The Ultimate Comeback

Tom Leighton, Akamai and a Return from the Edge

Few start-ups, already under intense pressure to survive, are faced with the tragic loss of one of their co-founders just a few short years into their existence. When Daniel Lewin, the 31-year-old American-Israeli math prodigy, was killed on board American Airlines Flight 11 on Sept. 11, 2001, his company, Akamai Technologies, a Cambridge, Mass.-based start-up, was already reeling after the bursting dot-com bubble pulled the rug out from under its heady beginnings. Of all the people left behind, F. Thomson (Tom) Leighton, Akamai’s cofounder and the M.I.T. mathematics genius who had mentored Lewin, was perhaps the most devastated. Nothing could have prepared him for the events of 9/11.

Leighton, now the chief executive of Akamai, traveled an unexpectedly circuitous route to the CEO suite. It took him 14 years to get there, only at the last moment realizing that he was ready for the job. As an M.I.T. math professor known for his breakthrough work on algorithms, Leighton attracted Danny Lewin, a brilliant former member of Israel’s most elite special forces unit, to Cambridge where the pair formed a symbiotic relationship that led to Akamai’s founding in 1998.

Together, Lewin and Leighton envisioned a method using algorithms to speed up the Internet at a time when the dishearteningly slow Web was known as the World Wide Wait. They were responding to a challenge from the Web’s inventor, Tim Berners-Lee, an M.I.T. colleague, who foresaw the potential congestion on the Internet. Lewin persuaded Leighton to join him in transforming theory into start-up, and the company was born.

In classic Alphonse and Gaston fashion, Lewin asked Leighton if he wanted to be CEO, a position Leighton quickly declined, and Leighton, in turn, asked Lewin the same question. Lewin also refused. Both decided that the fledgling company required a certified tech industry business leader to run things. They reached out to former I.B.M. executive George Conrades and former Time Inc. executive Paul Sagan, who both signed up with a mix of enthusiasm and skepticism. This was the height of the dot-com mania, when any cockamamie Web-based idea was reaping vast sums of venture money. Akamai, which had a breakthrough technology concept that actually worked along with a small cadre of brilliant M.I.T. computer scientists on board to develop the concept, looked to Conrades and Sagan as a bet worth making. Conrades became the founding CEO and Sagan, the president. Leighton was content to remain as cofounderer and chief scientist, and when Conrades stepped down in 2005, Akamai turned to Sagan to succeed him.

Flash forward to December 2012. Sagan, who ran the show for eight years, was stepping down, and a search was on for his replacement. After 14 years as Akamai’s chief scientist, Leighton decided, with some urging from Conrades and Sagan, that it was his turn to take the reins of the company.

The son of a nuclear engineer who designed nuclear warships, Leighton displayed remarkable math skills early in his childhood in Arlington, Va. He was a finalist in the Westinghouse Science Talent Search and ended up with a scholarship to Princeton, where he studied computer science and math and became fixated on the field of theoretical computer science, which married those disciplines. Inevitably, Leighton was drawn to M.I.T., where he did his postgraduate work and stayed on to teach. Ensconced in this fortress of virtuosity, Leighton emerged as a genius in the esoteric field of algorithms and became a lure for the best and the brightest young minds. He was content in academia and hadn’t considered leaping into the start-up fray until he met Lewin.

It is unusual for a founder to wait 14 years to become CEO of his company, but Leighton, characteristically, made his decision based on a logical assessment of the situation and of his own skill set, which he believed was now complete enough to allow him to take charge. It was no slam-dunk. The Akamai board of directors conducted an eight-month external search for Sagan’s successor and had to be convinced that Leighton was the right choice. Yes, he knew the company better than anyone, but he had never managed a business. With Conrades and Sagan, both board members, singing his praises, and Leighton’s own impressive self-evaluation in front of the board, the decision was made. On Jan. 1, 2013, Leighton became the company’s third CEO.

The analysts who track Akamai were not concerned about his lack of CEO experience. “He’s had a C-level position at the company for pretty much that entire period, and in addition to that he’s sat on the company’s board,” Scott Kessler, an analyst with S&P Capital IQ, told Bloomberg News at the announcement. “The understanding of the company is unequaled by any individual.”

Akamai’s history has been tempestuous. The company burst on the tech scene like a moon rocket, soared instantly to dizzying heights and just as quickly plunged back to earth. Along the way, it endured a tragedy on 9/11 that shook the foundations of the nascent company. Though it was hardly a “dot-com” venture, it came to symbolize for many the frenetic nature of that short and volatile era. Many industry pundits and naysayers wrote the company off, and even internally, there was a quarter-by-quarter survival watch. “How much cash,” employees wondered, “is left until the end arrives?”

Leighton endured the trauma and never considered returning full-time to his M.I.T. ivory tower. In so doing, he became a crucial player in a comeback story that bears retelling. Along with Conrades and Sagan, Leighton looked down into the abyss and never flinched. To some, his elevation to CEO seemed a bit odd. But from the inside looking out, it was inevitable.

Taking the Hill

Before Danny Lewin was to board American Airlines Flight 11 to Los Angeles on Sept. 11, he and Leighton stayed up most of the previous night planning Akamai’s next layoff. This was the painful, unexpected dark side of life at a start-up. Akamai was plummeting.

When the charismatic Lewin was killed on 9/11 along with nearly 3,000 other innocent civilians, the stunned employees at Akamai had no time to grieve. Government Web sites, suddenly flooded by millions of hits, turned to Akamai to help stay online through the crisis and beyond. With tears flowing down their faces, the team rallied.

“Everyone knew it was what Danny would have wanted,” Leighton said. “It was a culture of ‘Take the hill, get the job done, make it work, do the impossible, never give up.’ It was Danny’s spirit. Yeah, we were devastated, no question about that. But we were fighting for our lives as a company at that time, and on top of that, the government needed us. We had to keep them online because it was chaos.”

What happened to Akamai was complex on the one hand, but a simple case of guilt by association on the other. In the midst of the dot-com run-up, Akamai emerged as one of the brassiest new players. Akamai wasn’t a Web site selling pet supplies or toys. It was addressing a serious need — using complex algorithms to dynamically map Web traffic to avoid congestion on the Internet — and had the imprimatur of M.I.T. When Akamai filed for an IPO just over a year after its founding, investors were wowed by the young company’s bold initiative and imposing management team, as well as the unparalleled enthusiasm that Lewin brought to every sales call and customer encounter (see book review, page 70).

For Sagan, meeting Lewin and Leighton in their M.I.T. lab had been life-changing. The original “big idea” was built upon changing the Internet, making it work better and faster through the radical application of mathematics. Using its algorithms, Akamai would allow Internet service providers to host content on thousands of servers around the world and thus be able to handle the flood of traffic and provide fast, impressive content distribution and even data-intensive material like video and graphics, without crashing the sites.

“The big idea is what affected George and me,” Sagan said, “along with the incredibly smart people and the opportunity to work with them.”

With early customers like CNN, Apple, Disney and Microsoft, Akamai made a stunning debut. When it went public in October 1999, shares skyrocketed to $145 on the opening day, and both Lewin and Leighton were suddenly worth nearly $2 billion on paper. By the end of the year, the stock was trading above $340 a share, and euphoria enveloped the young company. But when the bubble burst on the dot-com era, harsh reality set in quickly. As the dot-com start-ups began to fall like dominoes, Akamai suffered as well. During the crash, panicked investors saw no distinction between a start-up with a viable raison d’être and the pretenders built with smoke and mirrors. Akamai’s shares tumbled so far—to less than a dollar a share —  that it was nearly delisted from the Nasdaq.

Losing Lewin was an emotional body blow. “Danny was irreplaceable,” Leighton said. But the darkest fiscal days for Akamai were still ahead. “We had nowhere near reached the bottom financially by 9/11,” Leighton said. “We reached bottom emotionally when Danny was killed, but financially things were going to get a lot worse before they got better.”

Akamai’s market capitalization sunk from $35 billion at the height of the run-up to $50 million by early 2002. The company had $300 million in debt, which was at junk status. “From the outside world, we looked pretty dead,” Leighton said.

The employee count, which had grown to 1,300, had to be slashed to around 500. Given the tight-knit nature of a start-up with its intense, take-no-prisoners culture, laying off so many was nightmarish. In situations like this, most founders simply cut their losses and head for the exit. But Conrades, Sagan and Leighton, driven by Lewin’s gritty spirit, refused to concede defeat.

“We never felt like we were going to turn out the lights and lock the door,” Leighton said. “We weren’t in La-La Land. We knew how many quarters of cash we had left, and we knew the challenges we were facing. But we also knew we had a plan. And we believed in the plan, and we were going to execute that plan.”

The Comeback

A week after 9/11, the company paused for a formal memorial service for Lewin at M.I.T.’s Kresge Auditorium. Speakers tearfully recalled the remarkable life force that was Lewin. Rumors were already spreading that Lewin had tried to stop the terrorists on Flight 11 but was stabbed to death before the plane hit the World Trade Center. If anything, the event reinforced the company’s determination to push on, as Lewin would have wanted, and find a path to success.

At the genesis of the comeback was a wellspring of talent at the top. No one could replace Lewin, but Leighton was more than his equal in understanding and propagating the underlying technology of the company’s offerings. Having had two cofounderers was a saving grace. Lewin had focused on customers and outward-facing issues, while Leighton had handled internal assignments. But because they had worked so closely together, with a shared skill set, Akamai’s complex technological underpinning was not compromised. Leighton, who had continued to teach in the algorithm unit at M.I.T.’s Computer Science and Artificial Intelligence Lab, gave up his academic role for several years and pulled his Akamai team together to fill in the gaps. Added to the seemingly short list of positives was the presence of Conrades and Sagan, seasoned business leaders who had made deep commitments to Akamai and were determined to spearhead a comeback. Conrades quickly outlined his plan and set it in motion (see sidebar page 58).

Conrades also made a public effort to display his confidence. With the shares at a dollar, he invested $1 million in Akamai stock. The employees saw it. Prospective customers saw it, as did investors. At the same time, “we suffered together,” Conrades said. The top three executives set their cash compensation at $25,000 each without a bonus. “Everybody knew this, and it was an important way to communicate our commitment,” he said. “It was a Dannyism that he got from the military: You have to suffer together.”

Leighton demonstrated his own commitment by sponsoring parties at the company’s offices, which he paid for out of his own pocket, to celebrate important milestones.

But all the internal morale-building would have been for naught without a restructuring of the business and a slow but steady influx of new customers. Ironically, it was Akamai’s performance on 9/11 and in the immediate aftermath that became the focus of its short-term sales pitch. During the worst chaos on that terrible day, Akamai’s networks had worked! Word spread quickly that Akamai was the place to go to make sure your Web site didn’t crash. On 9/11, CNN’s site was flooded with traffic. But its Akamai-based network was able to handle the crush without a glitch. According to a 2005 article in Business 2.0, the homeland security business that sprang up after 9/11 “helped resurrect the company.” The FBI, for example, had been outraged when part of its Web site went down during the attacks. So the bureau reached out to Akamai and by March, the FBI had become one of Akamai’s first major government customers. In short order, the new Department of Homeland Security, four branches of the military and the Internal Revenue Service signed on as well.

The promise of uninterrupted Web service also drew in new business from companies like ETrade, FedEx, L.L. Bean and MTV.

Behind the scenes, Conrades and Sagan drew up a new organization chart and divided the key responsibilities. The important thing was to put the right people in the right boxes. “We made choices,” Conrades said. “Who was on the team we would go with? We’d hold them accountable in transparent ways with frequent reviews. We had quarterly meetings, and we’d put them all in a room and reviewed the commitments they had made and whether that commitment had been met. ‘We’re going to acquire so many accounts, ship so many servers, get so many new networks.’ We worked this as a team.”

Leighton’s team had to bulletproof the company’s offerings and make sure new releases were timely and reliable. On this precarious perch, glitches could be fatal. Inevitably, they came anyway. After one software rollout onto a live network, a malfunction hit some of the company’s big customers such as General Motors.

“G.M. had a practice of calling in vendors who failed them,” Conrades recalled. “Tom and I had to go to Detroit to talk to them and tell them exactly what had happened.” When they were called into the conference room with G.M.’s CIO Ralph Szygenda, Conrades began by thanking him for the opportunity to explain what happened. Szygenda, a bit of a taskmaster, laughed. “You know, George, all the years we’ve been doing this, nobody has ever thanked us for this opportunity.”

Conrades replied, “We know what went wrong, and we want to explain this to you and how we do it.” He then introduced Leighton, who deftly explained how a human error had affected the rollout and that it would not happen again. Szygenda was more than mollified—he was so impressed that he asked the pair to wait and sit in on another meeting to explain how Akamai could update software in a running system. Conrades invited him to Cambridge for a further meeting, and G.M. upped its commitment to Akamai. It remains a loyal customer to this day.

A New CEO

With a slow but steady pace of progress, Akamai emerged from its crisis as the major player in a small but crucial market sector. Its customer list is now populated by hundreds of multinational enterprises such as Merck & Co., Airbus, the British Broadcasting Corporation, BMW, Best Buy, Nintendo, National Public Radio, Charles Schwab, Dow Chemical, Fox Broadcasting, Hitachi, I.B.M. and Verizon Wireless.

At any given time, 15 to 30 percent or more of Internet traffic flows through Akamai servers around the world. When he reached his self-determined targets of growth and profitability in 2005, Conrades, in his late 60s, stepped down. Sagan reluctantly, and with strong persuasion by Leighton and Conrades, took over. Former head of Time Inc.’s New Media division, Sagan steered Akamai to unprecedented growth, with sales jumping from $250 million to $1.5 billion and headcount up to 3,500 employees.

Sagan never intended to stay so long at Akamai’s helm, and when he announced his decision to step down in early 2012, Leighton had no designs on the CEO suite. “I was very happy during those 14 years working for those guys,” Leighton said. “I learned a ton from both George and Paul and worked very closely with them all through those years.”

As a national search for his successor was proving difficult, Sagan sat down with Leighton for a conversation. “Tom said, ‘Well, aren’t there more candidates out there?’” Sagan recounted. “I said, ‘There’s one obvious candidate and he’s right here. And that’s you... but you have to want it.’ He did the Tom thing. He thought about it and came back and said, ‘O.K., tell me everything I don’t really understand. What else is there?’”

Leighton’s reluctance had nothing to do with a lack of self-confidence. He had displayed business skills throughout his tenure as chief scientist, negotiating many of the company’s first network and colocation deals with customers. He was a shrewd negotiator and born leader, Sagan said. “People loved to follow him.”

More likely, his comfort in his long-held position coupled with the company’s culture kept him from throwing his hat into the ring. Self-reverence is anathema at Akamai. It is never about “I.”

“People who talked about themselves or thought about themselves in that way at almost any level here are rejected by the culture,” Sagan said. “You have to be all about the team.”

For Leighton, the leadership aspect of the job didn’t present a problem. “I was playing a leadership role throughout my 14 years here,” he said. “Plus George and Paul have always been fantastic mentors, so I got good advice from them. And I did go out and get some books on management.”

Leighton’s rise comes at a time of relative stability for Akamai, but he is acutely aware that economic volatility and the ever-present twists and turns of the technology sector make complacence impossible. There is no “hockey stick” of growth around the corner, and he must fend off ceaseless rumors of mergers and takeovers while keeping the product pipeline full.

For smart executives, a comeback never ends. Leighton accepted the baton from Sagan and made no fundamental changes in the company’s strategic planning. Leighton recognizes the importance of culture, and at Akamai the culture was forged through cataclysmic events with which few companies must contend. “Part of our culture is that we’re always looking up at the next hill,” he said. “Coming from M.I.T., there’s a sense that you can solve any problem if you work hard enough and smart enough, and as a team. It’s intense.”

A year into his tenure, Leighton is confident that he made the right decision. Eventually, he said, he realized “that this would be the best outcome for the company. Because I really care most about the company. You put more than 14 years of your life into something and you really believe it has the potential to change, maybe not the world, but at least the Internet for the positive. You want to do what is best for the company. And if I look back now, I’ve been training for this job for 14 years.”

Shaping a Comeback

Despite the crushing loss of Akamai co-founder Daniel Lewin and the company’s struggle to survive in the aftermath of the dot-com bubble, industry veteran and Akamai CEO, George Conrades committed to stay. “We shall return,” he told employees. He and Akamai President Paul Sagan had a plan. The focus would be on three key agenda items: productive revenue, reduction in spending, and keeping up morale.

By 2005, Akamai had revenues of $200 million, achieved sustained profitability and had a positive cash flow. Conrades stepped down and Sagan reluctantly took the helm, a post he held for the next eight years. Sagan then handed the keys to Leighton. Both Conrades and Sagan remain on the board.

Does the technology work?

Is it the right business model? If we grow revenues and keep costs down, we will be profitable. People need to understand this.

Do you still believe in the big idea?

The people. Look to your left, look to your right. Do you love these people? Do you want to work with them?