do they have enough information to do so with accuracy? If employees have a real say in the business, they will be far more engaged and productive. It’s also often the case that sending the question up the hierarchy not only slows things down but results in a less accurate decision.

On the other hand, pushing decision making too far down the organization can cause several problems: It can break communication across product groups. This results in a frustrating customer experience. For years, nearly every Google product had its own customer profile: I might be BenH on gmail, but I couldn’t seamlessly log in as BenH on YouTube. This led to a fractured experience for customers and prevented Google from understanding user behavior across its product lines. (Larry Page finally forced his groups to prioritize a common customer profile.) It can break communication between divisions. Which can lead to a company making great products that it has no capability to market or sell. Xerox’s Palo Alto Research Center (PARC) famously spawned a dizzying array of technological breakthroughs, including the Graphical User Interface, but it could never bring them to market effectively, because the rest of the company didn’t understand what they were up to. Eventually, Xerox realized this and spun PARC off as a wholly owned subsidiary. You can lose input from your very best minds. If you’re at Netflix, what decision wouldn’t benefit from Reed Hastings’s knowledge and experience?

Peacetime CEO knows that proper protocol leads to winning. Wartime CEO violates protocol in order to win. Peacetime CEO focuses on the big picture and empowers her people to make detailed decisions. Wartime CEO cares about a speck of dust on a gnat’s ass if it interferes with the prime directive. Peacetime CEO builds scalable, high-volume recruiting machines. Wartime CEO does that, but also builds HR organizations that can execute layoffs. Peacetime CEO spends time defining the culture. Wartime CEO lets the war define the culture. Peacetime CEO always has a contingency plan. Wartime CEO knows that sometimes you gotta roll a hard six. Peacetime CEO knows what to do with a big advantage. Wartime CEO is paranoid. Peacetime CEO strives not to use profanity. Wartime CEO sometimes uses profanity purposefully. Peacetime CEO thinks of the competition as other ships in a big ocean that may never engage. Wartime CEO thinks the competition is sneaking into her house and trying to kidnap her children. Peacetime CEO aims to expand the market. Wartime CEO aims to win the market. Peacetime CEO strives to tolerate deviations from the plan when coupled with effort and creativity. Wartime CEO is completely intolerant. Peacetime CEO does not raise her voice. Wartime CEO rarely speaks in a normal tone. Peacetime CEO works to minimize conflict. Wartime CEO heightens the contradictions. Peacetime CEO strives for broad-based buy-in. Wartime CEO neither indulges consensus-building nor tolerates d...

Your company’s culture should be an idiosyncratic expression of your personality, beliefs, and strategy—and it should keep evolving as your company grows and conditions change.

How do you build a culture that enables you to discover these problems sooner rather than later? It’s surprisingly difficult. There are several reasons employees don’t naturally tend to volunteer bad news: It seems to conflict with an ownership culture. A common management adage is “Don’t bring me a problem without bringing me a solution.” This idea encourages ownership, empowerment, and responsibility among the employees, but it has a dark side. For one thing, what employees are likely to hear is just “Don’t bring me a problem.” At a deeper level, what if you know about a problem but you can’t solve it? What if you’re an engineer who sees a fundamental weakness in your software architecture, but doesn’t have the authority or expertise to fix it? What if you’re a salesperson who believes one of your colleagues is making fraudulent promises? How can you solve that without help? If you encourage bad news, you must be careful not to disem-power people in doing so. The company’s long-term goals may not align with an employee’s short-term incentives. Imagine that your new product must ship this quarter. It’s so critical you’ve offered a shipping bonus to the engineers. Now imagine that the product has a dangerous security flaw. If you’re an engineer who discovers the flaw, but you need the bonus to buy holiday presents for your children, what do you do? Nobody likes to get yelled at. If you know about a problem, there’s a reasonable chance that you caused it and have no idea ho...

People take their cues from the leader, so if you’re okay with bad news, they’ll be okay, too. Good CEOs run toward the pain and the darkness; eventually they even learn to enjoy it.

Loyalty emerges from an expectation that the other party feels the same way; that your colleagues, and your company, are there for you. CEOs have different approaches to encouraging it. Here’s how Patrick Collison of Stripe thinks about loyalty: We obviously can’t offer lifetime employment. I hope what we can deliver is that in fifteen years, when people look back, they will think that they were able to do the most meaningful work of their careers here. In exchange, I expect two things: first, ethical integrity. Second, that they optimize for the company rather than for themselves. If they satisfy those two expectations, then they have our appreciation, respect, and loyalty. In other words, he will support them throughout their careers.