Craigslist

Thirty-two employees. In 2011, when Jim Buckmaster told a reporter from The Guardian that Craigslist operated with a staff of thirty-two, the number landed like a misprint. The site was, by then, the eleventh most popular in the United States and thirty-seventh globally, serving 65 million unique monthly visitors across 700 localized sites in seventy countries, generating an estimated $115 million in annual revenue — and it ran on fewer people than a mid-tier Applebee's. No banner ads. No venture capital. No product managers A/B testing button colors. No growth team. No IPO. No acquisitions (save one disastrous equity sale that would metastasize into the company's most consequential strategic wound). The entire operation — customer service, engineering, policy, the whole clattering infrastructure of the world's largest classifieds platform — was housed in a Victorian in San Francisco's Inner Sunset neighborhood, a building so modest that visitors sometimes assumed it was a private residence.

What Craigslist did, and continues to do, is process the raw material of American economic life — the apartment searches, the job hunts, the couch sales, the gig postings, the missed connections, the free piano that's yours if you haul it down three flights of stairs — at a scale that should be structurally impossible for a company of its size. By 2018, one consulting group estimated annual revenue had crossed $1 billion. The staff had swelled to fewer than fifty.

The disproportion is the point. Every meaningful decision Craig Newmark and Jim Buckmaster ever made about Craigslist was, in some fundamental sense, a decision to leave money on the table — to charge less, to hire less, to build less, to want less — and the cumulative result of that disciplined refusal was one of the most durable competitive positions in the history of the consumer internet. The paradox at the center of Craigslist is that its moat was built from the absence of the things other companies consider moats: technology, talent density, capital, product sophistication, data infrastructure. The moat was cultural, and it was reinforced by every dollar the company chose not to extract.

This is the story of a business that became a verb, a punchline, a weapon, a civic utility, and — against every law of Silicon Valley thermodynamics — a survivor. It is also the story of what happens when a company's greatest strategic asset is the one thing that can never be replicated by a competitor with more money: the willingness to be small.

Craig Newmark was a programmer at Charles Schwab in 1994, which is about the least romantic origin story the internet has ever produced. Born in 1952 in Morristown, New Jersey, raised by a single mother after his father died when he was thirteen, Newmark was the kind of person who described his own personality with clinical detachment — a self-diagnosed "nerd" who found social interaction taxing and gravitated toward systems that mediated human contact. He had a master's in computer science from Case Western Reserve. He'd worked at IBM for seventeen years. He wore pocket protectors without irony.

What Newmark saw at Schwab, though, was not a business opportunity. He was going around the company showing people the early internet, saying, as he later recalled, "Hey, here's how we're going to do business someday." But what struck him was the behavior on the WELL and Usenet — people helping each other out, sharing information, giving each other a break. By early 1995, he decided to give back in a small way: a CC list, sent to a handful of friends, about arts and technology events in San Francisco. The Anon Salon. Joe's Digital Diner. Nothing idealistic, he insisted later. Nothing altruistic. "Just giving back a little bit."

The list grew by word of mouth. People started suggesting additions — job postings, stuff for sale. Newmark added apartments, responding to San Francisco's chronic housing shortage. When the list outgrew email, he migrated to a list server called Majordomo, which required a name. Newmark wanted "sf-events." Friends who were, by his own admission, smarter than him said: "We already call it craigslist. Keep calling it that — it will keep it personal and quirky."

They were right. The name was the first and most important product decision, even if Newmark didn't recognize it as one. "Craigslist" embedded a specific promise in the brand — that this was a person's project, not a corporation's, and that the social contract between poster and platform was something closer to a neighborhood corkboard than a media property. It was a promise Newmark would keep, and it would become the company's defining constraint and its greatest competitive advantage simultaneously.

Through 1995, 1996, and 1997, craigslist existed in the liminal space between hobby and institution. Newmark was still a contract programmer. The list was growing, but he resisted formalization with a stubbornness that would become the company's animating philosophy. When approached in late 1997 about running banner ads, he declined. "Some things should be about money, some shouldn't, and I make enough doing contract programming," he said at the time. Other volunteers proposed running face-to-face parties, creating a nonprofit foundation. The organism was developing its own immune system before it even had a body.

By 1999, the volume was impossible to manage as a side project. Newmark devoted himself to craigslist full-time and incorporated. A Forrester Research report in February 2000 confirmed what San Francisco's HR community already knew: craigslist had become the most effective job site in the Bay Area. Newmark needed help.

He posted a job listing. On craigslist, naturally.

Jim Buckmaster found the posting. A biochemistry graduate of Virginia Tech who had attended medical school, studied classics, and made tofu at the University of Michigan before pivoting into web development, Buckmaster was the kind of person whose résumé read like a random walk through American intellectual life. He'd built a terabyte-scale, database-driven web interface at the Inter-university Consortium for Political and Social Research (ICPSR) in 1994–95, directed web development for two companies most people have never heard of, and was, at six foot seven, literally impossible to overlook. He was hired as lead programmer. Within a year, Newmark made him CEO.

The partnership worked because it was perfectly asymmetric. Newmark was the conscience, the customer service obsessive who took every complaint personally because the site bore his name. Buckmaster was the architect, the engineer who designed the multi-city infrastructure, the search engine, the community moderation system, the self-posting process, the homepage design that would remain essentially unchanged for a quarter century. He added discussion forums, personals categories (including missed connections), and best-of-craigslist — the curatorial layer that gave the site its strange, enduring personality.

The business press would later describe Buckmaster as "anti-establishment," "a communist," and "a socialistic anarchist." He wore jeans to interviews and gave answers so measured they sounded like silences. Asked about revenue, he would deflect. Asked about growth strategy, he would talk about user experience. Asked about competition, he would note, almost serenely, that craigslist existed to serve people's basic needs — shelter, employment, community — and that the commercial implications were secondary consequences.

This was not performance. It was doctrine.

To understand craigslist's competitive position, you have to understand what it chose not to build. In the early 2000s, as the web recovered from the dot-com collapse and the second wave of consumer internet companies began raising capital, craigslist made a series of decisions by omission that collectively defined its strategic identity:

No banner ads. No display advertising of any kind. The site would remain visually uncluttered — blue hyperlinks on a white background, organized by category and city, with a design vocabulary closer to a text file than a web application.

No algorithmic sorting. Posts appeared in reverse chronological order. No recommendation engine, no personalization, no behavioral targeting. The absence was deliberate — it preserved the classifieds' function as a commons, where every poster had equal visibility.

No mobile app (for years). No redesign. No venture-backed growth initiatives. The site looked the same in 2010 as it had in 2000. "It's like a shark that's never had to evolve," Jessa Lingel, a University of Pennsylvania professor who wrote An Internet for the People: The Politics and Promise of craigslist, told NPR in 2020.

No data monetization. According to Lingel's research, craigslist does not sell user data to third parties. In an industry where user data became the fundamental currency, this was the equivalent of an oil company refusing to drill.

No real-time payments. No escrow. No transaction infrastructure. Craigslist facilitated the connection; the exchange happened in the physical world — in parking lots, on sidewalks, in apartment lobbies. The site's refusal to insert itself into the transaction was, paradoxically, the source of both its greatest vulnerability (fraud, scams, safety concerns) and its greatest asset (zero take rate for most users, which meant zero switching cost from craigslist was also zero switching cost to craigslist — the platform's "free" price point was a floor no competitor could undercut).

The selective charges that did exist were minimal: job postings in a handful of major U.S. cities ($25 per listing, $75 in San Francisco), apartment listings in New York ($10). Buckmaster framed these not as revenue optimization but as quality control — charging a small fee reduced spam and improved listing quality. "Any extra profit accrued," he told The Guardian in 2011, "is an unintended secondary consequence."

The entity that felt craigslist's impact most viscerally was not a tech company. It was an industry.

Between 2000 and 2007, according to a study published in Management Science by Robert Seamans of New York University and Feng Zhu of Harvard Business School, craigslist diverted approximately $5 billion in classified advertising revenue from U.S. newspapers. The mechanism was straightforward: craigslist offered what newspapers offered — a local marketplace for jobs, housing, and goods — for free, or close to it. Classified advertising had been one of the highest-margin revenue streams in print journalism, a subsidy that funded newsrooms across the country. Craigslist didn't set out to destroy it. The destruction was a byproduct.

The irony was not lost on Newmark, who would later devote much of his fortune to propping up the institution his creation had undermined. By 2018, he had donated $20 million to endow the graduate journalism school at the City University of New York, which renamed itself the Craig Newmark Graduate School of Journalism. He gave millions more to ProPublica, the Sunlight Foundation, the Columbia Journalism Review, Common Sense Media. In 2022 alone, he donated $81 million to causes including journalism, cybersecurity, and veterans' services — enough to land him on the Chronicle of Philanthropy's list of the fifty biggest donors in the country.

Jeff Jarvis, who directed the Tow-Knight Center for Entrepreneurial Journalism at CUNY and had been a key influence on Newmark's understanding of the news industry, rejected the blame narrative directly: "Craig didn't invent the internet. He created the most prominent example of what the internet could do in directly connecting buyers and sellers, reducing inefficiency in a market."

But the structural damage was real, and it created a paradox that defined craigslist's public reputation: the company was simultaneously celebrated as a democratic utility — free, accessible, community-driven — and condemned as the destroyer of the business model that funded local accountability journalism. Same company. Same decision. Same refusal to charge.

When craigslist expanded to 690 cities worldwide in August 2009, adding 140 new sites including towns as small as Susanville, California (population 18,000) and cities as large as Shenzhen, China (14 million), Jamie Kimmet, the advertising director of the Daily Free Press in Elko, Nevada, hadn't even heard the news. "Boy, that's news to me," he told the New York Times. "I don't really see it impacting us." He was almost certainly wrong.

The only significant equity transaction in craigslist's history was a wound the company spent a decade trying to heal.

In 2004, a former craigslist employee sold a 28.4% stake to eBay. The details of the transaction were never fully public, but the structural implications were immediately clear: eBay, the world's largest online marketplace, now held a substantial minority interest in the company that was, in many product categories, its most effective competitor. The motivations were misaligned from the start. eBay saw an investment in a high-growth platform that could be leveraged, monetized, or eventually integrated. Newmark and Buckmaster saw a passive financial partner. What they got was something closer to a hostile guest.

The relationship deteriorated. In 2008, craigslist's board diluted eBay's stake, and eBay sued. Craigslist countersued. The litigation dragged through the Delaware courts for years, consuming management attention that a 32-person company could not afford to waste. In 2015, eBay finally sold its stake back to craigslist as part of the corporate restructuring surrounding eBay's separation from PayPal. The ownership returned to where it had always philosophically resided: with Newmark and Buckmaster.

But the episode revealed something essential about craigslist's governance structure. The company had no board of directors in any conventional sense, no outside investors with governance rights (after the eBay exit), no reporting obligations, and no accountability mechanism beyond the internal convictions of two people. This was either the purest expression of founder-led vision in the history of the consumer internet or the most complete absence of checks on strategic drift, depending on your perspective. It was probably both.

If you want to understand the last fifteen years of venture-backed consumer technology, you could do worse than to study what craigslist chose not to build.

The thesis that craigslist was "a bundle of classified verticals" waiting to be disaggregated by specialized startups became one of the dominant investment frameworks of the 2010s. Andrew Parker, a partner at Spark Capital, published a visualization in 2012 that mapped every craigslist category to a venture-backed vertical competitor. The image went viral in venture circles and became a founding document of the "unbundling" thesis.

Jobs? Indeed, LinkedIn, ZipRecruiter. Housing? Zillow, Apartments.com, Trulia. Short-term rentals? Airbnb. Used goods? OfferUp, Letgo (later merged), Facebook Marketplace. Cars? CarGurus, Carvana, Autotrader. Gigs? TaskRabbit, Fiverr, Upwork. Dating? Tinder, Bumble, Hinge. Even the "free stuff" category spawned apps like Freecycle and Buy Nothing.

The unbundling was real. Billions of dollars in venture capital flowed into companies whose founding insight was, essentially, "we can do one thing craigslist does, but better — with photos, with payments, with trust and safety infrastructure, with a mobile-native experience." Airbnb, which filed its S-1 in November 2020 before going public at a $47 billion valuation, was perhaps the most spectacular success story. Brian Chesky and his cofounders had literally started by posting air mattress listings.

And yet. Craigslist survived the unbundling. Not only survived — continued to operate profitably, at scale, with fewer than fifty employees, while the companies that unbundled it collectively consumed tens of billions in venture capital, went through multiple pivots, and in several cases failed entirely (Letgo merged into OfferUp; numerous Craigslist clones folded; even OfferUp struggled with profitability).

The survival defied every framework. If craigslist was a bundle, and bundles are supposed to be vulnerable to focused vertical attack, why didn't the verticals kill it?

The answer lies in the nature of craigslist's bundle. It was not a product bundle — a set of features stitched together by a single application. It was a behavioral bundle. People came to craigslist because craigslist was where you went when you needed something local, now, cheaply. The apartment search might lead to a furniture purchase which might lead to a gig posting which might lead to a missed connection. The breadth was the product. A user looking for an apartment on craigslist was already primed to sell their old couch on craigslist, because they were already there. The cross-category liquidity generated a gravitational pull that no single-vertical competitor could replicate.

Facebook Marketplace, launched in 2016, came closest. By 2024, Fortune reported that Facebook Marketplace had four times the customers of Amazon in the secondhand goods category. But Facebook Marketplace was a feature embedded in a social network, not an independent product — it benefited from Facebook's distribution, not its own network effects. And it still couldn't kill craigslist.

The feature that drew the most regulatory heat and public scrutiny was also the one that revealed craigslist's deepest structural vulnerability: the tension between openness and harm.

Craigslist had, from its early days, a category for erotic services — created, Buckmaster told reporters, "at the request of our users" for legitimate massage, escorts, and exotic dancers. But the category became a vector for prostitution and, in some cases, sex trafficking. Law enforcement authorities across the country — from Nassau County to Cook County to Seattle to Jacksonville — used the site in sting operations, arresting dozens. The Nassau County police alone arrested more than seventy people in a single year.

Buckmaster's response was that a 24-person staff (as it was then, in 2007) could not patrol 20 million new postings per month and relied on community flagging to remove objectionable content. Under Section 230 of the Communications Decency Act, web platforms were legally shielded from liability for user-generated content. Craigslist was operating within the law.

But "within the law" and "morally comfortable" are different addresses. The controversy intensified for years. In 2010, craigslist shuttered the adult services section. Then, in March 2018, after Congress passed the Fight Online Sex Trafficking Act (FOSTA-SESTA), craigslist removed its entire personals section — not just erotic services, but all of it: dating, missed connections, the categories that had been among the site's most culturally distinctive features.

The personals shutdown was a defensive amputation. Rather than risk liability under the new law's erosion of Section 230 protections, craigslist simply cut the limb. It was a decision that revealed the limits of the "keep it simple, keep it free, trust the users" philosophy: when the users included bad actors, and the legal framework shifted, the company's minimal infrastructure left it with no tools between "fully open" and "fully closed."

The financial reality of craigslist is almost perversely simple. The company charges for a narrow set of listing categories — primarily job postings in select U.S. cities and broker-listed apartments in New York — and offers everything else for free. The fees are small: $10 to $75 per listing, depending on category and city. There are no banner ads, no display ads, no sponsored placements, no premium listings, no freemium upsells. The revenue model is a narrow spigot attached to an ocean of free activity.

And the ocean is enormous. By the late 2010s, external estimates (craigslist does not disclose financials) placed annual revenue at over $1 billion, with operating margins that private technology companies would envy — the consequence of generating massive revenue against a cost structure that included fewer than fifty salaries, no marketing budget, minimal infrastructure costs (the site's design simplicity translates directly into minimal server load), and no sales team.

Forbes has listed Craig Newmark as a billionaire, though Newmark has never commented precisely on his wealth. His operational involvement with craigslist has been minimal for years — he describes himself as a "semi-retired" customer service representative. His focus since 2015 has been the Craig Newmark Philanthropies, through which he has donated hundreds of millions to journalism, cybersecurity, veterans' services, and — with characteristic whimsy — pigeon rescue. In December 2025, he signed the Giving Pledge, committing to donate the majority of his fortune.

Buckmaster, who has led the company since 2000, remains possibly the longest-tenured active CEO in Silicon Valley — a distinction he noted with quiet amusement in his 2011 Guardian interview, placing himself second only to Jeff Bezos. He has run the company through the dot-com crash, the eBay litigation, the newspaper extinction event, the unbundling, the erotic services controversy, FOSTA-SESTA, the rise of Facebook Marketplace, and the mobile revolution — and the company's fundamental proposition has changed so little that a user from 2003, if transported to 2025, would find the interface instantly recognizable.

The longevity is the mystery. Twenty-five years in internet time is geological — the equivalent of centuries in other industries. Companies that were dominant when craigslist incorporated in 1999 — Yahoo, AOL, Excite, AltaVista — are dead, absorbed, or irrelevant. Companies that were supposed to replace craigslist — Backpage (shut down by the Department of Justice in 2018), Oodle, Recycler, dozens of local clones — are gone. Even the successful unbundlers coexist with craigslist rather than replacing it.

The durability has three structural sources, and they are not the ones that conventional competitive analysis would predict.

First: the price floor. When your product is free for 99% of users, the only way a competitor can undercut you is to pay users — which is what some did, burning through venture capital in the process. But you cannot sustain negative pricing indefinitely, and craigslist's cost structure meant it could sustain zero pricing forever.

Second: local network effects. Craigslist is not one marketplace. It is hundreds of local marketplaces, each with its own supply-demand dynamics. A competitor that launched in one city had to build liquidity from scratch in each subsequent market — a process that craigslist had completed over years, organically, city by city. The compounding was slow and durable in exactly the way that venture-funded blitzscaling was fast and fragile.

Third — and this is the one that the unbundling thesis consistently underestimated — the anonymity premium. Craigslist does not require real names, social graphs, or identity verification. For a vast category of transactions — selling a couch you'd rather not have associated with your LinkedIn profile, seeking a roommate without broadcasting the search on Facebook, posting a gig you'd prefer your employer not see — craigslist's anonymity was not a bug. It was the feature. The privacy-hostile trajectory of the modern internet made this more valuable over time, not less.

Craig Newmark, at seventy, was recovering from a heart ablation procedure and talking about hospital pudding. "Hospital pudding is really good," he told CNBC in September 2023. "And since they served my lunch cold, I asked them and got more containers of pudding."

This is the man whose website obliterated a $19 billion industry. Whose refusal to monetize, in the face of what one consultant estimated was a $10 billion-plus revenue opportunity, constituted the largest single act of consumer surplus creation in the history of classified advertising. Who then spent years trying to repair the collateral damage, giving away hundreds of millions to sustain the journalism his platform had inadvertently defunded.

His philanthropic logic, articulated over dozens of interviews, had an appealing circularity: "I learned in high school history that a trustworthy press is the immune system of democracy." He funded journalism schools because journalism mattered. He funded cybersecurity because the country was "under threat from people who wish us harm." He funded veterans because "there are people giving up our great deal to protect me and my family." He funded pigeon rescue because — and here the logic broke down in the best possible way — he loved birds and had a sense of humor.

The giving was not strategic in the way Silicon Valley philanthropy is usually strategic. There was no attempt to align charitable activity with corporate interests, no PR apparatus, no impact metrics. Newmark described his role as "money, influence, and, perhaps the most difficult of contributions, getting out of people's way." The modesty was genuine, or at least had the practiced consistency of something held for decades.

There is a page on craigslist called "best-of-craigslist" — a curated collection of user-nominated postings that, taken together, constitutes something like the site's collective unconscious. A free Volvo 240DL, no key. A "Haunted Thai Puppet" in Atlanta. "Will you split a Costco pie with me?" in Phoenix. "I need someone to retrieve a hidden obelisk" in Vancouver. A "mean cat that bites you" in Portland.

These posts are not accidents. They are artifacts of a platform that preserved the human voice on the internet at a moment when every other platform was engineering it out. Craigslist's mission statement — written in an era when mission statements were not yet brand exercises — declared its purpose as "restoring the human voice to the Internet, in a humane, non-commercial environment." No one uses the word "humane" in a corporate mission statement anymore. It sounds naïve, or possibly subversive.

The site's design — those blue hyperlinks, that white background, the ASCII art header on the best-of page — has become its own kind of monument. Not to technological sophistication, but to a theory of the internet that lost. The theory that said the network's purpose was to connect people with what they needed, locally, directly, without intermediation, without extraction, without surveillance, without someone optimizing the connection for their own revenue. The theory that said simplicity was not a feature to be iterated past but a principle to be defended.

Craig Newmark found that principle awkward, personally. He still finds it awkward that such a visible site is named after him, his official biography notes, "but he'll get over it." Jim Buckmaster, the tallest CEO in Silicon Valley, writes intermittently on the craigslist blog and publishes the occasional haiku. The company's homepage design — Buckmaster's own creation, circa 2000 — has not been materially altered.

Somewhere in San Francisco, fewer than fifty people keep the whole thing running. The mean cat in Portland is still available.