In brief
Wikipedia is an encyclopedia written by volunteers that sells neither advertising nor subscriptions. It was born on January 15, 2001, as a side experiment inside the company Bomis, because the main project, Nupedia, with its seven stages of review, was producing only a handful of articles. Today it is supported by a nonprofit foundation: in fiscal year 2024 the foundation received $185.4 million, and in 2025 $205.4 million, mostly in donations. The story is interesting not for the encyclopedia but for the economics: how a project with no revenue from users grew from $1.51 million in FY2006 to $205.4 million in FY2025.
How it started (the founders)
Jimmy Wales owned Bomis, a small internet company, and put its money into a project for a peer-reviewed encyclopedia, Nupedia. In early 2000 he hired Larry Sanger as editor-in-chief: according to Sanger, Wales was looking specifically for a philosopher, and the very idea of an open encyclopedia, Sanger also says, was entirely Jimmy's. Nupedia worked on a strict scheme: articles were written by experts and went through seven stages of review. The first article to pass through the whole process came out in June–July 2000.
The pace was far too slow. On January 2, 2001, according to Sanger, an old acquaintance, Ben Kovitz, explained the idea of a wiki to him over dinner. On January 10 Sanger sent the Nupedia mailing list a message titled "Let's make a wiki": he wrote that a wiki could be set up in literally ten minutes, and that Wales, as he put it, thought many people would dislike the idea. A week later, on January 17, he asked colleagues to go to wikipedia.com and add a little article in five to ten minutes. Sanger gives January 15 as the launch date.
The domain was .com because, according to Sanger, Bomis saw advertising as a way to pay for Sanger's salary, the servers, and the programmers. But the advertising never appeared: by the end of 2001 Bomis was laying people off, and Wikipedia was bringing in no money at all; Sanger describes the advertising market of early 2002 as almost at rock bottom, and Bomis did not sell a single ad.
Year-by-year timeline
- 2000: Nupedia mailing lists from spring to fall; the first article to pass through the whole process, June–July fact
- 2001-01-10: Sanger proposes a wiki on the Nupedia mailing list fact
- 2001-01-15: wikipedia.com launches fact
- 2001-01-20: a traffic spike after an announcement on Slashdot (Wales's message concerns the Nupedia site; a link to Wikipedia is not proven) estimate
- 2001-07 – 2001-10: 6,000 articles as of July 8, 13,000 as of October 4; Google indexes more and more pages fact
- 2001-12 – 2002-01: Bomis cuts its staff from 10–12 to 4–5 people, and Sanger's salary is cut in half fact
- 2002-02-26: Spanish contributors leave for the Enciclopedia Libre fork after talk of advertising fact
- 2002-03-01: after Sanger's resignation, Wales states that the advertising plans are canceled for now, and that the next step is a foundation and donations fact
- 2003-06-20: Wales announces the Wikimedia Foundation: Bomis's domains and rights pass to the foundation, which does not yet have a bank account fact
- 2005-04: the IRS recognizes the foundation as an educational organization (tax deductibility) estimate
- 2005-12-15: a Nature study: in 42 natural-science articles, reviewers found 123 errors in Britannica and 162 in Wikipedia; 4 serious ones in each fact
- FY2004–FY2006: revenue of $80,000 → $379,000 → $1.51 million, expenses of $23,000 → $178,000 → $792,000 fact
- FY2024: revenue of $185.4 million, expenses of $178.6 million fact
- FY2025: revenue of $205.4 million, expenses of $189.7 million, donations of $187.6 million
Lesser-known but significant facts
- By his own account, Sanger heard the wiki idea over dinner on January 2, 2001, and eight days later wrote to the mailing list, while Wales, according to Sanger, expected objections. The "Let's make a wiki" message arrived on January 10, five days before the launch.
- Wikipedia was on .com because of the planned advertising. Sanger links the move to .org in 2002 to Wales's assurance that there would be no advertising.
- The first announcement of Wikipedia on the Nupedia mailing list was not a launch announcement but a request to humor him: to go there and add an article in five to ten minutes. The big waves of new contributors came later, after Sanger's pieces on Kuro5hin were featured on Slashdot (July and September 2001).
- At its creation, the foundation did not accept money. In a message dated June 20, 2003, Wales asks people not to donate yet: there is no tax status and no bank account.
- In FY2005, PayPal fees ($11,091) came to two-thirds of the payroll ($16,930): donations were taken through PayPal. In FY2006, the fees were $45,617 against $107,122 in salaries.
Legend vs. the record
- Legend: Wikipedia was founded by one person. In March 2005 Wales wrote that he had founded Wikipedia and that Larry had simply worked for him, and he credited the wiki idea to Bomis employee Jeremy Rosenfeld (the quote is known from a comment on Slashdot, low reliability); Sanger calls himself a co-founder. The record: the wiki proposal and the launch of the site come from Sanger (the messages of January 10 and January 17, 2001), while the money, the company, and the idea of an open encyclopedia came from Wales; Sanger himself writes that the credit for Bomis investing in the project belongs entirely to Jimmy. Verdict: neither "one person" nor "equal founders"; the documents show a division of roles, with the money and ownership on Wales's side and the proposal and organization on Sanger's. The "co-founder" status remains a matter of dispute estimate.
- Legend: Wikipedia lives without money. The record: the foundation's financial statements for FY2006 show expenses of $791,907, including $189,631 for hosting, $107,122 in salaries, and $45,617 in PayPal fees. FY2024 revenue was $185.4 million, and expenses $178.6 million. Before the foundation, Bomis paid for the servers. Verdict: the money comes not from paying users but from donors; this is not an absence of expenses but a different model of who pays.
- Legend: Wikipedia rejected advertising from the very beginning. The record: Sanger writes that advertising was planned to pay for his work, and the plan was announced in February 2002; on March 1, 2002, Wales wrote that the advertising plans were canceled for now and that advertising might become necessary later. No advertising was ever sold. Verdict: the refusal was tactical and temporary, not a principle; as Sanger remembers it, it became a principle with the move to .org in 2002 estimate.
- Legend: Nature proved that Wikipedia is more accurate than Britannica. The record: in 42 science articles, reviewers found 162 errors in Wikipedia and 123 in Britannica, on average about 4 versus 3; four serious errors in each; Britannica, according to the Signpost's account, disputed the methodology estimate. Verdict: close to Britannica, not more accurate.
The first growth lever
The first lever was the format itself: editing after publication instead of review before it. In a year Nupedia produced a handful of articles estimate; over 2001 Wikipedia reached 13,000 by October and 20,000 by its first anniversary. The outside push came from Sanger's pieces "Britannica or Nupedia?" (July 25, 2001) and "Wikipedia is wide open…" (September 2001), which Slashdot put on its front page; search traffic grew in their wake, since the more pages there were, the more often Google indexed the site, and the more contributors arrived. It cost almost nothing: a Bomis server, the paid time of one editor, and an open license.
The paired story
The pair is Encarta (history-encarta): Microsoft's paid encyclopedia versus a free volunteer one.
- Contrast in the models. Wikipedia means volunteers, a copyleft license, and money from donors; Encarta was a paid Microsoft product: a disc from 1993, and from 1998 a free concise version with 16,000 articles plus a subscription at $6.95 a month. Wikipedia's bet was speed and openness instead of review before publication.
- The shared plot. Both became an answer to the question of who pays for knowledge: for Encarta it was the buyer of the disc and then the subscriber, for Wikipedia first Bomis and then donors.
- The outcome. In January 2009 Hitwise counted 97% of Americans' visits to online encyclopedias going to Wikipedia and 1.27% to Encarta; on March 30, 2009, Microsoft announced that the Encarta sites would close by October 31, without naming Wikipedia and saying only that the category of the traditional encyclopedia had changed. The link between the closure and Wikipedia is a conclusion drawn by the press, not a statement by the company.
Parallels today (projects from the catalog)
- AI Directories (
ai-directories) is an open database on GitHub to which anyone can submit an edit; according to its founder, the free list brought 100+ downloads and $40 in donations within a day, before there was a paid tier. It is the same scheme: open content first, money later. view this project's dossier → - Awesome Indie (
awesome-indie) is a directory where upvotes determine position and money does not move it; Wikipedia has the same rule, that reputation is driven by shared work, not by payment. The difference is the era: the directory has a paid slot tied to a launch date, while Wikipedia has no paid positions. view this project's dossier → - Blume (
blume) is a free open-source tool with no monetization from day one and a fast release cycle; like Wikipedia, it chooses zero friction at entry over revenue. view this project's dossier →
What a builder can take from this in 2026
- A side experiment can overtake the main project. In a year Nupedia produced a handful of articles, while Wikipedia produced thousands over 2001; set up a "wiki version" of your idea, the cheapest version that launches in ten minutes.
- Don't lose sight of who pays. Bomis funded the project as long as it could, and in 2001–2002 it shrank from 10–12 people to 4–5 at the bottom of the advertising market. If you fund an open project from your main business, decide in advance what will happen when the main business slumps.
- A decision on monetization is better announced publicly and in advance. The announcement about advertising at the end of February 2002 became the trigger: the Spanish contributors left for the Enciclopedia Libre fork. If you have a community, do not discuss monetization in passing.
- Donations are a separate payment product. As early as FY2005, PayPal fees ($11,091) came to two-thirds of the payroll ($16,930); if an open project asks for money, check the cost of accepting payments before you launch the fundraiser.
- Accumulated traffic is an index. Growth in 2001 came from Google, which indexed more and more pages, not from advertising. For a project built on user-generated content, every new page is an entry point from search.
Discrepancies and what we could not verify
- The launch date: January 15, 2001, according to Sanger; the public announcement on the mailing list came on January 17, 2001 estimate.
- Wired's figure of $250,000 and 12 articles in 18 months comes from Wales's words, is disputed by Sanger, and is not in the timeline rumor.
- Jeremy Rosenfeld's role in the wiki idea is known only from a comment on Slashdot and from the Signpost; there is no primary document rumor.
- Britannica's objection to the Nature methodology is known only from the Signpost's account estimate.
- FY2011 revenue: $24.8 million (Wikiversity) versus $27.2 million (Form 990, ProPublica); the second figure is used.
- The Form 990s for 2005–2010 were not found; the audited financial statements for FY2004–2006 and the Wikiversity table were used.
- The Nupedia launch date (March 2000): there is no primary document, and the earliest snapshot of the site in the Wayback Machine is from April 7, 2000 unverified. The first donation campaign and the number of editors were not checked unverified.
- Wales's message about the traffic spike of January 20, 2001, concerns Nupedia; a link to Wikipedia is not proven.
Sources (primary first)
- http://web.archive.org/web/2002/http://www.nupedia.com/pipermail/nupedia-l/2001-January/000676.html — Sanger, "Let's make a wiki," 2001
- http://web.archive.org/web/2002/http://www.nupedia.com/pipermail/nupedia-l/2001-January/000684.html — "Wikipedia is up!" 2001
- http://web.archive.org/web/2002/http://www.nupedia.com/pipermail/nupedia-l/2001-January/000695.html — Wales, "Huge traffic influx," 2001
- http://web.archive.org/web/2003/http://mail.wikipedia.org/pipermail/wikipedia-l/2002-March/001597.html — Wales, "Advertisements," 2002
- http://web.archive.org/web/2004/http://mail.wikipedia.org/pipermail/wikipedia-l/2003-June/028376.html — Wales, founding of the foundation, 2003
- https://web.archive.org/web/2007/http://upload.wikimedia.org/wikipedia/foundation/2/28/Wikimedia_2006_fs.pdf — the foundation's financial statements, FY2004–2006
- https://web.archive.org/web/2006/http://www.nature.com/nature/journal/v438/n7070/full/438900a.html — Nature, Jim Giles, 2005
- http://web.archive.org/web/2005/http://features.slashdot.org/article.pl?sid=05/04/18/164213 and
…sid=05/04/19/1746205— Sanger's memoirs, 2005 - https://projects.propublica.org/nonprofits/organizations/200049703 — Form 990, ProPublica data
- https://networkcultures.org/cpov/2011/01/15/spanish_fork/ — interview with Edgar Enyedy, 2011
- https://meta.wikimedia.org/wiki/Wikimania05/Paper-AS1 — paper on the Spanish fork, 2005
- https://signpost.news/2005-04-25/The_Sanger_memoirs and https://signpost.news/2005-12-19/Nature_study — retellings, 2005
- https://en.wikiversity.org/wiki/Statistical_analysis_of_Wikimedia_Foundation_financial_reports — summary table
- https://www.theregister.com/2009/03/30/microsoft_kills_msn_encarta/ and https://www.networkworld.com/article/797036/lan-wan-hitwise-wikipedia-squashes-encyclopedia-rivals.html — the Encarta shutdown and Hitwise data, 2009
- https://news.microsoft.com/source/1998/05/11/microsofts-new-and-improved-encarta-online-offers-exciting-resources-for-teachers-and-students/ — Microsoft press release on Encarta Online, 1998