← Back to the ledger
Entry · 1995–present

Craig Newmark

A programmer emails a dozen friends about local events, keeps his day job four more years — and accidentally drains billions out of American newspapers.

FieldMarketplaces & Classifieds
CompanyCraigslist
Read7 min
Job at launchKept day job
Starting capital$0
Tipping pointScratch own itch
RouteBuilt from zero
Industry knowledgeAdjacent
By the numbers
Starting capital
$0
Time to first dollar
Years — it was deliberately unmonetized
Peak scale
~$1B revenue estimated (2018) with roughly 50 employees
What nearly killed it
The adult-services scandal and a shareholder war with eBay

Every entry is researched against 30+ structured fields

CH. 01 — The Setup

Craig Newmark spent seventeen years at IBM as a programmer and systems engineer, starting in 1976 straight out of a computer science master's at Case Western. In 1993 he moved to San Francisco to work at Charles Schwab, on security architecture.

He was forty-two, new to the city, and by his own description not a natural at the social parts of life — he has called himself something like patient zero for nerds, taped glasses and all. The relevant fact is not the self-deprecation. It is that he was a middle-aged engineer who had just arrived somewhere he did not know anyone, in a city then filling up with people building the early web.

There is no origin myth here. He wanted to know what was going on, and he wanted to tell people about it.

◆ THE PIVOT — The Tipping Point

In early 1995 he started copying a group of friends — accounts vary between ten and a dozen — on emails about arts and technology events around San Francisco. Small gatherings. Things worth going to.

The tipping point is not the first email. It is what he did when the list started behaving in ways he had not designed.

The list spread by forwarding. By the middle of 1995 it had grown to around 240 names, and in June the ad-hoc email approach collapsed under its own weight, so he installed Majordomo listserv software and turned it into a real mailing list. Then people began posting things he had never invited. Job openings first — technical roles, because these were software people. Then things for sale. Then apartments, because San Francisco's housing market was already brutal.

Newmark's actual decision was to keep saying yes to whatever people showed up wanting to do, rather than defending the thing he had originally built. Every major category on craigslist exists because users started posting it unprompted, not because anyone planned it.

He did not even name it. He wanted to call it "sf-events." People were already calling it Craig's List, and a friend told him he had inadvertently created a brand. He deferred to them, and has said since that he still finds it awkward that the site carries his name. The website followed at craigslist.org in 1996, again because users asked for it.

CH. 02 — Getting Started

The part most worth studying is what he did not do, which is quit.

From 1995 to 1999 craigslist was a hobby running alongside paid programming work. Four years. No investors, no salary from the project, no runway, and consequently no clock forcing decisions before they were ready.

In late 1997, with the site growing, advertisers started approaching him. He turned them down, and the reasoning he gave on craigslist's own history page is the closest thing this story has to a thesis: some things should be about money and some should not, and he made enough doing contract programming.

Read that as a business decision rather than a moral one and it is still sound. His costs were near zero, so his break-even was near zero, so he could decline revenue that would have cost him user trust. That is a structural advantage that money would have destroyed. Every competitor with investors had to monetize aggressively; he did not.

He incorporated in 1999 and went full time — not because of a revelation but because the site had outgrown what a side project could absorb. He has never taken venture capital. He has estimated, in his own framing, that he would have been worth something like $11 billion had he played it the conventional Silicon Valley way, and has said plainly that he did not need billions.

CH. 03 — The Build

In late 1999, a programmer named Jim Buckmaster posted his résumé on craigslist. Newmark hired him, and in November 2000 promoted him to CEO — then stepped down himself.

His explanation is unusually blunt for a founder. He has said he was a bad manager: bad at hiring, bad at hard calls, unable to fire anyone, reluctant to make bold decisions like opening new cities. So he took the title Founder and Customer Service Representative and went to work on the front line, handling spam and scam and fraud complaints personally. He was still doing it more than twenty years later.

Monetization stayed deliberately narrow. Craigslist charges for job postings and, later, for broker apartment listings in New York. Nearly everything else is free. Job fees started at $25 in New York and Los Angeles in August 2004, expanding city by city over the following decade. The interface never modernized; the site in 2026 looks close to the site in 1999.

The scale that produced is genuinely strange. Analysts have estimated revenue crossing $1 billion in 2018 — a figure the company has never confirmed — with roughly fifty employees, serving around 700 cities in 70 countries.

The externality was enormous. Economists Robert Seamans and Feng Zhu measured what happened to local newspapers when craigslist entered their markets: classified ad rates fell 20.7%, subscription prices rose, circulation dropped, and classified advertisers saved an estimated $5 billion between 2000 and 2007 — which the authors describe as a conservative figure, since it stops at 2007. That money came out of the newsrooms that classifieds had been quietly subsidising for a century.

LEDGER NOTES — What to Take From It

The complications are real and belong in the record.

Craigslist's erotic services section became a national scandal, tied in public argument to prostitution and trafficking and to a murder case that attached the site's name to a killer. Newmark's company replaced it with a paid, verified "adult services" section in 2009 — analysts estimated those ads alone brought in tens of millions of dollars a year — and shut it down entirely in September 2010 after seventeen state attorneys general demanded it, briefly replacing the link with a black bar reading "censored." The Personals section went in 2018 after the FOSTA legislation.

Then there was eBay, which bought a minority stake from a departing shareholder in 2004, launched a competing classifieds product, and spent seven years in litigation with craigslist before selling the stake back in 2015. The Delaware Court of Chancery struck down craigslist's defensive measures along the way.

What survives all of that is the structural lesson. Newmark built something with enormous economic leverage by refusing, repeatedly, to extract the available value. No ads, no venture capital, no redesign, no expansion into every monetizable category. The restraint was not sentimental — it kept the trust that was the entire product, and it kept the cost base low enough that he never needed the money.

He also matched his role to his actual skill instead of his title, handing the company to a better manager within five years and going back to answering complaints.

▸ THE PLAYBOOK — Run It Yourself

The framework: build something useful with a cost base near zero, keep your income while it grows, and let users tell you what it is.

Move 1 — Start with the smallest useful thing for a group you are already part of, and pick a format with no fixed costs. His was an email to a dozen friends. Yours can be a newsletter, a Discord, a spreadsheet, or a recurring local meetup. The requirement is that shutting it down costs nothing, because that is what lets you keep going when it grows slowly.

Move 2 — Watch for the uninvited use and follow it. Craigslist's biggest categories — jobs, housing, for sale — all came from users posting things nobody asked for. Once a month, look at what people are actually doing with your thing rather than what you built it for, and move resources toward the surprise. If nobody is misusing it, it is not being used enough.

Move 3 — Keep your income until the project cannot be absorbed by evenings any more. He waited four years and he waited too long by conventional standards, which is precisely why he never had to accept money that would have changed the product. Set the trigger as workload, not revenue: quit when the thing is breaking for lack of attention, not when you hit an arbitrary number.

Budget line: a domain and a mailing list run about $50 to $200 a year, which is genuinely all he was risking. The real cost is four years of evenings and weekends — call it eight to ten hours a week — while a salary continues to arrive. That is the trade, and it is available to almost anyone with a job.

Sources & verification corroborated
  • craigslist's own mission and history page
  • Internet Hall of Fame official biography
  • Seamans & Zhu, 'Responses to Entry in Multi-Sided Markets', Review of Economic Studies
  • Delaware Court of Chancery rulings in eBay Domestic Holdings v. craigslist
  • Contemporaneous coverage of the 2010 adult-services shutdown (NBC News, Forbes)

Craigslist is private and has never confirmed a revenue figure. Every dollar amount here is a third-party estimate, most of them from the analyst firm AIM Group, and they vary widely. Newmark's employment between 1993 and 1999 is genuinely muddled in the record — some accounts have him at Charles Schwab throughout, others have him leaving in 1995 and doing contract work and a stint at Bank of America. What every account agrees on is that he had paid programming work the whole time, so that is what is claimed here. eBay's stake is reported as both 25% and 28%.

Last verified 2026-09-01

Keep going

Get one of these a week.

Same structure every time — the tipping point, the first ninety days, and a playbook with real numbers attached.