The $26 Billion Résumé
On the morning of June 13, 2016, a line began forming outside LinkedIn's Mountain View cafeteria at 9:45 a.m.—unusual for a Monday, unusual for a company whose employees were accustomed to the rhythms of Wednesday all-hands meetings. Jeff Weiner, the CEO who had been running the company for nearly eight years, watched the queue snake through the building from across the campus and momentarily forgot the reason for it. "I was looking to myself, I wonder who is here speaking today?" he later told the assembled crowd. "Like a rockstar or something. And I was like, oh yeah." The last time LinkedIn had held an all-hands on a non-Wednesday was the day of its IPO, broadcast live from the Empire State Building in May 2011. This time, the news was different in kind: LinkedIn, the professional social network with 433 million members and $3 billion in annual revenue, had agreed to be acquired by Microsoft for $26.2 billion in cash—a 50% premium over the previous Friday's closing price, and, at the time, the largest acquisition in Microsoft's forty-one-year history.
The number itself carried a strange resonance. LinkedIn had been, for most of its existence, the unglamorous sibling of Silicon Valley's social media family—never as culturally electric as Facebook, never as addictive as Instagram, never as politically combustible as Twitter. It was the platform people visited not for entertainment but for obligation: to update a résumé, to accept a connection request from a colleague they barely remembered, to search for a job when the current one had curdled. And yet that mundanity—the sheer workaday utility of the thing—was precisely what made it irreplaceable. By the time
Satya Nadella wrote the check, LinkedIn had become something no other social network had managed: the professional identity layer of the internet, the place where your working self existed in digital amber, visible to recruiters, employers, and the algorithmic machinery of the global labor market. The acquisition price valued each of LinkedIn's members at roughly $60 apiece—the going rate, it turned out, for owning the world's most comprehensive database of who works where, doing what, for how much.
What Microsoft bought was not merely a social network. It was a system of record for human capital on a planetary scale, a machine that converted the anxious energy of career management into recurring revenue across three distinct business lines. The story of how that machine was built—from a philosopher's apartment in Mountain View to a $17 billion subsidiary of the world's largest software company—is a story about the nature of professional identity in the network age, about the paradox of building a monopoly on something nobody particularly enjoys using, and about what happens when the keeper of the world's careers becomes a division of the company that already owns the world's productivity software.
By the Numbers
LinkedIn at Scale
1B+Members across 200+ countries (2024)
$17.1BAnnual revenue, FY2024
67MCompanies with LinkedIn pages
$26.2BMicrosoft acquisition price (2016)
41KSkills listed on the platform
~20,000Employees worldwide
140KSchools with LinkedIn pages
The Loner Who Built the Network
Reid Hoffman was born in 1967 at Stanford Hospital, during the Summer of Love, to parents who promptly fell apart. His father, a law student, and his mother, also a future lawyer, married and separated while still in their early twenties, leaving Hoffman an only child shuttled between California, Alaska, and his grandparents' house in Sunnyvale. He ended up in Berkeley with his father, who had entered a series of relationships and whom Hoffman describes less as a parent than a co-traveler: "We all grew up together, in some way. It was not idyllic. It was intense, vibrant, sometimes oppressive." The loner kid who didn't meld much in school would, within three decades, build the world's largest professional network—a biographical irony so neat it almost demands psychoanalytic interpretation. When asked whether growing up without siblings, without a stable family, without roots in any neighborhood, had driven his obsessive focus on connections, Hoffman shrugged. "Is that the psychological origin story for my focus on networks? Maybe."
His formative encounter was not with computers but with games. At nine, Hoffman discovered Dungeons & Dragons. By middle school in Berkeley, he had talked his way into Chaosium, a game company in nearby Emeryville, where he corrected errors in their published role-playing scripts and wrote reviews for their gaming magazine. The kid was testing rule systems, finding bugs, proposing patches—the essential pattern of his entire career. He applied to the Putney School, a boarding school in Vermont, without telling his parents. "Vermont was the farthest place from California I could imagine that still seemed feasible." When a bullying campaign began there, he solved it with game logic: "The way you deal with bullies is you change their economic equation. Make it more expensive for them to hassle you."
Stanford, in 1989, ended the miseries. He enrolled in Symbolic Systems—a hybrid of philosophy, linguistics, psychology, and computer science—befriended
Peter Thiel (they served together on the student senate, Hoffman as the left-winger, Thiel as the right-winger), and met his future wife, Michelle Yee. A Marshall scholarship took him to Oxford for three years of philosophy. He returned to California convinced that academia was too narrow—"His professors spent their time thinking about highly specific problems and publishing for an audience of their peers"—and that the real philosophical project was building systems that organized human interaction at scale.
He was entranced by Neal Stephenson's Snow Crash, the 1992 novel depicting a virtual society called the Metaverse. The term "Internet" was not yet in general circulation. "Social network" was an academic concept used by psychologists to derive mathematical formulas. But Hoffman was already assembling the ingredients—fantasy gaming, computer technology, philosophy—and searching for the architecture that could, as he put it, "configure the space in which people would interact."
False Starts and the Real Name Problem
The path to LinkedIn ran through two failures and one of the most consequential companies in Silicon Valley history. Hoffman's first job was at eWorld, Apple's short-lived online service. Then WorldsAway, a Fujitsu-owned virtual chat community where users interacted through fictional avatars. In 1997, he started SocialNet, which let people connect for dating and other purposes using pseudonyms. SocialNet was acquired for a modest sum by Spark Networks, which later owned JDate and ChristianMingle—a destination so far from Hoffman's ambitions it reads like a punchline.
The lesson embedded in SocialNet's failure was deceptively simple: the most successful online networks would require people to use their real names. Anonymous connection turned out to be a feature of dating sites. Professional connection required identity, reputation, verifiability. The distinction seems obvious in retrospect. It was not obvious in 1997.
After SocialNet, Hoffman joined PayPal in 1999, where Peter Thiel had assembled a team that would become Silicon Valley's most productive alumni network. Hoffman's role was diplomat and negotiator—"Relative to the rest of the crew, I had a massively better idea of where another person was coming from and how to bridge the gap." He persuaded eBay not to cut off PayPal's access to its marketplace by invoking the specter of antitrust enforcement. He learned from Napster's cautionary tale that total defiance of existing institutions was suicidal; moderation in external relations, without abandoning aggressive behavior, was the winning strategy. And he helped pioneer a technique that would become standard across Silicon Valley: mobilizing a company's user base as a political lobbying force, flooding regulators with emails.
It's better to beg for forgiveness than to ask for permission.
PayPal also taught Hoffman a principle about growth that would become foundational. Early on, the company paid people five and then ten dollars for recruiting friends—a strategy that produced losses in the tens of millions annually but built the user base that made the business viable. The lesson: it wasn't yet clear that it was more important to build up a big user base than to make money, but the people who figured that out first won. When the Patriot Act damaged PayPal's secondary business of processing gambling transactions in October 2001, Hoffman helped arrange a quick sale to eBay. It was time to build the thing he'd been circling since Berkeley.
Founding the Professional Graph
In December 2002, Reid Hoffman sat in his living room in Mountain View and sketched the outlines of a new social network with four co-founders: Allen Blue, a product designer; Konstantin Guericke, a marketing professional; Eric Ly and Jean-Luc Vaillant, both engineers. The idea was SocialNet rebuilt from its ashes—but with real names, a focus on professional lives, and a structure modeled less on the chat room than on the résumé.
LinkedIn launched publicly on May 5, 2003. Within the first month, 4,500 people had signed up. The number feels quaint now, but the timing was exact: Friendster had launched a month earlier, MySpace was months away, and Facebook was still a year from existing in
Mark Zuckerberg's dorm room. The most popular employment sites, like Monster.com, didn't focus on social networks. The most popular social networks didn't focus on employment. Hoffman was betting that those two markets were actually the same market, and that the combination—a social graph organized around professional identity—would be more valuable than either half.
📅
The Building of LinkedIn
Key milestones from living room to global platform
2002Reid Hoffman conceives LinkedIn in his Mountain View apartment; recruits four co-founders.
2003LinkedIn launches May 5 with a free basic account and premium options. Sequoia Capital invests. 4,500 users in month one.
2004Reaches 1 million members. Introduces premium subscriptions—first monetization. Job listings for companies debut.
2005First advertisement runs on the platform. Membership reaches 6 million.
2006Profiles made partly public for Google indexing. 20 million members.
2007LinkedIn turns profitable for the first time. 15+ million members.
2008Over 32 million members. Sequoia and others buy 5% for $53M, valuing the company at ~$1 billion.
Growth was slow at first—intentionally. Hoffman concentrated on building density within professional clusters rather than chasing raw sign-ups. He wanted each new member to find enough existing connections to make the network immediately useful. This meant LinkedIn's early user base skewed heavily toward Silicon Valley, venture capital, and the technology industry—the very people most likely to understand, and evangelize, a professional network's value proposition.
Almost from the beginning, LinkedIn offered members the ability to upload their entire email contact list, generating large numbers of automatic invitations. Hoffman knew people found this annoying. It was a problem only if it impeded growth. "People may say, 'I'm getting all these fucking invitations,' " he acknowledged, "but you don't tune it too high or too low." The resulting flood of connection requests became LinkedIn's most powerful growth engine and its most persistent source of public ridicule—the endless "I'd like to add you to my professional network on LinkedIn" emails that defined the platform's cultural identity for a decade.
The Deliberate Awkwardness
There is a paradox at the center of LinkedIn that its founders understood from the beginning: the platform's value is inversely proportional to the pleasure of using it. Facebook is a place people visit because they want to. LinkedIn is a place people visit because they feel they must. Nobody opens LinkedIn to relax. You open it because you're looking for a job, or because you suspect someone is looking at you, or because a recruiter has sent a message that might be the beginning of a 30% raise. The emotional register of the platform—somewhere between anxiety and obligation—is its competitive moat.
This is not accidental. Hoffman believed that people would want to maintain separate professional and personal online identities, and he bet the company on that conviction when Facebook's explosive growth in 2008 (100 million members to LinkedIn's 32 million) raised the terrifying possibility that Mark Zuckerberg might simply absorb the professional use case into his social graph. LinkedIn's response was to double down on the résumé-as-identity model: detailed profiles, regularly updated, elaborately connected, functioning as the permanent digital foundation of a career. The profile was not a social media page. It was a public document—part CV, part reputation ledger, part signal to the labor market that you were open to possibility.
In 2006, LinkedIn made a decision that would prove strategically decisive: it made all profiles partly public, so that when you typed someone's name into a Google search, their LinkedIn profile appeared among the top results. This was genius and it was aggressive. It meant that LinkedIn became, for hundreds of millions of professionals, the first thing a stranger would see about them online. The platform had inserted itself between every human being and their professional reputation, and it had done so by giving away the product—the public profile—while charging for the tools that made the product actionable.
LinkedIn harnessed its members' competitiveness by listing connection counts up to a maximum of five hundred, creating a game mechanic that rewarded network-building. Every conference room in Hoffman's building was named after a canonical game: Pac-Man, Tetris, Space Invaders. "Business is the systematic playing of games," Hoffman said, and he had designed a game that millions of people played without quite admitting they were playing it.
The Weiner Years: From Startup to System
Jeff Weiner arrived at LinkedIn in December 2008, a former Yahoo executive with a biography that rhymed with the company's own trajectory—polished where Hoffman was rumpled, operational where Hoffman was philosophical, obsessively present where Hoffman was omnidirectionally scattered. He had spent six years at Warner Bros. before Yahoo, where he oversaw a $3 billion consumer-facing division with 3,000 employees. Hoffman hired him as president; within six months, Weiner was CEO.
The engineers were initially suspicious. Weiner wasn't one of them. But someone analyzed his activity on the LinkedIn platform and discovered that the only time he wasn't logged on was between 3:30 and 4 a.m. (His office later insisted this had improved to 11 p.m. to 5 a.m.) The discovery converted skeptics. Here was a leader who treated the product with the reverence engineers reserved for code reviews.
December 15th, 2008, marked the first day of the best job I've ever had. My rationale for joining LinkedIn was simple: The opportunity to work with Reid Hoffman, a founder I greatly admired and respected; to join an extremely talented and dedicated team; and to massively scale LinkedIn's membership and business.
What Weiner scaled was breathtaking. During his eleven-year tenure as CEO, LinkedIn's membership grew from 33 million to nearly 690 million. Revenue increased from $78 million to over $7.9 billion. The employee count expanded from 338 to more than 16,000. He took the company public in May 2011, pricing at $45 per share—the stock more than doubled on the first day of trading as investors, still hungry for the next tech and digital media success story, appeared to banish all memory of the 2000 dotcom crash. The IPO raised $353 million.
Weiner's critical strategic contribution was articulating and operationalizing what he called "the economic graph"—a vision to digitally map every member of the global workforce, every company, every job, every skill, and every educational institution. It was grandiose in the way that only Silicon Valley mission statements can be, and yet LinkedIn was arguably the only company on earth with both the data and the distribution to attempt it. The economic graph gave LinkedIn's product development a gravitational center: every new feature, every acquisition, every expansion could be evaluated against a single question—does this bring us closer to mapping the entire labor market?
The Lynda.com acquisition in 2015, for approximately $1.5 billion, was the economic graph's most expensive expression. If LinkedIn was going to be the platform where people managed their careers, it needed to be the platform where they acquired the skills those careers demanded. Hoffman published an essay called "Disrupting the Diploma," arguing that the future of education would not flow exclusively through universities. Lynda.com—rebranded as LinkedIn Learning—would become the upskilling engine attached to the career identity layer, creating a closed loop: identify skill gaps on LinkedIn, fill them through LinkedIn Learning, display the credentials on your LinkedIn profile.
The Day the Stock Cratered
The ascent was not uninterrupted. On February 5, 2016—four months before the Microsoft acquisition—LinkedIn's shares plunged more than 40% in a single day, erasing $11 billion in market value. The trigger was a revenue forecast that fell well short of expectations: online advertising revenue growth had decelerated to 20% in the fourth quarter of 2015, down from 56% a year earlier. Over thirty brokers downgraded their forecasts in the aftermath.
The crash exposed a structural vulnerability that LinkedIn's supporters had been minimizing for years. The platform's advertising business, while growing, could not compete with the targeting precision and scale of Facebook and Google. LinkedIn's ad inventory was constrained by the fundamental nature of the product: people did not spend hours scrolling LinkedIn the way they scrolled Instagram. Sessions were shorter, less frequent, more purposeful. That was excellent for the recruitment business—users who visited infrequently but maintained obsessively current profiles were perfect targets for recruiters—but it was a liability for an advertising model that rewarded engagement and time-on-site.
The stock collapse also highlighted a governance structure that gave Hoffman disproportionate control. He owned twelve percent of the company but held fifty-eight percent of the voting shares through a dual-class stock structure. When the board began quietly exploring strategic options in the weeks after the crash, Hoffman's consent was essential. The background of the merger section of LinkedIn's proxy statement—a surprisingly gripping piece of corporate literature—reveals that the first formal discussion between Weiner and Satya Nadella occurred on February 16, 2016, eleven days after the stock implosion, when Weiner met with the Microsoft CEO to discuss ways to enhance the commercial relationship between the companies. The concept of a business combination was raised at that meeting.
What followed was a four-month mating dance involving at least four suitors: Microsoft, widely rumored to be Salesforce (Party A in the proxy), and two others believed to be Google and Facebook. LinkedIn hired Qatalyst Partners and Wilson Sonsini. Salesforce brought in Goldman Sachs. By June, Microsoft had won—paying $196 per share, a 50% premium, in all cash. Salesforce CEO
Marc Benioff made a last-ditch effort to mount a competing bid; it didn't materialize.
The Microsoft Paradox
The conventional wisdom in 2016 held that large-company acquisitions of social networks were graveyards of value—that the bureaucratic metabolism of a $400 billion software company would inevitably smother the agile culture of a social platform. The conventional wisdom was, in this case, wrong.
Microsoft's approach to LinkedIn was, by the standards of technology acqui-hire culture, almost shockingly restrained. As LinkedIn co-founder Allen Blue told CNBC at Davos in 2020: "All the growth we've been seeing has been in the way we've been operating our own businesses." Nadella permitted LinkedIn to operate as a semi-autonomous subsidiary with its own CEO, its own culture, and its own strategic cadence. The integration was surgical rather than structural: LinkedIn identity woven into Microsoft's Outlook email client, LinkedIn Learning integrated into Microsoft Teams, LinkedIn data feeding the Dynamics 365
CRM platform.
The financial results were the acquisition's most compelling vindication. LinkedIn's revenue grew from approximately $3 billion at the time of acquisition to over $17 billion in fiscal 2024. The Productivity and Business Processes segment—anchored by Microsoft 365 subscriptions and LinkedIn—achieved operating margins exceeding 58% in recent quarters, up from 33% in 2017. LinkedIn, which had recorded a net loss of $166 million in its final full year as a public company, became one of Microsoft's highest-margin, fastest-growing divisions.
Whether it's worker displacement, the skills gap, youth unemployment, or socio-economic stratification, the impact on society will be staggering. I've said it on multiple occasions and believe it even more so every day: creating economic opportunity will be the defining issue of our time.
Nadella's strategic logic was both simple and non-obvious. Microsoft already owned the tools—Word, Excel, PowerPoint, Outlook, Teams—that constituted the operating system of professional life. LinkedIn owned the identity layer. Together, they could create something no competitor could replicate: a closed loop from professional identity (LinkedIn) to professional communication (Outlook, Teams) to professional productivity (Office 365) to professional learning (LinkedIn Learning) to professional recruitment (LinkedIn Talent Solutions) to professional sales intelligence (LinkedIn Sales Navigator). The user never had to leave the Microsoft ecosystem. The data never had to leave the Microsoft servers.
The Roslansky Succession
In February 2020, Weiner announced he would step down as CEO on June 1, transitioning to executive chairman. His successor was Ryan Roslansky, a product leader who had been Weiner's first hire at LinkedIn in 2009. Roslansky, born in the late 1970s, was a freshman in college in 1996 when the internet was just beginning—lucky timing, as he freely admitted. He taught himself to code, co-founded a startup, then landed at Yahoo as a junior product manager, where he met Weiner. He followed Weiner to LinkedIn and spent the next decade in leadership roles across nearly every part of the business: marketing solutions, the influencer program, the publishing platform, consumer products, the Lynda.com acquisition.
If Hoffman was the philosopher and Weiner the operator, Roslansky was the product thinker—the person who understood that LinkedIn's value was not in its code but in its data, and that data's value was a function of how precisely the platform could match supply (talent, content, ads) with demand (recruiters, learners, marketers). Under Roslansky, LinkedIn has more than doubled its revenue to north of $17 billion annually while growing the platform to record engagement levels. The membership has crossed one billion—a number that means less than it appears (LinkedIn does not disclose monthly or daily active users, a deliberate opacity) but that matters enormously as a signal to the labor market that LinkedIn is not optional.
In June 2025, Nadella expanded Roslansky's portfolio to include oversight of Microsoft's Office productivity software unit, including the M365 Copilot app—a signal that Nadella views LinkedIn and Office as converging products in the AI era. "Office is one of the most iconic product suites in history," Roslansky wrote in a LinkedIn post announcing the expanded role. "Productivity, connection, and AI are converging at scale."
The Architecture of Professional Anxiety
The deeper story of LinkedIn is not about technology or business models. It is about the transformation of work itself—and the platform's role in both reflecting and accelerating that transformation.
Hoffman's foundational thesis, articulated in his book
The Startup of You, was that the postwar social contract between employers and employees had irreversibly shattered. The era of lifetime employment at a single corporation—the world described in William H. Whyte's 1956
The Organization Man—was over. In its place, Hoffman argued, was the era of the Network Man: a world in which careers were portfolio projects, jobs were tours of duty lasting two to four years, and the keeper of your professional future was not your employer but your personal network. LinkedIn was the infrastructure for that world.
The thesis was prescient. It was also self-serving in a way that Hoffman's critics were not shy about pointing out. If the future of work was permanent impermanence—a life of serial gigs, endless upskilling, relentless personal branding—then the platform that monetized that anxiety was the primary beneficiary of the very disruption it claimed to facilitate. LinkedIn didn't just describe the networked economy. It built the toll road.
John Lilly, one of Hoffman's partners at Greylock, was characteristically blunt about the paradox: "Clearly, wealth is becoming more concentrated, and the network takes a larger and larger share." He suspected the twentieth century's middle class was an anomaly. "There was no middle class, then there was a middle class, now we're back where we started—it's hollowed out." Even Mike Maples, a self-described believer in "free people and the free market," reported that Glenn Beck had surprised him with the question: "What do you say to a guy like me? How do you answer the argument that there are forty million people in red states who are going to get displaced?"
Hoffman's answer was always the same: more networks, more entrepreneurship, more platforms. "I'm trying to get politicians to understand that solving this problem is about facilitation of a network, as opposed to"—and here his tone turned sarcastic—"the New Deal." The UN estimated the global economy would need six hundred million new jobs over twenty years. Existing businesses could provide ten to twenty million. The rest, by Hoffman's logic, would come from startups—which meant societies everywhere would need to make entrepreneurship easier.
The circular reasoning was elegant and almost airtight: the problem with the networked economy was insufficient networking, and the solution was more of the platform that sold networking. Reid Hoffman had built the world's largest professional anxiety machine and then written three books arguing that professional anxiety was, properly channeled, a feature rather than a bug.
A Billion Profiles and the AI Pivot
By 2024, LinkedIn had achieved something remarkable and slightly eerie: it had become the professional identity of record for over a billion people across more than two hundred countries. The Harvard Business School case study published that year noted that the LinkedIn profile was "well established as the professional identity of record on the Internet"—a phrase that captured both the platform's dominance and its peculiar ontological status. Your LinkedIn profile was not you. But for an increasing number of professional interactions—recruiting, sales outreach, background checks, partnership evaluation—it functioned as you. The profile had become the avatar, and the avatar had become the person.
The platform's chief product officer, Tomer Cohen, articulated a philosophy that distinguished LinkedIn from every other social network: "LinkedIn exists so people can reach out to professional communities and get their jobs done." The metric that mattered was not time spent on the platform—the universal currency of attention-economy social networks—but whether users got new jobs, acquired new skills, closed new deals. "We might be wrong, but we're not fucking confused," Cohen told The Verge in a formulation that delighted people who had grown weary of social media companies optimizing for engagement at the expense of everything else.
The AI era presented LinkedIn with both its most potent growth vector and its most existential threat. In January 2026, LinkedIn launched a skills verification system in partnership with AI tool makers—Descript, Lovable, Relay.app, Replit—that would validate users' proficiency based on real usage patterns rather than standardized tests. "Employers are no longer simply asking what degree a candidate holds," the company told Fortune. "They want to know what you can actually do." The feature was designed to make the LinkedIn profile not merely a self-reported résumé but a verified capability ledger—a shift from identity-as-narrative to identity-as-evidence.
LinkedIn had simultaneously released fifty free AI courses for all members and introduced AI-generated insights for premium subscribers. Roslansky's vision was explicit: LinkedIn would become the platform where the global workforce transitioned into an AI-powered future. The company's own AI at Work report, based on surveys of over 31,000 participants, documented the accelerating adoption of AI tools across industries and job functions. LinkedIn was positioning itself as both the mapper and the guide of this transition—cataloguing which skills were becoming obsolete, which were surging in demand, and selling the tools to bridge the gap.
Reid Hoffman, meanwhile, had co-founded Inflection AI in 2022 and joined the board of OpenAI (before stepping down), and published
Blitzscaling and a stream of AI-focused books, including
Impromptu and
Superagency. The philosopher-founder who had once discussed AI ethics with Catholic priests at a dinner arranged by McKinsey was now one of Silicon Valley's most prominent AI optimists, arguing that AI agents would amplify rather than diminish human agency. "The most prevalent coding language will be English," he predicted, "and we will all have the ability to add code to what we're doing."
The Network Graph on the Wall
On the wall of Reid Hoffman's office at LinkedIn, alongside photographs of himself with
Michael Bloomberg, Bill Clinton, and Barack Obama, hangs a framed "network graph"—a diagram produced by the company's data-analytics team showing all the connections he maintains through the platform. It is thousands of color-coded lines linking nodes, with Hoffman at the center, by far the most densely connected node. He calls himself, without apparent irony, the Ubernode.
The image captures something essential about LinkedIn that no financial metric can. The platform is, at its core, a map of economic relationships—not the casual bonds of friendship or the parasocial attachments of fandom, but the functional connections through which work gets done, deals get made, careers get built. That map, maintained and updated by a billion people who have every incentive to keep it current (because their livelihoods depend on it), represents a dataset of almost incomprehensible strategic value. It is the X-ray of the global economy's circulatory system.
LinkedIn ran a five-year experiment on nearly twenty million users, randomly varying the strength of connections suggested in the "People You May Know" feature, and confirmed a long-held sociological theory: people are more likely to get a new job through distant acquaintances—what sociologists call "weak ties"—than through close contacts. Users shown connections with whom they shared only ten mutual friends doubled their chances of a new job compared to those shown connections with twenty mutual friends. The experiment was published in Science magazine in 2022 and immediately drew criticism from privacy experts who noted that the unknowing participants whose algorithms were flooded with strong ties had, in effect, been denied weaker-tie opportunities. LinkedIn responded that the study relied on routine A/B testing for user experience purposes.
The controversy illuminated a tension that defines the platform's relationship with its users. LinkedIn's value proposition requires members to voluntarily disclose detailed, accurate, continuously updated information about their professional lives—their skills, their employers, their career aspirations. That data is the raw material from which LinkedIn extracts revenue through recruitment tools, advertising, and premium subscriptions. The exchange is implicit but not quite symmetrical: the user gets a profile that functions as a professional passport; LinkedIn gets a dataset that is, in aggregate, the most comprehensive map of human capital ever assembled. As one data privacy expert observed: "Most users, if you asked them, would say there's no way they would have consented to this kind of study."
But they do consent, every day, every time they update a headline or accept a connection. They consent because the cost of not being on LinkedIn has become, for a growing share of the global professional workforce, higher than the cost of being on it. That calculus—the quiet coercion of network ubiquity—is the moat.
Settlers of Silicon Valley
One evening in 2015, Hoffman arrived at Fuki Sushi, a popular Silicon Valley restaurant, hoping to gather friends for a game of Settlers of Catan—the board game in which players compete to build the fastest-growing settlements. He had produced a custom version for his circle called Startups of Silicon Valley, with the same rules but different nomenclature: products instead of settlements, disrupters instead of robbers, talent instead of wheat. Nobody was available, so he had dinner instead with James Manyika, a McKinsey partner who had grown up in Zimbabwe and become an engineering faculty member at Oxford.
The dinner covered the usual Hoffman territory at speed: a caterer specializing in African-diaspora food, a potential co-hosted dinner for a young entrepreneur from Dubai, an invitation to join Obama's Global Development Council, a meeting Manyika had arranged between Hoffman and a delegation of Catholic priests to discuss AI ethics. "I thought it would be only about social media," Hoffman said. "Instead, it was about A.I."
A waiter entered. "I have an algorithm," Hoffman announced. "If it's a good place, order the special. If it's a bad place, order what they can't screw up."
They ordered the special.
There is, in the granularity of Hoffman's networking rituals—the list-making at the start of every meal, the relentless brokering of connections, the game logic applied to every social interaction—a miniature of the platform itself. LinkedIn is, at scale, what Hoffman does at every dinner: it structures informal human relationships into a systematized, searchable, monetizable graph. It takes the thing that ambitious professionals have always done—maintain relationships, collect favors, signal availability—and removes the friction, the forgetting, the geographic accident. It makes the invisible visible. And then it charges for the view.
Hoffman lives in a four-bedroom house in Palo Alto, drives a Tesla, and does not own a private plane. His only obvious extravagance is the relentless construction of relationship capital—a currency that, unlike the three to four billion dollars in conventional wealth that puts him somewhere between twentieth and thirtieth place among Silicon Valley's richest, has no known ceiling and no diminishing returns.
The Startups of Silicon Valley board game sits somewhere in his house, its custom-printed tiles a monument to the conviction that everything—politics, philanthropy, friendship, the fate of the global labor market—is a game with discoverable rules and optimal strategies. LinkedIn is what happens when that conviction is encoded into software and distributed to a billion players, most of whom don't realize they're playing.
