Part IThe Story
A Dorm-Room Website and a Disciplinary Board
In the autumn of 2003, a Harvard sophomore named Mark Zuckerberg wrote a website called Facemash. It pulled photos of students from the online directories that Harvard's residential houses kept, showed them in pairs, and asked visitors to pick the more attractive one. The site spread across campus within hours, drew complaints from student groups, and was taken down. In November 2003, Harvard's Administrative Board looked into how he had collected the images. He was allowed to stay.
Facemash is usually told as an embarrassing origin story, and it was one. It also showed, very early, the things that would define Zuckerberg's career. He could build software quickly and alone. He understood that people are intensely curious about other people they know. And he tended to ship first and deal with the reaction afterwards. Each of those traits would make him one of the richest and most powerful people in the world, and each would get him into trouble again, on a much larger scale.
Zuckerberg was born on May 14, 1984, in White Plains, New York, and grew up in nearby Dobbs Ferry. His father was a dentist and his mother a psychiatrist. He started programming as a child. Around the age of eleven or twelve he built a messaging system, later nicknamed ZuckNet, that let computers in the family home and his father's dental office send messages to each other. He transferred to Phillips Exeter Academy for his last two years of high school, where he captained the fencing team. With a classmate he built Synapse, a music player that learned a listener's habits. Microsoft and AOL both showed interest in it. He turned them down and entered Harvard in September 2002.
By the time he arrived, he had already written software that strangers wanted to buy, and he had already said no to a large company once. Within four years he would say no to a much larger offer, and the pattern of keeping control rather than cashing out would shape everything that followed.
By the Numbers
Mark Zuckerberg and Facebook, 2004–2025
Feb 4, 2004Date TheFacebook went live for Harvard students
$500KPeter Thiel's 2004 investment, for about 10.2% of the company
$104BFacebook's valuation at its May 2012 IPO, which raised $16 billion
~$19BTotal price of WhatsApp in 2014, including restricted stock units
$5BPenalty in the July 2019 FTC privacy settlement
3.58BDaily active people across Meta's apps in December 2025
$200.97BMeta's revenue for fiscal year 2025
TheFacebook
On February 4, 2004, Zuckerberg launched TheFacebook, a directory for Harvard students. Members set up a profile with a photo, their classes, their interests and their friends, and could browse profiles of other students. It required a Harvard email address, so everyone on it was a real, verifiable person from the same community. That was the key difference from earlier social sites, which let anyone join under any name.
The site spread through Harvard quickly. Zuckerberg's co-founders, all classmates, were Eduardo Saverin, who handled the business side and put in early money; Dustin Moskovitz, who helped with engineering and expansion; Chris Hughes, who worked on publicity; and Andrew McCollum, who designed the early logo. In March 2004 the site opened to Stanford, Columbia and Yale. Each new campus was a closed, dense community where many students would join within days, so each launch looked like a small version of the original Harvard one.
That rollout was a deliberate way of managing network effects. A social network with a handful of scattered users is worthless, because nobody you know is on it. A network where most of your classmates joined last week is very useful. By going campus by campus, TheFacebook made sure every new market reached that level of density before moving on. Rival social networks of the period, including Friendster and MySpace, had grown broadly and loosely. TheFacebook grew narrowly and deeply, and its users were far more engaged as a result.
The launch also created Zuckerberg's first serious legal fight. Three Harvard students, Cameron and Tyler Winklevoss and Divya Narendra, had hired Zuckerberg in late 2003 to work on their own social site, HarvardConnection, later called ConnectU. They accused him of stalling on their project while building his own. The dispute went on for years. It was settled in 2008 for $20 million in cash plus Facebook shares, a package later valued at about $65 million, and was dramatized in the 2010 film The Social Network.
Palo Alto, Parker and Thiel
In the summer of 2004, Zuckerberg and a few friends rented a house in Palo Alto to work on the site. He did not go back to Harvard in the autumn. Sean Parker, the co-founder of Napster, became the company's president and helped it incorporate that July. Parker introduced Zuckerberg to investors, and one of them, Peter Thiel, put in $500,000 for about 10.2% of the company. It was Facebook's first outside investment, and Thiel joined the board.
Parker's lasting contribution, according to most accounts, was in how the company was set up. He had watched founders lose control of their companies to investors, and he helped arrange the early board and share structure so that Zuckerberg kept firm control. Parker left the presidency in 2005, but the principle stayed. Zuckerberg would go on to raise billions of dollars without ever giving up control of the company.
In April and May of 2005, Accel Partners led a $12.7 million round at a valuation of about $98 million. The company dropped "The" from its name that year and kept adding colleges, then high schools. It also started to look more like a business. Advertising brought in revenue, and the company hired engineers from established technology firms.
The early relationship with Saverin went badly. As the company raised money and reorganized, his stake was sharply diluted, and he sued. The case settled on confidential terms, with Saverin keeping a share of the company and credit as a co-founder. The episode showed a harder side of Zuckerberg's commitment to control: the people who shaped the company's structure early on would decide who kept influence and who did not.
Saying No to Yahoo
In 2006, Yahoo offered to buy Facebook. The talks reached as high as $1 billion, an extraordinary price for a company run by a 22-year-old that was still mostly used by students. Zuckerberg turned it down. Accounts of the period, including David Kirkpatrick's The Facebook Effect, describe several senior people at the company wanting to accept, and the decision strained relations inside the young management team.
His reasoning, as he later described it, was that Facebook was still at the beginning of its growth. It had only just started opening beyond colleges. If the network could become the place where most people online kept track of their friends, a billion dollars would look small. Turning down the offer was a bet that the curve was still rising steeply. Few founders make that bet, and fewer are right.
The same month, September 2006, Facebook made two changes that tested the bet. On September 6 it launched News Feed, which collected friends' updates, photo posts and profile changes into a single stream on the home page. Until then, users had to visit each friend's profile to see what was new. Many users hated the change. They felt that information they had shared on their own pages was now being broadcast to everyone they knew, and protest groups on Facebook itself quickly drew hundreds of thousands of members. Zuckerberg published an open letter apologizing for how the launch was handled, and the company added privacy controls. It kept News Feed.
Engagement soared. News Feed became the core of the product and eventually of the advertising business, since the feed was where people spent their time. Three weeks later, on September 26, 2006, Facebook opened to anyone aged 13 or older with an email address. The combination of a feed that kept people coming back and a network open to everyone turned a student directory into a mass platform. It also established a pattern: launch something users found intrusive, apologize for the handling, keep the substance.
Microsoft, Beacon and Sandberg
In May 2007, Facebook opened its platform to outside developers, letting them build applications that ran inside the site and spread through users' friend lists. Games and quizzes multiplied, and Facebook became a distribution channel for thousands of other companies. That October, Microsoft paid $240 million for a 1.6% stake, which valued Facebook at $15 billion. Microsoft also won the right to sell some of Facebook's advertising internationally.
A month later, on November 6, 2007, Facebook launched Beacon, an advertising system that reported what users did on partner websites, such as buying a movie ticket, back to their friends on Facebook. Many users had no idea it was happening until their purchases showed up in friends' feeds. There was an uproar, and Zuckerberg apologized publicly in December and made the program easier to turn off. A class-action lawsuit followed, and Facebook shut Beacon down in September 2009 as part of the settlement. Beacon was an early case of the same problem that would recur for the next two decades. The data that made Facebook valuable to advertisers was data users did not always expect to be shared.
Zuckerberg's biggest decision of this period was about management. He was an engineer who had never worked anywhere else, running a company that needed a serious advertising business. In March 2008 he hired Sheryl Sandberg, a Google advertising executive and former chief of staff at the U.S. Treasury, as chief operating officer. The two split the company in a way that lasted for fourteen years. Zuckerberg ran product and engineering, and Sandberg ran advertising sales, operations, policy and much of the company's public relations.
The arrangement worked commercially. Sandberg built an advertising organization that sold targeted ads to businesses of every size, and Facebook became cash-flow positive in September 2009. Marc Andreessen joined the board in 2008 and became one of Zuckerberg's closest advisers. By 2010, Facebook had hundreds of millions of users, Time named Zuckerberg its Person of the Year, and The Social Network was playing in cinemas. In November 2011, Facebook settled charges from the Federal Trade Commission that it had misled users about their privacy, agreeing to a consent order that required privacy audits. That order, finalized in 2012, would become very expensive seven years later.
The IPO and the Mobile Scramble
Facebook priced its initial public offering at $38 a share on May 17, 2012, and began trading on the Nasdaq the next day. It raised $16 billion and valued the company at about $104 billion, one of the largest technology listings ever. Zuckerberg held about 22% of the shares but, thanks to a dual-class structure that gave his Class B shares ten votes each, controlled about 57% of the votes.
The filing included a letter from Zuckerberg that set out how he intended to run a public company. He said Facebook would put its mission ahead of short-term profits and described "the Hacker Way", a culture built on rapid iteration, with sayings such as "Done is better than perfect" and "Code wins arguments." The company's motto of the time, "move fast and break things", appeared in the letter too.
— Mark Zuckerberg, letter in Facebook's IPO filing, 2012Facebook was not originally created to be a company. It was built to accomplish a social mission — to make the world more open and connected.
The listing went badly. Technical problems at Nasdaq disrupted the first day of trading, the stock soon fell below its offer price, and within months it had lost roughly half its value. The underlying worry was mobile. People were moving from desktop computers to smartphones, and Facebook's filing had warned that it earned no meaningful revenue from mobile. Its apps were slow, having been built largely with web technology, and it showed few ads on phones.
Zuckerberg's response was one of the fastest large-scale pivots in the history of the industry. He had the apps rebuilt natively, told teams to present mobile designs first, and pushed advertising into the mobile News Feed. By the fourth quarter of 2013, mobile made up 53% of Facebook's advertising revenue. On the earnings call in January 2014, he said that if 2012 was the year the company turned its product into a mobile product, 2013 was the year it turned its business into a mobile business. The stock recovered and then kept rising for years.
The pivot showed that Zuckerberg could see the innovator's dilemma coming at his own company and act before it hurt him. A desktop giant with no mobile business was exactly the kind of incumbent a younger rival could have displaced. He rebuilt the company around the new platform while the old one was still earning money.
Buying the Future: Instagram and WhatsApp
The mobile shift also brought new competitors. People on phones used apps built for phones, and several of those apps were growing faster than Facebook. In April 2012, shortly before the IPO, Facebook agreed to buy Instagram, a photo-sharing app with about thirteen employees, for roughly $1 billion. Many observers thought the price was absurd. It turned out to be one of the most successful acquisitions in the history of the technology industry.
Zuckerberg's reasons for the deal later became the core of a federal antitrust case. In 2020, the House Judiciary Committee published internal emails from February 2012 between Zuckerberg and Facebook's chief financial officer, David Ebersman. In them, Zuckerberg wrote that Instagram could hurt Facebook meaningfully without becoming a huge business. When Ebersman asked whether the goal was to neutralize a potential competitor, Zuckerberg partly agreed, then wrote a follow-up saying he had not meant to imply that Facebook would buy companies to stop them from competing. Regulators cited an even earlier email, from June 2008.
— Mark Zuckerberg, internal email, 2008, cited in the FTC's antitrust complaintIt is better to buy than compete.
In February 2014, Facebook announced it would buy WhatsApp, the messaging app founded by Brian Acton and Jan Koum, for $16 billion in cash and stock plus $3 billion in restricted stock units, about $19 billion in total. WhatsApp had hundreds of millions of users, especially outside the United States, and almost no revenue. A month later, in March 2014, Facebook bought Oculus, the virtual-reality headset maker, for $2 billion. Facebook also reportedly tried to buy Snapchat in 2013 for about $3 billion, an offer Evan Spiegel declined.
When Facebook could not buy a rival, it copied the rival's best idea. In 2016, Instagram launched Stories, a format of short-lived photo and video posts that Snapchat had popularized, and it soon had more daily users than Snapchat itself. In 2020, Instagram launched Reels in response to TikTok. The strategy of buying or copying protected Facebook's position as users moved between formats. It also left Zuckerberg exposed to the charge that he had kept a monopoly by removing competitors before they could grow. The relationship with WhatsApp's founders soured too. Acton left in 2017 and later told Forbes he had disagreed with Facebook's plans to make money from the service.
Cambridge Analytica and the Five-Billion-Dollar Fine
In March 2018, The Guardian, The Observer and The New York Times reported that Cambridge Analytica, a political consulting firm, had obtained data on tens of millions of Facebook users. The data came from a personality quiz app built by an academic researcher. Under Facebook's platform rules of the time, the app could collect information not only from the people who took the quiz but also from their friends. Facebook later estimated that the data of up to 87 million people might have been shared improperly.
The scandal came on top of criticism of Facebook's role in the 2016 U.S. election, including Russian influence campaigns on the platform, and it put Zuckerberg in front of Congress for the first time. On April 10 and 11, 2018, he testified before Senate and House committees for about ten hours in total. His prepared testimony took personal responsibility.
— Mark Zuckerberg, testimony to the U.S. Senate, 2018We didn't take a broad enough view of our responsibility, and that was a big mistake. It was my mistake, and I'm sorry. I started Facebook, I run it, and I'm responsible for what happens here.
The apology did not end the matter. The FTC concluded that Facebook had violated the 2011 privacy consent order. In July 2019, by a 3–2 vote, it approved a settlement under which Facebook paid a $5 billion penalty, by far the largest the agency had ever imposed for a privacy violation. Facebook also paid $100 million to settle Securities and Exchange Commission charges that it had misled investors about the risk of misuse of user data. The FTC order created a privacy committee on Facebook's board and required Zuckerberg to personally certify the company's compliance with its privacy program.
The period also produced more of Zuckerberg's large bets. In 2019, Facebook announced Libra, a planned global digital currency backed by a group of companies. Regulators and central banks objected almost immediately, partners dropped out, and the project, renamed Diem, was wound down in early 2022. In 2021, a former employee, Frances Haugen, gave internal documents to The Wall Street Journal and to Congress, and the resulting reports described internal research on harms associated with Instagram and Facebook. The company disputed how the documents were characterized. By the end of 2021, the company was facing antitrust suits, privacy regulation and political hostility from both major U.S. parties at the same time.
Meta and the Metaverse Bet
In October 2021, Zuckerberg renamed the parent company Meta Platforms. He said the company's future lay in the "metaverse", a set of connected virtual and augmented-reality spaces where people would work, play and socialize, and he committed to spending heavily to build it. The Oculus acquisition had become the Reality Labs division, which made the Quest headsets and was developing augmented-reality glasses.
Investors did not share his conviction. In February 2022, Meta reported its first-ever quarterly decline in daily users on Facebook, and the stock fell by about a quarter in a single day, erasing more than $200 billion in market value. Apple's 2021 privacy changes, which let iPhone users block apps from tracking them across other apps, had also made Meta's ad targeting less precise. TikTok was taking young users' time. And Reality Labs was losing enormous sums, $13.7 billion in 2022 alone. During 2022, Meta's stock fell more than 70% from its peak, according to Reuters.
Zuckerberg's control of the company meant shareholders could complain but could not stop him. The dual-class structure had been built for exactly this situation, and he had already defended it once. In 2016, Facebook proposed issuing a new class of non-voting shares so that he could give away most of his stock while keeping control. Shareholders sued, and in September 2017, just before the case went to trial, the company withdrew the plan. His control survived intact without it.
The metaverse spending continued. Reality Labs lost $19.19 billion in 2025, and its cumulative losses since the division was created ran into the tens of billions of dollars. The company kept releasing new headsets and, with the eyewear maker EssilorLuxottica, smart glasses that sold better than many expected. Whether the bet will pay off is still an open question. What is clear is that no other public company of Meta's size could have made it without the founder controlling the vote.
The Year of Efficiency
In November 2022, Meta laid off about 11,000 employees, around 13% of its workforce, the first large layoffs in its history. Zuckerberg said he had been wrong to expect the pandemic surge in online activity to last and had hired too many people as a result. On the earnings call in February 2023 he said the company's theme for the year would be efficiency. On March 14, 2023, he announced about 10,000 more job cuts in a memo that called the previous year a humbling wake-up call. He removed layers of management, closed projects, and told remaining managers to take on more direct work.
The results came quickly. Costs fell, profit margins recovered and the stock more than tripled from its late-2022 low over the following year. Zuckerberg had shown twice now, first with mobile in 2012 and again in 2022, that he could turn the company sharply when it was in trouble. Both times he did it by concentrating decision-making at the top and cutting projects that were not working.
He also went on the offensive against a rival. When Elon Musk's changes at Twitter unsettled many users, Meta launched Threads, a text-based app linked to Instagram accounts, on July 5, 2023. It passed 100 million sign-ups in five days, helped by the ability to import an existing Instagram network. Threads was a clear example of how Meta's existing user base, and the switching costs of rebuilding a social graph elsewhere, could let the company launch a new product at a scale no startup could match.
The efficiency drive did not mean spending less on the future. Zuckerberg moved money and people from the metaverse toward artificial intelligence, which by 2023 had become the most competitive field in technology. Meta already had a large AI research group, and its ranking and ad-targeting systems depended on machine learning. What changed was the scale of the ambition.
The Superintelligence Push
In 2023, Meta released its Llama family of large language models and made later versions available for outside developers to download and run. Zuckerberg argued that openly available models would become an industry standard, as Linux had, and that it was in Meta's interest to help that happen rather than to depend on a rival's closed system. Critics pointed out that Llama's licence carried restrictions and that Meta's openness served its own competitive interests. Both things could be true.
In 2025 the push became much more aggressive. In June, Meta agreed to invest $14.3 billion in Scale AI, a company that labels training data for AI models, for a 49% stake, valuing it at about $29 billion. Scale's 28-year-old founder, Alexandr Wang, joined Meta to lead a new group called Meta Superintelligence Labs. The New York Times described the deal as Meta's second-largest after WhatsApp. Meta also recruited researchers from rival labs with pay packages that were widely reported to be among the largest in the industry.
Spending rose to match. Meta's capital expenditures reached $72.22 billion in 2025, mostly on data centers and AI chips, and it guided investors to between $115 billion and $135 billion for 2026. The core advertising business was paying for it. Meta's revenue for 2025 was $200.97 billion, and 3.58 billion people used at least one of its apps every day in December 2025.
Zuckerberg also changed his approach to content rules. On January 7, 2025, he announced that Meta would end its third-party fact-checking program in the United States and move to a Community Notes system similar to the one on X. In a video he said the company's systems had been making too many mistakes and engaging in too much censorship, and that it was time to get back to its roots around free expression. He acknowledged a trade-off: the company would catch less harmful content, but it would also remove fewer innocent posts by mistake. Critics saw the change as an attempt to win favour with the incoming administration. Supporters saw it as a correction to years of heavy moderation.
The Antitrust Verdict and What Control Buys
In December 2020, the FTC and a coalition of states sued Facebook, arguing that it had illegally maintained a monopoly in personal social networking, largely by buying Instagram and WhatsApp. A federal judge dismissed the first complaint in June 2021 for failing to establish monopoly power. The FTC filed an amended complaint that August, and the case went to a bench trial in spring 2025. Zuckerberg testified, and the 2008 and 2012 emails were central evidence.
On November 18, 2025, Judge James Boasberg ruled for Meta. He found that the FTC had not proved Meta held a monopoly today, because the relevant market had to include TikTok and YouTube, which compete directly for the same users' time. The FTC appealed in January 2026, and the appeal was still pending at the time of writing. For now, Meta keeps Instagram and WhatsApp, the two acquisitions that defined its growth for a decade.
Zuckerberg's career is best understood through one decision he made over and over: keep control and use it. He turned down Microsoft and AOL as a teenager, turned down Yahoo at 22, built a share structure that let him overrule investors, and used that power to force through the mobile pivot, the big acquisitions, the metaverse bet, the layoffs and the AI spending. Some of those bets made Meta one of the most valuable companies in the world. Others cost tens of billions of dollars. Several produced scandals, fines and lawsuits that another chief executive might not have survived.
In 2015, he and Priscilla Chan announced the Chan Zuckerberg Initiative and pledged to give 99% of their Facebook shares to philanthropic causes over their lifetimes. By then he had spent more than a decade as the only person who could decide what Facebook would become. More than twenty years after TheFacebook went live, that is still true. Few founders of companies this large keep that kind of control for so long, and he has used it to run Meta in a way no board would ever have approved.
Part IIThe Playbook
Zuckerberg's record is unusually useful to study because it contains both extraordinary strategic calls and costly failures, often made by the same instincts. The principles below are drawn from what he actually did: how he grew a network, protected his control, bought and copied competitors, handled crises of trust and made long, expensive bets. Some are models to follow, and some are warnings.
The principles
- 1.Saturate one tight community before expanding.
- 2.Design the share structure before you need it.
- 3.Refuse the exit while the curve is still steep.
- 4.Hire a partner for the work you do badly.
- 5.Rebuild around the platform your users are moving to.
- 6.Acquire the fast-growing rival early.
- 7.Copy a winning format without embarrassment.
- 8.Treat user trust as a liability that compounds.
- 9.Fund long bets from a dependable profit engine.
- 10.Cut hard and fast when the plan was wrong.
- 11.Launch new products on top of an existing graph.
- 12.Give away the layer a rival wants to charge for.
Principle 1
Saturate one tight community before expanding.
TheFacebook did not try to sign up the whole internet. It started with a single university, required a campus email address, and moved to Stanford, Columbia and Yale only once Harvard was thoroughly hooked. Each new campus was a closed community where most people knew each other, so the network became useful to newcomers almost immediately.
This is how network products avoid the empty-room problem. A thin network spread across many places delivers little value to anyone, while a dense network in one place delivers a lot. Winning one small market completely also gives a team a repeatable launch playbook and proof to show the next market.
Tactic: Define the smallest community in which your product would be genuinely useful if most members joined, and do not expand until you have reached that level of adoption there.
Zuckerberg's early board and share arrangements, shaped with Sean Parker's help, and the later dual-class structure left him in control of the company through a failed IPO, a mobile crisis, a collapse in the share price and years of spending that investors disliked. Without that control, several of his most important decisions would almost certainly have been blocked.
Control is much easier to set up at the start than to win back later. Once investors hold large stakes and board seats, they will resist any change that reduces their power. The flip side is real: control removes a check on the founder's mistakes, and Facebook's governance became a lasting source of shareholder criticism.
Tactic: Before your first priced round, decide what decisions you must be able to make alone, and have a lawyer design voting rights and board composition around keeping them.
Principle 3
Refuse the exit while the curve is still steep.
In 2006, Yahoo's offer of up to $1 billion looked enormous for a company used mainly by students, and some of Zuckerberg's colleagues wanted to take it. He judged that Facebook was at the start of its growth, since it had only just begun to open beyond colleges. Within a few years, Microsoft's investment valued it at fifteen times that offer.
The question for a founder is where the company sits on its growth curve, not whether the price sounds large. An offer that looks generous against current revenue may be cheap against where the business will be in two years. The judgment needs evidence, such as engagement and growth in new markets, rather than optimism alone.
Tactic: When an acquisition offer arrives, model the company's value two and five years out under conservative growth assumptions, and compare the offer to those numbers rather than to today's.
Principle 4
Hire a partner for the work you do badly.
By 2008 Zuckerberg was running a fast-growing company with no mature advertising business and little management experience. He hired Sheryl Sandberg as chief operating officer and gave her advertising, operations and policy while he kept product and engineering. The split lasted fourteen years, and the business she built paid for everything else.
Founders often hire deputies who resemble them. The more valuable hire is someone whose strengths cover the founder's gaps, with a clear division of responsibility so the two do not compete for the same decisions.
Tactic: Write down the three parts of the business you are least qualified to run, and hire a senior leader with full authority over them instead of a subordinate who will defer to you.
Principle 5
Rebuild around the platform your users are moving to.
At its 2012 IPO, Facebook earned almost nothing from mobile. Zuckerberg had the apps rebuilt natively, made teams design for phones first, and put ads in the mobile feed. Within about eighteen months, mobile was more than half of ad revenue and the investor panic had faded.
Platform shifts are the moment when incumbents are most vulnerable, because their products and business models are built for the old platform. The companies that survive move before the old platform stops paying, even if the change costs them in the short term.
Tactic: Track what share of your users' time is moving to new devices or channels each quarter, and when it passes a set threshold, reassign your best team to rebuilding for that channel first.
Principle 6
Acquire the fast-growing rival early.
Instagram cost about $1 billion when it had around thirteen employees, and WhatsApp about $19 billion when it had almost no revenue. Both deals looked expensive at the time. Both removed a potential challenger and added a business that became central to the company. Zuckerberg's own emails showed he understood the competitive logic clearly.
This approach works best when a buyer can see which small companies are winning its own users' attention. It also carries serious legal risk. Meta spent years fighting an FTC suit over these deals, and the internal emails describing them were the heart of the case. It won, but it could have lost two of its most valuable businesses.
Tactic: Monitor the fastest-growing products competing for your users' time, and make an offer while the price is still small relative to your market value, with antitrust counsel involved from the first conversation.
Principle 7
Copy a winning format without embarrassment.
When Snapchat could not be bought, Instagram launched Stories in 2016 and soon had more daily users of the format than Snapchat did. When TikTok's short videos took off, Instagram launched Reels. Meta repeatedly let others discover a format and then brought it to its much larger audience.
Pride stops many established companies from copying an upstart. Users generally do not care who invented a feature, only whether it is available where they already spend their time. A large platform that adopts a proven format quickly can often neutralize the advantage of the company that invented it.
Tactic: Keep a running list of formats gaining traction on rival products, and give a small team a deadline to ship your own version when one of them starts drawing your users away.
Principle 8
Treat user trust as a liability that compounds.
News Feed, Beacon and Cambridge Analytica followed the same pattern. Facebook used or exposed data in ways users did not expect, apologized, and kept going. The accumulated damage turned into a 2011 consent order, then a $5 billion penalty in 2019, personal compliance certification for Zuckerberg, and a long period of political hostility.
Each individual privacy lapse can look manageable. Together they build a record that regulators, courts and the public use against the company later. The cost of lost trust arrives all at once, years after the decisions that created it.
Tactic: Before launching any feature that shares user data in a new way, ask a small group of real users to predict what it will do, and delay the launch until what they expect matches what actually happens.
Principle 9
Fund long bets from a dependable profit engine.
Meta's advertising business, generating about $201 billion in revenue in 2025, paid for Reality Labs' losses of tens of billions of dollars and a data-center building program costing more than $70 billion a year. Few companies could survive that level of speculative spending, and Meta could because the core business kept growing.
Long-term bets need a patient funding source that will not disappear in a bad quarter. A company that bets the future on money it has to raise from nervous outside investors is more likely to be forced to stop halfway.
Tactic: Cap spending on speculative projects at a fixed share of profits from your established business, and review that share annually rather than project by project.
Principle 10
Cut hard and fast when the plan was wrong.
Zuckerberg admitted he had wrongly assumed the pandemic boom would last and had over-hired. In November 2022 and March 2023 Meta cut about 21,000 jobs, removed management layers and shut projects. Margins and the share price recovered within about a year.
Leaders often cut too slowly because each small reduction feels like an admission of error. One decisive round, explained honestly, usually does less damage to morale and the business than a long series of smaller cuts.
Tactic: When a core assumption behind your hiring plan turns out to be wrong, recalculate the headcount the business needs and make the full adjustment in one announcement.
Principle 11
Launch new products on top of an existing graph.
Threads reached 100 million sign-ups in five days because people could log in with Instagram and bring their followers with them. A startup launching the same app would have had to persuade every user to rebuild their network from nothing.
A large user base is more than an audience. It is a ready-made set of relationships that any new product can inherit. Companies that own such a network can launch new products at a speed independent competitors cannot match, as long as the product itself is good enough to keep people once they arrive.
Tactic: For every new product, identify which existing user relationships it can import on day one, and make that import a single step in sign-up.
Principle 12
Give away the layer a rival wants to charge for.
Meta released its Llama models for developers to download and run, partly so that its products would not depend on a rival's closed AI system. Zuckerberg compared the aim to Linux becoming a standard that no single company controlled. Giving the technology away made it harder for competitors to charge a toll on a layer Meta needed.
Openness can be a competitive weapon when a company earns its money somewhere else. Releasing a component free can commoditize a rival's main product, attract outside developers and improve the technology faster. It works only if the giving company's own profits come from a different layer.
Tactic: Identify the technology layers you depend on but do not make money from, and consider opening your own version of one to reduce any single supplier's pricing power over you.
Free playbook
Get The Business Model Playbook
58 business models, one visual page each: how the money flows, the metrics that matter, and who runs it. Free when you join the Faster Than Normal email.
Free. No spam. Unsubscribe anytime.
Part IIIMaxims
- Curiosity about acquaintances is the product. Facebook worked because people want to know what classmates, colleagues and friends are doing.
- Votes matter more than equity. Owning 22% while controlling 57% of the votes meant investors could object but not overrule.
- Apologies do not reset the ledger. Each promise to do better was remembered when the next lapse arrived.
- Emails become evidence. Private notes about competitors can end up as the centerpiece of a government lawsuit.
- Incumbency is a head start, not a moat. TikTok showed that even a network of billions can lose young people's attention.
- Engineering culture shapes governance. A company built on shipping quickly will tend to discover its mistakes in public.
- Renaming signals intent. Calling the company Meta told investors and employees where money would go, whether they agreed or not.
- Judges define markets. Meta won its antitrust case largely because the court counted TikTok and YouTube as competitors.
- Founder power cuts both ways. The same authority that forced through the mobile pivot also kept Reality Labs funded through years of losses.
In Their Own Words
I was just trying to help connect people at colleges and universities. But I guess it grew bigger than I expected.— Mark Zuckerberg
The biggest risk is not taking any risk. In a world that's changing really quickly, the only strategy that is guaranteed to fail is not taking risks.— Mark Zuckerberg
Move fast and break things. Unless you are breaking stuff, you are not moving fast enough.— Mark Zuckerberg
I think a simple rule of business is, if you do the things that are easier first, then you can actually make a lot of progress.— Mark Zuckerberg
The question isn't, 'What do we want to know about people?', it's, 'What do people want to tell about themselves?'— Mark Zuckerberg
People can be really smart or have skills that are directly applicable, but if they don't really believe in it, then they are not going to really work hard.— Mark Zuckerberg
The thing that we are trying to do at Facebook, is just help people connect and communicate more efficiently.— Mark Zuckerberg
I started the site when I was 19. I didn't know much about business back then.— Mark Zuckerberg
My goal was never to just create a company. A lot of people misinterpret that, as if I don't care about revenue or profit or any of those things. But what not being 'just' a company means to me is not being just that - building something that actually makes a really big change in the world.— Mark Zuckerberg
The companies that work are the ones that people really care about and have a vision for the world so they will do something to make it happen.— Mark Zuckerberg
I think that people just have this core desire to express who they are. And I think that's always existed.— Mark Zuckerberg
Having two identities for yourself is an example of a lack of integrity.— Mark Zuckerberg
The biggest mistake we made as a company was betting too much on HTML5 as opposed to native... because it just wasn't there. And it's not that HTML5 is bad. I'm actually, on long-term, really excited about it. One of the things that's interesting is we actually have more people on a daily basis using mobile Web Facebook than we have using our iOS or Android app combined. So mobile Web is a big thing for us.— Mark Zuckerberg
Virtual reality was once the dream of science fiction. But the internet was also once a dream, and so were computers and smartphones.— Mark Zuckerberg
I'm trying to make the world a more open place by helping people connect and share.— Mark Zuckerberg
The real question for me is, do people have curiosity about the world around them, and can I do anything to foster that?— Mark Zuckerberg
We have a pretty ambitious goal for the world. We want to make the world more open and connected.— Mark Zuckerberg
Privacy is not just about hiding things. It's about having the freedom to be yourself.— Mark Zuckerberg
We're running the company to serve more people. That's always been our mission from the beginning.— Mark Zuckerberg
I think the right framework is that there are real benefits to people getting to share information and being able to connect, to find people who they want to connect with, and we've always believed that privacy and sharing are not opposing forces.— Mark Zuckerberg
We want to make it so that anyone, anywhere - a child growing up in rural India who never had a computer - can go to a store, get a phone, get online, and get access to all of the same things that you and I appreciate about the internet.— Mark Zuckerberg
Further reading
- The Accidental Billionaires: The Founding of Facebook by Ben Mezrich
