Part IThe Story
The first thing ever sold on the website that became eBay was broken, and the man who built the website wanted to be sure the buyer knew. The item was a laser pointer, and it went for $14.83. Its seller was a twenty-eight-year-old software engineer named Pierre Omidyar, who had spent a holiday weekend in 1995 writing code for what he called AuctionWeb. When the auction closed, he did something no auction house required and no economist would have predicted. He wrote to the winning bidder to ask whether he understood that the laser pointer did not work.
The buyer understood. The reply has been repeated so often that it has worn smooth, like a coin passed through too many hands, but it is worth reading again for what it actually says:
— The winning bidder on AuctionWeb's first sale, as reportedI'm a collector of broken laser pointers.
Omidyar, by the account in the encyclopedias, was astonished. It is worth pausing on the astonishment, because it contains nearly everything that followed. A seller discloses a defect he isn't obligated to disclose. A buyer reveals a desire nobody could have guessed. Between them a price is set, $14.83, which neither could have set alone. The value of the object turns out to have nothing to do with whether it works and everything to do with who wants it and why. And the transaction rests, from the first, on a small voluntary act of honesty between two strangers who will never meet.
Thirty years later the system that grew from that exchange handles some $73 billion a year in merchandise, rates its buyers and sellers against each other in perpetuity, and employs people whose job is to catch strangers lying to strangers. In May 2026 a video-game retailer offered roughly $56 billion for it, and the company's board dismissed the offer in four words. Omidyar was by then long gone from the board. He had become something else: a funder of investigative journalists, of antitrust scholars, of a Facebook whistleblower's press operation, of an AI company's equity, of the very idea that platforms like the one he built might be too powerful to be trusted. He had spent the first act of his life building machinery for strangers to trust each other. He has spent the second act paying people to distrust the machinery.
It would be too neat to call this a reversal. It is closer to a single preoccupation that changed scale. The question in 1995 was whether a buyer could trust a seller. The question in 2025 was whether anyone could trust the record itself — the reputation score, the news story, the encyclopedia entry, the founding myth. Omidyar's life has a way of returning to that broken laser pointer: the defect disclosed or concealed, the value set by the people who know, the story told by whoever gets there first.
By the Numbers
From a Broken Laser Pointer
$14.83Price of the first item sold on AuctionWeb — a broken laser pointer
$1.9BeBay's approximate market capitalization on its first day of trading, September 1998
$73BMerchandise volume transacted on eBay in 2023, across 132 million active buyers
$1.4BPrice eBay paid for PayPal on October 3, 2002
~$2BAwarded by Omidyar Network to 700+ organizations by 2025
$12.5BOmidyar's estimated net worth, per Forbes, August 2026
3Citizenships held: Iran, France, United States
The Letters You Drop
He was born Parviz Morad Omidyar on June 21, 1967, in Paris, to Iranian parents who had come to France to study. His mother, Elahé Mir-Djalali, became an academic in linguistics, a scholar of how meaning survives translation. His father, Cyrus Omidyar, was a surgeon. The boy who would eventually be listed in one encyclopedia's categories as an American billionaire, a French billionaire, a French Buddhist, a businessman from Paris, a businessman from Honolulu, and a person of Iranian descent began, in other words, as a name that would not survive the trip.
Parviz became Pierre. The family moved to the United States while he was still a child, and the itinerary of his schooling reads like a diplomat's posting history: two years at Punahou School in Honolulu; ninth grade at the Potomac School in McLean, Virginia, where he first took an interest in computers; a high school diploma in 1984 from St. Andrew's Episcopal School in Potomac, Maryland. Then Tufts, a computer science degree in 1988. He holds citizenship in three countries — Iran, France, the United States — which is either a sign of belonging everywhere or of belonging nowhere in particular, and the record does not tell us which way he would put it.
Names, for Omidyar, have always been provisional things. Consider the company. It launched as AuctionWeb, one feature among several on his personal website, which also hosted, for reasons that remain obscure even to the reporters who went looking in 2014, a section devoted to the Ebola virus. The domain on which all of this lived was ebay.com. In September 1997 the service took that name for itself. The derivation is the kind of story founders usually polish into legend and Omidyar apparently never bothered to. He had named his consulting firm Echo Bay Technology Group, after a recreation area near Lake Mead, Nevada, because, as he put it, the name "sounded cool." But echobay.com was already registered — to Echo Bay Mines, a Canadian gold-mining company — so he dropped three letters. The c, the h, the o. What remained became one of the most valuable brands of the first internet age.
A gold mine got the full name. The marketplace got the abbreviation.
There is a detail here that the encyclopedias record separately, in different articles, without connecting: Omidyar and his wife, Pamela, now own property in Henderson, Nevada, a short drive from the lake whose cove he once wanted to borrow for a company. The man who couldn't have the name ended up with the neighborhood.
And then there is the name story that wasn't true. For years the canonical origin of eBay held that Omidyar built the site so his fiancée could trade Pez candy dispensers. It is a lovely story — domestic, whimsical, legible to a morning-show audience — and it was fabricated in 1997 by a public relations manager trying to interest the press. Adam Cohen exposed the invention in his 2002 history The Perfect Store, and eBay confirmed it. The company that would build its fortune on verifiable reputation had, at the moment it most needed attention, outsourced its own founding to a fiction. The Pez story still circulates on YouTube under titles like "Guy Built $30B Empire Just to Help GF Sell Pez Dispensers!" — a counterfeit, in other words, still trading long after the authentication.
Even the birthday is contested. The encyclopedia article on eBay says the company was founded on September 3, 1995. The encyclopedia article on Omidyar says he launched the service on Labor Day, Monday, September 4, 1995. Labor Day fell on the fourth that year. Two entries in the same reference work, maintained by volunteers who do not necessarily read each other, disagree by a day about the origin of a company that recorded 132 million active buyers in 2023. Nobody seems troubled by this. It is the sort of small defect a collector might prize.
The General's Money
AuctionWeb was a hobby until it wasn't, and the moment it stopped being a hobby has a dollar figure attached. In February 1996, Omidyar's internet service provider informed him that traffic to his site had grown too heavy for a personal account. He would need a business account. The monthly bill would rise from $30 to $250.
What happened next has two versions, and the difference between them is instructive.
The familiar version is the bootstrapper's version: Omidyar began charging sellers a small fee on each sale, the checks started arriving in the mail, and the site paid for itself. That is true as far as it goes. AuctionWeb made $1,000 in its first month of fees, more than it cost to run, and $2,500 the month after. The company's first employee, Chris Agarpao, was hired to process the mailed checks — a detail that belongs in any honest history of e-commerce, since it means the first great online marketplace ran, at the start, on envelopes.
The less familiar version appears in the encyclopedia's eBay entry, sourced to a 2025 Czech-language architectural history by Reza Alirezaijan: the price increase "prompted Omidyar to obtain support from his mother's family." His maternal grandfather was General Mahmud Mir-Djalali, a former vice chairman of the general staff of Iran's imperial armed forces, described as the "father" of the country's defense industries and its tank forces, a man with close relationships, the same entry says, with Škoda Works and Fabrique Nationale Herstal — the Czech industrial giant and the Belgian arms maker. The encyclopedia credits him with a role in the 1921 rise of the Pahlavi dynasty and in the overthrow of Mohammad Mossadegh.
I want to be careful here, because the record is careful in some places and remarkably uncareful in others. The claims about the general are recent, single-sourced in the passages supplied, and attached to a book about Czechoslovak architects in the Shah's Iran. The Omidyar Network's encyclopedia entry goes further — it describes the 2004 foundation as "a continuation of earlier philanthropic efforts" of the general, and its infobox, as of the version I read, lists among the network's "Key people" both Mir-Djalali and an associate named Nasser Yeganeh, who the same article says died under suspicious circumstances in 1993. The article layers on Soviet military intelligence contacts and French generals, all from the same 2025 source. A man who died more than a decade before Omidyar Network existed is listed alongside its current chief executive as one of its key people.
What can be said with confidence is narrower and more interesting than any of the intrigue. The founding story of eBay, as the world was first told it, was a fabrication about Pez dispensers. The founding story as the record is now being rewritten contains a general, an arms network, and family money arriving to cover a $220 monthly increase. Somewhere between those two versions lies a 28-year-old with $1 million from a recent acquisition and a day job at General Magic. The most valuable thing a reputation system can do is let you check the claim against the transaction history. Omidyar built that system for other people. His own history remains open to edit.
A Hobby With a Fee Schedule
He did not quit his job. This is the part of the story that the founder mythology tends to skip, because it is boring, and because it contradicts the prevailing theology of the burned boat.
By 1995 Omidyar had already been through one startup cycle. After Tufts he had worked at Claris, an Apple subsidiary, on the team that upgraded MacDraw into MacDraw II — a drawing program, a tool for making the cursor do what the hand intended. In 1991 he co-founded Ink Development, a company built around pen-based computing, the dream that you would write on your computer instead of typing into it. The dream did not arrive on schedule. Ink Development pivoted to electronic commerce and renamed itself eShop. On June 11, 1996, Microsoft acquired it for less than $50 million. Omidyar's share was $1 million.
So the man who wrote AuctionWeb over a holiday weekend was not a naïf. He had watched a company bet on a future, miss, reinvent itself around commerce, and get absorbed by a larger company for a modest sum. He had a job at General Magic, the celebrated and ill-fated Apple spinoff. And he treated his auction site the way a careful engineer treats a side project: he waited to see whether it would carry its own weight.
It did. The small fee on each sale financed the expansion, and the revenue, by the encyclopedia's account, "soon outstripped his salary at General Magic." Nine months after launch, Omidyar went full-time. Nine months. Not a leap. A gestation.
The growth is one of those curves people describe as hockey sticks, though hockey sticks bend only once. By November 1996, fourteen months in, the site had hosted more than 200,000 auctions (the Omidyar entry puts it at 250,000; the encyclopedias, again, disagree by a rounding error), and Omidyar had signed a deal with Electronic Travel Auctions to sell airline seats, hotel rooms, and cruise berths. In the first month of 1997 the site hosted two million auctions. By the middle of that year it was hosting nearly 800,000 a day. Jeffrey Skoll, a Canadian with a Stanford MBA, came on as the company's first president in early 1996 — the first in a line of operators Omidyar would hire to run the thing he had built, a pattern that would define both his fortune and his temperament.
In 1997 the venture firm Benchmark put in $6.7 million. That year, roughly ten percent of all listings on the site were Beanie Babies.
It is hard to overstate how strange this is. The defining commercial internet company of the late 1990s was, for a stretch of its adolescence, a secondary market for plush toys stuffed with plastic pellets. Ty, the manufacturer, had set up its own site for collectors to trade used Beanie Babies, but there was no way to sort the listings. eBay could sort. So the collectors came — and here the laser pointer returns, because the Beanie Baby boom was the broken-pointer phenomenon at industrial scale: objects whose value had almost nothing to do with their function and everything to do with the intensity of a particular community's desire.
Omidyar seems to have understood this with unusual clarity, and to have been unusually willing to say so. In the risk-factors section of the annual report filed with the Securities and Exchange Commission in 1998, the company noted its dependence on the continued strength of the Beanie Babies market. It is the corporate equivalent of the email to the laser-pointer buyer: You understand this is broken, right? Most companies going public in 1998 were busy disclosing nothing. eBay disclosed the toy.
The Chair and the Keys
In March 1998 Meg Whitman became president and chief executive. The company she inherited had 30 employees, 500,000 users, and $4.7 million in United States revenue. Whitman was a Hasbro and Procter & Gamble veteran, an operator from the world of brands and distribution — the opposite, in almost every particular, of a pen-computing engineer who named his company after a cove because it sounded cool. She would run eBay until January 2008, nearly a decade, and leave to enter politics.
Six months after she arrived, in September 1998, at the height of what would later be called the dot-com bubble, eBay went public in an offering led by its chief financial officer, Gary Bengier. Shares opened on the Nasdaq at $18 and closed the first day at $53. The market capitalization reached nearly $1.9 billion. Omidyar and Skoll were billionaires. The next year Omidyar was named the national winner of Ernst & Young's Entrepreneur of the Year award.
And then he receded.
Omidyar served as chairman from 1998 to 2015, and the striking thing about those seventeen years, in the record supplied, is how little of eBay's subsequent history carries his fingerprints as an operator. The acquisitions, the international campaigns, the litigation — the entries attribute these to "eBay," to Whitman, to the board. He occupied the chair. Others held the keys.
This was a choice, and it was rare. The founders of his era who became household names mostly did so by refusing to step back, by fusing their personalities with their companies until the two could not be pried apart. Omidyar did the opposite. He had built a system designed to work without anyone watching each transaction; perhaps it was natural that he would build a company designed to work without him watching it. In 2008, his 178 million shares were worth about $4.45 billion. He was, by then, already spending much of his attention elsewhere.
What his company did in his relative absence is a case study in the difference between buying adjacency and buying ambition. In June 2000 eBay paid $312 million in stock for Half.com, a site for used books and media; it shut Half.com down in August 2017. On October 3, 2002, it paid $1.4 billion for PayPal, the payments company built by a group of young technologists who would later be called a mafia, and phased out its own payments product, Billpoint, in January 2003. PayPal solved eBay's oldest problem — the envelope, the check, Chris Agarpao at a desk — and became, for a dozen years, the most important acquisition the company ever made.
Then came the ambitions. In August 2004 eBay bought a quarter of Craigslist, the anti-corporate classifieds site, from a former Craigslist executive named Phillip Knowlton for $32 million. It ended in litigation: eBay sued in April 2008, claiming Craigslist had diluted its stake; Craigslist countersued, alleging eBay had used its board seat to gather intelligence for a competing classifieds business. In September 2010 a Delaware judge, William B. Chandler III, found that Craigslist founders Jim Buckmaster and Craig Newmark had breached their fiduciary duty of loyalty and restored eBay's stake. In June 2015 eBay sold the stake back. Eleven years. A quarter of a bulletin board.
In October 2005 eBay paid $2.6 billion for Skype, the internet telephone service. The logic was that buyers and sellers would want to talk. They did not, or not through Skype. In November 2009, after failing to integrate the service into its marketplace, eBay sold 70 percent to a consortium led by Silver Lake Partners and Marc Andreessen at a $2.75 billion valuation. Eighteen months later, in May 2011, Microsoft bought all of Skype for $8.5 billion — the same Microsoft that had once bought Omidyar's eShop for under $50 million. The asset was worth more to almost anyone than to the company that couldn't figure out what it was for.
In January 2007 eBay paid $310 million for StubHub; in February 2020 it sold StubHub to Viagogo for $4.05 billion. In June 2011 it paid $2.4 billion for GSI Commerce, the King of Prussia, Pennsylvania, e-commerce services firm founded by Michael Rubin, a man who had started selling sporting goods as a teenager and who would go on, after the sale, to build Fanatics into a multibillion-dollar business of his own. eBay rebranded GSI as eBay Enterprise in 2013 and sold it to a Permira-led consortium for $925 million in November 2015. Rubin, in other words, sold eBay a company for $2.4 billion, kept the piece that became Fanatics, and watched the rest of it come back on the market at well under half the price.
Rubin's biography rhymes with Omidyar's in an inverted key: a founder who sold to a bigger company and kept going, building the next thing faster than the acquirer could digest the last one. Omidyar had sold eShop to Microsoft and gone on to build eBay. The pattern repeats. The big company buys the past. The founder walks off with the future.
Local Knowledge
In 2003, eBay bought EachNet, the leading online auction platform in China, and with it an 85 percent share of the Chinese market. Within four years, its share had fallen to 7.7 percent. In December 2006 the company announced it would turn its Chinese site over to a joint venture controlled by Tom Online: eBay put in $40 million for 49 percent, Tom Online put in $20 million for 51 percent and management control. The company described the departure not as a failure but, in the phrase quoted in the record, as "an evolution of our strategy here in China."
The evolution had a name, and the name was Taobao.
The postmortems, assembled years later, read like a manual of what not to do. eBay charged listing and platform fees; Taobao was free. eBay migrated its Chinese platform onto its global technology stack, hosted in the United States, and the site became slow and crashed repeatedly, bleeding trust with every outage. Chinese buyers preferred to talk directly to sellers and build relationships before transacting; Taobao offered integrated instant messaging, and eBay did not. Chinese consumers favored fixed prices; eBay was built on auctions. And in replacing EachNet's local systems with global ones, eBay weakened its own acquisition's ability to respond to the market it had paid to enter. It failed, as one analysis put it, to localize — it copied its global approach rather than adapting to how Chinese consumers actually behaved.
It is impossible to read this without hearing an echo of the laser pointer. The whole insight of AuctionWeb was that value is set by the people who know — the collector of broken pointers, the Beanie Baby completist — and not by the person who built the platform. In China, the platform forgot it was a platform. It tried to tell the market what it wanted.
There is a body of thought that explains exactly this failure, and it surfaces, unexpectedly, in another encyclopedia entry supplied for this profile: the history of Wikipedia. Jimmy Wales, who built that encyclopedia, has called Friedrich Hayek's 1945 essay "The Use of Knowledge in Society" central to his thinking about how to manage the project. The essay argues that knowledge is dispersed — that each individual knows only a fraction of what is known collectively — and that decisions are therefore best made by those with local knowledge rather than by a central authority. Wales, a finance student when he read it, was fascinated by the incentives that led volunteers to build free software.
Omidyar's marketplace and Wales's encyclopedia are the two great Hayekian machines of the early web: one for pricing objects, the other for pricing facts. Both run on strangers trusting strangers, and both developed elaborate immune systems — feedback scores, edit histories, protection levels, arbitration — when the strangers turned out not to be entirely trustworthy. Wales, in September 2001, told the New York Times: "It's kind of surprising that you could just open up a site and let people work." It could have been Omidyar talking about 1995.
In August 2009 the two machines touched. Omidyar Network gave the Wikimedia Foundation $2 million over two years, and Matt Halprin, a partner at Omidyar Network, joined Wikimedia's board of trustees. The headline on the NBC Bay Area story was blunter than any press release:
— NBC Bay Area, August 25, 2009, under the headline “$2 Million Buys eBay Founder a Wikipedia Seat”The Omidyar Network, a fund established by eBay billionaire Pierre Omidyar, has pledged $2 million over two years to support the site, and in return landed fund partner Matt Halprin a seat on a shaken-up board of trustees.
The next day, on the discussion page attached to Omidyar Network's own encyclopedia entry, an editor posting as Thekohser complained that the article read like "PR puff," and then asked the question that would hang over the page for the next sixteen years: "what are we to do, with an Omidyar partner now sitting on the Wikimedia Foundation board of trustees?" Four days later another editor, Seth Finkelstein, traced the IP address behind much of the article's early text. It belonged, he reported, to "Omidyar Network Services, LLC."
Local knowledge cuts both ways.
The Liability of Strangers
A marketplace for strangers is also a marketplace for fraud, and eBay spent much of the 2000s discovering, through litigation and catalog, how many ways strangers can lie to one another. Sellers shilled their own auctions with dummy accounts. They took payment and shipped nothing; shipped something else; shipped stolen goods; listed devices with battery capacities physically impossible for their size. Buyers claimed nondelivery of delivered items, returned boxes full of garbage, stripped parts from goods before sending them back, broke nonrefundable items and blamed the seller. Users developed counter-techniques — "auction sniping," bidding in the final seconds — that were themselves a kind of distrust made into a habit.
In 2004 Tiffany & Co. sued, claiming that more than 70 percent of the Tiffany silver jewelry offered on eBay was fake. On July 14, 2008, a federal judge ruled that eBay had no legal responsibility to monitor its users for counterfeits; the Second Circuit affirmed in 2010. Europe saw it differently. In June 2008 a Paris court awarded LVMH €40 million over counterfeit bags and perfumes and permanently barred eBay auctions of LVMH perfumes, counterfeit or not. That same month a court in Troyes ordered eBay to pay Hermès €20,000 over the sale of two counterfeit bags in 2006 — and to post the ruling on the home page of its French site for three months. Two bags. Three months on the front page. The French court understood something about reputation that the American one declined to: the punishment that hurts a platform is not the fine but the disclosure.
There was also the matter of the "Buy It Now" button, which a company called MercExchange claimed infringed its patent. A jury found eBay guilty of willful infringement on May 28, 2003, and awarded $35 million. The case went to the Supreme Court, which ruled on May 15, 2006, that a permanent injunction is neither required nor automatic in patent cases — a decision that reshaped patent law far beyond eBay. The company that had built its identity on auctions ended up in the history books defending the right to sell things at a fixed price.
Through all of it, the American legal architecture tended to treat the platform as a conduit rather than a seller. When the Justice Department sued eBay in September 2023 for allowing the sale of devices that defeat automobile pollution controls, restricted-use pesticides, and paint strippers containing methylene chloride, the court dismissed the case on the ground that the Communications Decency Act shielded eBay from liability — the same Section 230 that, in limited ways, has protected Wikipedia from defamation suits. The law, in short, agreed with the original design: the platform sets the table, and the strangers are responsible for what they serve.
Then the platform served something itself.
In June 2020 federal prosecutors charged six former eBay employees, including executives, with an aggressive cyberstalking campaign against Ina and David Steiner, the editors of EcommerceBytes, a small newsletter in Natick, Massachusetts, that eBay's leadership regarded as critical. The deliveries ordered to the Steiners' home are now part of the public record, and they read like the inventory of a nightmare: a preserved fetal pig, a bloody pig Halloween mask, a funeral wreath, a book on surviving the loss of a spouse, pornography, live cockroaches. The defendants vandalized the couple's home. They pleaded guilty, and most went to prison. Devin Wenig, the chief executive the newsletter had frequently targeted, was found by investigators to have had no knowledge of the harassment and was not charged; he had left in September 2019 under pressure from the activist fund Elliott Investment Management, with a $57 million severance package. Steve Wymer, the chief communications officer, was fired for cause over alleged involvement, was not charged, and was hired by the local chapter of the Boys & Girls Clubs.
A company founded on an email disclosing a defect to a stranger had sent a stranger a funeral wreath for writing about its defects.
On September 10, 2020, the Wall Street Journal reported that Pierre Omidyar was stepping down from eBay's board, part of a broader overhaul; he retained the title director emeritus. Sixteen days later, on September 26, the New York Times published its long account of the affair under the headline "Inside eBay's Cockroach Cult." The record supplied draws no connection between the two events, and I will not invent one. But the calendar has its own way of editorializing. The founder left the room in the same month the room's worst story was told in full.
Philanthropy With a Cap Table
By the time he left the board, Omidyar had been giving money away for sixteen years in a form that confused nearly everyone who tried to describe it.
Pierre and Pam Omidyar founded Omidyar Network in 2004. It is, in its own early phrase, a "philanthropic investment firm" — a hybrid of a 501(c)(3) foundation and a limited liability company, able to make grants to nonprofits and equity investments in for-profit companies from the same pool of intent. It was an early version of what the philanthropy press would later call the "LLC giving" model, adopted afterward by other tech fortunes. Its stated purpose, in the encyclopedia's paraphrase, was harnessing the power of markets to create opportunities for people to improve their lives. The founder of the most famous marketplace on the internet had decided that markets were also a delivery mechanism for charity.
The early bets were in microfinance, including the organization Unitus; the network was, according to one 2013 academic paper, an advocate of microlending as a way to extend financial access. The controversies followed the capital: in July 2010 the New York Times ran a story under the headline "Rich I.P.O. Brings Controversy to SKS Microfinance," about the moment when lending to the poor turned out to be lucrative for the lenders. In 2012 the network became a partner of the Better Than Cash Alliance, which encourages governments to distribute money digitally rather than in cash. The committed totals rose and rose — $992 million, then more than $1.5 billion, then, by 2025, almost $2 billion to more than 700 organizations.
What distinguishes the network from a conventional foundation is less its size than its metabolism. It incubates things and then sets them loose. Democracy Fund began as an Omidyar Network initiative in 2011 and became an independent foundation three years later. Spero Ventures, a venture fund, spun out in March 2018. Later that year the Governance & Citizen Engagement initiative became Luminate, a global philanthropy for civic empowerment, digital rights, financial transparency, and independent media. In 2019 the Financial Inclusion initiative became Flourish. A year later the education portfolio became Imaginable Futures, and in January 2020 the property-rights work became a nonprofit called PlaceFund. The structure is the eBay chairmanship translated into philanthropy: build it, hand over the keys, keep a seat at the table, or don't.
The exits were as revealing as the spinoffs. In 2010 Omidyar Network opened an operation in India to invest in startups; by July 2023 it had $673 million in investments under management there, in companies including 1mg, Vedantu, DealShare, and Pratilipi. In 2022 it was reported to be under investigation by India's Central Bureau of Investigation over alleged violations concerning overseas donations. In December 2023 it announced it would cease operations in India by the end of 2024. Thirteen years. Like China, a market entered at scale and left by announcement — though this time the departing party was the investor, not the platform.
And then, in April 2024, Omidyar Network, together with Ford and Cummings, bought a stake in Anthropic, the generative AI company. Later that year it reorganized its strategy around building consensus-based guardrails for digital technology, artificial intelligence included. Mike Kubzansky, a former Monitor Group partner who ran the network from 2018 to 2026, gave an interview that summer under the headline "AI Opened a Window." In March 2026 he was succeeded by Michele Jawando, who had come in as senior vice president for programs.
The network's encyclopedia entry has its own small history of hands on the keyboard. In 2014 an editor slapped an "advert" tag on it and wrote, on the discussion page, "As written, this currently reads like a PR piece. This is embarrassing to Omidyar, Wikipedia, editors, and readers." In June 2022 an editor called BlueRoses13 arrived with a disclosure: "Omidyar has hired me to fix these issues in line with Wikipedia's policies." The fixes were proposed in numbered lists, sourced to the Times and CNBC and Barron's, and accepted by unpaid volunteers. In September 2025 another disclosed paid editor, working through the agency Porter Novelli on behalf of The Omidyar Group, submitted a fresh round. A volunteer reviewer accepted most of them and declined some, noting that a proposed passage about the network's history "sounded a bit too WP:PROMOTIONAL to me."
This is, if you squint, the system working exactly as designed: a party with an interest discloses the interest, makes its claims in the open, and lets strangers rate them. It is the email to the laser-pointer buyer, scaled to an institution. You understand that I'm being paid to say this, right?
The Insurgency
In 2010 Omidyar launched an online investigative news service in Hawaii called Honolulu Civil Beat, devoted to civic affairs. It was named the state's best news website in 2010, 2011, and 2012. On September 4, 2013 — eighteen years to the day after the Labor Day launch of AuctionWeb, by one encyclopedia's reckoning — Civil Beat began a partnership with the Huffington Post to launch HuffPost Hawaii.
Six weeks later, everything got larger. In October 2013, prompted by Edward Snowden's leaks of National Security Agency documents, Omidyar announced he would create and finance a journalism venture that would come to be called First Look Media. Forbes reported he was ready to spend $250 million on it. On February 10, 2014, First Look launched The Intercept, built around the journalists who had broken the Snowden story — Glenn Greenwald, Laura Poitras, Jeremy Scahill — with Dan Froomkin, John Temple, and the New York University press critic Jay Rosen also drawn in. Later Omidyar's name appeared among the executive producers of two films about institutions concealing their defects: Merchants of Doubt (2014), about the professional manufacture of scientific uncertainty, and Spotlight (2015), about the Boston Globe's investigation of abuse in the Catholic Church. Spotlight won the Academy Award for Best Picture.
Andrew Rice, profiling Omidyar for New York magazine in November 2014, called the piece "The Pierre Omidyar Insurgency" and subtitled it "How the Snowden Leaks Gave Pierre Omidyar a Cause — and an Enemy." His method was the method his subject's money was paying others to use.
— Andrew Rice, “The Pierre Omidyar Insurgency,” New York, November 2, 2014It is possible to begin to discern Omidyar's motivations with a little online surveillance.
There is a wryness in that line that the whole decade earned. The founder of a company that, in 2020, was caught port-scanning its users' computers through obfuscated JavaScript — probing their machines for remote-access software, presumably to catch fraud — had become the patron of a publication founded to expose surveillance. Both things were true at once. Neither canceled the other.
The insurgency widened. Omidyar bankrolled the Open Markets Institute, the anti-monopoly think tank, and the digital-rights group Public Knowledge, in what Politico described as a fight against big tech companies he regarded as overly powerful and destructive to democracy. Omidyar Network distributed widely read papers laying out antitrust cases against Facebook and Google. It funded the Center for Humane Technology. In 2020 it invested $150,000 in Whistleblower Aid. In February 2021 it hosted a series on whistleblowing in the tech industry — months, Politico noted, before the Facebook product manager Frances Haugen came forward. When she did, Omidyar's philanthropy Luminate handled her press and government relations in Europe. The precise lines mattered to the organizations involved, enough that a paid editor later asked to have the network's encyclopedia entry corrected, citing a Luminate spokesperson's statement to Insider: "We are the only organization from The Omidyar Group that is directly supporting the work of Frances Haugen and her team." The volunteers agreed and deleted the sentence.
In April 2019 Politico published a piece under the headline "How the Intercept Is Fueling the Democratic Civil War." Omidyar is a long-term, major donor to Democratic candidates and organizations. His publication was, by that account, at war with them. Insurgencies are hard to steer once you've paid for them.
The 2022 Reporters in Residence program — monthly stipends, one cycle advertised at $8,000, for freelance journalists covering economic issues — is the quietest expression of the same idea. Not a newsroom, not a manifesto. A check, mailed to a stranger, so that someone, somewhere, keeps looking at the defects.
The Island and the Pledge
He went back to the island.
Omidyar spent only two years at Punahou School as a boy, but he now sits on its board of trustees, and Hawaii has become the center of gravity for the private life the record otherwise barely describes. He and Pam own property in Honolulu as well as Henderson. A 2009 Honolulu newspaper story about his move was headlined "A 'wonderful rediscovery' in isles." He is, by Forbes's state-by-state reckoning, the wealthiest person in Hawaii. He founded the Ulupono Initiative there; Civil Beat covers the statehouse there. The boy who was Parviz in Paris and Pierre in Virginia and a temporary pupil in Honolulu returned to the place he had been briefly, the way some people return to a word in a language they only half-learned as children.
The rest of the private record is a short list of affiliations, each a kind of public promise. He is described as a follower of the Dalai Lama. He sits on the Berggruen Institute's 21st Century Council. He is listed among the people associated with the Santa Fe Institute, the home of complexity science, where researchers study how order emerges from decentralized agents — the Hayekian marketplace, rendered as mathematics. He invested in the Montage resort in Laguna Beach. He was named in the Paradise Papers. In 2010 he joined Bill Gates and Warren Buffett as a signatory of the Giving Pledge, declaring his intention to give away most of his wealth during his lifetime. In 2019 he gave away roughly $500 million. That same year Forbes put his net worth at $13.1 billion; by August 2026 it was $12.5 billion, 255th in the world. The pledge moves slower than the market. Most pledges do.
Every year for two decades, the company he founded hosted a charity auction for lunch with Buffett — eight people at Smith & Wollensky in Manhattan, proceeds to the Glide Foundation in San Francisco. Over 21 auctions between 2000 and 2022 it raised $53.2 million, the final lot going for $19 million. It was the laser pointer in reverse: an object of no intrinsic value — a steak, a table, an hour — priced by the intensity of a single collector's desire, the money sent to strangers.
eBay itself carried on without him. It paid its first dividend in March 2019, under investor pressure. It sold its classifieds business to Adevinta for $2.5 billion in cash plus shares, sold its South Korean business to Emart for $3 billion, began authenticating sneakers in 2020 and later watches, handbags, jewelry, and apparel. It bought companies whose entire purpose was detecting fakes — 3PM Shield, which uses artificial intelligence to flag fraudulent goods and sellers, and Certilogo, which issues AI-powered digital IDs for fashion. The platform that had once argued in court it had no duty to monitor counterfeits now sold certainty as a feature. In 2025 it reported $11.1 billion in revenue, 12,300 employees, operations in 190 markets. On May 3, 2026, GameStop, already holding about five percent of the company through shares and derivatives, proposed to buy eBay for about $56 billion in cash and stock. eBay rejected the bid outright as "neither credible nor attractive." It is the kind of phrase a reputation system might produce, if reputation systems could talk.
Omidyar did not appear in the coverage. He is director emeritus. His name is no longer on the company's board; it is on a network, a group, a fund at Tufts for microfinance, a line in a Wikipedia infobox next to a general who may or may not have been there.
I went back, at the end, to the encyclopedia page for Omidyar Network, because it is the closest thing in the record to a transaction history for the second half of his life — every claim made, every claim disputed, every paid editor disclosing the payment. In September 2025 the disclosed editor working on the Omidyar Group's behalf had asked, among a dozen other changes, that the article's first sentence be rewritten. The cited source, the request said, did not support it. In November a volunteer marked the request "partly done" and closed it.
In the version I read, the first sentence still stands, and the two words the request had asked to remove are still the first two words after the network's name. Omidyar Network, it says, is a self-styled "philanthropic investment firm." It is listed with its defect. Somebody, presumably, is collecting.
Part IIThe Playbook
What follows is not a portrait of a heroic operator, because Omidyar was, by design, rarely the operator. It is an extraction of the patterns embedded in what he built and how he let go of it — the AuctionWeb fee schedule, the risk-factor disclosure, the chairman's restraint, the LLC structure, the spinoffs, the insurgent check. Some of these principles come from his successes; several come from failures that happened on his watch but not by his hand. Both are instructive. The tactics are written for founders, operators, and allocators who build systems for other people to use.
The principles
- 1.Disclose the defect before anyone asks.
- 2.Let the users tell you what the product is.
- 3.Let revenue, not conviction, set the date you quit.
- 4.Hire the operator; keep the chair.
- 5.Name your fragile dependency in writing.
- 6.Defer to local knowledge, or lose the market.
- 7.Buy complements, not ambitions.
- 8.Make the legal structure match the intent.
- 9.Incubate, then spin out.
- 10.Fund the critics of your own category.
- 11.Kill the useful myth before it hardens.
- 12.Your name stays on the building after you leave.
Principle 1
Disclose the defect before anyone asks
The founding gesture of eBay was not a feature. It was an email asking a buyer whether he understood the laser pointer was broken. Nothing in the auction's mechanics required it; the buyer had bid on a listing, and the listing was a listing. But Omidyar's instinct was that a marketplace between strangers survives only if the people running it behave as though the counterparty deserves the truth — and that the platform's founder must model the behavior he hopes users will adopt.
The same reflex shows up three years later, at a far higher stake, when eBay told the SEC in its 1998 filing that its business depended on the Beanie Babies market. It reappears, almost comically, in the French court's order that eBay post a ruling against itself on its own home page — a disclosure imposed rather than volunteered, and therefore far more damaging. The difference between those two disclosures is the whole lesson. When you announce your defect, it reads as integrity. When a judge announces it for you, it reads as a verdict.
The counterexample is the stalking scandal, in which a company built on disclosure tried to suppress a newsletter that disclosed its defects. Defensive concealment is not merely unethical; for a trust business, it is self-annihilating.
Tactic: Before your next launch, write down the single most embarrassing true thing about the product and put it where your first customers will see it.
Principle 2
Let the users tell you what the product is
Omidyar did not design eBay as a Beanie Baby exchange, an airline-ticket broker, or a market for broken electronics. Users made it all three. The laser-pointer buyer taught him that value is a property of desire, not function; the Beanie Baby collectors taught him that a platform's real advantage might be as humble as letting people sort listings, something the toy's own manufacturer had failed to do. His role was to notice, to be astonished, and to get out of the way.
This is harder than it sounds, because founders typically arrive with a thesis, and theses resist evidence. The discipline is to treat strange early demand not as noise to be filtered but as the most valuable signal you will ever receive — the moment the market reveals what it actually wants before you've had the chance to convince it otherwise.
Tactic: Keep a running log of every early use of your product that surprises you, and review it monthly as a candidate roadmap rather than a list of edge cases.
Principle 3
Let revenue, not conviction, set the date you quit
The culture of startups prizes the burned boat — the founder who quits the job, drains the savings, and commits before the evidence is in. Omidyar did the opposite. He kept his job at General Magic. He charged a small fee on every sale from early on. He watched the revenue until it exceeded his salary, and then, nine months after launch, he left. When the ISP raised his bill from $30 to $250, the business had to justify the cost; the first month of fees, $1,000, did.
He also arrived with a buffer from a prior exit — $1 million from Microsoft's eShop acquisition — which matters. The lesson is not that founders should be timid. It is that a fee schedule is a truth-telling device. Free products can hide whether anyone values them for years. Paid ones tell you within weeks.
Tactic: Charge something — anything — from the first transaction, and set an explicit revenue threshold that triggers your full-time commitment.
Principle 4
Hire the operator; keep the chair
Omidyar brought in Jeffrey Skoll as president in early 1996 and Meg Whitman as chief executive in March 1998, when the company had 30 employees and $4.7 million in revenue. He then served seventeen years as chairman while others ran the business. The IPO six months after Whitman's arrival made him a billionaire; the company she ran for a decade made him far richer.
The pattern recurs throughout his career. At Omidyar Network he hired CEOs; at First Look he hired star journalists; at Luminate, Flourish, and Imaginable Futures he handed the keys to independent leadership. Few founders of his era could separate their identity from their company. Omidyar seems to have understood early that the skills of the person who invents a system and the person who scales it are rarely found in the same body — and that pretending otherwise is a vanity the shareholders pay for.
The cost is real: a chairman who doesn't operate cannot fully govern what he doesn't see. The acquisitions that went wrong — Skype, EachNet, the Craigslist stake — happened in that gap.
Tactic: Write down the specific moment (a headcount, a revenue line, a market) at which you'll hire someone better than you to run the company, and decide now which seat you'll keep.
Principle 5
Name your fragile dependency in writing
Most companies going public in the 1998 bubble disclosed as little as legally possible. eBay's annual report named the Beanie Babies market as a dependency. That is an unusual thing to tell investors about a company being valued at nearly $1.9 billion — and an unusually useful thing to tell yourself.
Concentration risk is the most common silent killer of platforms: one category, one supplier, one channel, one partner. Naming it publicly forces the organization to treat it as a problem to be solved rather than a secret to be protected. It also inoculates the company's reputation against the moment the dependency breaks, which, for Beanie Babies, it did.
Tactic: In your next board deck, add a slide titled "What we'd lose if this one thing stopped" and update it every quarter.
Principle 6
Defer to local knowledge, or lose the market
eBay held 85 percent of China's online auction market after buying EachNet in 2003. Four years later it held 7.7 percent. The causes are a catalog of central planning: global fees imposed on a market that expected free, a U.S.-hosted technology stack that crashed, an auction model in a country that preferred fixed prices, no instant messaging for buyers who wanted to build relationships before buying, and the dismantling of the acquired company's local systems in favor of global ones.
The irony is acute, because eBay's original insight was Hayekian: the people closest to the object know its value best. Wikipedia's founder cites the same essay as central to his project. In China, eBay stopped being a platform for local knowledge and became an authority imposing a global template. Taobao, which listened, won.
⚖
Global Template vs. Local Knowledge: eBay in China
How a platform built on dispersed knowledge lost a market by centralizing.
| Dimension | eBay's approach | What the market wanted | Outcome |
|---|---|---|---|
| Pricing to sellers | Listing and platform fees | Free (Taobao's model) | Lost sellers |
| Infrastructure | Migrated to U.S.-based global stack | Fast, reliable local service | Crashes, lost trust |
| Communication | No integrated messaging | Direct buyer–seller chat | Weaker trust |
| Format | Auction-centric | Fixed-price purchasing | Mismatch |
| Acquired systems | Replaced EachNet's local systems | Local responsiveness | 85% → 7.7% share |
Tactic: When you acquire or enter a new market, freeze every local system for a year and require written evidence before replacing any of them with your global default.
Principle 7
Buy complements, not ambitions
eBay's acquisition record under Omidyar's chairmanship divides cleanly into two kinds. The complements fixed a problem the core business already had: PayPal, at $1.4 billion, replaced mailed checks and an inferior in-house product and became the company's most valuable asset; StubHub, at $310 million, extended the marketplace into a category with obvious secondary demand and sold for $4.05 billion. The ambitions tried to graft a different business onto the marketplace: Skype, at $2.6 billion, was sold four years later after the company failed to integrate it — and promptly fetched $8.5 billion from Microsoft. The Craigslist stake produced eleven years of litigation. GSI went in at $2.4 billion and out at $925 million.
$
The eBay Acquisition Ledger
Complements compounded; ambitions leaked value.
| Acquisition | Price in | Exit | Type |
|---|---|---|---|
| PayPal (2002) | $1.4B | Spun off 2015 as independent public company | Complement |
| StubHub (2007) | $310M | Sold to Viagogo for $4.05B (2020) | Complement |
| Skype (2005) | $2.6B | 70% sold at $2.75B valuation (2009); Microsoft paid $8.5B (2011) | Ambition |
| GSI Commerce (2011) | $2.4B | Sold as eBay Enterprise for $925M (2015) | Ambition |
| Craigslist stake (2004) | $32M | Litigation 2008–2010; sold back 2015 | Ambition |
| EachNet (2003) | Undisclosed in source | 85% share fell to 7.7%; China site shut 2007 | Ambition |
The test is simple: does the target make your existing users' transactions easier, safer, or cheaper? If the honest answer requires a slide about synergies, it is probably an ambition.
Tactic: Before any acquisition, write one sentence naming the specific friction it removes for your current customers — and walk away if you can't.
Principle 8
Make the legal structure match the intent
Omidyar Network combines a 501(c)(3) foundation with an LLC, letting the same capital flow to a nonprofit grant or a for-profit equity stake depending on what the problem requires. The structure was early; the "LLC giving" model it prefigured was later adopted by other technology fortunes. It allowed the network to fund microfinance institutions, newsrooms, and, in 2024, an equity stake in Anthropic — a purchase no traditional foundation could have made as easily.
The flexibility has costs. Hybrid structures are harder to explain, easier to criticize, and invite scrutiny from regulators and editors alike — as the network discovered in India and on its own encyclopedia page, which for years described it as "self-styled." But the alternative is letting the tax code dictate strategy, which is how many foundations end up making grants to problems that needed investors and investments in problems that needed grants.
Tactic: Map each initiative to the capital form it actually needs — grant, equity, debt, or advocacy — before choosing the legal vehicle, not after.
Principle 9
Incubate, then spin out
The most distinctive feature of Omidyar Network is how much of it has left. Democracy Fund, Spero Ventures, Luminate, Flourish, Imaginable Futures, PlaceFund — each began as an internal initiative and became independent, with its own leadership and mandate. It is the eBay chairmanship pattern applied to philanthropy: build the thing, prove it, then hand over the keys rather than accrete an empire.
↗
The Spinout Sequence
Omidyar Network's habit of building programs and releasing them.
2004
Pierre and Pam Omidyar found Omidyar Network as a foundation-plus-LLC hybrid.
2011
Democracy Fund launched as an Omidyar Network initiative; independent by 2014.
2018
Spero Ventures spun out in March; Governance & Citizen Engagement becomes Luminate.
2019
Financial Inclusion initiative becomes Flourish.
2020
Education portfolio becomes Imaginable Futures; property rights becomes PlaceFund (January).
2024
Stake in Anthropic; strategy refocused on guardrails for digital technology and AI; India exit completed by year-end.
Spinouts force clarity. A program that must survive on its own needs a mission sharp enough to recruit leaders and funders who aren't you. Programs that stay inside indefinitely tend to drift toward the parent's politics and the parent's habits.
Tactic: Give every new internal initiative a written "graduation test" — the conditions under which it becomes independent — on the day you launch it.
Principle 10
Fund the critics of your own category
The founder of a platform that benefited from Section 230 and won Tiffany v. eBay went on to fund the Open Markets Institute, Public Knowledge, antitrust papers against Facebook and Google, the Center for Humane Technology, Whistleblower Aid, and — through Luminate — the European communications of Frances Haugen. He also financed The Intercept, an outlet devoted to exposing surveillance, despite owning a fortune made by a company that would later be caught port-scanning its users.
One can read this as atonement or as hedging. It is more usefully read as an understanding that platforms need an immune system they cannot supply themselves. Internal critics get managed; external ones get funded or they don't exist. The eBay stalking scandal is the dark mirror of this principle: a company that treated an outside critic as an enemy to be destroyed rather than a function to be tolerated.
Tactic: Identify the most credible critic of your industry and ask what it would cost to make their work more rigorous — then consider paying it, at arm's length.
Principle 11
Kill the useful myth before it hardens
The Pez dispenser story worked. It got eBay press in 1997, when the company needed attention. It was also false, invented by a public relations manager, and when Adam Cohen exposed it in 2002 the company had to confirm the fabrication. Two decades later the myth still circulates in YouTube titles. Useful myths are like counterfeit goods on a marketplace: cheap to produce, durable once listed, and corrosive to the reputation system they exploit.
The true story — an engineer, a holiday weekend, a broken laser pointer, a buyer who collected broken laser pointers — was better than the myth all along. It was stranger, more specific, and more revealing of what the company actually did. Most true founding stories are. They are rejected not because they are weak but because they require the teller to trust the audience.
Tactic: Audit your company's origin story for anything that is "basically true," and replace it with what actually happened.
Principle 12
Your name stays on the building after you leave
Omidyar stepped down as eBay's chairman in 2015 and left its board in September 2020, keeping the title director emeritus. Every article about the company still begins with his name. The stalking scandal, the counterfeit lawsuits, the China retreat, the GameStop bid — all of it is filed under a founder who, by design, ran very little of it.
The same holds for his philanthropy. Omidyar Network's encyclopedia entry has, at various points, been tagged as advertising, traced to the network's own IP address, rewritten by disclosed paid editors, and expanded with single-sourced claims about his grandfather. None of these were his keystrokes. All of them are his reputation. Founders who step back from operations do not step back from attribution; they merely lose control over what gets attributed.
Tactic: Before you hand over the keys, write down the three cultural norms you'd want enforced in your absence — and get the operator to sign it.
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 IIIQuotes and Maxims
In their words
Omidyar has left surprisingly few lines in the record assembled here. The voices that define his story belong mostly to others — a buyer, a board, a spokesperson, an editor — which is perhaps fitting for a man whose life's work was building rooms for other people to talk in.
— Pierre Omidyar's reason for wanting the name Echo Bay, as reportedIt sounded cool.
— eBay, rejecting GameStop's roughly $56 billion takeover offer, May 2026Neither credible nor attractive.
— A Luminate spokesperson, to Insider, October 2021We are the only organization from The Omidyar Group that is directly supporting the work of Frances Haugen and her team.
— BlueRoses13, a disclosed paid editor, on the Omidyar Network Wikipedia talk page, June 2, 2022Omidyar has hired me to fix these issues in line with Wikipedia's policies.
Maxims
- Disclose the defect. Trust between strangers begins when the seller volunteers what the buyer couldn't have known.
- Value lives with the collector. Price is set by the intensity of a particular desire, not by whether the thing works.
- Charge from day one. A fee schedule tells you the truth about demand faster than any survey.
- Quit on evidence. Leave the day job when revenue exceeds the salary, not when conviction exceeds the doubt.
- Keep the chair, hand over the keys. Inventors and operators are rarely the same person; pretending otherwise is a vanity shareholders pay for.
- Name the fragile thing. Write your most dangerous dependency into the filing before the market writes it into your obituary.
- Local knowledge beats global templates. The moment a platform tells a market what it wants, it has stopped being a platform.
- Buy complements, not ambitions. If an acquisition doesn't make your existing users' transactions easier, it's a story you're telling yourself.
- Fund your critics. A platform's immune system has to be built outside the platform.
- The record outlives the founder. Whoever edits your history after you leave is writing your reputation — so leave the true story where they can find it.
