No Better Time

by Molly Knight Raskin

  The remaining critical question before Lewin, Leighton, and Seelig was their business model. How would they deliver this service? What would they charge? Even by midsummer, it was clear that no Internet Service Providers were willing to take a chance on Akamai. If they wanted to make the company work, they had only one choice: build the technology themselves. “We ran out of options,” explained Leighton. But a clear business plan was still eluding them. For help, they turned to Todd Dagres, a shrewd, Boston-based venture capitalist at Battery Ventures.

  Dagres had been introduced to Akamai before the 50K by a colleague at Battery, Scott Tobin, when the company was still Cachet, and Lewin, Leighton, Seelig, and Nijhawan were seeking out the support of investors. Initially, Dagres admitted he wasn’t that impressed. “I remember thinking that they didn’t really look like much,” said Dagres. “Tom was nice and older and scholarly, and Danny—well, he did not look at all like a genius. He looked like a big kid, stocky but not overly imposing. They were academics and kind of nerdy.” But the more Dagres got to know them, the more he grew to respect them. “These guys were very smart and if they didn’t know something they would find someone who did,” he concluded. What Dagres knew well, with his eyes on the trends in the high-tech sector, was how to take a great idea and turn it into a living, breathing business.

  Favoring designer denim over finely tailored suits, and with a closely shaved head and athletic physique, Dagres looked more like a rock star than a financier. At Montgomery Securities and Solomon Smith Barney, Dagres had made millions as a technology analyst covering communications and the Internet, still in its infancy, later taking a job with Battery Ventures in 1994. Over time, he began to convince Lewin, Leighton, and Seelig that building and selling software wasn’t going to work. Instead, Dagres was convinced they should offer a service on their own. “I thought they had something special and something they could use to build a recurring business,” said Dagres. To him, the company’s real powerhouse could be found in its algorithms—they were proprietary and so sophisticated they would be almost impossible for any competitor to replicate. “Instead of giving other guys the software or Akamai’s ‘secret sauce’ and letting them skim the cream off the market, I thought Akamai should be the cream by building an annuity business with a big margin on this unique software,” Dagres explained.

  This made perfect sense, but it was a daunting proposition. It required capital, and lots of it, to build out a global infrastructure and buy and program all the servers. Installing Akamai servers in data centers around the world would cost hundreds of millions. But the most daunting aspect of the service model was not the cost; it was the customer expectations. To service these Web content companies, Akamai would have to provide service 24/7, 365 days a year. And it would have to do so with nearly 100 percent reliability.

  Labor Day came quickly. And before they knew it, Leighton, Lewin, and Seelig were back at LCS with the decision before them. They each brought with them good news. Leighton had a functioning prototype. Lewin had fine-tuned the business plan. And Seelig had secured the support—both tentative and firm—of potential partners including data centers, content providers and hosting companies. “I spent a lot of time asking: ‘If we can prove to you that your site will not crash and people all over the world will have a better experience, then would you be a customer?’” Seelig recalled. “In a sense we had to promise stuff that wasn’t quite there yet—we were selling a vision and the challenge was that we needed both constituents—the content providers and the data centers—to nibble to make it work. It was like tap dancing.”

  The decision to go forward did not come easily. As Seelig put it, “I remember feeling like we’d proven enough that we could succeed; we vetted the technology and the market and had what I thought was a real prospect for success. But at the same time, I had no illusions that a prototype on the fifth floor of MIT’s lab for computer science would guarantee it.”

  Then there was the competition already beginning to creep into the market. While no one was doing exactly what Akamai proposed, a handful of companies were pushing services that were similar enough to pose a threat. One of them was Adero, Inc., formed by David Crosbie, Akamai’s ill-fated presenter at the 50K. Adero’s business plan, which offered “infrastructure services” to content providers, was disconcertingly similar to Akamai’s.

  Logistically, the decision to start the business was easiest for Seelig; he could move back to his Cambridge apartment whereupon his parents would loan him some funds as a stopgap until Akamai could pay him a salary. For Lewin, however, the decision was weighty. For as much as he believed in Akamai, he also knew he had a lot to lose if it didn’t succeed. He would have to continue taking at least one course to remain in MIT housing with Anne and the boys, but he’d lose his stipend, his only earnings. He’d also have to put his PhD candidacy on hold. Leighton knew that could be a perilous move for a student like Lewin, who was on the fast track towards a professorship. “He had a lot of doubts,” said Leighton. “Once he committed to something he would plow forward, but this was not an obvious decision, and he agonized over it.” It wasn’t a totally effortless choice for Leighton, either. He was content where he was, with a steady income and job. But he had never taken a sabbatical. So maybe, he thought, it was time to take a chance.

  In September 1998, Lewin and Leighton moved out of their offices at MIT and into a rented space at One Kendall Square, just down the block from LCS. Akamai Technologies was officially co-founded by Lewin and Leighton, Seelig and Randall Kaplan, the California businessman. Leighton still recalled an afternoon that fall when he and Lewin were walking from LCS to their new office at One Kendall. Lewin used the occasion for an impromptu pep talk, offering up all the reasons he believed Akamai would succeed. He told Leighton that it wasn’t just their great technology, smart staff, or exemplary business plan that would guarantee success. “He said we had all those things, of course, but that wasn’t why we were going to succeed,” said Leighton. Instead, Lewin told him, “We’re going to succeed because we’re tenacious as hell.”

  Despite Lewin’s bravado, he and Leighton felt some ambivalence. It didn’t take rigorous math to understand that the odds were against them. According to the rule of venture capital, only one out of every sixty new businesses succeeds. “It was scary,” recalled Leighton. “Danny was worried about it, and I was worried for him.”

  But in 1998, if you didn’t move fast, you’d miss the moment. You’d be just another smart entrepreneur with a great idea left standing in the wake of the dot-com boom. Lewin charged forward at full speed.

  Chapter 6

  Kings of Cache

  “We have an interface that makes you think, ‘Click, and result,’ so we want everything. And we want it now.”

  — JAMES GLEICK,

  Faster: The Acceleration of Just About Everything

  ON A WINTER DAY at the start of 1998, shocking news broke over the nation’s front pages: President Clinton was accused of having an affair with a young woman in her twenties, Monica Lewinsky, a Pentagon employee and former White House intern.

  With its salacious mix of power, sex, and political intrigue, the story gripped the media and the country, gathering steam when Independent Counsel Kenneth Starr (formerly assigned to the Whitewater investigation) took over the case. As the lurid details emerged, the public clamored for more.

  On September 11, 1998, Congress conceded, releasing a 445-page report on the affair by the Independent Counsel Kenneth W. Starr. To release the entirety of the report (often called The Starr Report), law makers decided to use the Internet. It was the first time a widely read government document debuted online, and getting it there wasn’t easy. Staffers at the House Oversight Committee had to take all the content of the report off the floppy disks it was saved on and transfer it all into Web files. When they were finished, they copied the Web files onto hundreds of CD-ROMs and delivered them to news organizations and government Web sites, leaving them to decide ho
w best to manage the load. The media predicted the online release would cause a massive, nationwide meltdown of the Web. There was no meltdown, exactly. But when hundreds of thousands of readers eager for salacious details about the president tried to access the report, most of the government servers and news sites they logged onto froze or crashed completely. Even before the Washington Post actually posted the report, its Web site crashed as traffic soared to three times its load compared to August 17, its busiest day, when Clinton testified before the start of the grand jury.{31}

  It was a record-breaking day for the Internet, and, in some ways, defined the new medium the way the JFK–Nixon debates defined television. CNN.com recorded a historic three hundred thousand hits a minute, and MSNBC witnessed a one-day record with two million hits the first day it posted the report.{32} But it also brought one of the Internet’s biggest weaknesses squarely into focus as dozens of major Web sites buckled under the traffic—and Akamai was uniquely poised to fix it.

  One of Akamai’s first checks came from Seelig’s parents, Michael and Julie, for the sum of $50,000. The money went fast with rent to pay, staff to hire, and a system to begin building. It disappeared so fast that they scrimped and borrowed and began the relentless push to secure significant funding. Seelig wasn’t too concerned; he had enough savings and family support to wait it out. Kaplan was in a similar situation, having come to Akamai from a lucrative position at SunAmerica. Leighton needed a salary, but he, too, had a safety net—a savings account, his wife’s salary, and the option of returning to MIT if Akamai failed. In contrast, Lewin was broke. And with no guarantee of a profit in the near future, financial fears continued to needle him.

  In October of 1998, Lewin drove to New York to attend the wedding of Greenberg to actress Stacey Nelkin, who was already pregnant with the couple’s first child. Greenberg has never forgotten the eve of the ceremony, when Lewin showed up on the doorstep of his apartment, elated that he’d made it in time. Greenberg and Nelkin remembered Lewin having no qualms about asking to crash on the floor; he didn’t have much of a choice. With no credit card and no funds for a night in a hotel, he seemed content to just curl up inconspicuously in a corner. Greenberg suggested that Lewin take a room in a new apartment the couple planned to move into, even though it was unfurnished, and handed him the keys. When Greenberg asked if Lewin wanted a blanket or something to sleep on instead of the bare hardwood floor, Lewin hesitated and said: “I guess I’ll take a pillow.” And off he went. The next day, Lewin held up the chuppah at the wedding for his best friend, then drove back to Cambridge.

  At some point upon his return, Leighton helped Lewin work out a loan from Akamai’s coffers to cover some of his expenses and keep him and Anne from falling into deep debt. Lewin knew that Akamai needed at least one or two of the investors who’d been flirting with the company to come through.

  One of them was Arthur “Art” Bilger, a Los Angeles-based businessman who had just left a successful career in media as president and COO of New World Communications Group to invest in early stage companies. Bilger, who first heard about Akamai from Kaplan at a board meeting for an L.A.-based charitable organization, wasn’t the most likely candidate for an angel investor. He didn’t understand the math that formed the company’s foundation, and he wasn’t technologically savvy; he didn’t even own a personal computer. But he liked Cambridge and was intrigued enough by what he heard from Kaplan to arrange to meet the team at MIT.

  Over lunch, Lewin eagerly described Akamai to Bilger, a soft-spoken, bespectacled man. “I didn’t know what the hell they were talking about,” admitted Bilger. “But from the first meeting with them, one of my motivations was that, even in the worst case of it not working out, I’d learn a lot.” Eager to learn more, Bilger returned to the West Coast and did some research, followed by a few trips to Cambridge to check on Akamai’s progress. Bilger formed a close relationship with the small team. Over a short period of time, he became convinced Akamai would succeed. “In that moment, I decided to take their word for it,” Bilger said. He knew they needed help with a first round of financing, so Bilger committed to investing in return for a personal stake in the company.

  Around the same time, Akamai was fortunate to secure a meeting with Gilbert Friesen. A close family friend of Marco Greenberg, Friesen was a legend in the movie and music businesses. A gregarious charmer, he worked his way up from the mailroom of Capitol Records to found A&M Records with Herb Alpert and Jerry Moss. Friesen was responsible for the careers of music greats like Sting and hit films including the 1985 cult classic The Breakfast Club. Friesen grew A&M from a small operation out of a home garage to the largest indie label in the country. Greenberg had mentioned Akamai to Friesen, who made a career out of investing in early talent. Friesen was intrigued enough to meet with Seelig several times during his summer in CA and to introduce Akamai to some key contacts. Suddenly, Friesen said he began to think more about Akamai in late 1998: “The Internet technology drumbeat became intense; it was clearly a rare moment of tectonic change.” Friesen added, “I just loved the name—Akamai. It was extremely sexy and different—a damn good name.” Friesen had lunch with Randall Kaplan, who enthusiastically pitched Akamai, telling him he had so much faith in the company that he was investing $100,000 of his own money (in addition, he said he had convinced both his parents to invest). Friesen liked what he heard. He later picked up the phone, called Greenberg and asked, “When can I meet the guys at MIT?” A date was set, and sometime in September, Friesen—who knew so little about Akamai’s technology he often mistakenly referred to it as “logarithms” instead of “algorithms”—flew to Cambridge to learn more. Despite the fact that he spent most of his time in the company of celebrities, Friesen found the academic atmosphere in Cambridge exhilarating. “I was out of my element, and it was exciting,” related Friesen.

  In that meeting with Akamai, Friesen became one of the first members of what had become a de facto fan club of Lewin and his performance in front of a whiteboard. In many ways, it was his canvas. With a pen in hand, he could stand in front of it for hours at a time and cover it with academic ideas—strings of (sometimes incomprehensible) math or business strategies. And when he did, his presentations could only be described as theatrical. The more he talked and scrawled, the more animated he became, hopping around and grinning from ear to ear as his ideas came to life on its smooth, white surface. As if on cue, he would intermittently turn and look out on his audience, gauging their interest and level of understanding. Friends liken Lewin’s theatrics at the whiteboard to a freight train gathering steam until that stopping point when nearly everyone in the room sat silently wondering what, exactly, had just hit them. It was one of Lewin’s most effective weapons.

  Stunned by what he saw in Lewin at the whiteboard, Friesen said he felt a familiar sensation—the same one he experienced when he watched some of his top musical talents at the start of their careers. “Danny was like ambition and intellect on steroids,” Friesen said. “His belief in this thing was so profoundly convincing that I believed, too.” On the spot, Friesen pulled out his checkbook and handed them a check for half a million dollars. “I remember walking out of the building and thinking ‘Friesen, what have you done?’” he said. “When people asked me what, exactly, does Akamai do, I’d say, ‘I don’t know.’ But the thing I did know, when I made that commitment, was that Danny Lewin was a star.”

  Friesen became one of Akamai’s key contacts on the West Coast and a willing point of contact for every potential investor or customer in his entertainment and media circles. He’d already connected Seelig with media companies and investors at Sequoia. In no time, he brought several more angel investors on board, including the publishing magnate Jann Wenner, Hard Rock Café co-founder Peter Morton, and the heavy-hitting New York attorney Joel S. Ehrenkranz.

  With secured angel funds, Jonathan Seelig circled back to a company called Exodus Communications in San Francisco, California. Founded in 1994, Exodus had a massive glob
al network of Internet data centers. Seelig had met with a team from Exodus several times over the summer, so he set up one more meeting with the company’s president, Ellen Hancock. Lewin flew out to San Francisco so that he and Seelig could combine forces. They slept on the floor of a friend’s apartment in the city and planned their pitch. Yet Exodus’s services weren’t cheap; it was the five-star data center complete with HVAC cooling systems, earthquake protection, a security system, and on-site power with multiple backup generators.{33} In the meeting with Hancock, Lewin presented Akamai’s technology on the whiteboard. Hancock was so impressed she offered them a deal: Akamai could have space in almost a dozen Exodus service centers to install and unlimited connectivity for three months. If Akamai had raised startup funds by the end of the trial and was ready to launch, Exodus would begin charging for the space.

  The injection of funds and the space to put servers allowed Akamai to begin the business in earnest: building the software, planning the deployment of servers with Akamai’s technology, and finalizing agreements with a core group of beta testers. They also hired their first employees, including a handful of Leighton’s Hacker Haven students, some other programming masterminds, and a network systems engineer named Bill Bogstad. By October, they were well on their way to launching Akamai’s debut service, called FreeFlow. But to secure more money in a first official round of financing, they needed a business veteran. Or, as Todd Dagres of Battery Ventures put it, they needed a “risk mitigater.” Akamai was still lacking in business gravitas; thus, in meetings with venture capitalists, their inexperience was often obvious. Seelig and Kaplan had some experience in the real world, but they were young. Lewin and Leighton were green. Before Battery Ventures would invest, Dagres said, Akamai needed “an adult with business experience to help them run a service.”