How google works
By Eric Schmidt, Jonathan Rosenberg, Alan Eagle
- Eric Schmidt, ex-CEO of Google, current Executive Chairman, Alphabet
- Jonathan Rosenberg, ex-head of products at Google; current advisor to the CEO, Alphabet
- Alan Eagle, director of executive communications at Google
In 2010 we decided to write a book about the changing rules of management in the 21st century, based on the humbling and exhilarating experience of helping to build Google over the previous 10 years. In September, 2014, we published How Google Works.
As we were working on the project, we collected all of our stories, principles, and sayings, and started to write a manuscript. The material was great, but when Eric asked the questions, “What’s the book about? Why do people need to read it, now?” we didn’t have the answers.
So we went over all of our notes and ideas and boiled them down to the bare essentials. When you strip away our stories and words, what you have left is our ideas - 50 of them. This is what How Google Works is about.
Let’s start by asking ourselves, what’s different now?
(1) Three converging technology platforms - Internet, mobile, and cloud - are transforming virtually every business sector. The Internet provides all of the world’s information and media, and massive amounts of data, to a global audience, mostly for free. Mobile devices and networks make global reach and continuous connectivity available to every person on the planet. And cloud computing puts practically infinite computing power, machine learning capabilities, and a host of sophisticated applications at everyone’s disposal, on a simple, pay as-you-go basis.
As a result, important resources that were previously scarce (information, data, connectivity, computing) are now abundant and cheap. Incumbents in many industries have built barriers to entry based on the scarcity of these resources. Now those barriers are going away, making many incumbents highly vulnerable.
(2) This transformation is happening at an unprecedented speed, and it is accelerating. General purpose technologies such as the steam engine, electricity, and the computer historically have taken several decades to go from invention to massive impact. It took about 150 years from the steam engine to move from a little device pumping water out of mines to driving the industrial revolution and the settling of the American West. Modern computers were invented in the 1940s, but didn’t scale broadly and globally until the PC era of the 80s and 90s.
But today’s most important general purpose technologies, the worldwide web and smartphones, are different. Their growth is driven by the principles of Moore’s Law, which states that processing power of computer chips will double roughly every 18 months and has held true for nearly 50 years. Similar exponential improvements are also occurring in technologies such as storage and bandwidth. The worldwide web was only invented in the 80s and early 90s (with the advent of TCP/IP, HTML, and browsers), while smartphones were invented in the late 90s and popularized with the launch of the iPhone in 2007. Both were causing massive upheaval in businesses such as media, retail, and communications, within a decade.
(3) The tools of innovation have been democratized. Access to these technology platforms has been democratized: the masses can build new inventions and systems on them for free or for very reasonable prices. Open source software and application programming interfaces (APIs) lets engineers build on each others’ work, most notably in the area of machine learning. Robust protocols and standardized programming languages allow for fast development. Large data sets on things such as traffic, weather, economic transactions, human genetics and business and product ratings are available for analysis. These raw tools of innovation are widely available and cheap. This gives people plenty of opportunities to solve big, old problems in clever new ways that were simply impossible only a few years ago.
In economic terms we are approaching “perfect information” (search, the web) and more “frictionless markets with near zero transaction costs” (reputable vendors, low shipping and delivery costs). So as clever people use the tools of innovation to invent new things, it is easier to get them to market and into the hands of customers.
(4) The balance of power has shifted from companies to customers, and expectations are skyrocketing. Consumers have never been better informed or had more choice. Anyone considering a purchase can get price comparisons, reviews, specifications, photos, videos, product availability, and recommendations from friends and experts, often via their phones while standing in a store aisle. Shelf space is infinite online, so people no longer have to wait to get what they want (same day shipping will be the norm for online shopping within a few years) or be forced to choose a not-quite-right product because the perfect alternative is out of stock.
When customers are completely empowered, the only certain way for businesses to achieve success is by consistently creating superior products, services, and experiences. Companies will no longer be able to power bad products to success by controlling distribution and shelf space or with clever marketing.
(5) Individuals and small teams can have a huge impact.
Technology is also transforming the workplace, by putting near infinite amounts of information, data, and computing power in every employee’s hands. Individual employees and small teams can create products that can reach the world in an instant. They can move quickly too: It is easy to prototype new ideas and try them out, so the cost of experimentation (and failure) has dropped dramatically. Lengthy, costly R&D cycles are being replaced by a much faster, more iterative and flexible product development processes. Fast-moving employees can create massive value in short periods of time.
(6) Product people are more valuable than business people. The most valuable employees today are the ones who combine technical knowledge (engineers, scientists, designers), business expertise (executives, MBAs), and creativity (creatives). We call these people “smart creatives”. Armed with abundant data, access to customers, and powerful analytical tools and machine learning platforms, these “smart creatives” have everything they need to design, build, and improve products rapidly. They also can thrive in the highly dynamic, uncertain conditions that characterize more and more industries.
(7) Most companies today are designed and run to minimize risk, not maximize speed. Most companies are organized and run according to management principles designed in a time when information and technology were expensive and the cost of experimentation high. These principles were designed to funnel information into the hands of a few executive decision-makers, an approach that deliberately slows things down in order to minimize risk. As a result, in most companies, disparate skill sets are segregated - engineers, creatives, and business people are in different departments - and management has plenty of opportunities to check what is going on and intercede if there are problems. So in a time when the speed of business transformation is accelerating, and moving quickly is a vital component of a healthy product development process, most companies are deliberately designed to slow things down so that costly mistakes can be avoided.
(8) The only way for businesses to consistently succeed is to attract the best smart creatives and create an environment where they can thrive at scale. This requires a very different approach to management than what is traditionally practiced today. Business leaders must cede a high degree of control to employees at all levels and give the best ones the chance to thrive without limits, while maintaining a structure that scales globally. Management’s job is to stay out of the way, to limit the total cost of errors but not to limit the total number of errors. Prevent the $10,000,000 mistake, not the $100,000 mistake.
Ok, so the key to success is to attract smart creatives and create an environment where they thrive at scale.
How, exactly, do you do that?
First, you have to get them to want to work with you, and that starts with your company culture. Here are some ideas about that.
(9) Define your culture early. Once it is set selection bias sets in and it becomes very difficult to change. This has been proven again and again by start-ups, whose culture (and success) is often derived directly from the personalities of the founders. This is why founder-led companies often stay successful, while non-founder led companies can stagnate. To emulate this level of success, actively plan the company’s culture at the outset; it is as important as a clever product or strong strategic foundation. Then live that culture - lead by example, not phrases - and communicate it frequently and authentically (no “messaging”).
(10) Build around small teams. Small teams can get more done than big ones, and because people on those teams are so closely aligned with each other’s interests, they tend to spend less time politicking and worrying about who gets credit. Within small teams, social norms inspire higher performance and keep poor behavior in check; there’s nowhere to hide.
(11) Keep them crowded. Offices should be designed to maximize energy and interactions, not peace and quiet. When you can reach out and tap someone on the shoulder, there is nothing to get in the way of communications and the flow of ideas.
(12) Working from home kills innovation. Make sure your offices have everything employees need to be happy, healthy, and productive. Then expect them to work there.
(13) Listen to the best ideas, not the hippos. A hippo is the highest paid person’s opinion. But pay level is intrinsically irrelevant and experience is valuable only if it is used to frame a winning argument. Unfortunately, in most companies experience - the hippo - is the winning argument. Strive to create a meritocracy where the smartest arguments carry the day.
One way to control hippos is to keep them busy with a lot of direct reports, The “rule of 7” stipulates that a manager have at least 7 people working for them, which prevents most would-be hippos from micromanaging their staff. Management’s job is to stay out the way of the smart creatives with the best ideas. The rule of 7 helps accomplish this.
(14) Organize the company around the people whose impact is the highest. There is no perfect organizational design, ever. So the best approach is to put aside preconceived notions about how the company should be organized, determine which people are having (or deserve to have) the biggest impact, and organize around them. How do you discover who they are? Performance and passion. Performance can be measured by numerous methods, while passion comes down to a more subtle point. The most passionate people are the ones who are going to do things not because they are told, or paid, or given permission, but because they think it is the right thing to do.
(15) Do all re-orgs in a day. There is no perfect organizational design, so when doing a re-org there’s no need to try to achieve perfection. Do all re-orgs as quickly as possible, solving the most important issues and communicating the most important principles, and let the team figure out the details.
(16) Exile knaves but fight for divas. Knaves are people who exhibit low integrity (they will do whatever it takes to get ahead), while divas exhibit high exceptionalism (they are better than everyone else). Always be firm with knaves, those people who violate the basic interests of the company. But as long as their contributions match or exceed their egos, accommodate divas. Great people are often difficult in many ways, and cultural factors can conspire to sweep them out. Fight that -- let divas be divas, and they will produce great things.
(17) Overwork people in a good way. When you identify the highest impact people, give them more to do, but also give them the freedom to manage their own lives and achieve their own balance. The right work-life balance equation for the best people is one that’s ridiculously over-laden on both sides.
(18) Cultivate fun, even if (especially if) it comes at management’s expense. Like anything innovative, a fun culture cannot be ordained or planned, it needs to be supported and allowed to grow organically. Structured, corporate fun events usually aren’t very fun, but give smart creatives enough freedom and they will create plenty of fun on their own. Often this will come at management’s expense, which is great. Humor is a very healthy way for organizations to surface and discuss dissent. Open dissent is a hallmark of a healthy organization, and allowing it to happen through humor and satire is often better than other modes.
But great people need more than a great culture. They need to know that the business is built for long-term success. They need a smart strategy.
(19) Businesses should be built on a strong set of strategic principles, not an MBA-style plan. Traditional style business plans are bound to be obsolete within weeks or months. Companies that have solid foundations but lack full scale business plans can give employees the freedom to build on the foundation, without being constricted by the plan. This is precisely the type of freedom that the best smart creatives seek and crave.
(20) Superior products start with unique technical insights. Great new products usually start by applying one or more technology, design, or business insights in such a way as to significantly shift the cost-performance curve of a product. History is full of companies achieving success by applying technical insights and Google’s own history is full of product successes based on unique technical insights (and failures lacking them): from Search, AdWords, and Gmail, to Translate, Knowledge Graph, and Self-driving Cars. Products that aren’t based on unique technical insights will only achieve incremental success at best. No great technical insight has ever come from market research. The lab coats are more important than the suits.
(21) Optimize for growth. The most successful businesses will be the ones who understand how to create and grow platforms. Competition is more intense than ever and most competitive advantages are fleeting, so you have to have a “grow big, fast” strategy. An effective way to do this is to default to open.
(22) Know your competition, but don’t copy it. Copying competition leads to only incremental improvement. It can shift market share, but doesn’t grow the market. While you are busy fighting over share points, someone else will come in and create a new platform for the entire industry.
(23) Ask the hardest questions. For leaders who are trying to initiate new ventures within existing enterprises, the biggest challenge is often gaining a realistic perspective on the industry and the company’s positioning. One simple way to address this issue, and take the first step to establishing a fresh strategic foundation, is to ask the hardest question you can. E.g. When Eric was at Sun, it was clear that the Wintel PC was on a price / performance curve that would soon surpass Sun workstations, so the hardest question would have been what to do about that.
Once you ask the hardest question, use that challenge to assemble a few smart creatives from the company to address it. How can you use the principles of technical insights and scaling through open platforms to develop a new path?
And of course, it isn’t enough to attract great people. You need to get them on board.
(24) Hiring is the most important thing you do. Hiring defines a company and creates a network effect: great people attract more great people, while good people attract only good and mediocre people. Business people at all levels do not invest enough time hiring; they talk about how important it is, but then let someone else do it. For incumbents, it can be very hard to change the culture from within, but relatively easy to change hiring practices and shift the talent and culture of incoming employees.
(25) Interviewing is the most important skill you can develop. It is very difficult to do well, so do it as often as possible, regardless of your level. Recognize and reward your best and most prolific interviewers.
(26) Hire learning animals, not specialists. In dynamic industries conditions change frequently, so experience and ability to perform a particular role is not as important as the factors that define a smart creative: technical knowledge, business expertise, and creativity. When considering candidates for a role, favor the ones with a track record of learning new things over the ones with a track record in that particular role. The learners will successfully adapt to new roles, but role specialists will not.
The exception to this rule is when you are developing products that require deep and unusual technical expertise. There are some technical and scientific areas where you need people with PhDs and post-docs. But very few business areas.
(27) Check your biases at the door. Great talent often doesn’t look and act like you. Homogeneity in an organization breeds failure, so when building a team bring in as wide a variety of viewpoints and backgrounds as possible.
(28) Make hiring and promotion decisions by committee, based on data. In dynamic industries roles and organization charts are always in flux, and new hires will likely change managers several times during their career with the company. Hiring and promotion decisions are therefore too important to entrust to a single manager. Instead, these decisions should be made by committee, with several different levels represented on each committee. Hiring and promotion committees remove much of the politicking from these processes (friends can’t just hire friends).
A committee-driven process only works if each committee member is given the same, comprehensive set of information. Every hiring and promotion decision needs to be accompanied by this data packet, so that all committee members have the same information and decisions can be based on data and not emotion. This is a labor-intensive process, but it is necessary to ensure high quality hiring even as the company scales.
(29) Reward top performers disproportionately. Smart creatives are highly leveragable and have an inordinately big impact, so reward the great ones disproportionately and give them more responsibility.
(30) Force the best people to rotate. The best people are usually the ones to get bored, and bored people leave. Meanwhile, the formation of silos in organizations is death to innovation and energy. Fortunately, there is one answer to both of these problems. Create rotation programs, and force the best people to use them. This guarantees a steady stream of talent moving around the company, building relationships and learning new skills and expertise. And it keeps the talent challenged and energized. Beware: most managers won’t want their best people to rotate.
Nice work! You have attracted and hired a great crew of smart creatives and they are eager to start changing the world.
Now you need to give them an environment where they can succeed at scale. That starts with your approach to making decisions. Decision-making done right lets smart creatives know that they can make a difference. Done wrong, it kills their spirit.
(31) Consensus isn’t about getting everyone to agree; it’s about making the best decision for the company and getting everyone to rally about it. This requires dissent, and for everyone’s voice to be heard (especially those who don’t always speak up). People need to disagree and debate their points in an open environment, otherwise there won’t be full buy-in on the ultimate decision. Smart creatives don’t always need to carry the day, but they do need to be heard.
(32) Make all decisions with data. It is much easier to get accurate and comprehensive data now than it used to be, so decisions once based on subjective opinion and anecdotal evidence can now be based on data. That data is usually understood best by the people closest to the issues, who don’t seek to convince by saying “I think,” but rather by saying “Let me show you.” Presentations shouldn’t be used to run a meeting or argue a point. They should contain the data, so that everyone in the meeting has the same facts.
(33) Meet every day. Leaders can’t dictate decisions, but they can control the calendar. So for the most important decisions, meet every day. Setting daily meetings lets everyone know how important that particular issue is, and eliminates the need to re-hash material from the previous meeting, so the time can be spent working the issue.
(34) You’re both right. To gain true consensus on complex issues, you need to touch people’s hearts, not just win the argument. To get everyone on board with a decision, the people who disagreed with all or part of it need to feel that their opinions and data were heard and considered. Telling them that they were right accomplishes that, and it is usually true since it is rare to have the smart people working on an issue be completely wrong.
(35) Every meeting needs an owner. Meetings should have a single decision-maker who owns the meeting and drives the process. This person, preferably the most senior person in the room, needs to be hands on, organizing the meeting, setting the agenda, determining the participants, and quickly producing the minutes. Meetings should be small and easy to kill, and if you aren’t needed then consider not attending (which is easier when the leader is producing good minutes).
(36) Everyone needs a coach. The world’s best athletes all have coaches, but most business people think they don’t need them. They are wrong. Coaches make individual execs better, and they make their teams better too.
Communication is just as important as decision-making, and it’s something that most leaders think they are good at. They are mostly wrong.
(37) Default to open. The job of leaders is to maximize the velocity and volume of information flow across the company and to guide it to the right people. Most companies are organizationally and culturally structured to hoard information into silos; information guarding becomes the norm as companies grow. The only way to combat this is to aggressively promote transparency and open communication. It needs to be actively practiced by upper management. Otherwise, the natural tendency is towards opacity.
(38) Be an effective router. A leader’s job isn’t to hoard information, it is to route it as quickly and efficiently as possible. This means relentlessly staying on top of your inbox (since email is the principal form of communications inside of businesses) and being thoughtful about who should see what. When you thoughtlessly send a long email to a big distribution list, you are wasting a lot of valuable people’s time.
(39) Start conversations. Conversations are still the richest, most valuable form of communication, but in today’s technology-driven work environment they are often the most neglected. Leaders need to create opportunities for genuine conversations, and shouldn’t limit themselves to topics related directly to work.
(40) Repetition doesn’t spoil the prayer. When in doubt, over-communicate. It is a leader’s job to cut through layers and communicate critical messages. Smart creatives require context as well as prescriptions (why to do something, not just what to do). Leaders communicate this in a variety of ways: speeches and talks, meetings, and email are the most prominent. The overall objective is to over-communicate a consistent set of values and messages.
These communications need to be authentic and at times personal. They can be developed in partnership with communications professionals, but can’t be 100% outsourced. Smart creatives will easily spot and discount insincere, “corporate” communications.
(41) Value relationships over process and hierarchy. In a successful business the business outruns process, so chaos is the normal state. (If you don’t have chaos, something is wrong.) In chaos, you depend on relationships as much as process to get things done, so the best managers invest time in developing relationships across all levels and functions of the company. Your best insights will usually come from these relationships, and not from formal processes.
When you have done all these things right, when you have hired a great team of smart creatives and established good decision-making and communications practices, you are on the verge of achieving that most sacred of business buzzwords: innovation.
On the verge, but not there yet.
(42) The CEO should be the CIO (Chief Innovation Officer). Innovation can’t be ordained and owned from the top or managed via committees and processes. Innovative people do not need to be told to be innovative, they need to be allowed to do it. The Chief Innovation Officer position usually fails because it doesn’t have enough power to create an innovative culture and environment. The CEO is the only person with this power.
(43) Think 10X, not 10%. Global scale is available to just about everyone. But too many people are stuck in the old, limited mindset. Thinking big gives people much more freedom, since it pushes them to remove constraints and spurs ideas that were previously not considered. And it is a powerful tactic to attract and retain the very best people, who are usually drawn to the biggest challenges.
(44) Focus on the user, and the money will follow. The primary objective of product teams is to create new, surprising, radically better products. Do this, and any smart company will figure out how to make money from it. Products that are highly valuable and differentiated on a scalable basis will generate great revenue.
(45) Empower smart creatives over managers. Ideas come from anywhere, so give all smart creatives the resources and freedom to work on what they want, without interference from an imperial manager. Google’s famous 20% time policy, which allows employees to spend 20% of their time working on projects of their choosing, is a great example of this. In practice, the time that employees invest on these projects is usually their personal time, but they are able to use company resources as they develop new ideas, and no one can tell them no. The point of 20% time isn’t the hours of the day that people can commit to their own ideas, it’s the freedom to work on those ideas at all.
(46) Set unattainable goals and share them widely. Every team and employee should set audacious goals and share them publicly within the company. These goals should be so big that achieving 100% of them is considered a failure of small thinking. Model this behavior at the most senior level: company leaders should set big goals every quarter, and then regularly explain to the team why they didn’t achieve 100% and how they are going to adjust.
(47) Optimize for speed. Product cycles need to be quick: ship new products and features as quickly as possible, then gather market data to rapidly iterate and improve. When you optimize for speed things will go wrong, so tolerate some messiness and empower smart creatives to adjust for problems. When you fail, do it quickly and without stigma and salvage the technology and expertise gained in the effort.
(48) Fail well. Failure is permissible as long as the team was focussing on the user and thinking big. But it must be done well: fail quickly (be ruthless, don’t throw good money after bad), learn from your failure, morph any valuable assets into other projects, and don’t stigmatize the team that failed.
(49) Imagine the unimaginable. Ask what could be true in five years: What thing that is unimaginable when abiding by conventional wisdom is in fact imaginable?
So there you have it, the 49 biggest ideas from our book, How Google Works. Follow these principles and you have a good chance of creating something that can solve big problems.
And, unfortunately, there are a lot of big problems in our world. Which brings us to our 50th, and final, idea:
(50) We are technology optimists. We believe that most big problems are information problems. We believe that with enough data and the ability to crunch it, virtually any challenge facing humanity today can be solved. And we believe that computers will serve at the behest of people— all people— to make their lives better and easier.
Thank you, and good luck!