Decisions - The True Meaning of Consensus - How Google Works (2014)

How Google Works

Decisions—The True Meaning of Consensus

In December 2009, we learned that Google was under attack from hackers. That we were under some form of attack wasn’t unusual, in fact it happened practically every day. But this time was different. The sophistication of the attack was something we hadn’t experienced before, and so was its objective. A criminal (or, more likely, team of criminals) had somehow found a way to access Google’s corporate servers. Up until then, most bad guys who attacked us were intent on disrupting Google’s services, to shut us down or make it harder for users to access us. This time the bad guys wanted our confidential information.

Sergey immediately started working on stopping the attack and figuring out who was perpetrating it and how. In a matter of hours he formed a team of the smartest computer security experts he could find, and gathered them in a nondescript building near our Mountain View headquarters. Over the next couple of weeks, the team set up systems that ultimately allowed them to watch the attacks as they were in progress, and what they found was chilling. The hackers weren’t just stealing intellectual property, but were also trying to access Gmail accounts, including those of human rights activists. And the attacks originated from within the nation with the fastest-growing major economy in the world: China.

It was about five and a half years earlier, in mid-2004, that we began to get involved in the Chinese market. From a business standpoint, entering China was not a controversial decision. China was (and is) a huge market, with more people than any other country, tens (now hundreds) of millions of Internet users, and an economy that was growing very quickly. There was a local competitor, Baidu, who had already developed a formidable presence in search, and Yahoo was also gaining momentum. Larry and Sergey visited the country and came away very impressed by all the innovation and energy they witnessed. They had always wanted to hire all the best engineers in the world, and a lot of those engineers were in China.116

But while the business indicators all pointed to a slam dunk decision to get involved, the don’t-be-evil indicators were much more mixed. Information did not flow freely across the Chinese Internet. We knew this from direct experience: On most days, Chinese citizens were allowed to access our US site,, and get its unfettered (albeit English) results. But occasionally, Chinese traffic would drop to zero, and people from China trying to get to would instead be routed to Baidu (and its filtered results). Would opening up a localized site in China be better for the Chinese people, even if we would have to abide by local regulations, or would it make us complicit in the government’s censorship, something that ran counter to the essence of our company’s culture and values? Would establishing ourselves as a local business give us a chance to improve access to information and shed light on the questionable (and nontransparent) practices of the other search providers in China?

From the get-go, Sergey Brin was squarely in the “stay out” camp. His family had immigrated to the United States from the Soviet Union when he was a child, so he had firsthand experience with Communist regimes and he didn’t want to support the one in China in any way. But many others on Eric’s staff disagreed, and the business factors—plus the hope of being able to change the information climate in China—tipped the scale in favor of entering. Sukhinder Singh Cassidy, who was running our Asia operations at the time, moved quickly, and within a few months established a Google China subsidiary. We set up a business office in Beijing, and we grudgingly decided to comply with local censorship regulations, but with a twist: We would inform users when results were being blocked. They couldn’t access the censored information, but at least they would be informed that censorship was occurring.117

One thing that surprised us was that many of the censorship requests we received were intended to suppress links to content that didn’t violate any clear, written law. Sometimes these requests were an attempt to mitigate spats between various government departments (one agency censoring the public statements of another agency) or to suppress scandals that had been planted online. For example, rumors started circulating that the sparkling new Beijing headquarters of CCTV (China Central Television) had a design based on rather salacious images. So we received, and complied with, a request to censor searches related to, among other things, CCTV, genitalia, and porn jokes. (And for all of you who just Googled those terms, (1) shame on you, and (2) we hope you’re not at work!)

In January 2006, we launched our localized Chinese site,, with in-country servers, and a few months later Eric visited Beijing to promote the site. During one of his press interviews he somehow ended up sitting directly below a framed picture of Mao Zedong and Ho Chi Minh. The US press, which was already ambivalent about Google entering China, had a field day with that one. But things went well after that inauspicious beginning: Our local engineers helped the product get much better, and traffic and revenue grew steadily between 2006 and the end of 2009.


With the hacking attacks, all that progress was suddenly in danger. Eric had always believed that engaging in China was not only the right business decision, but the right moral decision as well. While Sergey had always disagreed, Larry had sided with Eric. In light of the attacks, though, Larry was changing his mind. The behavior we were seeing was evil, he told Eric, and wasn’t going to stop; in fact, the harassment would likely get worse. Eric agreed with this assessment, but was surprised that self-eviction was our answer. Both founders were now firmly against censoring our results on

For leaders, decisions are when the hard work begins; there’s a reason why the word “tough” is so often followed by “decision.” (In recent decades it’s also often followed by “love,” but the implementation of that policy is beyond the sphere of this book.) Google’s decision to leave China was emblematic of how we reach decisions, how our process works. Formulating a strategy, hiring the right people, and creating a unique culture are all preliminaries to the fundamental activity of all businesses and business leaders: decision-making.

Different institutions take different approaches to decision-making based on their hierarchical structure. The Marines (top-down) keep it simple: One guy gives the orders to take the hill; everyone else takes the hill. “Dammit, there’s only one guy in charge here so put on your helmet and get going.” Most big corporations (bureaucratic) have far more analyses to perform before they can decide the best course of action. Do they have all the data they need? Have the analysts crunched it? Did they calculate pro forma revenues and EBITDA?118 Weeks go by, the seasons change, and the hill stays before them, untaken. “Maybe next quarter; the hill is definitely one of our stretch goals.” And in the hip start-up (enlightened), the CEO proclaims that she works for the employees so decisions are made by consensus. Everyone gets a say and the arguments are collegial, considerate, and last forever. “Let’s everyone go chillax, grab a cappuccino, and meet back here in a half hour to see where we stand, hill-wise.”

So who’s right—top-down Marines, bureaucratic corporations, or enlightened start-ups? The pace of business change in the Internet Century dictates that decisions be made quickly; the Marines win in that regard. More demanding and informed customers and increased competition dictate that they be as well informed as possible; the corporations may have an edge there. And having a team of smart creatives dictates that everyone gets a say; hello, start-ups. So all of them are right, of course. And they are all wrong too.

The answer lies in understanding that when it comes to making decisions, you can’t just focus on making the right one. The process by which you reach the decision, the timing of when you reach it, and the way it is implemented are just as important as the decision itself. Blow any of these, and the outcome will likely be negative. And since there’s always another decision to be made, the impact of a poorly executed decision-making process can reverberate past that one issue.

As Sergey and his team continued their investigation throughout the latter part of December 2009, Eric knew that one of the most important decisions in the company’s history was at hand. Although he believed that staying in the China market was the best thing for the company, he also knew that both of the founders now disagreed with him. They no longer felt that our presence in the market was helping change government censorship practices, and didn’t want to participate in any way in that censorship. It would be an uphill battle to change their minds, so Eric’s focus shifted. It wasn’t just about making the best decision for the company, but about orchestrating the process so the company reached that decision in the best possible way. There would be other crises and other important decisions, and the smart creatives who populated his staff and ran the company would be paying attention to and learning from how this one was handled. It was especially challenging, given that he was reasonably confident he would disagree with the outcome.

Sergey and his investigative team conclusively confirmed the origin and scale of the attack in early January, and the news was bad. Not only were the hackers trying to steal source code, they had also attempted to compromise the Gmail accounts of several Chinese political dissidents. Sergey felt it was important to announce the attack, and how Google would react, very quickly. There was little disagreement on that point. In Eric’s staff meeting that first week of January, Sergey forcefully made the argument that, as a response to the hacker attacks, we should stop complying with government censorship policies. He wanted us to stop filtering search results on, even if it meant that the government would likely shut down the site, reversing much of our hard-won progress in the market. He stood up in the meeting to deliver his point; usually Sergey stands in meetings only when he’s wearing his Rollerblades. Eric was traveling that day and attending the meeting via video conference, so he counseled his team to consider all the data and come to the next meeting prepared to express and defend a position on what the company should do.

Because of the urgency of the situation, Eric convened the next team meeting for the following Sunday afternoon—January 10, 2010—at four p.m. It started with Sergey conducting a detailed technical review of the situation for well over an hour. He then reiterated the position he had expressed earlier in the week: We should stop filtering our results. Eric knew that Larry was on Sergey’s side, which meant that the decision was effectively made. But it was critically important that all of the members of his team be heard and have a vote. Everyone would have to pull together and rally behind the decision, regardless of where they stood on the matter. So the meeting continued for several hours. We reviewed the facts and had a lengthy, sometimes heated discussion. Finally, Eric called for a vote. The sentiment in the room was clearly favoring Sergey’s position, and the vote wasn’t really necessary, but Eric felt it was important that each person get a chance to record his or her position. Some agreed with Eric that leaving China was tantamount to disengaging from that market for the next hundred years. The majority sided with Sergey, who believed that the Chinese government would eventually change their behavior because their current model would not be sustainable, leaving the door open at some point in the future for Google to reenter the market.

The ultimate decision, which the weary team reached around nine p.m. that evening, wasn’t to pull out immediately. Rather, we would disclose the hacking attack with as much transparency as possible; to the best of our knowledge, of the numerous companies that were affected, we were the only one to go public with the details. And we would announce our plans to stop censoring results on We would not make this change immediately, instead giving ourselves time to—as our lead attorney, David Drummond, put it in the blog post announcing the decision—“[discuss] with the Chinese government the basis on which we could operate an unfiltered search engine within the law, if at all.” On Monday, Eric discussed the decision with the board, and on Tuesday, January 12, 2010, we announced it publicly.

The morning we made the announcement, we got several calls from government officials to our Beijing office wondering if it was some sort of joke. No one does this, one of them told us. Everyone just leaves quietly.

We were not leaving quietly. It was a public ultimatum, and Eric had complete clarity on what was going to happen. We would continue to talk with Chinese officials, to see if we could find a solution that was consistent with both our new public position and Chinese law, but that would fail. Google wouldn’t back down from its public stance, and China wouldn’t repeal its laws. So, as expected, in March we took the preordained step of shutting down search on Users visiting that page who tried to perform searches were directed to our site in Hong Kong, From that point on, Google search results would be subject to being blocked by the Great Firewall of China. Our traffic dropped precipitously.


The TGIF of January 15, 2010, was dominated by discussion of the Chinese issue. Sergey and the security team presented in great detail what had happened, and reviewed the process by which the management team had made its decision. But before he could even get started, Googlers gave the entire senior team a long and thunderous standing ovation. The response from employees in China was of course very different. They feared for their jobs and even their security. Head of Engineering Alan Eustace, along with several dedicated team members in China, was instrumental in steering morale back on track, ensuring that the China team remained safe, engaged, and successful throughout that turbulent time. As a result, the legacy of the China decision was a giant dose of goodwill from Googlers around the world, and the legacy of the thoughtful process by which it was made was the reaffirmation of a set of principles governing how all tough decisions should be made.

Decide with data

One of the most transformative developments of the Internet Century is the ability to quantify almost any aspect of business. Decisions once based on subjective opinion and anecdotal evidence now rely primarily on data. Companies like ours aggregate anonymous signals from mobile phones to provide accurate traffic data in real time. London’s water pipes are monitored by thousands of sensors, reducing leakage by 25 percent.119 Ranchers embed sensors into their cattle that transmit information about the animals’ health and location; each cow transmits about 200 megabytes of data per year,120 allowing ranchers to fine-tune what, when, and how much they feed their cattle. That’s a cattle list for change!

John Dewey, an American philosopher and writer, said that “a problem well put is half solved.”121 In Dewey’s time, which spanned the latter half of the nineteenth century and first half of the twentieth, putting a problem well would usually entail an opinion and an anecdote. But as Berkeley political science professor Raymond Wolfinger once observed, “the plural of anecdote is data,”122 which means, by our interpretation, that if you don’t have data, you can’t decide. (Wolfinger went on to note that the singular of data is datum, then dismissed class early because he had a date.)

This is why most conference rooms at Google have two projectors. One of them is for videoconferencing with other offices or for projecting meeting notes. The other is for data. When discussing options and opinions, we start the meetings with data. We don’t seek to convince by saying “I think.” We convince by saying “Let me show you.”

A bias toward data is a great way to kill the death-by-PowerPoint syndrome. How many meetings have you been in where the first dozen or so slides are full of words, and the person stands up there and repeats the words? People who are presenting a point of view in a meeting shouldn’t need the crutch of slides to present that argument, only to support it. Slides should not be used to run a meeting or argue a point. They should just contain the data, so that everyone has the same facts. If the data is wrong or not relevant, you can’t fix it with fancy slides. Edward Tufte, the uber-guru of data presentation and visualization, advocates putting more data on fewer slides: “Visual reasoning usually works more effectively when relevant information is shown side by side. Often, the more intense the detail, the greater the clarity and understanding.”123

It should go without saying—but it usually doesn’t, so we’ll say it—that data is best understood by those closest to the issue, which is often not management. As a leader, it is best not to get lost in details you don’t understand, but rather trust the smart people who work for you to understand them. When making financial decisions, for example, don’t worry about the ABCs of the MBAs’ and CPAs’ EBITDAs, ADRs, and RPMs; focus on what matters, which is usually cash and revenue. (A frequent Eric aphorism during financial discussions: “Revenue solves all known problems.”) This applies to technical and product decisions as well. Eric was once in a meeting with the CEO of one of Google’s partners. The executives were debating some technical issues, and doing a rather poor job of it. So a young Googler who had been listening from the corner stepped up and presented several data points to clarify Google’s position. In a meeting crowded with impressive titles, this young woman with the least seniority was obviously the best-informed person in the room. She ultimately carried the day simply by having the best grasp of the facts.

Beware the bobblehead yes

You may be familiar with the bobblehead dolls baseball teams often give away at games—Jonathan has a bobblehead of San Francisco Giants catcher Buster Posey124 in his office. But you may not know that bobbleheads are also prevalent in conference rooms, where they sit around a table nodding their heads in almost rhythmic unison. Tim Armstrong, AOL CEO and former Googler, dubbed this phenomenon the “bobblehead yes.” (When Eric was CEO at Novell, he gave it his own name: the “Novell nod.”) Bobblehead yessers are different from your classic “yes-men” because, unlike them, bobbleheads have a nasty tendency to complain and whine and not do or support the very thing to which they just agreed as soon as they walk out of the meeting. This is something bobblehead Buster Posey would never do.

Getting everyone to say yes in a meeting doesn’t mean you have agreement, it means you have a bunch of bobbleheads. Many leaders strive for “consensus-driven” decisions, but they fundamentally misunderstand the meaning of consensus. For those of you who skipped Latin, it stems from the Latin cum, meaning “together with,” and sentire, meaning “to think or feel,” so it literally means “to think or feel together.” Note that this implies nothing about unanimity; consensus is not about getting everyone to agree. Instead, it’s about coming to the best idea for the company and rallying around it.

Reaching this best idea requires conflict. People need to disagree and debate their points in an open environment, because you won’t get buy-in until all the choices are debated openly. They’ll bobblehead nod, then leave the room and do what they want to do. So to achieve true consensus, you need dissent. If you are in charge, do not state your position at the outset of the process. The job is to make sure everyone’s voice is heard, regardless of their functional role, which is harder to achieve when the top dog puts a stake in the ground.

As General Patton famously said, “If everyone is thinking alike, then somebody isn’t thinking.”125 If you’ve hired well, there’s good news: There is dissension in the ranks. Lots of people are thinking. Smart creatives, especially at the most senior leadership level, should and usually do think of themselves as owners of the business, rather than leaders of just their particular area. Therefore they should have opinions, and quite possibly valuable insights, even about decisions that fall outside their realm. Encourage this, since it helps build a stronger bond among the team and stronger support for the ultimate decision.

Using data can be helpful to get everyone to weigh in, since it’s not personal.126 Be especially aware of the quiet people; call on the ones who haven’t spoken up yet. They may be dissenters who are afraid to disagree with you in public (but need to get over that fear), or they may be of the shy but brilliant type. Or perhaps they truly have nothing to say, in which case maybe they shouldn’t be at the meeting in the first place. One technique is to throw out a few “stupid softballs” that let people dip their toe in the water of disagreeing with the boss. (“I think we should all pour hydrochloric acid on ourselves. Thoughts?”) Do your best to surface all potential dissent early in the process; there is a natural (and valid) bias toward rejecting dissent the later it surfaces in the decision-making process.127

Once everyone weighs in with an opinion, then the argument will be on, and everyone can participate in the decision-making process and have their voice be heard. A proper consensus-driven process has elements of inclusion (involving all the stakeholders in a participatory manner); cooperation (aiming for the best decision for the group, sometimes at the expense of a minority or individual); and equality (everyone on the team counts and can at least temporarily engage in blocking behavior). Above all it is solution-oriented: The right decision is the best decision, not the lowest common denominator decision upon which everyone agrees. And it’s not always your solution. As Coach Wooden once said, “Be interested in finding the best way, not in having your own way.”128

Know when to ring the bell

This conflict-based approach works only if it is managed by a single decision-maker who owns the deadline and will break a tie. Often there is too much data, or the data is inconclusive. When that happens, people can debate for hours, a time sink that often ends in mediocre compromise and always incurs a hefty opportunity cost, since there are always better things for smart creatives to be doing than rehashing a decision for the umpteenth time. There is a point at which more analysis won’t lead to a better decision. This is the most important duty of the decision-maker: Set a deadline, run the process, and then enforce the deadline. It’s like the kids on the playground at recess; they will play forever, but when the bell rings they know they have to wrap it up and head back to class. (Hopefully employees are better behaved and less prone to hogging the monkey bars.) The decision-maker gets to decide how long recess lasts, then ring the bell.129

Our coach and mentor Bill Campbell told us a story of how he had just joined Intuit as its CEO when he heard of an important product decision that had stalled. The executive in charge of the product had gathered plenty of data, but the numbers were inconclusive. So he ordered up more research. Then, when the new set of data was still not helpful, he ponied up for yet another feeding at the data trough. Bill got wind of this and ordered a stop to the dillydallying. “Do something,” he told the executive, “even if it’s wrong.”

Tom Peters would call Bill’s attitude in this situation a “bias for action,” and his book In Search of Excellence lists it as a top common attribute of the companies he studied.130 Many designers also believe a bias for action is a positive force, nothing short of “a core… mindset of design thinking,” according to the Stanford design school (aka the, because “design school” isn’t designy enough).131 It promotes a hands-on, trial-and-error approach: If you’re not sure if a course of action is right, the best thing you can do is try it out and then correct course.132

But some behavioral economists believe that a bias for action can be deleterious, since it can favor hasty, poorly thought-out decisions, and in some situations we agree. In a negotiation, for example, Eric’s “PIA” rule can help get the best outcome: Have patience, information, and alternatives. P is especially important. You want to wait as long as possible before committing to a course of action. This is true in fields beyond business too (or should we say “pitches”): Soccer goalies who are facing penalty kicks can double their save rate by simply doing nothing at the moment the shooter kicks the ball, rather than following the common bias-for-action practice of guessing in advance to which side the kick will go and diving in that direction.133 In that way, goalies could learn from pilots, who are trained not to act instantly in emergencies but to take a moment to assess the situation before deciding what to do.

The job of the decision-maker, then, is to get the timing just right. Exhibit a bias for action, to cut off debate and analysis that is no longer valuable, and start moving the team to rally around the decision. But don’t be a slave to a sense of urgency. Maintain flexibility until the last possible moment.

Make fewer decisions

When Eric joined Google, he was well aware of the not-so-good history of CEOs being hired by founders into their companies. Typically, the founder hires the CEO, eventually they disagree on something fundamental, the board backs one of them, and the other leaves. Steve Jobs’s hiring of John Sculley, a Pepsi executive, to succeed him as CEO at Apple in 1983 is the classic example. The two clashed and Sculley (backed by the board) fired Steve in 1985.134

To avoid a similar fate when he joined Google, Eric decided he would let Larry and Sergey do what they did best and he would focus more on the stuff needed to build the company at such an incredible pace, so it could continue to operate effectively and efficiently. The scenario of having a ruling triumvirate was so unique that Larry and Sergey described it in some detail in the letter that accompanied Google’s IPO in 2004. In fact, codifying the who-does-what working process of the trio was very helpful. The letter stated that Eric “focuses on management of our vice presidents and the sales organization. Sergey focuses on engineering and business deals. [Larry focuses] on engineering and product management,” and that the three leaders were meeting daily (which continued throughout most of Eric’s stint as CEO). Most important, it said that the arrangement “works because we have tremendous trust and respect for each other and we generally think alike.”

This all worked very well as long as the three agreed on key issues, which was most of the time. But it did occasionally lead to some difficult situations; when you have three strong-willed leaders, they will sometimes disagree. When that occurred, Eric’s process to get to a good resolution was similar to his general decision-making process: Identify the issue, have the argument (alone, just the three of them), and set a deadline. And he often added a corollary: Let the founders decide.

The tendency of a CEO, and particularly (speaking from experience) of a new CEO trying to make an impact in a founder-led company, is to try to make too big an impact. It is hard to check that CEO ego at the door and let others make decisions, but that is precisely what needs to be done. In general, when you are CEO you should actually make very few decisions. Product launches, acquisitions, public policy issues—these are all decisions that CEOs should make or heavily influence. But there are many other issues where it is OK to let other leaders in the company decide, and intervene only when you know they are making a very bad call. So a key skill to develop as the CEO or senior leader in a company is to know which decisions to make and which to let run their course without you.

This skill is even more important when you find yourself in the situation Eric did, running a company in the presence of two very active, respected, smart founders. For example, there was one product review meeting where Eric, Sergey, and Larry ended up disagreeing about a key feature of a new product. There were about twenty people in the meeting, and after a few minutes Eric suspended the argument and then resumed it later that afternoon with just the three of them. It was there he discovered that the two founders not only disagreed with him, but with each other as well. So Eric said fine, he would let the two of them decide, but they had to decide by the next day. When he dropped by the office they shared in building 43 the next day at noon, he asked them, “Which one of you won?” And the response was typical: “Actually, we came up with a new idea.” It turned out to be the best solution, and the decision was made.

Meet every day

One of the frustrating aspects of being a leader of smart creatives is how little power you actually have. Look at this chapter so far. Even if you are the CEO of a company, it says, you can’t just pound your fist on a table and dictate decisions (well, you can, but if that’s your modus operandi you will quickly lose most of your smart creatives), and in fact you shouldn’t even make many decisions. Instead, you have to analyze data and orchestrate consensus by encouraging debate and then knowing, through some divine skill, exactly the right time to cut off that debate and make the decision. Sort of makes you yearn for those days a long time ago, when Darth Vader could unilaterally crush someone’s throat with the power of the Force and then destroy a planet.

But there is one thing that leaders can still control, and that is the company’s calendar. When faced with a critical decision, there is real signaling value in using your convening power as a leader to hold regular meetings. If the decision is important enough, the meetings should be daily. Scheduling meetings with this frequency lets everyone know the importance of the decision at hand. And there is another simple benefit: When you have daily meetings, you spend less time in each meeting rehashing things that were discussed at the previous meeting, since everyone’s memory is still fresh. That leaves more time to consider new data or opinions.

Eric used this approach to good effect in 2002, when Google was negotiating a deal with AOL to be the popular portal’s search and ads engine. It was a difficult negotiation, and Eric was particularly concerned about the financial commitment Google was potentially taking on. AOL had a number of advertisers on their platform that were not as yet advertising with Google, so the deal had tremendous strategic value: It would bring those advertisers to our platform. Nevertheless, Eric felt the commitment was too big for a small company like ours to take.

Omid Kordestani, our head of sales, led the negotiations with AOL, which had merged with Time Warner in early 2001 and was eager for the revenue this deal would bring it. Omid agreed with Eric that we shouldn’t accept AOL’s terms. But Larry and Sergey wanted to take the risk; they had always felt that being aggressively generous with partners on revenue share would ultimately benefit the company (“If it doesn’t bankrupt us first,” Eric thought when they expressed this point). David Drummond, the company’s lead attorney, agreed with them, as did the board of directors, who felt we could always borrow to cover any cash shortfalls. There was an honest disagreement, and the team wasn’t making a lot of progress in its meetings. So Eric acted. He set up even more meetings, and he set a deadline. For the next six weeks, the team would get together every day at four p.m. to review the AOL deal. By the end of that time period, they would come to a decision and conclude the negotiations with AOL, one way or another.

At first they didn’t make much progress. But the sheer drudgery of repeating the same argument every day helped spur the team to delve even deeper into the data we had on how our ads engine was performing, and over the weeks we performed analyses that demonstrated that the deal wasn’t as risky as we had originally thought. We started to realize that we could afford it, and we were right. We did the deal, basically on AOL’s terms, and our performance exceeded all of the guarantees. But no one knew this at the time we were in negotiations; we got to the right answer through a rigorous, time-intensive process of considering all the details. It was a critical decision, and when you are considering something that is fundamental to the existence of the company, you should meet every day.

“You’re both right”

There is a mistake technical and scientific people make. We think that if we have made a clever and thoughtful argument, based on data and smart analysis, then people will change their minds. This isn’t true. If you want to change people’s behavior, you need to touch their hearts, not just win the argument. We call this the Oprah Winfrey rule. (It’s also the way that good politicians operate, but Oprah does it better than anyone.)135 When companies are run by smart creatives and product people, they need to learn the Oprah rule. Otherwise they are apt to make smart decisions, but to fail to execute them well.

There is a simple trick to getting this right. When ending a debate and making a decision that doesn’t have 100 percent support, remember these three words: “You’re both right.” To emotionally commit to a decision with which they don’t agree, people have to know that their opinion was not only heard, but valued. “You’re both right” accomplishes this. It tells the person whose argument lost that there are elements of truth amidst the rubble of their failed position. It provides an emotional boost—people like hearing that they are right. And fortunately, it is often true, since in a group of smart creatives there are usually elements of truth in everyone’s position. It’s rare for a good person to be completely, 100 percent wrong.

Then, after reassuring the argument’s losers and articulating what needs to be done, the decision-maker must ensure that everyone who was involved does one of two things: disagree but commit, or escalate publicly. If it’s the latter, then the escalator must let the decision-maker know the reasons for her objections, and how and to which higher-up she plans to escalate. (“I’m sorry, I still don’t think this is the right decision because of.… How about we see what Barack thinks?”) Public escalation is a valid option and should be encouraged, because if you don’t it will just happen anyway, only with a lot more rancor.

Every meeting needs an owner

The forum for decision-making is almost always a meeting, which may be the most hated of all business practices, except for Secret Santa. People complain about meetings and how they are a great waste of time, but in fact a well-run meeting is a great thing. It’s the most efficient way to present data and opinions, to debate issues, and yes, to actually make decisions. Note the italics on well-run though, since most meetings are anything but. A badly run meeting—we probably don’t need to tell you this—is a giant, demoralizing time waster.

Computer scientists hate inefficiency, so over the years Eric’s team developed a series of rules for meetings that we found to be quite effective:

Meetings should have a single decision-maker/owner. There must be a clear decision-maker at every point in the process, someone whose butt is on the line. A meeting between two groups of equals often doesn’t result in a good outcome, because you end up compromising rather than making the best tough decisions. Include someone more senior as the decision-maker.

The decision-maker should be hands on. He or she should call the meeting, ensure that the content is good, set the objectives, determine the participants, and share the agenda (if possible) at least twenty-four hours in advance. After the meeting, the decision-maker (and no one else) should summarize decisions taken and action items by email to at least every participant—as well as any others who need to know—within forty-eight hours.

Even if a meeting is not a decision-making meeting—for example it’s designed to share information or brainstorm solutions—it should have a clear owner. Again, that owner should ensure that the right people are invited to the meeting, that there’s a clear agenda, that the necessary prep work has been done in advance, and that action items are circulated promptly.

Meetings are not like government agencies—they should be easy to kill. Any meeting should have a purpose, and if that purpose isn’t well defined or if the meeting fails to achieve that purpose, maybe the meeting should go away. The decision-maker needs to ask the hard questions: Is the meeting still useful? Is it too frequent / not frequent enough? Do people get the information they need?

Meetings should be manageable in size. No more then eight people, ten at a stretch (but we would seriously discourage this). Everyone in the room should be able to give their input. If more people need to know the result of the meeting, make sure you have a process for communicating it rather than bringing them in as observers, which lowers the quality of the meeting and people’s ability to talk openly.

Attendance at meetings is not a badge of importance. If you aren’t needed, leave, or better yet, excuse yourself ahead of time. This is especially true of meetings with customers or partners. Many times we have walked into an “intimate” meeting with a senior executive from one of our customers or partners, only to find the room full of people. We can’t help it if customers feel the need to bring their entire org chart to the meeting, but we try to control our side. Fewer people is almost always better.

Timekeeping matters. Begin meetings on time. End them on time. Leave enough time at the end to summarize findings and action items. If the meeting has accomplished its goal before its allotted time runs out, then end it early. Remember, we are human: Schedule time for lunch and bio breaks, and be respectful of employees working in different time zones. They like to spend time with their families too. These common courtesies get forgotten too often. Paying attention to them will earn the respect of employees and colleagues.

If you attend a meeting, attend the meeting. Multitasking doesn’t work. If you are in a meeting and using your laptop or phone for something not related to the meeting, it’s obvious your time is better spent elsewhere. Everyone attending a meeting should focus on the meeting, not other work. And if people have so many meetings that they can’t get work done, then there is a simple solution: Prioritize and go to fewer meetings.

Among all of these rules, this last one has been the most challenging for us to implement. In our own team meetings, people so often ignored our edicts to close their laptops that we had to give up. But it’s still a good rule!

Horseback law

Lawyers are, by training, backward looking. This makes sense, since so much of the law is determined by precedent: What happened before dictates what is OK going forward. They are also highly risk-averse. This also makes sense, because so many business lawyers practice in law firms and the job of a corporate law firm is to keep its clients out of trouble. So when you ask most lawyers to assess a situation, and if that situation is 99 percent good and 1 percent questionable, they will spend most of their time with you reviewing the questionable.

This sign is a good example. Jonathan snapped a picture of it one day when he walked across the street to check out the athletic fields that Google had just opened. The sign includes a nice map of the fields, but a quarter of its space is taken up by a legal disclaimer that basically says if you get hurt using these fields, don’t sue us. (Some lawyer reading this is about to correct our interpretation of the carefully worded legal phrasing. Please stop.) A well-meaning, backward-looking, risk-averse lawyer decided that even though the Googlers using the fields are intelligent grown-ups, there was still an infinitesimal chance that one of them would step on the field, twist an ankle, and sue Google. Hence, yet more brain-dead obvious legalese clouding our landscape.


Lawyers can be smart creatives too, which is why we were so surprised to see this sign at Google. The backward-looking, risk-averse approach to the law, which is so common in corporate America, doesn’t work in the Internet Century, when business evolves at a pace that is several orders of magnitude faster than the pace of legal change. A smart creative–fueled business that is trying to innovate will be lucky to be right 50 percent of the time, which can be a problem for a lawyer whose risk tolerance is in the single digits.

This is why, when they were building Google’s legal department, David Drummond and his colleagues Kulpreet Rana and Miriam Rivera set out to create an environment where lawyers approached their jobs differently. Our current general counsel, Kent Walker, likes to call this approach “horseback law.” Take a look at any old Western movie (we like Butch Cassidy and the Sundance Kid; Eric is Butch because thinkin’ is what he’s good at, and Jonathan is Sundance because he’s fast on the draw—but unfortunately not as accurate as the Redford-played bank robber). There is always the scene where a cowboy rides up on his horse and comes to a stop, surveying the situation and deciding what to do next. Kent advises his lawyers to do something similar: In certain situations, it’s often enough to ride up on a horse (figuratively speaking, usually), make a quick assessment, then mosey on. While many decisions (e.g., a major acquisition, a legal compliance question) may call for detailed analysis, don’t feel that you always have to dismount and spend weeks writing a fifty-page legal brief (ha!) of all the things that could possibly go wrong and what would happen if they did. In the early stages of a new project, the analysis won’t be 100 percent correct anyway. In those situations, it isn’t the lawyer’s job to cover every possible angle in detail; it’s his job to look into an unforeseeable future and provide educated, quick guidance to the business leaders making the decisions. Then saddle back up, pardner.

Horseback law works only if the lawyer is an integral part of the business and product teams, rather than just summoned occasionally. It works only with the right mixture of lawyers, which is why, in our early days, we tried to hire more generalists than specialists and spread our recruiting efforts across firms, businesses, and even nonprofits (but we rarely hired lawyers straight out of school). And since legal issues are bound to crop up when you are moving quickly and changing industries, it always helps to be doing the right thing by consumers and customers.

Spend 80 percent of your time on 80 percent of your revenue

One of the most important decisions any business leader makes is how to spend his or her time. When Eric became CEO of Novell in 1997, he got some great advice from Bill Gates: Spend 80 percent of your time on 80 percent of your revenue. But this rule can be deceptively hard to actually follow. At Novell, the company’s core business was the NetWare software suite, which enabled local area networking between PCs and workstations. Eric and his staff, though, were excited about growing a new product (NetWare Directory Services, or NDS) that provided a central point of management and access to network resources ranging from people and groups to printers and workstations. NDS clearly had big growth potential as networks proliferated, and it was hard for Eric and the team to resist spending more time on it.

Leadership teams often underestimate how long it takes for revenue from a new product area to ramp up. That shiny new stuff can be much more interesting than the boring old core business stuff, but it’s the core stuff that pays the bills, and if you make a mistake there, you probably won’t be able to recover. Even though Eric thought he was heeding Bill’s advice, in retrospect he should have been spending more time on NetWare.

You have to focus on your core business. You have to love it.

Have a succession plan

Loving a business means having a plan for leaving it, but leaders often neglect to think about who will succeed them. In most companies, your successor is already there, you just haven’t figured out who it is. (Eric’s experience, where his successor was the person who hired him, is rare!) Many companies get the idea right, but their timing is off: They identify the brother, someone ready to take over in the next few years, whereas they should be looking for the son, someone who could take over in a decade. Or they try to lock in the hundred most senior people in the company, not the hundred with the highest potential. The right approach is to look for the outstanding smart creatives who are already progressing rapidly through the ranks. Ask the question, Could one of these people be running the company in ten years? When the answer is yes, give them a lot of compensation and make sure their career doesn’t bog down. Losing these high-potential employees (especially to competitors) is very costly to the company, so be proactive and aggressive in your efforts to keep them happy. It may not always work out, but the benefits of the successes far outweigh the misses.

Then there’s the interesting experience of actually executing the succession plan. Those rising superstars tend to get smarter with the passage of time, but the generation at the top still looks at them as brash and inexperienced, and certainly not wise enough to take over. The solution for this is for the leader to remember what he was like back in the day.

When Google was preparing for its IPO, Eric, Larry, and Sergey made a commitment to each other that the three of them would work together for at least the next twenty years. Eric had always assumed that either Larry or Sergey would end up running the company—probably Larry, as he had previously been CEO. It was just a question of when. That time came in early 2011, when Eric, Larry, and Sergey decided that Larry would resume as Google’s CEO. It was the right decision for the company and the trio, but still, Eric was a little uncertain. After all, he was so much older and wiser! But then Eric mapped his age onto Larry’s: At the time, Larry was nearly thirty-eight years old, and when Eric was that age he felt he was ready to run a company (he was forty-one when he took over the reins at Novell). It was a bit of a surprise, but by following this thought process Eric realized that Larry was quite ready, and would be very successful as Google’s CEO.

The World’s Best Athletes Need Coaches, and You Don’t?

In the summer of 2002, when Eric had been on the job as Google CEO for about a year, he wrote a self-review of his performance and shared it with his team. The document included highlights (“developed proper business processes”), objectives for the next year (“run the clock faster without compromising the future”), and areas where he could have performed better. The last category included several points, but one self-critique stands out as the most important:

Bill Campbell has been very helpful in coaching all of us. In hindsight, his role was needed from the beginning. I should have encouraged this structure sooner, ideally the moment I started at Google.

This was a 180-degree turnaround from a year earlier: When Eric started at Google, board member John Doerr suggested that he work with Bill as his coach. Eric’s reply? “I don’t need a coach. I know what I’m doing.”

Whenever you watch a world-class athlete perform, you can be sure that there is a great coach behind her success. It’s not that the coach is better at playing the sport than the player, in fact that is almost never the case. But the coaches have a different skill: They can observe players in action and tell them how to be better. So why is it that in the business world coaches are so unusual? Are we all like Eric when he started at Google, so confident of ourselves that we can’t imagine someone helping us to be better? If so, this is a fallacy. As a business leader, you need a coach.

The first ingredient of a successful coaching relationship is a student who is willing to listen and learn. Just like there are hard-to-coach athletes, there are hard-to-coach executives. But once they get past that initial reticence, they find there are always things to learn. Business coaches, like all coaches, are at heart teachers, and Bill Campbell, the best coach around, tells us he believes that management is a skill that is completely learnable.

For Jonathan, class began right around the time when Larry Page was calling the regimented product plan that he created “stupid.” The following week, Jonathan was sitting in Coach Campbell’s office, wondering why he had ever joined this chaotic start-up and contemplating quitting. Don’t quit, Bill implored him. Stick it out. Maybe you’ll even learn something.

For that, and everything else you have done for us, thank you, Coach.

Book Title Page