Viral Loop

by ADAM L PENENBERG

  [ THE NAVIGATOR ]

  Over the next few months they explored ideas. Clark’s first plan involved interactive television, but with no hardware, there was no market for software. The second involved an online gaming network for the Nintendo 64, but there was too much risk in being dependent on one manufacturer; this became especially clear after the game console shipped late. Also, sufficient numbers of gamers weren’t yet online, although it was clear the Web and Mosaic-enabled Internet were becoming part of people’s everyday lives. Clark saw that in the near future the law of large numbers would take hold. If he and Andreessen could figure out a way to make a little money off enough users, they could have themselves a very healthy business. But how? Churning with his signature impatience, Clark was tired of talk without action.

  One night in March 1994, after dinner and a second bottle of burgundy at Clark’s house, Andreessen changed his mind, suggesting they create a “Mosaic killer.” The university was making a business out of the idea that he had conceived and coded with his friends. All they would need to do was release a better version of the browser and let it take over. Andreessen was sure he could line up the original Mosaic team to come onboard.

  He emailed his former comrades in code, telling them to pack their bags because he and Jim Clark—the Jim Clark—were coming to town. In the months since Andreessen had taken off for sunny California, life for his friends, who had taken to calling management at NCSA the “Politburo,” had become downright gloomy. Before Mosaic, NCSA administrators couldn’t have picked out the young programmers in a police lineup. Now they were called to meetings with forty people, and everyone, it seemed, outranked them. They didn’t need much convincing to join Andreessen.

  Clark incorporated Mosaic Communications in Mountain View, California, committing $3 million, about 20 percent of his net worth, and renting office space. The programmers separated into three teams—one each for Unix, Mac, and Windows—and got down to work. One of the key differences between Mosaic and the new browser was the kind of equipment it would run on. With Mosaic, both its creators and most of its users raced around on powerful computers whose information traveled over thick pipes, high-speed T1 and T3 lines paid for by government, universities, and corporations. To become a mass-market product, the new browser would have to course over phone lines and into PCs with sluggish 14.4 modems, which Clark likened to “soda straws.” Speed and economy would be crucial, and each team competed to see who could come up with the fastest program, using a stopwatch to clock which version downloaded a page the fastest. No matter which was slowest, they all slaughtered NCSA’s Mosaic.

  They lost track of the hours as they sat transfixed by their screens, sitting around in T-shirts and underwear and arguing about everything from politics to sports. In other words, it was a typical Silicon Valley tech start-up. Andreessen was anointed “vice president of thinking up stuff” and practically never left the building, exhorting his engineers to press on. Summer seeped into fall, and the company continued to expand not only its personnel but the number of products it would offer—server and security software, an e-commerce application, and, of course, the browser, which was the linchpin.

  The big debate was over pricing. The head of marketing hailed from Apple and proposed charging $99 for the browser, but Andreessen wanted to give it away so it would spread virally and create a huge installed base. “Ubiquity,” he argued, was key to jump-starting the company and wiping Mosaic off the face of the Web. For inspiration he turned to Microsoft, the dominatrix of the desktop, which had leveraged an early lead in the DOS and Windows operating systems into PC hegemony. This coaxed application vendors to create applications for Microsoft’s operating system and ignore the rest, which, starved of compatible software, faded away. Andreessen saw it as a simple equation: market share today equals revenue later; without market share, you don’t generate revenue, but whoever achieves and hangs on to it wins.

  It didn’t bother him that the conventional wisdom of the time was that nobody could make money off the Internet. Somebody sold the first train, the first telephone, the first car, he argued. A network has to begin somewhere. Only then would formidable business opportunities ensue. He was sure the same would happen with the Web, which had achieved a self-sustaining momentum that was quickly turning into a self-fulfilling prophecy. In the end, Andreessen came up with “free but not free.” They would give away the browser to students and educators and charge everyone else $39 (then $49 a year later). Even then, they offered a ninety-day free trial, which was never enforced, so businesses would primarily be the ones to pay.

  The engineers pressed on, and the browser was shaping up to be a significant improvement over the original Mosaic. Not only was theirs ten times faster (according to in-house tests), more secure, and less prone to crashes, users could create far more complicated page layouts and encrypt credit card numbers, which was necessary for commerce to bloom on the Web. All of this would serve to juice its virality. In the way that including pictures was key to the original Mosaic’s success by transforming a gray world of text into a multicolor extravaganza, fostering individuality and expression for users’ webpages encouraged webizens to tag their sites with buttons that said, “Best viewed with Netscape” along with a link to download.

  [ “THE SPARK THAT TOUCHED OFF THE INTERNET BOOM” ]

  Late on October 13, 1994, engineers posted the beta version of the browser online. A few stomach-churning minutes later someone in Japan downloaded it. A trickle became a stream, then a geyser. Someone jerry-rigged an alarm system to keep track—a bell for Macs, a moo for PCs, an explosion to indicate Unix. This quickly turned into a drinking game, with mooing drowning out the rest as the PC version pulled away. At the same time, their archrival Mosaic continued to spread, accounting for 60 percent of all Web traffic, and the university was licensing it to companies for $100,000, with Fujitsu and SpyGlass basing their own applications on it.

  But NCSA didn’t stop there. It retained lawyers who claimed that Andreessen and Clark were infringing on its intellectual property, implying that Andreessen had stolen the software he himself had created, and demanded a 50 cent per copy royalty for every download. Clark hired a forensic software expert to comb through the code to compare it to the original Mosaic. After the expert found “no similarity in form, only function” between the two, Clark rejected their demands. It was becoming clear, however, that naming the company Mosaic Communications Corp. was a liability, and battling over it only diverted their attention away from the big prize: an IPO. The last thing Clark’s young company could afford was a lawsuit hanging over its head. He offered to drop Mosaic from the name and pay NCSA $3 million or fifty thousand shares of stock. NCSA took the money and settled.

  Two months later, the newly named Netscape Corporation officially released Version 1.0 of the Navigator browser. Around midnight on December 15, 1994, the engineers gathered once again, this time rigging servers so a cannon fired every time a browser was downloaded. At first, virtually all of the activity occurred in Japan and Australia, since it was during their business hours. Within a few hours, ten thousand copies of the browser had been downloaded. With virtually no advertising or marketing, more than 6 million copies of Netscape Navigator were in use by spring 1995, and 10 million by the summer. The “browser,” as John Cassidy, author of Dot.Con: How America Lost Its Mind and Money in the Internet Era, quipped, “was spreading like one of the filthy jokes that Clark liked to tell”—at a time when only 45 percent of Americans had even heard of the World Wide Web.

  As Netscape’s market share rose, Mosaic’s plummeted to 5 percent. Businesses showered Netscape with licensing requests and purchases of server software, and by March 1995 the company generated $7 million in revenue. With newly installed CEO Jim Barksdale, a former executive at AT&T, handling day-to-day operations, Clark pushed the next step on his agenda: an initial public offering. Conventional wisdom dictated that a company needed to have three-quarters of robust revenue grow
th before going public, but Clark, who viewed an IPO as a media event, was in a hurry. SpyGlass, which licensed Mosaic’s code for its browser, filed paperwork to go public in May 1995, which prompted Clark to initiate June’s board meeting with a suggestion that Netscape follow suit. John Doerr was all for it. “Put the puck on the ice!” he said. But Barksdale thought it might be too early. While deliberating, they agreed the big risk was Microsoft, which planned to bundle a browser in its next Windows operating system upgrade. To fight Microsoft would take ample resources. They would either need to go public or go back for more venture capital, which would further dilute their shares. They decided on the IPO.

  On August 9, 1995, interest was so keen on Netscape’s first day of trading that Charles Schwab reset its phone welcome greeting: “Welcome to Charles Schwab. If you’re interested in the Netscape IPO, press 1.” Morgan Stanley added an additional line to handle the intense call volume. There was so much demand the stock—initially priced at $28—opened late so the market could set a price, which it eventually did: $71 a share. After peaking at $75 it would end the day at $58. Almost instantly, Clark and Andreessen had created dozens of millionaires. Each of the original Mosaic coders from Illinois cleared several million. Clark’s stake was worth more than $663 million that first day. Andreessen, who had worked late the night before, rolled out of bed, logged in to check the stock, then learning he owned $70 million in stock, promptly went back to sleep. “It’s probably a Midwestern kind of thing,” he said later.

  Editors at Fortune would look back in time and call Netscape’s IPO “the spark that touched off the Internet boom.” They weren’t alone. The Netscape IPO was largely credited with instigating the mania in technology. A sixteen-month-old company with picayune revenues and few products was suddenly worth close to a billion dollars. It stunned bankers, stock analysts, and the media.

  [ DISRUPTING NETSCAPE’S VIRALITY ]

  A year earlier, when Mosaic was first taking off, Bill Gates noted that the Web was free and couldn’t see any way Microsoft could make money on it. He changed his mind by the time Netscape Navigator squashed Mosaic and grabbed 75 percent of the browser market. In a memo titled “The Internet Tidal Wave” that Gates sent his executive staff and their direct reports a month before Netscape made the decision to go public, he said the Web was the “most important single development” since the IBM PC. “I have gone through several stages of increasing my views of its importance. Now I assign the Internet the highest level.” In fact, he called it “critical to every part” of Microsoft’s business. What spooked him was that after ten hours of browsing, Gates hadn’t come across a single Microsoft file on a website—no Microsoft Word, no nothing. He vowed to take on the company’s newest competitor: Netscape. “We have to match and beat their offerings,” he wrote. But that didn’t mean Microsoft would give away the browser, a suggestion one of his subordinates made weeks later at a company retreat. “What do you think we are,” Gates asked, “communists?”

  In November Netscape’s stock was topping out at $171 per share. When Goldman Sachs downgraded Microsoft stock because of concerns over the impact of the Internet on the software maker’s future, Gates rolled into action. He licensed the Mosaic browser from SpyGlass, which had also recently gone public, and made the creation of Microsoft’s own browser a priority, throwing a battalion of programmers at it. The irony wasn’t lost on Andreessen, who had conceived and coded both browsers, noting he would be competing against himself. In August 1995, with Netscape enjoying an 80 percent share of the browser market, Microsoft released Internet Explorer. Although it was greeted with a collective yawn, hammered by critics, and didn’t make a dent in Netscape’s advantage, the so-called browser wars had begun.

  With Netscape riding high and revenue flying to $346 million in 1996, Andreessen appeared on the cover of Time on a throne and with bare feet. Meanwhile Microsoft kept chipping away, releasing Internet Explorer 3.0 a year after it first entered the browser market. Microsoft often took three tries before it got something right, which was true of Windows and true of the new browser. But Microsoft hadn’t simply closed the technology gap. Gates did an about-face, deciding to give away the browser and bundling it with every new computer that ran Windows. Any PC maker that didn’t offer Explorer as the default browser on every machine would lose their license to run Windows. A computer without an operating system was like a man without a brain. Years later a federal court would rule against this monopolistic behavior in a case that eventually settled. Too late. The damage was done.

  Six months after Internet Explorer 3.0 hit the Web, Microsoft’s market share jumped 10 percent up to 22 percent, increasing to 32 percent in the next six months. While Explorer climbed, Netscape sank and its revenues slipped—the company had $132 million in operating losses that year. By the time AOL bought Netscape in 1998 for $4 billion in stock (worth $10 billion by the time the deal closed), Internet Explorer had a 50 percent market share and Netscape was en route to flaming out.

  Andreessen had achieved two viral loops in the span of a few years, but Gates found a hole in his strategy. With everything depending on spreading the browser, Gates simply disrupted Netscape’s virality, “choking off Netscape’s air supply,” software publisher Tim O’Reilly once said. Studies have shown that it is very difficult to induce people to switch browsers once they have selected one, so by pushing Explorer as the default option on PCs, Gates made his the de facto standard.

  But Netscape was hardly the ultimate dot-com flop. Not only did all those associated with Netscape walk away with riches—Clark made billions, Andreessen hundreds of millions, and his engineers tens of millions—the browser wars of the 1990s resulted in a higher purpose. By the end of the decade, about 400 million people were on the Internet, virtually all using either Netscape or Internet Explorer that ran on code that Andreessen conceived and created.

  By seeding the Internet with navigation tools, Mosaic and Netscape made possible today’s Web—and there was no turning back.

  3

  The Spreadable Product as New Business Paradigm

  Viral Plain, Ning’s Double Viral Loop, Your Digital Self, and Crushing Crushlink

  To understand how a viral expansion loop works, there’s no better place to begin than with Marc Andreessen’s latest company, Ning, located in Silicon Valley a couple of blocks from Facebook and a few clicks down the road from Google. The company has been growing automagically from the moment it launched its “Social Networks for Everything”—a free platform for do-it-yourself social networks—in February 2007. Within four months there were 60,000 Ning nets, and by six months, 80,000. At year’s end there were 150,000, and roughly 325,000 of them eighteen months into its run, increasing by about 2,000 a day. As of April 2009, Ning counted almost 29 million users (adding 2 million a month), more than 1.3 million social networks (with 3,500 new groups being formed each day) and 2.7 billion page views the previous month. By the end of 2009 it estimates it will have more than 60 million members and close to 3 million groups. About 40 percent of Ning networks originate outside the United States, and members from almost two hundred countries have signed up, with the service already available in several languages, including Chinese, Japanese, Spanish, and Dutch. At this rate, within a few years Ning will host millions of social networks with tens of millions of members serving up billions of page views daily.

  On its social networks, users post comments, questions, photos, and videos and topics run from parenting to Pez dispensers, movies to motorcycles to motherhood, TV shows to customized cars to Thai kickboxing. There are groups for horse enthusiasts, gamers, and gays, health, information technology, ska music, and urban living. (A few dealing with marijuana and stoner culture have been, perhaps not surprisingly, forgotten.) Over a two-month period almost one hundred thousand people joined the official fan site for The Twilight Saga, a vampire-themed book series cum hit movie. Another popular social net that belongs to hip-hop mogul 50 Cent boasts hundreds of thousands of
members (and is growing every day).

  Chris “Broadway” Romero, creative director of ThisIs50.com and a producer at his G Unit Records, describes it as “an entertainment industry news/rumor/editorial blog in the vein of tmz.com, combined with unparalleled access and interaction with the celebrity.” Romero uses the site to cast parts for music videos and film projects, and one day hopes to release music and video directly to the public, bypassing record companies completely. To Romero, it’s nothing less than “a new entertainment platform, period.” A single Ning group can, in theory, serve as an artist-owned and-managed foundation for an entire business; and collectively, the networks represent an ever-expanding commercial universe. Indeed, even within networks viral growth dynamics persist. When a group registers 150 users, it also reaches a point of nondisplacement, continuing to grow even if organizers do nothing to promote it.

  That’s because someone setting up a social network has no choice but to invite friends, family, colleagues, and like-minded strangers to sign up, and each new member brings others with him. That pushes Ning’s “viral coefficient”—the number of additional members each person brings in—way above 1. That’s key, because if the virality coefficient is 1, the start-up will grow, but at a linear rate, eventually topping out. Above 1, it achieves exponential growth. The table, created by Jeremy Liew, a venture capitalist with Lightspeed Venture Partners, an investor in RockYou, illustrates the difference a tiny increase in the viral coefficient can make, showing relative growth rates based on a viral coefficient of .6, .9, and 1.2. Liew started with a base of ten members and defined time as the period it takes for a member to invite others, which he estimated could be anywhere from two and eight weeks. Starting with 10 members and a viral coefficient of .6, you flatten out at 25 people, a gain of 15 users. At .9, you end up with 75 new members and growth slows dramatically. With a viral coefficient of 1.2, however, those same 10 people yield 1,271 additional users (see Table 1). Expressed in a line graph, a viral coefficient of 1.2 takes on the form of an exponential curve (Fig. 3).