“Everybody’s money is green,” says Internet entrepreneur Jonathan Seelig. “Any source can pay for computers and salaries. There’s nothing strategic about that, and it’s important to be strategic in where you go for money … We wanted to find investors who would give us credibility in the marketplace, help us recruit customers, and help us find the management expertise we knew we didn’t have.” Seelig, a founder and vice president for strategy and corporate development at
Akamai (Hawaiian for intelligent, clever and cool, according to the firm) is a 24/7 service company whose customers are Internet content providers like Yahoo, Reuters, Lands’ End, CNN and MSNBC.
What Akamai sells is the cyberspace equivalent of a by-pass route that promises to get data to end-users more quickly and more reliably than has been possible to date on the world wide web. Or as Seelig put it: “We alleviate the load at the source and get around the bottlenecks of peering and backbone by placing servers at the edge of the Net to serve the last mile. We have 6,000 servers in 335 networks in 54 countries around the world.”
Seelig defined the terms he used. “Most people tend to think of the Internet as some kind of cloud which has content providers on one side and end users looking at that content on the other,” he said. “They don’t think about how the content gets through the ‘cloud.’ It’s not easy.”
He explained the four stages of transmission: The “first mile,” which is getting information from the content provider’s web server onto the web; “peering,” which refers to going through exchange points to reach networks; “the backbone,” a network of networks [of which there are approximately 7,000, all independently owned, each determining its own interconnections and peering rights, according to Seelig], and the “last mile,” connecting to the end user’s computer.
Despite the efforts of many companies to improve the web’s connections, said Seelig, there are frequent overloads that result in data “traffic jams.” These produce delay (the famous “world wide wait”) and, sometimes, complete breakdowns. The three founders of Akamai saw a market opportunity in the frustration of content providers when they find themselves unable to get their content to end users quickly or at all.
Akamai’s unique product is a mathematical algorithm to route data to the company’s servers which then forward the content to the end user. The algorithm was developed by Tom Leighton, a mathematics professor at the Massachusetts Institute of Technology’s Laboratory for Computer Science, who specialized in using parallel algorithms to optimize the operation of complex networks. For three years, Leighton led a mix of MIT graduate and undergraduate students in the search for a way to improve traffic flow on the network known as the web.
The result was a patent on the algorithm, which is owned by MIT and leased to Akamai in return for an equity stake in the company.
Leighton founded Akamai in August 1998 by with Daniel Lewin, one of his grad students, and Seelig. Seelig was then studying for an MBA at MIT’s Sloan School of Business, but decided to drop out to join Leighton and Lewin in developing a business plan. Seelig had previously graduated from Stanford University with a degree in physics and worked for ECI Telecom in Tel-Aviv leading teams in network design.
The group, which spent six months coming up with a business model, was a finalist in the 1998 MIT $50K Entrepreneurship Competition. In addition to the technology background provided by Leighton and Lewin, the group made use of Seelig’s contacts in the telecom industry and network infrastructure world.
What the would-be entrepreneurs lacked was sufficient cash to build the infrastructure they believed the service would require. “We estimated we needed at least $6 million just to get started and decided to go after $8 million.”
And this is where what Seelig called “our value-added strategy” came in. “We sought out venture capitalists,” he said, “who could not only provide cash but could also help us in those areas where we weren’t strong. Our biggest concern was how do we support the operation. One firm, Battery Ventures, knew that a local exchange server firm had recently been bought out. Many of its employees did not want to move to Denver to take jobs with the new owners. “Those employees,” he said, “wanted a new opportunity and weren’t scared of 24/7.” The same venture capital firm also provided two seasoned executives who became Akamai’s CEO and COO. “Taking money from Battery gave us great value,” Seelig said.
In addition, Seelig noted, by applying to venture capitalists they knew had connections to potential customers for their service, Akamai sometimes got new customers “even when the venture capitalists chose not to finance.”
“The venture capitalists would ask a company in their portfolio to talk to us, as part of doing due diligence,” he explained. “They’re thinking: ‘If my customer likes you, that’s an objective opinion. It’s not coming from the entrepreneur’s second cousin.’”
Some investors, famous in their own right, provided the added value of “market cachet and buzz,” said Seelig. Akamai got this from Apple Computer, Cisco Systems and Microsoft. “We worked with those companies to develop a platform they could use and they also invested in the company. Taking that kind of investment validated us as a market leader.”
An audience member asked Seelig how to calculate the amount of stock one should keep when negotiating with investors. Seelig’s response: “As much as you can. Dilution is bad. But it’s more of a problem in the early rounds of money-raising.” Seelig said that Akamai stock is not only distributed to investors but also to employees. “Twenty-five percent of the stock is reserved for non-founding employees.”
Akamai’s most recent strategy to raise money was going public in October 1999. “It gave us increased credibility,” he said. “Our salespeople reported that they were getting calls returned that hadn’t been returned before. And going public gave us currency for growth through acquisition.”
The stock price at the IPO was $26 per share but by the end of the day investors had bid it up to $145 and a month later it was selling at $327. Seelig reported that AKAM was down to $28 per share the day of his talk but was unperturbed by the figure. “I’m happy to see Wall Street focusing on fundamentals once again.” The stock closed at $31 on December 18.
Akamai today has 1,300 employees and is no longer run by a committee of the founders and others. “The challenge for us now,” said Seelig, “is decentralizing authority.” The company has not yet turned a profit but Seelig is confident it will do so in the third quarter of next year. In the nine months ended September 30, Akamai lost $583 million on sales of $52 million.
Seelig explains the losses in terms of the heavy expenses the company has had to bear. “Infrastructure costs money,” he said. “We paid for 6,000 servers and licenses for software. We own a building in Cambridge [Mass.]. But now every customer we add makes us that much more gross margin positive. Wall Street respects that. We only started to provide our service in April, 1999. That’s less than a year and a half ago.”
Seelig had one final word of advice on financing a new venture: When you estimate how much capital you will need, be realistic about how long it will take to get customers. If you assume it will take three months to get a customer to write a letter of interest, it will turn out to be six to nine months. Working with customers to spend real dollars always takes longer than anticipated.”