No Better Time

by Molly Knight Raskin

  For most employees, the around-the-clock commitment to the company came easily. In addition to the looming prospect of an IPO, the atmosphere in the office was often exhilarating. Every day seemed to mark another milestone. Most of Akamai’s employees spent more time at 201 Broadway than they did at home, staying until all hours of the night and returning early the next morning juiced up on coffee and the adrenaline of being at the center of the boom. “All of these amazing things were happening and you wanted to be there,” said John Sconyers. “You never knew what would happen, but you knew if you weren’t there you might miss it.”

  The heart of Akamai’s headquarters was, and still is, the Network Operations Command Center, known internally as the NOCC. The NOCC still looks like something out of NASA command—a dimly lit room filled with banks of flickering computers. Larger screens line its walls, displaying what appear to be impressive numbers like “1,507,193 hits per second.” From the NOCC, Akamai boasts a bird’s-eye view of global Internet traffic; at the time, it was a perspective no one else in the world could boast. In that one room, the company has the capacity, using data from its global network, to gather information about congestion before most ISPs even know traffic is mounting. The center of the NOCC is a digitally rendered image of a spinning globe, which twinkles with thousands of tiny lights resembling stars, each one representing a city where Akamai has servers in one or more locations.

  Outside the NOCC, Akamai’s home at 201 Broadway was nondescript—clusters of cubicles and offices—but it had all the trappings of the trendy startup. MIT whiz kids who were barely old enough to order a beer came to work on rollerblades and skateboards. Every Thursday, a delivery truck pulled up and stocked the kitchen with Ben & Jerry’s ice cream, popcorn, soda, and frozen pizzas. A group of programmers, otherwise known as the “Java Weenies” for their caffeine-fueled all-nighters, spent their time producing the interface and graphics for the system and taking naps in a hammock suspended from the ceiling. Will Koffel, a student at MIT, was one of them. He recalled juggling the coursework for his dual degree at MIT with a part-time job at Akamai, where he worked the overnight shift overseeing operations in the NOCC. Koffel would attend class from 8:45 a.m. to 4:00 p.m., study from 4:00 p.m. to 7:00 p.m., and work at Akamai from 9:00 p.m. until 5:00 a.m. “My rule was that if I fell asleep before the sun rose, then it wasn’t an all-nighter,” related Koffel. He said Lewin was in the office so often and at such odd hours that he finally began to wonder if he ever slept at all. One night, he recalled, Lewin was at the office at 3:00 a.m. in what Koffel called “Field Marshal mode,” coordinating all kinds of efforts and keeping everyone awake and on track. Koffel asked him, “Danny, how long have you been here?” Lewin replied: “Three days.” When Koffel expressed his surprise and asked him what his secret was, Lewin told him a story about his time in the army, when his commanders would make everyone in his unit stand in full gear and a backpack for twenty-four hours straight. Every time someone would flinch or collapse, they’d add another hour to everyone’s time.

  Koffel said the office was “hopping” at all hours, with people rolling around in office chairs from one desk to another, tossing footballs, and microwaving a seemingly endless supply of burritos. “We’d be microwaving all night,” he said. “There was just so much energy that you didn’t even realize how exhausted you were.” It was not uncommon to see two employees tossing a Frisbee across the room or playing miniature golf on a makeshift par-seven course set up between a few desks. The atmosphere at Akamai was so fun and so intoxicating that it became easy for people to lose track of time within its walls.

  Melanie Wynkoop was one of the first few women to join Akamai’s sales group. At age twenty-six, she came to interview when the company had fewer than one hundred people after hearing about Akamai’s young founder who was being compared to the likes of Bill Gates, but with more charisma. She came from Price Waterhouse Coopers. “I read about the company and its background, and when I came to interview, I’d never seen anything like it,” she recalled. “It was small, booming, and fabulous.” Wynkoop arrived just as the sales craze began, the company securing contracts by the dozens. “It was almost like we were order takers,” she noted. Despite the rush, Wynkoop said Lewin always pushed the sales staff. “He’d come to sales meetings and say, ‘You people suck,’” she said. “But he wasn’t trying to be mean, he was just going for the jugular. That’s what he did.”

  But if Akamai was lacking anything, it was a chief executive officer. The company was growing exponentially, gaining customers by the handfuls and generating buzz as the next big thing, but none of this guaranteed success. For any Internet company, the paradox of making it in the dot-com boom was that, in order to survive, businesses had to explode onto the scene and keep growing rapidly; but the very pace that brought them to the top could take them out in one fell swoop. Akamai needed a seasoned executive at the controls, one with enough experience and battle scars to maintain calm in the boardroom, keep an eye to the long-term forecasts, and steer them through the inevitable ups and downs. The median age of the company’s employees was thirty, and many, including three of the company’s top executives, had come straight from academia. Akamai needed an elder statesman.

  Leighton recalled having a conversation with Lewin early on about the top role, which either of them could have easily assumed. Lewin didn’t want to be CEO, but said he’d work for Leighton if he wanted the job. Leighton said he didn’t want the job, but that he would work for Lewin if he did. This set off a formal search for a CEO that, by the spring of 1999, was well into its third month. A top executive search firm regularly sent candidates to Cambridge, but when it came time to decide on someone to fill the job, Lewin couldn’t commit. The reason, Leighton said, was that Lewin had already decided on the man for the job. In fact, he’d made up his mind before Akamai even incorporated, drafting a list of dream CEOs for his dream company. Topping the list was George Conrades.

  At age sixty, Conrades was comfortably settled into a plumb job as a Venture Partner at Polaris. After four decades on the frontlines of two of the most successful computing companies—International Business Machine (IBM) and Bolt, Beranek, and Newman (later BBN Technologies)—Conrades had retired from the career-climbing race. Or so he thought.

  Conrades first laid eyes on a computer as a student in math and physics at Ohio Wesleyan University. Conrades, who graduated from the university in 1961, was then somewhat of a star on the university’s campus. Handsome and charismatic, he served as president of both his fraternity and the student body. He played the drums in a rock ‘n’ roll band, which he described as “full of testosterone,” and spent a lot of time in the physics lab trying to blow things up. But Conrades was also a standout student with a gift for both math and physics. So when the head of the school’s math department decided to offer a computing course, Conrades signed up. Before he knew it, he’d learned to program one of IBM’s earliest models, the 650. In his day, Conrades said, “No one had even heard of computers.” With a love for motorcycles, hot rod cars, and “anything that moved,” Conrades said that he took one look at the colossal machine—with its blinking red lights and rotating magnetic drum—and fell in love. With its relatively low cost and ease of programming, the 650 was marketed as a teaching computer to science and engineering schools across the country.{40} Conrades became so skilled at programming that, before he graduated, some of the faculty members at Wesleyan approached him with a list of computer companies, including IBM, Honeywell, and NCR (National Cash Register Company). “They told me that I should interview at every one of them, and that I should be in technical sales,” Conrades remembered. “I told them that I really wanted to make it as a rock star.”

  Reason prevailed, however, and Conrades landed an entry-level job at IBM, which by the 1960s had burgeoned into a $1 billion business. Conrades began as a systems engineer, and at the same time earned an executive MBA from the University of Chicago School of Business. Over the course
of 31 years, he rose up through the ranks at IBM to become senior vice president of IBM North America, a $24 million business. In 1992, after a dispute with then-chairman of IBM John Akers, Conrades left the company. By 1994, he had become president and CEO of BBN Technologies, formerly Bolt, Beranek, and Newman, the technology and research firm that helped build ARPANET. When Conrades joined BBN, the company was a think tank that survived on government contracts. Conrades leveraged the tremendous brainpower at BBN to transform the company into one of the world’s largest Internet Service Providers. In 1997, GTE Corporation{**} purchased BBN for $616 million, or $29 a share, more than double the stock’s value when Conrades came aboard. But the corporate culture of GTE didn’t agree with Conrades, and after a year spent as the president of GTE Internetworking, he left and took a year off work for the first time in his career. Conrades could have comfortably retired, but he soon realized that slowing down was not for him, or his wife, Patsy, whom he met in college. “One day, I opened the freezer and suddenly everything was falling out—chicken and steak all over the place,” recalled Conrades. “And Patsy looked at me and said, ‘George, stay out of my freezer. And get a job.’” The fact was, Conrades had spent too much time at the forefront of the digital age to sit on the sidelines of the dot-com craze. He wanted to play a part in the next big thing, so he looked to venture capital, which he said was raising its head in anticipation of the digital gold rush. In August 1998, he joined Polaris.

  That summer, Conrades first made the acquaintance of Lewin and Leighton, agreeing to meet them at the suggestion of Battery’s Todd Dagres. Conrades recalled sitting through the pitch session and thinking that, as much as he liked the idea, he was initially uncertain about Akamai’s business model. “I understood what they were saying—not at the level of the algorithms—but about making the Internet feasible for e-commerce and robust audio and visual interaction,” Conrades explained. “I lived the problem at BBN, even though we threw a lot of money at it. That’s when I learned that you can’t throw enough money at the Internet to make it work right.” By the end of the meeting, however Conrades was so impressed with Akamai’s solution that he felt Polaris had a potential gold mine. He helped convince Polaris to invest in Akamai, and decided to put in some money of his own, too. According to Conrades, his investment strategy was informed by four basic elements: the idea’s greatness, the technology’s potential impact, the business model’s strength and, most importantly, the employees’ overall caliber. For Conrades, the people always came first. As for Akamai, he had no concerns about the people. Then he considered the idea. To him, the concept of what he called an “agnostic” network—one that provided a synoptic view of the Internet—was brilliant. “It was a big idea,” Conrades said. “That’s a word I use for something that’s not incremental in its impact, but transformative. They were promising a better Internet.”

  When Polaris put $4 million into Akamai’s first round of financing, Conrades agreed to sit on the company’s board. At the time, he had no idea that Lewin had a plan to lure him into the role of Akamai’s CEO. “It was a whole seduction to get George,” recalled Leighton. “Of course we knew he wouldn’t be our CEO on day one, so Danny worked to get him on our board. Then we got him on the board and kept working on getting him as CEO.” Conrades knew Akamai was actively searching for a CEO, but was unaware of Lewin’s subtle efforts to court him. So when Lewin and Leighton invited him for breakfast at Harvest restaurant in Cambridge in late March 1999, Conrades wasn’t expecting them to offer him Akamai’s top job. “We told him point blank, we need you to be CEO—what’s it going to take?” remembered Leighton.

  Conrades laid out some terms, and Lewin said they’d get back to him later that day. On the way home from lunch, Leighton noted, instead of celebrating what was sure to be a win for Akamai, Lewin suggested they push back on Conrades’s terms. “I’m driving away from the restaurant, and Danny is saying how he wants to negotiate,” observed Leighton. “I said, ‘Danny, what are you doing? Just say yes!’” Leighton added: “That was Danny, always wanting to negotiate the best possible deal.”

  In the end, there wasn’t much negotiating, and, on April 7, 1999, George Conrades became Akamai’s first CEO, a fitting position for a man who had long been associated with speed. “To me, Akamai is like Fed Ex,” explained Conrades. “Fed Ex changed the game on the postal service, and we changed the game on the Internet.” The media seized on Conrades’s move to Akamai as another example of many seasoned executives taking chances on Internet startups in exchange for equity. Reporting on this trend, the May 1999 issue of Forbes singled out Conrades, Richard Frank, the ex-chairman of Walt Disney Television who signed on to head Food.com, and James Cannavino, former CEO of Perot Systems who joined the small network security firm CyberSafe. The article posed this question: “Why are these elder statesmen, with little left to prove, now pulling twelve-hour days to run baby firms barely on their second round of venture funding?” While the story speculated that such moves were motivated largely by money, Conrades said he was genuinely thrilled to assume the role at Akamai.{41} “I was still full of energy, and I thought the people were just fantastic,” said Conrades.

  Conrades came on board just in time for Akamai’s second round of financing, led by Baker Communications Fund LP of New York. On May 7, Akamai secured $35 million total from Polaris and Battery Ventures, bringing the company’s total venture funding raised to more than $43 million. Todd Dagres told the media it was more than the company needed, and a sizeable sum even for hot startup, which averaged about $20 million in a first round in the late 1990s. Yet Akamai was less than six months old, and its market value had already multiplied tenfold. The day the second round closed, Conrades issued a statement saying Akamai had sufficient capital and had no immediate plans for an additional round of equity or any private or public offering.{42} First, Conrades said he needed to step back and give order to what was the chaos of a startup company moving too fast to position itself clearly. “My job was not to understand the algorithms and technology but to build the framework of a business,” Conrades observed. “We had to hire and establish a culture; it had to be codified for everyone.” One of the first questions he posed to the small company was this: Akamai exists to do what? It seemed so simple, but, as Conrades still recalls, “there was a lot of uncertainty.” With the help of everyone at Akamai, Conrades conceived of the company’s list of guiding principles, which still stands true today. One of them, which he and Lewin agreed on, was never to dismiss an idea from any source without first giving it consideration. “There was a highly argumentative culture in place, which was either inspired by Danny’s Israeli traits or the atmosphere at MIT,” said Conrades. “That was good—we reveled in arguing assumptions. But when we did we made a point of listening to everyone.” At IBM, Conrades concluded: “there was no ability to do this.”

  Some ideas, however, were not up for consideration, no matter how vehemently they were argued. Just after Conrades came on board, he asked Randall Kaplan—who was still working out of Los Angeles—to relocate to Boston. Kaplan pushed back. As the only employee based outside Boston, Kaplan argued that he had originally left his lucrative position at SunAmerica to join Akamai on two conditions: that he could remain on the West Coast, and that he would report to the company’s CEO. Conrades offered Kaplan a second option: remain on the West Coast, but report to Earl Galleher, the VP of Sales. Kaplan refused, and decided to resign. His decision didn’t go over well, particularly with Lewin. In the months leading up to Kaplan’s departure, tensions between the two were beginning to mount. Although he acknowledged Lewin as one of the most brilliant and talented people he has ever met, Kaplan said he disliked him on a personal level for being what he called too “abrasive.” Coworkers said Lewin came to dislike Kaplan, partly because he felt like Kaplan’s heart wasn’t in the company. Kaplan left the company with a very good deal for himself, one the Wall Street Journal referred to as “a boatload” of stock options—ones h
e could cash. “Even by the outrageous standards of Silicon Valley, Randall Kaplan is one lucky guy,” noted the newspaper. “Unlike his ex-colleagues at Akamai—who have to wait around for four years—Mr. Kaplan already owns all of his shares.” By the end of 1999, Kaplan’s 2.4 million shares were worth approximately $633 million. {43}

  Akamai used some of its new funds for hiring, or, as Lewin told reporters, “a massive mind-suck from America’s top universities,” to create a sixty-person research and development group. In the process, Akamai fended off complaints from a few MIT professors for creating such a powerful incentive for truancy. Even before Akamai’s IPO, there were employees who requested a leave of absence from their undergraduate studies at MIT to work full-time at the company. They were not the first wunderkinds to prioritize computing over a conventional education: Microsoft’s Bill Gates dropped out of Harvard, Apple’s Steve Jobs dropped out of Reed College. And in the late ’90s, with soaring markets propelled largely by Internet stocks, the tech sector became even more alluring, particularly to financially-strapped students inspired by the overnight millionaires of the time. In the late ’90s, as MIT began to spawn more successful startups, Michael Dertouzos, then director of LCS, voiced his concern that faculty, or students, might “call in rich.” In response, Akamai decided to take a firm line on the issue, partly because key members of its management team were also professors at MIT. To discourage overzealous undergraduates from abandoning their studies, the company instituted a strict stay-in-school policy.

  They also tried to encourage those who opted to join Akamai instead of a PhD or Master’s program to return to school and earn their higher degrees. But this became more of a challenge around the time of the IPO in the fall of 1999. Just months earlier, the Wall Street Journal ran a story by reporter Amy Marcus under the headline “Class Struggle: MIT Students, Lured To New Tech Firms, Get Caught in a Bind.” The story featured the dilemma faced by Will Koffel, who was working at Akamai when he was a junior at MIT. Koffel was handed an assignment for his computer systems engineering course from one of his professors: find a way to speed up the delivery of Web pages. Koffel had been working on the problem as one of the “Java Weenies” at Akamai, yet he was bound by a non-disclosure agreement (NDA) not to reveal any specifics about the job, particularly surrounding the design and engineering of the company’s technology.