I
As the Internet economy grew, several distinct markets developed within it, and in each of them there was intense competition. The markets were arranged in horizontal fashion, one on top of the other. At the bottom was the market for telecommunications gear, with firms like Cisco Systems, Ascend Communications, and Lucent Technologies as the major players. On top of this was the market for Web browsers, which Netscape dominated, with Microsoft trying hard to get a foothold. On the next level up was the market for Internet access, which included dozens of Internet service providers, as well as online services such as America Online and CompuServe. Finally, there was the market for content, which attracted hundreds of entrants every week. The new Web sites offered news, shopping, entertainment, and anything else that could be reduced to ones and zeroes.
Each market was different. The fiercest battle was in the Web browser market, where Microsoft was trying to protect its unparalleled profit margins against encroachment from the Internet. For years, Bill Gates’s company had been making things for a dollar and selling them for ten dollars, thereby demonstrating the vast potential for any company that could combine digital technology with market power. Both Microsoft Windows and Microsoft Office benefited from economies of scale and monopoly positions. Once they had been developed, the cost of replicating them and putting them on CDs was negligible. Yet Microsoft was able to charge hundreds of dollars for some versions of Microsoft Office, having seen off most of its competitors.
This business model had turned Gates into the world’s richest man, and he took a surprisingly long time to realize how the Internet challenged it by threatening to make Microsoft’s products obsolete. Nathan Myrhvold, a Cambridge University–trained physicist who was Microsoft’s technology guru, influenced Gates. Like many technically adroit people, Myrhvold was beguiled by the information superhighway, and he was slow to shift his gaze to the Internet. In the first edition of Gates’s 1995 book The Road Ahead, the Internet hardly featured at all. In May 1995, Gates finally woke up. He e-mailed his colleagues about the coming “Internet Tidal Wave” and the possible creation of “something far less expensive than a PC which is powerful enough for Web browsing.”1 But even at this point, Gates still seemed to think he could seize control of the Internet by creating the Microsoft Network, an online network with its own set of technical standards that would replace those laid down by the inventors of the Internet and the World Wide Web. It wasn’t until the Netscape IPO and the failure of the Microsoft Network to make much of an impact that Gates realized he had lost control of events.
In the fall of 1995, Marc Andreessen said that Windows might soon be reduced to a “poorly debugged set of device drivers.”2 Baiting Gates was unwise. On December 7, the anniversary of Pearl Harbor, the Microsoft chairman told hundreds of reporters and analysts who had gathered at the Seattle Convention Center that from now on Microsoft would be “hard core about the Internet.”3 The Internet Explorer, Microsoft’s struggling Web browser, would be improved, expanded, and distributed free. Microsoft would also give away the accompanying Web server software, for which Netscape was charging more than $1,000. It would license the Java programming language and generally do everything in its power to grab market share from Netscape. Andreessen and Jim Barksdale, Netscape’s chief executive, tried to laugh off Microsoft’s counteroffensive, but Wall Street wasn’t amused. In one day, Netscape’s stock fell by almost $30. The “Browser Wars” had begun.
A couple of months later, in February 1996, Andreessen appeared barefoot on the cover of Time, the first Internet entrepreneur to be accorded this honor.4 Time portrayed Andreessen as a modest tycoon, who rushed home from work at five every evening to walk his dog. As is often the case, the cover story proved to be a sell signal for investors. When somebody makes the cover of Time, he or she is usually at or past his peak. Andreessen was already stumbling in his battle with Gates, who had virtually unlimited resources to call upon. Microsoft’s research budget alone was almost five times the size of Netscape’s annual revenues, and the software giant had more than eight hundred people working on an upgraded Web browser. Microsoft isn’t the most creative company in the world, but it is expert at taking other firms’ products and improving them. When Internet Explorer 3.0 appeared in August 1996, most independent experts agreed that it was at least as good as Netscape Navigator 3.0.
Microsoft’s strength isn’t based on money and resources alone. More than nine out of ten PC users see a Windows screen when they switch on their machines. In the past, Microsoft had ruthlessly exploited this monopoly to cripple competing products like WordPerfect and Lotus 1-2-3. After Gates’s Pearl Harbor Day speech, it did the same thing to Netscape. In March of 1996, Microsoft persuaded America Online to choose Internet Explorer as its default browser. Steve Case, America Online’s chief executive, was no friend of Gates—just a few months earlier Microsoft had tried to put AOL out of business by setting up the Microsoft Network—but he couldn’t resist the offer of an AOL icon on every Windows desktop. “I could never have imagined myself standing there with Gates because of our history with Microsoft, but I thought it was best for AOL, so we did it,” he said.5
Two days after the deal with America Online, Microsoft signed a similar one with CompuServe, the second largest online service. Gates also exerted pressure on the PC manufacturers, especially Compaq Computer, which had previously agreed to install Netscape Navigator on its popular Presario models. When Microsoft started shipping Internet Explorer with its Windows 95 operating system, Compaq replaced the Microsoft Web browser with Netscape Navigator. Gates found this intolerable. Early in 1996, Microsoft sent Compaq a letter threatening to end its agreement to license Windows 95 unless it agreed to keep the Internet Explorer on the desktop. This was effectively a threat to put Compaq out of business. Without Windows loaded onto its machines, no PC manufacturer could hope to survive. Compaq quickly buckled and restored the Internet Explorer to its desktops. For a few months, it shipped the Presario with both Internet Explorer and Netscape Navigator. Then it dropped Netscape Navigator.
Partly as a result of these tactics, Internet Explorer’s share of the browser market increased sharply. In August 1996, Netscape’s lawyers contacted the Justice Department and claimed that Microsoft was once again using anticompetitive tactics. In 1995, Microsoft had signed a consent decree with the government relating to previous misbehavior in other areas. Netscape had a strong case to make, but even if it succeeded in attracting the Justice Department’s attention, which it did, any legal action would likely take years. Netscape didn’t have years. It was loosing market share every day. Forced onto the defensive, it tried to reposition itself as a supplier of software to companies setting up internal computer networks known as Intranets. This was a reasonable strategy—businesses were already Netscape’s biggest customers, and unlike private users they tended to pay for their Netscape Navigators—but it was also an admission of weakness. Netscape was no longer trying to displace Microsoft: like everybody else in the computer industry, it was merely trying to survive alongside the beast of Redmond. And survival was by no means assured.
In December 1996, Jim Allchin, a senior Microsoft executive, sent an e-mail to his boss, Paul Maritz, arguing that “we must leverage Windows more” to defeat Netscape.6 The best way to do this, Allchin went on, was to “think about an integrated solution,” by which he meant making the Internet Explorer a part of the Windows operating system, so that every customer who bought a Windows PC would automatically get a Microsoft Web browser. If Microsoft was allowed to do this, Netscape would be forced to persuade computer buyers to download a second Web browser, which is a bit like trying to sell people a second CD player to go with their stereo system. After some internal discussion, Gates agreed to go along with the integrated solution, a decision that would soon lead to a major antitrust case.
II
The competition in the market for Internet access had already produced a clear winner: America Online. In April of 1996, Steve Case, America Online’s thirty-seven-year-old chairman, appeared on the front of Business Week lying atop a pile of America Online diskettes, which were fast becoming as familiar to Americans as Nike sneakers and Oreo cookies. “It is Case’s America Online that has shown how to turn a community of cybernauts into a mass market and how to successfully turn a computer network into a new medium for entertainment and news,” the accompanying article declared. “With more than five million customers and 75,000 joining every week, AOL is the most potent force in cyberspace.”7
It hadn’t always been thus. For years, the digital cognoscenti had looked down on Case’s firm. “I never knew anybody who took AOL seriously as an Internet company,” Michael Wolff wrote in a fairly typical comment.8 The Internet insiders sneered at America Online for many reasons. It was based in Vienna, Virginia, an unfashionable suburb of Washington, D.C. It didn’t have much proprietary technology. It relied on others for content. It was slow providing access to the Internet. And many of its customers were computer neophytes who liked its blandly reassuring features, such as the recorded voice that said, “You’ve got mail.” But Case wasn’t seeking to impress Silicon Valley. He was trying to create a mass-market product, and when people referred to America Online as the “Kmart network,” he took it as a compliment.
If Marc Andreessen and Jerry Yang represented the frontiersmen in American history, Case represented the merchants and lawyers who followed in the settlers’ wake. He was born and raised in Honolulu, where his father was a corporate attorney. As a teenager, Steve and his elder brother, Dan, set up a mail-order company that sold everything from greeting cards to watches. At Williams College, in Massachusetts, Case was a B-minus student. He majored in political science—“the closest thing to marketing,” he said later—and his least favorite subject was computer science.9 (He was also a disc jockey on the college radio station and sang in new wave bands.) After graduating, Case joined Procter & Gamble as a management trainee. Two years later, he moved to Pizza Hut, where his job involved trying to come up with appetizing new pie toppings. At twenty-three, Case was just another corporate grunt, seemingly destined for a life of schlepping around the country from sales meeting to sales meeting. If anybody in his family was heading for the top, it seemed to be Dan, who had become a Rhodes scholar and an investment banker at Hambrecht & Quist in San Francisco.
In 1983, Dan invited Steve to the annual Consumer Electronics Show in Las Vegas and introduced him to Bill Van Meister, a flamboyant entrepreneur who was starting an online video game company. Bored by his job and sick of living in Wichita, Kansas, Case decided to join Van Meister’s venture, which was called Control Video Corporation. It was a fateful decision. Control Video Corporation ran into trouble, resulting in Van Meister’s ouster, but another firm, Quantum Computer Services, arose from its ruins to provide online services for two PC companies, Commodore and Apple. In 1989, Quantum changed its name to America Online, a name that Case, who by now was president of the company, had dreamed up.
Case’s humdrum character was arguably the key to America Online’s subsequent success. As a nontechnologist, he knew the fear of computers that lurked inside most ordinary mortals. His motto was “make it easy and make it fun.”10 America Online was the first online service to adopt the point-and-click technology pioneered by Apple, and it introduced other user-friendly features, such as the “key word” system that eliminated the need to type in Internet addresses. Case’s model was television, a complicated technology that had been reduced to a few buttons. America Online presented its content in channels, such as Today’s News, Personal Finance, Travel, Education, and Entertainment. The most popular channel, by far, was Personal Connection, the electronic doorway to hundreds (and later thousands) of online chat rooms.
America Online marketed itself as a family service, but some of the activities on its network were hardly family fare. Among the names of the chat rooms that its subscribers created were “submissive men,” “married but flirting,” and “crossdressers2.” The Internet, which allowed people to communicate with one another while retaining their anonymity, was ideally suited to sexual fantasy, and from the early years a good proportion of the traffic on it was taken up by pornography. America Online quietly encouraged some of this traffic to migrate to its network. Certain members of Congress criticized America Online for allowing sexual activity on its system, but in this aspect, too, Case was in tune with the majority of his countrymen. In public, Americans tend to be puritans. In private, as the success of the adult video industry had already demonstrated, there are few activities that they won’t enthusiastically countenance.
According to an article in Rolling Stone in October 1996, up to half of all the chat on America Online was sex-related, and the sex chat rooms were bringing in more than $80 million a year in revenues.11 The popularity of the sex chat rooms was a major reason that America Online overtook Prodigy, the online service owned by Sears and IBM. Prodigy didn’t allow sex chat rooms, and it checked all notices posted on its message boards for inappropriate content. America Online had no such compunctions. “That’s why AOL has eight million members and Prodigy failed to a shadow of its former self,” a senior executive at Prodigy told Kara Swisher, the author of aol.com, an informative history of America Online published in 1998.12
Clever marketing also played an important role in America Online’s success. Case’s experience at Procter & Gamble and Pizza Hut had taught him that AOL couldn’t be sold like soap powder and pizza. Going online is what economists call an “experience good,” which means that people only know if they like it when they try it. Many consumers were put off online services by the initial cost of the software. In 1993, Jan Brandt, America Online’s marketing director, suggested distributing the AOL software for nothing, along with a few free hours of service. Case approved Brandt’s plan, and America Online started sending out free diskettes. They were sent to people’s homes; they were given away with magazines; they were given away on airlines, in football stadiums, even in frozen steaks. Eventually, about 250 million discs would be distributed—almost one for each person in the country. The marketing campaign was expensive, but it was also phenomenally successful. In July 1993, America Online had about 250,000 subscribers. A year later, there were a million of them. In February 1996, the number reached 5 million. Along the way, America Online raced past Prodigy and CompuServe, and also defeated the threat from the Microsoft Network.
A steadily rising stock price accompanied this rapid growth. Although many of America Online’s subscribers didn’t get full access to the World Wide Web until 1996, the company traded as an Internet stock. Between March 1992 and March 1996, its market value rose from $70 million to $6.5 billion. Support from Wall Street enabled Case to finance his ambitious expansion plans. Following its successful IPO in 1992, America Online sold more stock in 1993, and again in 1995. The latter issue, which raised more than $100 million, demonstrated how far perceptions of the company had changed. America Online’s IPO was underwritten by Alex Brown & Co. and Robertson, Stephens. In 1995, the underwriters included Morgan Stanley, Goldman Sachs and Merrill Lynch. Mary Meeker, Morgan Stanley’s Internet analyst, was one of Case’s biggest supporters. From 1993 onward, she retained a buy recommendation on America Online’s stock. In The Internet Report, she wrote that “nobody does a better job of making general information available for a mass market audience than America Online,” and argued that the firm should be viewed not just as an online service provider but as a “consumer media/entertainment company.”13
Despite its remarkable growth, American Online faced some serious problems. In the spring and summer of 1996, these problems came into open view. The firm’s network hadn’t been built to deal with the traffic it was handling, and delays in connecting became so lengthy that it acquired the nickname “America On Hold.” In August, the network crashed for nineteen hours, prompting headlines such as “America Goes Offline,” and “CHAOS@AOL.” Case apologized, but many subscribers, who had been unable to retrieve their e-mails for a full day, were furious. Somebody set up an Internet newsgroup, alt.aol-sucks, where people could vent their anger toward the company. Web sites such as AOL Watch and Why AOL Sucks appeared. The vitriol shocked Case, but there wasn’t much he could do about it in the short term. The only solution was to invest in upgrading the system.
On Wall Street, too, criticisms surfaced, and America Online’s stock price suffered. Between May and September it fell from above $70 to below $30. For years, America Online had been bolstering its profits by using suspect accounting practices. The firm treated its marketing expenditures as an investment, the cost of which could be written off over a number of years, instead of being counted immediately against profits. Given that America Online was deluging the country with free diskettes, the sums involved were far from trivial. Case initially tried to ignore the Wall Street critics, but in October 1996, with the stock price still sagging badly, he finally faced up. America Online wrote off $353.7 million in what it called “deferred subscriber acquisition costs.” This was more money than the firm had made in its entire existence. With proper accounting introduced, America Online looked suspiciously like every other loss-making Internet firm that had raised a lot of money in the stock market and blown it on a big marketing campaign. “The Emperor hath no Pants, AOL hath no profits,” one wag posted on The Motley Fool.14
The most serious threat to America Online came from its outdated pricing structure. While most Internet service providers were providing unlimited access to the Internet for $19.95, Case’s company was still charging its subscribers by the hour if they stayed online beyond a certain time limit. Many heavy users, who could easily run up monthly bills of hundreds of dollars, were switching to cheaper alternatives. America Online’s “churn rate”—the percentage of its subscribers who left every month—reached 6 percent, a development that alarmed investors. If it continued, the majority of AOL subscribers would leave within a year. Case had to do something. He first introduced a slightly less restrictive pricing plan—twenty hours for $19.95, with additional charges after that—but that didn’t satisfy subscribers. Unlimited access was becoming the Internet standard. America Online couldn’t resist the trend and hope to maintain its leadership position. Case, who spent a lot of time online gauging the attitudes of his customers, was aware of this, but he also knew that moving to flat-rate pricing would be a tremendous challenge; America Online had found it hard enough to make a profit when it was charging heavy users by the hour. In October 1996, Case had his hand forced. The Microsoft Network adopted a $19.95 unlimited usage plan. Case couldn’t allow Bill Gates to steal his customers. A couple of weeks later, America Online announced that it, too, would henceforth charge its subscribers a flat rate of $19.95.
Case coupled the pricing change with an internal reorganization. He hired Robert Pittman, the cofounder of MTV, to run America Online’s service. Pittman was known in New York media circles as an ebullient and ambitious figure. Despite losing an eye while he was growing up in Mississippi, he flew his own plane, He also socialized with celebrities, and employed a slick public relations adviser, Ken Lerer, who had helped to mastermind disgraced financier Michael Milken’s public defense during the 1980s. After joining America Online, Pittman compared the Internet to the early days of cable television, when most of the cable networks lost money. Eventually, networks like MTV and Nickelodeon developed powerful brands and made big profits. Pittman insisted that the same thing would happen on the Internet, and that America Online would be one of the big winners.
Pittman had a point. America Online’s success was based on what economists call vertical integration. It provided both access to the World Wide Web and its own Web content. This allowed it to exploit the self-reinforcing network effects that had spurred the growth of the Internet in the first place. As its subscriber base grew, America Online could afford to provide more features, which in turn attracted more subscribers. Moreover, the new subscribers provided a wealth of content of their own, in the form of chat room messages. America Online was a media company that got much of its programming for nothing. If its subscriber base kept growing, this was a business model that was likely to generate profits before too long.
III
Men such as Gerald Levin, the chairman of Time Warner, Michael Eisner, the chairman of Walt Disney, and Rupert Murdoch, the chairman of News Corporation, looked jealously at America Online’s growth. Walt Disney, Time Warner, and News Corporation each owned a broad range of assets—newspapers, magazines, television stations, broadcast networks, film studios, cable networks—but many of these properties were now widely regarded as antiquated. The old media was itself full of pieces about how the old media was doomed. Many of these pieces were ill considered, but they prompted the media barons into action.
In September of 1993, News Corporation bought Delphi Internet Services, a Massachusetts-based firm that provided access to the Internet to scientists and researchers. The purchase took place well before the incorporation of Netscape, a testimony to Murdoch’s sharp eye, but it never had much chance of working out. The News Corporation executives, who were responsible for The Simpsons and The New York Post, didn’t know what to make of the Internet folk, with their strange customs; and the feeling was reciprocated. “We backed the wrong people,” Murdoch said philosophically some years later. “They were very clever, very nice, but they were not very commercial. And we didn’t put enough resources into it.”15 In May 1996, News Corporation sold Delphi back to its founders. By then, Murdoch had also abandoned his second Internet venture—a much more ambitious and expensive project called I-Guide.
The idea behind I-Guide seemed plausible: a TV Guide for the Web. News Corporation already owned TV Guide, so why not replicate its success in cyberspace? Murdoch dispatched Anthea Disney, TV Guide’s editor, to Silicon Alley, where she set up a 75,000-square-feet newsroom staffed with hundreds of young recruits, along with some more seasoned journalists from Murdoch’s Fleet Street newspapers. From start to finish, I-Guide was a fiasco. Initially, it was going to be a subscription-based online service, much like America Online and Prodigy. Then the focus shifted to the Internet. The launch date was repeatedly delayed. By the time that I-Guide finally launched, in February 1996, Yahoo! and other search engines had taken the market it was aimed at. A few days before I-Guide started operating, News Corporation restructured it and fired most of its employees.
Murdoch could console himself that his rivals weren’t faring any better. In October 1994, Time Warner launched Pathfinder, a Web site that featured online editions of Time, People, and several other magazines that the company owned. Pathfinder provided Levin with an escape route from the disastrous interactive television experiment in Orlando; its ambitious launch indicated that Time Warner had decided the Internet was the future. Paul Sagan, a cousin of Carl Sagan, the scientist and author, was hired as president of Time New Media. Walter Isaacson, a well-known Time journalist, was placed in charge of creating content for Pathfinder. Dozens of staff were hired. Technology was purchased. Consultants were retained.
To begin with, Pathfinder appeared to be a success. Compared to most Web sites, it provided a cornucopia of features—more than ninety of them, including the “Netly News,” question-and-answer sessions with Time Inc. journalists, and details of upcoming products from Warner Music and the Warner Bros. film studio. Practically overnight, Pathfinder turned into the most popular site on the Web; and it also became the first site to take in appreciable advertising revenues (about $2.5 million in 1995). Unfortunately, the early promise soon faded. Pathfinder’s share of the Web audience shriveled, and its ad response rates remained disappointing. It tried to introduce user charges—about $5 a month—but shelved them in the face of resistance from users. In its first year, Pathfinder lost more than $10 million. In its second year, Don Logan, the head of Time Inc., referred to it as a “black hole.”16 In early 1996, Isaacson returned to Time as managing editor, and the editor of Pathfinder was ousted. Later in the year, Sagan resigned. Levin had a second new media embarrassment on his hands.
In normal circumstances, the failure of Pathfinder and I-Guide would have persuaded other media companies to steer well clear of the Internet. But circumstances were far from normal. With the country increasingly fixated on the World Wide Web, every self-respecting media executive—even the most old-fashioned ones, such as the owners of The Washington Post and The New York Times—was forced to come up with an “Internet strategy.” In part, this reflected a crisis of confidence on the part of the old media. Many people who worked on newspapers and at broadcast networks tacitly accepted the idea that the Internet would eventually put them out of business. In fact, the old media’s prospects remained good. Newspaper readership and the share of the television audience captured by the broadcast networks were both declining, but that had been true for decades. There was little sign that the Internet was accelerating the declines. Indeed, a strange thing was happening on the way to the graveyard: the old media was making more money than ever before. Newspapers like USA Today and The Wall Street Journal were fat with advertising, much of it from technology companies. On prime-time television, the cost of thirty-second spots on popular shows like Seinfeld and E.R. was reaching unprecedented heights, in some cases as much as half a million dollars.
This didn’t happen by accident. As audiences splintered under the impact of the new media, the remaining mass-market mediums became more valuable, even if their numbers were falling. When Ford wanted to launch a new car, or Paramount Pictures wanted to promote an upcoming movie, they needed to reach the entire country, which wasn’t possible via the Internet. The major networks, while they didn’t deliver as many eyeballs as they had done in the 1970s, still had tens of millions of viewers. USA Today sold more than 2 million copies a day. Herbert Simon, a Nobel Prize–winning economist at Carnegie Mellon University, captured what was happening when he wrote, “A wealth of information creates a poverty of attention.”17 For this reason, anything that could hold the attention of a large part of the population was practically priceless. As a result, the Super Bowl and the Oscars, two quintessentially old media events, were turning into advertising extravaganzas, with dozens of expensive commercials specially created for them.
The media executives could see all this happening, but still they refused to sit out the Internet frenzy. In July 1996, NBC teamed up with Microsoft to launch MSNBC, a round-the-clock interactive news service that would compete with CNN on television and with Pathfinder and other Web sites on the Internet. Each side committed $200 million to MSNBC, with no clear idea how the money would ever be repaid. “We’ll be doing our best to get advertising income, and we expect the Internet to expand dramatically, and that’s partly why we see this as a profitable venture,” was the best explanation that Bill Gates could offer.18 Jack Welch, the chairman of General Electric, which owns NBC, was even more vague. “This is a big deal for GE because commerce is never going to be the same again in the next decade,” he said. “Commerce in the next decade will change more than it’s changed in the last hundred years. Business will be done differently. Distribution will be done differently. People will buy products differently.”19
Welch’s statement may have been inane, but he had a well-earned reputation as one of the shrewdest businessmen in America. The launch of MSNBC allowed NBC (and General Electric) to align itself with the Internet, while having somebody else, Microsoft, pick up part of the cost. The cable channel that MSNBC replaced—America’s Talking—had been struggling anyway, so there wasn’t much to be lost by switching formats. From Welch’s perspective, the real audience for the MSNBC deal was Wall Street, which was agog about the coming “convergence” of media, computers, and telecommunications. By joining up with Microsoft, NBC appeared to be positioning itself for this convergence.
There was no reward for being financially conservative. When News Corporation turned its back on the Internet after the failure of I-Guide, its stock price lagged behind its competitors’. The perception that Rupert Murdoch didn’t “get” the Internet was much more damaging to News Corporation’s stock price than the earnings write-offs from Delphi and I-Guide had ever been. Eventually, Murdoch was forced to do more Internet deals in order to keep investors happy. A similar thing happened to Michael Eisner. For years, Disney avoided making big investments in the Internet. In August 1995 it spent $19 billion on an old media asset, ABC. The merger of Disney and ABC created an immensely powerful company, but it also helped create the impression on Wall Street that Disney had been left out of the multimedia age. In order to counter this impression, Disney would eventually spend hundreds of millions of dollars to launch its own updated version of Pathfinder, Go.com, with similarly dismal financial returns. But by that stage losing money on the Internet would be nothing to be ashamed of. To the contrary, it would be viewed as the inevitable and, indeed, desirable cost of moving into the twenty-first century. In such an environment, Eisner, Levin, et al. would have a fiduciary duty to their shareholders to waste large sums on dubious online ventures.