What You Do Is Who You Are

by Ben Horowitz

  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.

  When Tom Coughlin coached the New York Giants, from 2004 to 2015, the media went crazy over a shocking rule he set: If you are on time, you are late. He started every meeting five minutes early and fined players one thousand dollars if they were late. I mean on time. Wait, what?

  At first, the “Coughlin Time” rule went over poorly. Several players filed grievances with the NFL and the New York Times wrote a scathing critique:

  In the player-relations department, the reign of Giants Coach Tom Coughlin started poorly and is already showing signs of unraveling one game into the season.

  On the heels of Sunday’s 31–17 loss to the Eagles, the N.F.L. Players Association confirmed that three Giants had filed a grievance against Coughlin for fining them for not being early enough for a meeting.

  A few weeks ago, linebackers Carlos Emmons and Barrett Green and cornerback Terry Cousin, all free-agent acquisitions in the off-season, were fined $1,000 each after showing up several minutes early for a meeting, only to be told they needed to arrive earlier.

  Coughlin’s response to the reporter didn’t make him seem more sympathetic, but it did solidify his rule: “Players ought to be there on time, period,” he said. “If they’re on time, they’re on time. Meetings start five minutes early.”

  Was the rule memorable? Check. Did it beg the question “Why?” He had players asking everyone from the league to the New York Times “Why?” so, check. Did they encounter it daily? Yep, they ran into it every time they had to be somewhere. But what was he trying to achieve?

  Eleven years and two Super Bowl wins later, backup quarterback Ryan Nassib explained the cultural intention to the Wall Street Journal:

  Coughlin Time is more of a mindset, kind of a way for players to discipline themselves, making sure they’re on time, making sure they’re attentive and making sure they’re ready to work when it’s time to start meetings. It’s actually kind of nice because once you get out in the real world, you’re five minutes early to everything.

  In business, creating partnerships that work is a difficult art. Success stories such as the partnership of Microsoft and Intel or of Siebel Systems and Accenture become legendary, but for every success there are a hundred failures. It’s difficult enough to align interests in your own organization, where everyone works for you, but doing it between companies is close to impossible.

  In the 1980s, the business literature promoted the concept of win-win partnerships. Unfortunately, the idea was pretty abstract. How do you know if a deal is win-win? Can you actually determine when it’s fifty-fifty? The idea also failed to address the cultural adjustment required: if everything in a business culture is about winning, what behavior changes are necessary to achieve a win-win mindset? Finally, its meaning was easy to twist. Devious negotiators routinely said, “We want this to be a win-win.”

  In 1998, Diane Greene cofounded a virtualized operating system company, VMware, whose success depended on her partnership strategy. But she was entering a field that had witnessed the biggest win-lose partnership ever—Microsoft winning total dominance by “partnering” with IBM on the desktop operating system. VMware’s potential partners would be extremely skeptical of any independent-operating-system company proposing a similar “win-win.”

  So Greene came up with a shocking rule: Partnerships should be 49/51, with VMware getting the 49. Did she just tell her team to lose? That definitely begs the question “Why?”

  Greene said, “I had to give our business development people permission to be good to the partners, because one-sided partnerships would not work.” Her rule was actually met not with resistance but with relief. Her people wanted to create mutually beneficial partnerships, and Greene’s rule gave them permission. It was of course no easier to measure an exact 49/51 split than a 50/50 “win-win,” but Greene’s employees understood her underlying point: “If you’re negotiating something on the margin, it’s okay to give it to our partner.” VMware went on to create a stunning set of partnerships with Intel, Dell, HP, and IBM that propelled the company to a market capitalization of more than $60 billion.

  One of the most distinctive large-company cultures is Amazon’s. It promulgates its fourteen cultural values in a number of ways, but perhaps most effectively through a few shocking rules. One value, frugality, is defined as Accomplish more with less. Constraints breed resourcefulness, self-sufficiency, and invention. There are no extra points for growing head count, budget size, or fixed expenses.

  That’s a nice definition, but how do you drive home that you mean it? Here’s how: desks at Amazon were built by buying cheap doors from Home Depot and nailing legs to them. These door desks weren’t great ergonomically, but when a shocked new employee asked why she had to work at a makeshift desk, the answer pinged back with illuminating consistency: “We look for every opportunity to save money so we can deliver the best products for the lowest cost.” (Amazon no longer gives everyone a door desk, as the culture has now been set—and as there are cheaper alternatives.)

  Some of Amazon’s values are fairly abstract. Dive deep, for instance, encourages leaders to operate at all levels, stay connected to the details, audit frequently, and investigate more thoroughly when metrics and anecdotal evidence disagree.

  Great idea—but how do you drive this kind of thoughtfulness into the culture? The shocking rule that helps is No PowerPoint presentations in meetings. In an industry where presentations rule the day, this rule definitely counts as shocking. To convene a meeting at Amazon, you must prepare a short written document explaining the issues to be discussed and your position on them. When the meeting begins everyone silently reads the document. Then the discussion starts, with everyone up to speed on a shared set of background information.

  Amazon executive Ariel Kelman explains that the rule makes meetings much more efficient:

  If you have to talk about something complicated, you want to load the data into people’s brains as quickly as possible so you can have an intelligent, facts-based conversation about the business decision you’re trying to make.

  So, say you’re meeting to figure out pricing for a new product, you’ve got to talk about the cost structure, how much is fixed, how much is variable, and then there might be three different pricing models, each with pros and cons. That’s a lot of information. Now, you can sit and listen to someone pitch all of this information, but most people don’t have the patience to pay attention long enough to be effective in absorbing all of this data and it typically takes too much time. There’s been a lot of research done on this that shows that most people’s brains can absorb new information several times faster and more effectively by reading information versus listening to it. Also, asking people to present their plans in written format forces them to express their ideas with a deeper level of detail.

  A culture is a set of actions. By requiring thoughtful action before every meeting, Amazon moves its culture in the right direction every day.

  In the early days of Facebook, Mark Zuckerberg was keenly aware that the more people he got on his network, the better his product would be. As MySpace had far more users, Facebook had to outgrow them by building better software—software that had better features, was more user-friendly, and that excelled at identifying potential new Facebook users. Zuckerberg knew that he didn’t have much time: if MySpace got big enough, it might transform from an entertaining application into an invincible utility.

  Speed was the number one virtue he needed, so he created a shocking rule: Mo
ve fast and break things. Imagine you are an engineer hearing that for the first time: Break things? I thought the point was to make things. Why is Mark telling us to break things? Well, he’s telling you so that when you come up with an innovative product and you are not sure whether it’s worth potentially destabilizing the code base to push the product along, you already have your answer. Moving fast is the virtue; breaking things is the acceptable by-product. Zuckerberg later observed that the reason the rule was so powerful was that it stated not only what Facebook wanted, but what it would give up to get it.

  After Facebook caught and passed MySpace, it had new missions to pursue, such as turning the social network into a platform. At that point, the move fast virtue became more liability than asset. When outside developers tried to build applications on Facebook, the underlying platform kept breaking, which jeopardized the businesses of Facebook’s partners. So in 2014 Zuckerberg replaced his by-now-famous rule with the boring but stage-appropriate motto Move fast with stable infrastructure. Cultures must evolve with the mission.

  When Marissa Mayer became CEO of Yahoo! in 2012, its reputation was of a company whose workforce didn’t give its all. She knew that to compete with her old company, Google, she would need a better effort from the team. She began by trying to lead by example, working relentlessly long hours. Yet she kept arriving at work to see an empty parking lot.

  So in 2013, Mayer created a rule so shocking that it created massive backlash not only inside but even outside the company: during work hours, you must be at work. Nobody is allowed to work from home. But this was the technology industry—the industry that had invented the tools that enabled people to work from home! As the world exploded in anger, Mayer calmly explained her position. She had examined the virtual private networking logs of employees who were working from home; they had to use the VPN to securely access their work files. The logs showed that most people “working from home” had in fact not been working at all.

  She shocked people because she had to make a dramatic cultural change. It’s worth nothing that while Mayer succeeded in building assiduousness back into Yahoo!’s culture, she never quite turned the company around. That’s the nature of culture—it helps you do what you are doing better, but it can’t fix your strategy or thwart a dominant competitor.

  Dress for Success

  When Mary Barra took over as the CEO of General Motors in 2014, she wanted to dismantle the company’s powerful bureaucracy. It stifled employees and disempowered managers: rather than communicating with employees and giving them guidance, the managers relied on the extensive system of rules to do the job for them. The ten-page dress code was the worst example. To shock the system and change the culture, Barra reduced ten pages to two words: dress appropriately.

  She told the story at the Wharton People Analytics Conference:

  The HR department started arguing with me, saying, it can be “Dress appropriately” on the surface, but in the employee manual it needs to be a lot more detailed. They put in specifics like, “Don’t wear T-shirts that say inappropriate things, or statements that could be misinterpreted.”

  Barra was perplexed.

  “What does inappropriate, in the context of a T-shirt, even mean?” she asked the audience, half-jokingly.

  So I finally had to say, “No, it’s two words, that’s what I want.” What followed was really a window into the company for me.

  Barra promptly received an email from a senior-level director:

  He said, “You need to put out a better dress policy, this is not enough.” So I called him—and of course that shook him a little bit. And I asked him to help me understand why the policy was inept.

  The director explained that some people on his team occasionally had to deal with government officials on short notice, and they needed to be dressed appropriately for that.

  “Okay, why don’t you talk to your team,” I replied. He was an established leader at GM, responsible for a pretty important part of the company, with a multimillion-dollar budget. He called me back a few minutes later, saying, “I talked to the team, we brainstormed, and we agreed that the four people who occasionally need to meet with government officials will keep a pair of dress pants in their locker.” Problem solved.

  The change sent a lasting visual message to GM’s entire management team. Every time a manager saw an employee, it would trigger the thought, Is he dressed appropriately? And, if not, What’s the best way for me to manage that? Do I have a good enough relationship with him to communicate effectively on this sensitive issue? The new code empowered—and required—managers to manage.

  When Michael Ovitz ran Creative Artists Agency, Hollywood’s leading talent agency, he, too, had no explicit dress code. But he absolutely had an implicit one. “In the mid-seventies, we lived in a world coming off sixties culture, where everyone wore jeans and T-shirts,” Ovitz recalled. “That’s what I needed to counter-program.” The dress code he landed on came from the culture of authority he sought: “If you walk into the room wearing an elegant dark suit, you pick up unbelievable positioning power. If you want respect, carry yourself in a way that commands it.”

  Ovitz wore elegant dark suits every day, leading by example. He never explicitly asked anyone to follow his lead. That didn’t mean there weren’t consequences if you didn’t. “There was a downpour in LA, and some people came in in rain boots and jeans. I went up to one agent and said, ‘Nice outfit. Are you working on set today?’ And that rattled through our business.” Ovitz was giving him the hip-hop ultimatum: Are you a hustler or a customer? Are you a world-class agent or a wannabe actor? This steely but largely unspoken approach soon shifted CAA to nearly complete dress code compliance. “The only exception was our music department, because musicians don’t like guys in suits.”

  The results of the code on the culture were profound:

  It became part of our ethos: we were classy, elegant, conservative businesspeople. It spoke to everything we wanted to be without our having to say it out loud. Through our culture we built our business to a place where people respected it due to the culture itself.

  How you dress, the most visible thing you do, can be the most important invisible force driving your organization’s behavior. Ovitz sums it up: “Cultures are shaped more by the invisible than the visible. They are willed.”

  Incorporate Outside Leadership—Hey, Motherfucker!

  When I was CEO of LoudCloud, I had to shift the company from a high-flying cloud services company into a grind-’em-out enterprise software company so we could survive. After the dot-com and telecom crashes of the early 2000s, the market for cloud services had gone from nearly infinite to nearly zero overnight. After we squeaked through the transition as a new company called Opsware, we found ourselves getting killed in the software market by a competitor named BladeLogic. I knew that to compete with them we needed a major cultural change.

  At LoudCloud, we began with unlimited demand and built a culture oriented around fulfilling it. So we were focused on empowerment, removing bottlenecks to growth, and being a great place to work. To succeed as an enterprise software company, selling our platform to big businesses, we would have to become a culture distinguished by urgency, competitiveness, and precision. I needed to bring in a leader with those attributes.

  The person I hired as our head of sales, Mark Cranney, was not a cultural fit with the rest of us. In fact, he was a complete cultural misfit. Our employees were mostly irreligious Democrats from the west coast who dressed casually and who were cordial and easygoing. We assumed that everyone had the best intentions. Cranney was a Mormon Republican from Boston who wore a suit and tie, was deeply suspicious of everyone, and was one of the most competitive people on earth. But over the next four years he not only saved the company, but got us to an outcome nobody would have believed.

  I knew why I hired Mark: when I interviewed him, I could tell he had the urgency, the know-how, and the discipline we needed. But I did not understand why he took the job.
He knew we were losing and, given our granola-eating demographics, that we were probably losers. So what made him take the risk? I recently asked him, and his reason surprised me:

  I had risen as far as I could at an east-coast-based company called PTC; they had nepotistic politics at the top level. I must have looked at forty sales jobs in Boston and there was nothing good.

  The Opsware recruiter called several times and I finally called him back and said, “I’m not going to California. In California the real estate sucks, the culture sucks, and they don’t appreciate the sales side. Plus, isn’t that the company the BladeLogic guys call Oopsware? What do you think, I’m fuckin’ stupid?”

  He keeps calling and finally, I say “Fine, I’ll go meet Marc and Ben, but that’s it.” [Marc Andreessen was the company’s cofounder.] Then I look at my BlackBerry when I land in San Francisco and I see there’s a whole fucking crew I have to interview with.

  So I go over and you come out of your cubicle and I’m like, Fucking cube company. The cubes confirmed my initial suspicions: soft, beach boys, consensus, everyone gets a say. That’s fine for engineering, but sales and marketing have to go to battle every day, so people in those groups need to fall in line. Then I’m looking at the conference room names—Salt-N-Pepa, Notorious B.I.G.—and I’m like, What the fuck are these? When I realize they’re the names of rappers I think, Jeez, this is not going to go well.

  We sat down and I said, “Ben, before we get started I need to know what your process is and what your decision criteria are. You’ve got all these people interviewing me and if they all get a vote then it kind of explains why you’re in the situation you’re in.” You got up out of your seat and said, “Hey, motherfucker! I’m the CEO. I make the decisions.” When you said “Hey, motherfucker!” I said to myself, Wait a minute. Maybe I can do this.