One way to look at how much the industry is changing is to spend some time with an executive – and a company – that has stuck to the same formula and industry view for almost 20 years.

Phil Harvey, Editor-in-Chief

August 16, 2018

10 Min Read
Cogent CEO Dave Schaeffer: The World on a Fiber

If people rattled when they shook their heads, the hotel ballroom following Dave Schaeffer's presentation at the 1999 OptiCon would have sounded like preschool maraca lessons.

That conference was one of the many venues that year where Schaeffer pitched the idea of a radical new ISP to investors and industry folk. He didn't have a network yet. He didn't even have PowerPoint slides. He got behind a podium, with no notes, and talked about how he could solve some vexing network problems by throwing more bandwidth at them -- a LOT more bandwidth.

Schaeffer tuned out the proverbial rattling. "I think the reaction, in terms of the business was, you know, we're not sure that a narrow-focused business, with limited products, kind of a very standardized offering can survive," he told Light Reading earlier this month, when we visited his office in Washington, DC.

Schaeffer's company, Cogent Communications Holdings Inc. (Nasdaq: CCOI), is now one of the world's largest ISPs, carrying more traffic on its network than most telcos. During its most recent earnings report, Cogent said its quarterly traffic growth came in at 10%, while year-over-year traffic growth hit 44%. Revenues, earnings and customer numbers are also heading in the right direction (see chart on page 3). Investors admire the company's diligence, but constantly wring their hands because they know Cogent is in a commodity business, by Schaeffer's design.

The company's more than 65,000 on-net customer connections and its nearly 2,600 on-net office buildings and carrier-neutral data centers send traffic over its all-IP-over-DWDM network, protected at Layer 3, using Ethernet as its network interface. The distinction of "on-net" is important. Those are the connections that Cogent owns from end-to-end, where customers are physically connected to its network and it isn't dependent on local telephone companies or cable TV companies to help serve those customers. They're the most profitable customers for Cogent, and the ones for whom the ISP continually lowers prices, relative to its competitors.

From the beginning, Schaeffer was clear about his network vision: Cogent would have no legacy network, and that rankled the telecom equipment vendors back in 1999. "I think from the equipment vendors, their doubt was about using enterprise-class equipment, as opposed to carrier-class equipment, and then scaling that on a global footprint," Schaeffer said.

Figure 1: Big bandwidth, at the speed of light. Big bandwidth, at the speed of light.

The beauty of Schaeffer's plan was in the simplicity, if only this new ISP could manage its finances well enough and avoid being clobbered by the competition. "Cogent's strategy is straightforward: Keep costs down, and speeds high, by owning facilities, multiplexing single strands of fiber and simplifying administration and provisioning by offering just one product," Light Reading wrote in March 2000. (See Cogent Banks on T1 Replacement.)

"I mean, I think that [lack of product variety] was a real concern," Schaeffer recalled. "Even to this day, I think a lot more legacy telecom service providers scratch their heads and don't quite understand how we're still here."

We checked: Cogent is still here, but it's not standing still.

The company is perpetually in a race. It's racing to provide the most bandwidth at the lowest possible price, which means it's in a race to run its network at the lowest possible cost, which means it's in a race to take every advantage of new optical networking and routing technologies, as soon as they're available. "We divide the network into four big technology regions -- edge routing, core routing, metro transport and long-haul transport," Schaeffer said. "In all of those functional areas we are on our third generation of equipment -- we've done two complete forklift upgrades in 19 years -- and, you know, I'm sure we'll go to a fourth generation soon."

Next page: The Whole World's Throughput

Why the frequent upgrades? It's the only way to catch the brief advantage new tech can offer before the market catches up. The average price per megabit for a Cogent customer declined sequentially by 3.6% last quarter and about 25% from the year-ago quarter. "We understood that bandwidth was a commodity and therefore you need to have the lowest cost of production, and you always needed to refresh your technology to capture the advances in the industry," Schaeffer explained. "So, while we sell customers Internet access, the products that we really produce are interface-routed bit miles connected to other networks, and we had to really be focused on driving down that cost."

With Cogent's network architecture, and its ability to make strategic, bargain-priced acquisitions (13 of them, so far), the company is primed to keep up with the soaring traffic demands of its customers -- and your snot-nosed teenagers.

"Ninety-five percent of our traffic -- and what drives the technology in Cogent -- is actually the residential consumer business," Schaeffer noted, "even though Cogent has no direct exposure to that industry -- our exposure is indirect -- and we have really two ways to serve that industry. One is we sell upstream to about 6,300 ISPs around the world. As their end-user base spends more minutes [online] a day, or [consumes] more bits per minute, we end up selling them more services at lower and lower prices. And on the flip side of that transaction, we also sell to the content publishers -- whether they're the direct publisher, like a Google, Facebook, Netflix or an Amazon, or an aggregator like an Akamai or a Limelight. (See Cogent Gearing for Another Peering Battle.)

"In both sides of the transaction, two things are true about Cogent. One is, we're very transparent with customers, telling them where their traffic is going and disclosing that we might be getting paid on both sides of a transaction -- and we reflect that in our pricing. The second thing is that we try to help the customers analyze traffic patterns in a way that a lot of times they can't do easily," Schaeffer said.

He noted that Cogent usually has "very detailed and granular visibility into where customers traffic enters the network, how far it goes, where it's exiting, and to what networks it's exiting, what time of day, and that incremental information helps our customers, on these bulk purchases, to optimize their network."

The payoff? "Even though we're a dumb pipe, there's kind of some network intelligence that our architecture can deliver that doesn't really exist anywhere else," Schaeffer said.

How much more bandwidth can Cogent's network support? It's meant as a far-out kind of question, but Schaeffer answer quickly. It's the kind of thing he thinks about a lot. "When we think about where we can go today, it is possible to run 160 channels in the EDFA (Erbium Doped Fiber Amplifier) bands -- that's using both C and L [bands] and 50-gigahertz spacing," he said. Then, after talking a bit about the various attenuation issues in optical networking and how to overcome them, he sticks the landing: "Practically speaking, you could take the entire throughput of the world and run it on one pair of fibers today and still have excess capacity with that technology," Schaeffer said.

"What's even more interesting is that we're only starting now to explore utilizing other bands. So, EDFA was just simple and easy for amplification, but there's no reason to think that there aren't other exotic materials that can be doped into fiber to create other bands of amplification, which would then dramatically increase the throughput on a single pair of glass," Schaeffer said. "So, I think there's just a very long way to go before we either hit exhaustion or before the current outside plant proves to be obsolete."

Figure 2: Dave Schaeffer, CEO, Cogent Dave Schaeffer, CEO, Cogent

Next page: 'We Only Sell Things People Want'

We do need fiber everywhere to provide big bandwidth, Schaeffer said, but we don't necessarily need a bunch of fiber everywhere. "It's like a swimming pool. You swim in the top ten feet, so nobody cares if your pool is 1,000-feet deep. In fact, a 1,000-foot-deep pool costs a lot more to maintain. Right? So, owning surplus fiber is actually not an asset. It's not necessarily economic; it's more expensive to maintain." (See Cogent Hedges Fiber Bets.)

Why, then, does fiber seem so plentiful in some places, but not others? Regulation, Schaeffer said. The world's developed countries have created hostile regulatory environments for building privately-funded, modern networks. Said Schaeffer: "I think everybody agrees on two points: One, the HFC plant -- the coax plant -- needs to go away. And two, everyone wants [network modernization] funded with private capital, not public. Those are both very laudable goals."

However, where regulators are screwing up, Schaeffer said, is that they try to create facility-space competition in what is a natural monopoly business. "It creates a recipe for investors to overbuild and that's a recipe to destroy value. If Google and Verizon can't do it, then nobody can do it, because their cost of capital is cheaper than anybody else's," he said.

"So, when I look at it, I think we're going to get there. We're going to have a completely fiber-developed world at some point, but it's going to take longer, it's going to destroy more value, and it's not going to address the needs of the underserved as efficiently as it could if we just had some rational regulation."

Cogent, of course, has to tread carefully. It overbuilds incumbent carriers in most locations and has to be "super selective" about where it expands and what markets it takes on, Schaeffer said. But not being everything to everyone has paid off. "I've started seven companies from scratch and Cogent is the first company that is the same now that I'm running it as it was when I thought of it."

Thinking back to what Light Reading wrote about Cogent in 2000, I check the company's 2017 annual report for a description of its updated strategy. "We intend to become the leading provider of high-quality, high-speed Internet access and private network services and to continue to improve our profitability and cash flow," the report said. That kind of consistency is rare in telecom.

"We're pretty disciplined. That's why we've grown our dividend for 24 consecutive quarters and guys like Frontier and CenturyLink and Windstream are cutting their dividends," he said. In the last 52 weeks, the company's shares have performed roughly in line with the S&P 500 -- both are up more than 14%.

Figure 3: Source: Cogent Source: Cogent

Meanwhile, outside Cogent's Washington D.C. offices, every crevice of the legacy telecom industry is being reshaped by virtualization, new services, content bundling and network intelligence. Cogent's barebones approach to delivering one thing -- really fast Internet access -- remains the exception that proves the rule. So long as physics, finances and regulation are still on its side, Cogent keeps adding buildings to its network and converting customers whose data usage is soaring compared to what it was just a few years ago.

"Look, we don't sell things that people don't want," Schaeffer said, smiling. "We only sell things people want, and we sell it cheaper. It's not really complicated."

Not complicated, said the man who argued against artificial intelligence in his network and recounted all the ways you manage amplification and dispersion in optical networks. Maybe the key to understanding Cogent is that it works exceptionally hard to hide complexity -- in its business and the industry it plays in. (See Dumb Networks, Not Just Dumb Pipes.)

Cogent plans to be here as long as there's a demand for high-speed Internet connections, delivered as cheaply as possible. Maybe it's really not that complicated. Perhaps everyone else is making this industry too hard to understand. As I exited Schaeffer's office, and looked outside, it was pouring rain. I take the last sip of my vending machine coffee from a styrofoam cup, wince and shake my head.

— Phil Harvey, US News Editor, Light Reading

About the Author(s)

Phil Harvey

Editor-in-Chief, Light Reading

Phil Harvey has been a Light Reading writer and editor for more than 18 years combined. He began his second tour as the site's chief editor in April 2020.

His interest in speed and scale means he often covers optical networking and the foundational technologies powering the modern Internet.

Harvey covered networking, Internet infrastructure and dot-com mania in the late 90s for Silicon Valley magazines like UPSIDE and Red Herring before joining Light Reading (for the first time) in late 2000.

After moving to the Republic of Texas, Harvey spent eight years as a contributing tech writer for D CEO magazine, producing columns about tech advances in everything from supercomputing to cellphone recycling.

Harvey is an avid photographer and camera collector – if you accept that compulsive shopping and "collecting" are the same.

Subscribe and receive the latest news from the industry.
Join 62,000+ members. Yes it's completely free.

You May Also Like