“Wikipedia is something special”
In the mid-1990s, Jimmy Wales had a lot in common with Jeff Bezos and Peter Thiel—an early career in finance, libertarian-leaning politics, youthful fascination with science and computers, and a determination to make it rich from the Web. What he lacked, however, was their elite educational pedigree and Wall Street experience. And that would make all the difference for Wikipedia, the brilliant project Wales has shepherded since its inception in 2001.
The encyclopedia anyone can edit, Wikipedia appeared to offer a golden ticket to a Web 2.0 entrepreneur like Wales with its oodles of user-generated content. But Wikipedia has taken a different path: no advertising appears next to the articles its editors have written or the photos they have posted; there are no suggestions of products to buy based on your search; the project doesn’t track its visitors, and hasn’t once thought to acquire a rival online encyclopedia. Wikipedia is a shining outlier in the commercialized Web, a site that has benefited enormously from the network effect but has never sought to profit from it. And Wales, the Web entrepreneur, has watched it all happen.
Jimmy Wales was a precocious kid, born in Huntsville, Alabama, in 1966. His early education came at a small school started by his mother and grandmother, though he later attended a private high school. Huntsville was a scientific hub, the home to a space research center led by the infamous Wernher von Braun, who had masterminded the Nazis’ rocket program. In high school, Wales learned to program a PDP-11 minicomputer, one of the Digital machines that were beloved by the early hackers. Wales kept up his interest in computers and programming as a side hobby while he studied finance at Auburn and the University of Alabama, getting bachelor’s and master’s degrees. He left a PhD program in finance at Indiana University to become a futures and options trader in Chicago.1
The Netscape IPO in late 1995 caught Wales’s attention, and he made another career change: Web entrepreneur. The next year, Wales started his own online business called Bomis, which included a simple search engine and an index of Web sites assembled by registered users, who were called “ringmasters.” What these human indexers produced were “rings” about a specific topic, one site joining the next. In classic Web 2.0 fashion, Bomis was planning to profit from the indexes it solicited by surrounding them with advertisements. The company made an appeal to ego and self-interest to spur contributions. “If your ring is really good, we may choose to include it in our search engine and tree structure,” the Web site said. “This is a great honor because it means that we think you’ve gotten at the essentials of some topic, and that we think you have been reasonably fair and unbiased in your selection and descriptions. (Of course, putting your own page first in the ring is PERFECTLY fine with us!)”2 This was not an ideal situation for runaway growth, however; Bomis was highly dependent on the work of humans, either its small group of contributors or its even smaller staff. There was little artificially intelligent about the company, and thus little chance of its scaling.
Wales was open to any new idea for encouraging users to produce material that he could then sell ads against; like many sites of this era, Bomis wasn’t shy about helping its users find porn.3 In January 2000, when the company had about a dozen employees and tech stocks were at their peak, Wales proposed that Bomis publish an online encyclopedia created by users, which he called Nupedia. For the first editor of Nupedia, Wales hired Larry Sanger, a technically proficient philosophy PhD, whom Wales had met through an online discussion group about Ayn Rand, the “virtue of selfishness” philosopher.4 As any sensible encyclopedia operators would do at that time, Sanger and Wales established a thorough review system for articles, one so thorough that after a year barely twenty articles had been given the green light to be published. At that rate, Nupedia would challenge the Encyclopedia Britannica in a few thousand years.5
This system obviously couldn’t last, and the dot-com crash, again, had a way of bringing clarity to the situation. Bomis hadn’t discovered an automated way to unlock profits from the Web, as Google or PayPal or Amazon had. Wales and his partners had no rich friends to lean on in tough times, either. His company, which was based in San Diego and then Florida, was “on the periphery of the Internet,” recalled Terry Foote, an old friend of Wales’s who was Bomis’s advertising director. “Jimmy and I, each in our own different ways, pounded on the metaphorical doors of all the Silicon Valley big shots, and mostly what we got was a solid wall of silence.”6 Under pressure to rethink Nupedia, Sanger recalled proposing a way to speed up the review process, but “everything required extra programming,” he said. “By then, Jimmy Wales was worried about keeping costs low, the dot-com boom was turning to bust, and so whatever we did to solve the problem had to involve no more programming.” That’s where “wikis” came in. A friend told Sanger about the free wiki software that allowed users to collaborate online on a single article, roughly along the lines that Berners-Lee had planned for the original Web browser.7
The Web site Wikipedia.com began operating on January 15, 2001, and, like Mosaic and Google before it, found an audience almost immediately. With its software innovation, Wikipedia had six thousand entries after six months. By the end of the year, there were twenty thousand articles, most in English but hundreds of others in more than a dozen languages, including German, Spanish, Polish, and the artificial tongue Esperanto.8 When the original wiki software was unable to handle this growth, a German volunteer created the more capable version that is still used today.
In addition to its enthusiastic volunteers, many of whom were originally drawn to Nupedia, the Wikipedia project found crucial assistance from the new Google search engine, whose algorithm didn’t care a whit about offline reputation. Google knew only what could be conveyed through Web links, and according to that calculus Wikipedia could have a loftier reputation than the Encyclopedia Britannica, whose articles didn’t freely circulate online. Speaking to a Stanford class almost exactly a year after Wikipedia was created, Sanger described the positive feedback loop his project was experiencing through Google. “We write a thousand articles; Google spiders them and sends some traffic to those pages,” he said. “Some small percentage of that traffic becomes Wikipedia contributors, increasing our contributor base. The enlarged contributor base then writes another two thousand articles, which Google dutifully spiders, and then we receive an even larger influx of traffic. All the while, no doubt in part due to links to our articles from Google, an increasing number of other websites link to Wikipedia, increasing the standing of Wikipedia pages in Google results.”9
Those first twelve months, shattered by the 9/11 attacks, included the absolute trough of the dot-com collapse. Under those conditions, Bomis had yet to make money from its runaway hit, Wikipedia. The company laid off half of the staff, including Sanger in February 2002.10 Without Sanger’s salary, the costs of running Wikipedia were surprisingly low—by 2004, total expenses for a significantly larger Wikipedia were still less than $25,00011—but even those sums were a drain on Bomis when it could least afford one. Sanger understandably agitated for Bomis to raise some money to restore his salary as the editor in chief of Nupedia and “chief organizer” of Wikipedia, jobs he described as the best he’d ever had. In a letter to the Wikipedia community explaining that he would have to reduce his involvement in the projects while he looked for full-time employment, Sanger opened Pandora’s box, writing hopefully that “Bomis might well start selling ads on Wikipedia sometime within the next few months, and revenue from those ads might make it possible for me to come back to my old job.”12
The mere mention of advertising was so disturbing to the volunteers who were building Wikipedia that the project was nearly destroyed by the backlash. Why this group of content generators—as opposed to people listing their résumés or reviewing restaurants—would be particularly offended by advertising is hard to explain. Was their anger based on the fact that writing an encyclopedia article is fundamentally altruistic and thus shouldn’t be exploited for profit? Maybe it was Wales’s laid-back posture toward the project that gave them such self-confidence? He didn’t come off as a typical Silicon Valley entrepreneur, convinced how brilliant he was, treating users like idiots to be exploited. Bomis knew it was bringing very little to the table when it came to Wikipedia—the inmates were running the asylum, and that seemed to be working out fine. Perhaps the best explanation is that Wikipedia didn’t begin with ads, which meant that they would have to be introduced, which gave the opposition something to fight against.
There was a final factor: Wikipedia was created under a so-called “free” software license, and those terms applied not only to the software but to the articles as well. While there certainly was no rule against running advertising next to material made under a free license—free software’s most prominent activist, Richard Stallman, is fond of saying, “free as in free speech, not free beer”—this license gave anyone the right to copy all the material that appeared on Wikipedia and begin his own project, a process known as “forking.” An exact duplicate of Wikipedia—a fork—could appear under a new name at a new Web address; the only condition was that the new project operate under the same free license. Immediately after Sanger made his idle suggestion about advertising, a group of contributors to the Spanish version of Wikipedia forked the encyclopedia. The articles were copied and stored on servers at the University of Seville, where the project was renamed Enciclopedia Libre Universal en Español.
The Spanish contributors were on the left politically and already annoyed to be part of a for-profit business. Furthermore, Edgar Enyedy, one of the leaders of the Spanish fork, resented having to go through Bomis to make improvements to his own community’s Wikipedia. “We were all working for free in a dot-com with no access to the servers, no mirrors, no software updates, no downloadable database, and no way to set up the wiki itself,” he said. “Finally, came the possibility of incorporating advertising, so we left. It couldn’t be any other way.” The threat that other foreign-language Wikipedias might fork if there were ads—or potentially that the English Wikipedia itself would—was enough to get Wales to take notice and say no to advertising.13
With the best opportunity for profiting from an online encyclopedia cut off, Wales in 2003 shifted ownership of Wikipedia to a charitable foundation; at least the project wouldn’t be a financial drain anymore. A few prominent articles in the tech press spread the word about the new Wikimedia Foundation and donations from individuals and institutions almost immediately managed to cover whatever expenses were connected to running the site. In 2004, when English Wikipedia had 150,000 articles, the foundation raised more than $80,000 to cover $23,000 in expenses.14
Late in that year, Wales channeled his thwarted entrepreneurial energy into Wikia, a for-profit venture built on wiki principles. The articles on the Wikia site are like Wikipedia articles, created and edited by visitors, but they reflect the enthusiasm of a ’zine rather than the neutral view of an encyclopedist, which is why the San Francisco–based company recently renamed its Web site “Fandom powered by Wikia.” The company has followed a traditional strategy for a Web 2.0 company, running targeted advertisements on article pages and seeking investments from venture capital firms, as well as Amazon.15
To this day, the Wikimedia Foundation is never short of funds, even as the size of its staff and the number of articles within its dozen or so wiki projects has grown substantially. Wikipedia alone has 40 million articles, which can appear in any of 290 different languages. In the 2016 fiscal year, expenses at the foundation were $65 million, including paying for a staff of more than 280 people. In that period, total revenue at the foundation was more than $80 million, with $77.7 million coming from donations and contributions.16
The unconventional development of Wikipedia presents an interesting contrast to how Google grew. Google’s founders, as we’ve seen, were academics fundamentally opposed to advertising, who eventually succumbed to the start-up-friendly vortex surrounding Stanford. As a result, Larry Page and Sergey Brin had the resources to continue operating during dark times until they discovered a way to make advertising work. Jimmy Wales, on the other hand, was committed to Web advertising from the start and would have loved nothing better than to bring advertising to Nupedia and its new incarnation, Wikipedia. He has spent his entire entrepreneurial life trying to create user-generated content to place ads against, yet when he finally hit a gusher he was unable to capitalize.
Had Wales not stumbled on the idea of Wikipedia in absolutely the worst time for dot-coms—or if he had a network of Stanford friends like the PayPal mafia to call upon—perhaps he could have pushed through and operated Wikipedia as a business. He could have ignored the Spanish fork and introduced advertising. With the appropriate resources, he then could have hired programmers to add artificially intelligent features to make the Bomis version of Wikipedia—ads and all—such a superior experience to a forked, ad-free version that few would turn away.
Wales would occasionally return to the idea that the site might accept ads, if not for company profits, then to benefit a charity. “That money could be used to fund books and media centers in the developing world,” he said in 2004. “Some of it could be used to purchase additional hardware, some could be used to support the development of free software that we use in our mission. The question that we may have to ask ourselves, from the comfort of our relatively wealthy Internet-connected world, is whether our discomfort and distaste for advertising intruding on the purity of Wikipedia is more important than that mission.”17 After the community again objected to such a flirtation with advertising, Wales gave up on the idea forever. Not that advertising is evil, he wrote later: “But it doesn’t belong here. Not in Wikipedia. Wikipedia is something special. It is like a library or a public park. It is like a temple for the mind. It is a place we can all go to think, to learn, to share our knowledge with others.”
In that same statement, he glosses over the early, contentious history of the project, declaring, “When I founded Wikipedia, I could have made it into a for-profit company with advertising banners, but I decided to do something different.”18 The narrative that Wales offers in regard to Wikipedia, of being tempted by, and rejecting, a lucrative opportunity to run advertisements is actually closer to what happened at Craigslist. In late 1997, you will recall, the site’s founder, Craig Newmark, was approached about accepting banner ads for a Microsoft service on his site, which had already gone through a growth spurt. He turned the offer down, knowing he had “stepped away from a huge amount of money,” because it wasn’t in the best interest of Craigslist’s users.
Wikipedia and Craigslist are indeed the exceptions that help define what is standard Silicon Valley behavior. They both have scaled to become worldwide phenomena, but not by automating every experience through artificial intelligence. Newmark welcomes personal interactions with his users and proudly describes himself as a customer service representative for the site, albeit one who also relies on automated tools.19 Wikipedia likewise has teams of volunteers who help new users and watch out for vandalism, also aided by automated tools. Truly, they are what Berners-Lee expected the Web to be from its inception, before it took a sharp turn toward commerce. They are decentralized and collaborative; regular folks are given an important voice. Doubtful we will see their like again.
Starting in the 2000s, Thiel, Hoffman, Andreessen, and their ilk shifted from cashing checks to writing them. In their role as investors and mentors, they made sure that the next generation wouldn’t make the same mistakes they had. They warned founders to keep control, pushing back meddling VCs by being sure to retain a majority of voting shares. And they advised start-ups to be ruthless in acquiring market share and then protecting that share—Thiel’s stealth monopolist strategy. When Mark Zuckerberg visited the Bay Area the summer after his sophomore year at Harvard, this crew found their prodigy, like the weathered boxing trainer Cus D’Amato being introduced to the thirteen-year-old Mike Tyson, whom he would shape into a fearsome champion. Zuckerberg was a talented programmer with both a hacker’s belief that computers could save the world and a Thielian ruthlessness about using the network effect to replace the Web with his own service, which would soon be renamed Facebook.