The information-centered network - MCI CIO Lance Boxer says company plans new integration services - Company Business and Marketing

Communications News, Sept, 1997 by Ripley Hotch

Lance Boxer's position as CIO of a major carrier puts him at the center of network development in data and voice communications. He creates networks, manages them, and directs the U.S. arm of MCI's own worldwide network. His organization numbers nearly 8,000 employees, and is involved in everything from network integration to creating software to managing MCI's billing transactions.

He's also aware of the pressures on designers and managers of modern communications networks, who have to meet the constant demands for greater bandwidth, ease of use, customer service, and lower costs. And he has an unusual vision of where data will be kept in tomorrow's networks.

The decentralized nature of current network management responsibility has spurred him to come up with a new term for all those positions that have to do with creating and maintaining communications in a large organization: "We call them `enterprise specialists' now, and within that there are subtitles. So rather than call them network managers, we have enterprise specialists that deal in security or Web hosting or whatever." Boxer thinks that is better than coming up with "more goofy titles."

Boxer came up on the network engineering side, so he feels comfortable with MCI's move into systems integration. It is, he says, a natural role for the large carriers, and a natural outgrowth of dealing with all kinds of enterprises of all sizes.

Mid-sized companies, for example, cannot afford to hire engineers and planners, so they look to solve business problems, and "they come to companies like MCI to provide an end-to-end solution. They're not sure whether you need ATM or frame relay or IP switching. They put a problem in front of us -- to solve bandwidth or response times or they need global access -- and we really are obligated at that point to become the integrator."

Becoming an integrator is a way for the carriers to hold and increase their market share: "There's very strong evidence right now that the more product you sell, the longer people stay with you," Boxer says. "And with transmission costs becoming, unfortunately (from our point of view) a fairly low-margin product, you need to think of all the corollary services that you put behind them.

"If you become an integrator, you take the issue of transmission almost off the table, because now you're really dealing with a lot of value-added services that have the right margins. That's how we've been able to rationalize buying MCI Systemhouse and getting into the integration business." And Boxer characterizes it as "a huge business."

It may be large in part because of the cost of making a decision that can't be reversed -- and finding out that it was the wrong decision. Asked his take on ATM, Boxer says that MCI has definitely seen a slowdown. He thinks there are several reasons for the cooling of what looked to be the hottest technology of the decade.

"We're using ATM today more for bandwidth aggregation and I would call it the high-end ATM -- this is like the 10 Gbps hub," Boxer says. "It's typically for the consolidation of our data services. We're taking SMDS, we're taking frame, and aggregating it up to ATM cells. I think at the carrier level ATM is becoming a reasonable tool to help us manage bandwidth better."

Below that level, however, Gigabit Ethernet is making serious inroads into what looked like the inevitable march of ATM to the enterprise desktop: "Gigabit Ethernets are dollar-for-dollar a very effective way to move a lot of data around. And people understand it, it doesn't require them to hire a whole new skill set. They're not upgrading the infrastructure" as they are for ATM.

What also is cutting into the acceptance of ATM is the success of frame relay, Boxer says, "where we're able to put fun T1s and we're looking at DS3 interfaces now on frame."

So the buyers of ATM are either carriers "buying them to fork-lift-upgrade their backgrounds," or "a Fortune 500 company that wants to do a lot of campus connectivity."

Looking at this kind of technology, Boxer says, should be done with care, because once you make the transition, there's no going back. "When you dance with a gorilla, you sort of dance when the gorilla wants to dance," he says.

The lesson in ATM is that people need to be careful, to not get turned on by all of the whiz-bang technology that keeps getting thrown in front of them," Boxer says.

"One of the things that we keep sort of disciplining ourselves to here is that while we have a very competent team of technologists, over time we have tinkered a lot less with technology. There are a lot of options out there, and I think it's all risk management.

"I think that you have to really think about what you're willing to risk as a business to save a few bucks on certain capacity. And ATM is a great example. You think you save money on transmission, but you need new skills, new training, new infrastructure."

Boxer and MCI have gotten excited by what began as a simple branding effort: the Information Centered Network. The idea is to put as much of the intelligence in the network as possible, rather than on the customer premises.