Cisco, 12/27/99

By any measure, Cisco dominates the market for enterprise data network equipment. Nobody argues that fact. But the landscape is changing, with talk of adding voice to data nets and an emphasis on professional services to help users make sense of it all. At the same time, Cisco is trying to mount a credible assault on the carrier equipment market, which many prognosticators, not to mention Cisco itself, consider essential for the company's growth.

So in the struggle to retain its spot atop the network equipment market, it's really Cisco vs. everybody else, most notably Nortel Networks, Lucent Technologies and 3Com. But coming up on the outside are a slew of well-heeled start-ups fresh from wildly successful IPOs. This bunch includes Foundry Networks, Juniper Networks, Redback Networks and Sycamore Networks.

Can any of these competitors take Cisco down and claim the top spot? Not likely, but that doesn't mean these competitors won't enjoy success of their own. Cisco won't keep its position unless it continues to stay focused on customer requirements. However, that is the single thing the company does better than any other.

Cisco earned its commanding lead in the enterprise data market by building a loyal customer base. Many customers are so entrenched they won't go with a product from a competing vendor even if it can outperform a Cisco box.

"I'm less interested in another company's product that's slightly faster in packets per second," says Russ Davis, assistant director of network services for Ernst & Young in Lindhurst, N.J. Ernst & Young uses myriad Cisco products, including routers, remote access equipment, switches, load balancers and firewalls.

Introducing another vendor's gear at this point would only present interoperability problems and require his staff to learn how to manage another vendor's equipment. Davis, like many users, wants no part of it.

Kathryn Korostoff, president of Sage Research, a market research firm in Natick, Mass., recently conducted focus groups and asked users to match automobile logos to different network equipment vendors. "The majority say Cisco is the Mercedes - it's the brand you want to be seen driving," she says.

Even charges that Cisco products are overpriced don't sway users. "Price is important, but it's not number one," Davis says. "You really have to look at the total cost of ownership," which means factoring in training time, time to repair problems and vendor support. When you consider the hidden costs associated with other vendors, Cisco comes out looking pretty good.

But how long can Cisco continue realizing 40% to 50% growth rates in enterprise sales? Thomas Nolle, president of the technology assessment firm CIMI Corp. in Voorhees, N.J., thinks the enterprise market will peter out as more users flock to public data network offerings, just as voice traffic fled private networks in favor of software-defined network services a decade ago.

"No steps Cisco could take to sustain the enterprise market for its products would be able to offset the trend away from enterprise networking," Nolle says. "By 2003, unless Cisco has an incredibly effective public data position with the incumbent facility carriers, its growth will be problematic."

Cisco's best growth opportunity is SNA. "SNA is the largest source of corporate data traffic in the wide area, much larger than IP," Nolle says. "If Cisco can get SNA buyers to convert their SNA networks to IP, it would gain a significant revenue advantage." Cisco has taken the appropriate steps toward making that happen, he says, noting that it bought the bulk of IBM's Network Hardware Division.

Peter Alexander, vice president of enterprise marketing at Cisco, dismisses concerns about growth rates for the company's enterprise data business. "If you look at steady state demand growth, just what is expected of data growth today, there is a large opportunity for us and our competitors on the order of 50% per year," he says.

Sage Research figures seem to bear that out. Based on results of a survey of 600 U.S. companies, Sage says only 14% have pure IP networks today and only 36% will by year-end 2001. It stands to reason that, as companies migrate to all-IP nets, they will in the process buy lots of switches and routers. Will voice and IP mix?Even as Alexander spouts figures to demonstrate growth potential in enterprise data, he notes that Cisco firmly believes migration of voice and video to IP networks is inevitable and key to fueling future growth.

Most other observers, be they users, vendors or analysts, agree that convergence is coming. But not Nolle.

"Voice over IP is just a dumb idea. It's been a dumb idea from the start. It's never been anything other than a dumb idea and it's a dumb idea today," Nolle says. He has long argued that ATM will be the transport of choice among carriers while IP will merely ride on top, transporting data as it was intended to, but not voice, for which Nolle says IP is ill-suited . He cites SBC Communication's November announcement of a $6 billion, "100% ATM" network upgrade as evidence that the voice-over-IP movement has no legs.

Alexander says that's hogwash. And he's got the figures to prove it: In the 18 months between May 1998 and November 1999, Cisco shipped more than one million packet telephony ports, mostly in gateways that connect PBXs to data networks. (If a gateway supports a T-1 connection, that counts as 24 ports.) And, while acknowledging that not many users have IP telephony wares in production, Alexander says more than 600 customers are trialing the company's Architecture for Voice, Video and Integrated Data (AVVID) products.

Columbia Energy Group, in Columbus, Ohio, is one of the companies testing Cisco's IP phones - as well as its quality-of-service (QoS) protocols. Cisco is the sole provider of switches and routers to Columbia but, even so, the company has already put Lucent IP telephony products through their paces, says George Yeager, manager of architecture and design in the Enterprise Multimedia Communications Department. "We're sold on voice over IP, but not necessarily on a particular approach," Yeager says.

His initial impression is that Cisco's products are more fully evolved than Lucent's. While emphasizing that Lucent's products were beta versions, he says they lacked support for Dynamic Host Configuration Protocol in the phones and had compression problems.

"Almost every customer we're talking to is investigating IP telephony," says Ray Keneipp, director of internetworking for The Burton Group in Charlottesville, Va. "It's going to be very big in the enterprise, but not for at least two to three years."

The key drivers behind IP telephony are economics and applications, Alexander says.

In a typical enterprise, two-thirds of the combined voice and data budget is spent on voice, but that buys each user only a 64K-bit/sec link. By contrast, the one-third spent on data often buys a 100M-bit/sec connection. "Two-thirds the cost for one-hundredth of capacity is a glaring economic opportunity," he says.

As for applications, the ones cited most often have to do with Web-enabled call centers and unified messaging, neither of which actually requires IP telephony. But the theory is that convergence will make application integration more seamless.

Another potentially big and largely untapped revenue stream for Cisco and its competitors is in professional services, meaning providing expertise to help users get the most of out e-commerce efforts, for example.

In mid-1998, Cisco founded its Internet Business Solutions group to do just that. This past August, it made a $1 billion investment in the consulting firm KPMG LLP with the same end in mind. And in November, the company announced Cisco Content Networking, a set of enhancements intended to add intelligence to Cisco hardware and software such that they can better handle differing applications, including e-commerce.

But service is an area where observers think competitors Lucent and Nortel have an edge. Lucent especially, because it just acquired the big consulting and management services firm International Network Services. Nortel likewise acquired Clarify, a maker of customer relationship management software that has a professional services organization of about 500 people.

New kids with big switches

The carrier market is everything

Netting it out