Level 3 brings an engineer's approach to Networx

Level 3 relishes odd duck status on Networx contract as it pushes the advantages of its fiber network.

An engineer by training and choice, Ed Morche has spent his working life in telecommunications — at GlobalOne, Sprint Corp. and a string of other telecoms before joining Networx-Enterprise carrier Level 3 Communications Inc. eight and a half years ago.

He is reserved — even a little stiff — and chooses his words with the care one might expect of an executive responsible for all of the company’s federal government and federal systems integrator business. So the passion with which he speaks of the company, its executive team, and the science and technology of telecommunications networks comes as something of a surprise.

“One of the reasons we exist is, don’t forget, we had the first all-IP network in the world,” he said. “Today it’s easy to forget what a big deal that was. We were a bunch of engineers arguing about how do you deploy MPLS, this cool thing that exists only on paper. We deployed the Martini draft. We developed it; we rolled it out and said, 'Done! If you guys want to follow us, that’s fine.’ ”

Washington Technology editor Nick Wakeman and writer Sami Lais talked with Morche, Level 3’s senior vice president and general manager of federal markets, at WT offices recently on a wide range of topics. Here are some of the key issues discussed.

WASHINGTON TECHNOLOGY: Level 3 is kind of the odd duck among contract holders on the General Services Administration’s Networx contract. I mean, no one has a Level 3 Communications phone.

MORCHE: (laughs) Yeah, right, right. We have no end-users on our network, which means we don’t sell antiquated technologies, no Centrex, no frame relay. Now, if a customer still has services on frame relay, and they really need to keep it, we can do it. But we don’t go out and actively sell it. We tend to have a real engineering bent, so we’re more interested in newer, more efficient technologies.

WT: Level 3 was started in 1998 and didn’t start laying cable for this network until 1999, right?

MORCHE: Yes, but there are a couple of good things about being the last guy at the party. One of them is you’ve got the most modern fiber of all the networks. Some of the earlier networks out there are starting to have issues; they can’t perform the way they used to. Like the rest of us, we start to age, the knees start to hurt, and things just don’t work as well.

We also recognized when we were laying fiber that the biggest cost of a network was not the fiber and not the optronics; it was digging trenches, getting permits and crews, and going out and digging up roads. So we buried bundles of conduit so we could pull multiple networks in the future if we had to. Our concept was to make the network continuously upgradable.

WT: While Level 3 was spending $12 billion assembling 77,000 miles of fiber, other companies also were laying fiber, and, according to a Wall Street Journal story just this February, investors didn’t see the profits roll in as expected. Have you gotten your money back on that network?

MORCHE: We’re continuing to drive toward profitability. Could we have been more profitable? Absolutely. We could have not acquired other companies and kept that money for ourselves. We could have slashed headcount and been wildly profitable. The problem is that it would have been a good near-term solution and kept everybody happy, but then we’d have no long-term strategy. So we continue to work with our investors and look at what the right thing is for the company.

We’ve always been a technology leader, we’ve always been set up as the largest of the independent neutral providers, and you can’t do that if you’re not investing and growing.

WT: And that investment continues?

MORCHE: Yes. For example, we’re continuing to upgrade our long-haul [inter-city] network, which we own and operate. If you’re going to keep up with price compression, which you’re going to have — some years worse than others — with wavelengths and high-speed IP, which are the technologies that are the poster children for price compression, you have to keep [working at it].

You can just squeeze your margin profile, adjust your opex budget accordingly and reduce your headcount, or you can keep driving down your cost structure to try to keep your margin profile as attractive as it was before [the compression], and that’s what we do. We keep looking at newer technology at a lower incremental cost structure.

WT: With all the change in technology, in the business and economic environment, the changing opportunities, how do you stay focused on your core capabilities and business goals? 

MORCHE: That’s a really good question. The temptation is always there. Could you go and do wireless resale and make a couple of points on it? Yes, you could. Could you go out and sell a bunch of boxes and help pipes connect to them and make a couple points on top? Yes you could. Could you go out and sell bodies to do things and make a couple points? Yes you could but that’s not what we do. 

I think we’re able to do what we do because we are a more nimble company. We’re extremely flat in terms of hierarchy. There’s just a handful of us at the executive level of the company, and we have a good checks-and-balances system. When there’s a decision to be made, my knee-jerk reaction is to go and grab the other executives on our team and say, "Here’s the opportunity; what do we think? Is there synergy anywhere else in the company?"

We would consider doing something if it benefited multiple business units. But I’m not going to do something just because it benefits the government unit. It’s got to be part of our core strategy.

WT: But reaching out for new technologies also is central to that core strategy?

MORCHE: We see ourselves as a technology leader. We’re not going to sit back and wait to see what happens. [After deploying the first-ever all-MPLS network], the next thing we focused on was 100 gigabits/sec Webs. How fast can we get that out there? How much will that cost? That’s one of the really interesting things about the company. We’re willing to live at the edge, not in a risky business way, but if a technology looks real and we know we’re going to have it, let’s hurry up that move.

WT: How does that tech-forward approach work with Networx? With agencies so far behind on the transition from FTS2001, are many of them interested in new technologies?

MORCHE: It depends on the technology and the customer. The big agencies? Yes, but they’re the minority; smaller agencies just don’t have the experience in new technology.

Voice over IP is a proven technology commercially, but it’s still relatively new in government. So agencies will say, "Yes, we want VOIP, but we want Centrex, too. Yes, we want the benefits for DR/COOP, but we still want that PBX in the basement."

WT: The buzz in the private sector is all about the cloud and video. What are you hearing from government?

MORCHE: We’re doing a lot with [content delivery networks], which includes cloud — actually, we call it a kind of cloud computing. That’s mostly commercially so far, but we’re doing it all over the world. We’re selling the service as either a dedicated or — if your needs are at a predictable level most of the time, with [an occasional ballooning] need — a burstable solution. Because we own and operate our own network, it’s also easy for us to do a burstable solution geographically.

WT: What do you mean by geographically?

MORCHE: You might want to do training on a regional basis or have a project that just includes the New York team. There’s no need to pay for national coverage if all you need is specific regional areas.

You can offer that if you own and operate your own network, which also speaks to the security issue. How else can you guarantee that security?

WT: The Defense Department’s new Cyber Command will oversee DOD cybersecurity efforts but not federal civilian networks. They will remain the responsibility of the Homeland Security Department. How could DHS similarly protect civilian agencies’ networks? Is there a role for Networx providers?

MORCHE: Level 3 regards DOD as an example of how to solve the cybersecurity concerns for the civilian agencies. DOD created a shared and controlled infrastructure for the services to use that gives them great control over what is allowed in or out. If a branch of service were to be compromised, DOD could lock out the infected part, whether that’s a whole branch, a base or some other component, until they can determine the appropriate next steps.

In our opinion, this set of capabilities is what is still missing in the civilian [Trusted Internet Connections] response and one that Level 3 is well positioned to help solve. DHS has the thought leadership for this issue on the civilian side, but it doesn’t have the kind of charter DOD has. It will take someone with significant clout, like the new cybersecurity czar, to orchestrate such a mandate for civilian agencies.

We think the incoming czar could start the process by doing three things:

  • Ask the CIOs of the agencies who have opted to self-provide for TIC to explain their rationales for doing so.
  • Ask DOD to explain the benefits of a shared and controlled network infrastructure.
  • Provide the agency and DOD feedback to the Networx suppliers and ask them to provide proposed solutions.

WT: Level 3 withdrew from consideration to provide TIC services under GSA’s Managed Trusted IP Service contract modification. But you’re still interested in TIC?

Morche: MTIPS is a suitable application layer response, but what about actual network infrastructure? What are we doing to help the civilian agencies to create an infrastructure-based solution? If you don’t know where the network is and you don’t operate or see the network yourself, how are you going to close it down when you need to? How do you ensure resiliency? How do you ensure survivability? How do you ensure all these things when you allow everyone to go and procure whatever network they think they need?

What happens when an infection starts to spread through all these agencies? How would you shut it down? You can’t. There has to be a planned network connectivity between them.

What happened and how do I fix it is the other issue? But you can’t ask that question or answer it unless you have an infrastructure-based solution to start with.

WT: So in the same way that DOD has the Defense Information Systems Network, civilian agencies would have CISN — by leveraging the Federal Aviation Administration’s new network?

MORCHE: That’s one possibility.

FAA’s approach was to ensure that they had two or three ways in and out of every market that they operate in for DR/COOP. If there’s a fiber cut, it’s not always a malicious act. Because human beings run networks, they go down. People touch them, and they go down, and they’re all that way. I’ve worked at several of them, and none of them are impervious.

You go look at, say, Cleveland. Maybe there are three ways in and out of Cleveland, but on three separate networks, [there are] three separate physical sets of assets.

If something or someone brings down the network, there are still two or three other ways in and out.

We really enjoy this conversation. We applaud that kind of thinking. Most folks don’t think that way until something bad happens. And then they say, ‘Why didn’t we do something about it? Why didn’t the carrier tell us?’ Well, first, unless you tell the carrier what the problem is, they can’t solve it for you. You walk in and say, 'This is my situation, and I want this CLIN. And I want two of them, and I want it done by this date.' That’s not engaging somebody in an engineering solution, that’s order-taking. And that’s appropriate sometimes, but not in this segment.

WT: Going back to Networx, why do you think the transition from FTS2001 is so slow?

MORCHE: I don’t think there’s a single reason, but I would contrast it with the transition from [Washington Interagency Telecommunications System] 2001 to WITS 3. Agencies have until January 2010 to make that transition, and GSA has said there will be no crossover contracts. In the first six or seven months, something like more than half the statements of work have been submitted. That’s going well for us.

WT: Is it going to end up being worth the money you spent to be on Networx?

MORCHE: Once the transition is complete, then yes. There are a number of reasons why it’s good to be on the Networx contract in addition to any profit we might make — the prices government is getting on Networx are as good or better than what any other customer is getting. 

And we expect to get more awards as we get closer to the deadline. There are a lot of statements of work in progress, but it is slow, and that does cost. I think I can speak for all of the carriers when I say that we all put a lot of money, a lot of resources into Networx. We staffed up, we set aside capital, we built network on ramps, and it’s all continuing to cost. If [the transition] doesn’t speed up, we may have to let people go, get rid of the set-aside.

WT: GSA has funds to offset agencies’ costs in switching carriers, should they determine that would be profitable. Do you see that making a difference in agencies' willingness to switch carriers?

MORCHE: What we’re seeing is that, even where agencies could get the same services on Enterprise as on Universal, it’s easier for them not to debundle services. The Energy Department debundled services. They’re doing multiple rounds of statements of work. But most agencies aren’t.

This is a mistake, and I say this not just because it cuts us out of a chance to bid but also because it eliminates a chance for agencies to get a better price. If they stay with Universal only, they get three companies bidding for their business. If they include Enterprise, they get five companies.

But there’s nothing to say the agencies have to do it. There needs to be someone who can direct them to do it or provide further incentives in some other way.

There’s nothing to say we’d get any business from it, although it most likely would save agencies and taxpayers money. But that’s not enough of an incentive or we’d be further along with the transition, which also will save a lot of money.