Network Computingwww.networkcomputing.com

Many within the industry think vendors are just throwing in the towel, giving up the big fight against the Cisco Systems marketing machine. Others think this is all a big hand-waving exercise; indeed, some of these vendors wave so high and hard they're about to take flight. But for all that hand waving, a common thread ran through conversations with managers at all three vendors: shareholder value. Investment analysts are driving these vendors' strategies, taking precedence over short-term profit-and-loss statements and customer satisfaction with product offerings.

And yet, investors' reaction has been anything but enthusiastic. Of the three, only Lucent's stock price reacted positively to its recent announcement, according to Abid Mansoor, an investment analyst from Zacks Investment Research. Mansoor speculates that even Lucent's upticks reflected the fact that the stock was underperforming prior to the announcement. Still, with the completion of the spin-off, Mansoor expects the old Lucent business to show greater margins and profitability, thus improving stock performance.

These vendors might eventually win their battles at the expense of customer trust. Certainly the vendors aren't helping their enterprise image by spinning off separate companies. Why shouldn't customers drop these companies' products and head to Cisco's more solid ground? What possible benefit could these announcements and spin-offs have for you, the enterprise-computing customer? That's exactly what we asked the executives of Cabletron, Lucent and 3Com.

3Combobulation

On March 20, 3Com announced it would be "narrowing its focus" toward markets, products and technologies in which it could hold a leadership position. The company is departing from several key spaces in which it previously held significant market share. Instead, 3Com is targeting consumers, small and midsize businesses, and network service providers, with IP telephony products, wireless and broadband access devices, and Web-based solutions.

We asked 3Com's president and COO, Bruce Claflin, why the company is moving out of the enterprise market space. He cited two key reasons. First, 3Com's share in the enterprise market is very low, especially when compared with "some large incumbent vendors"--which is Cisco, in case you've been living in a box the past few years. We agree that's probably a good decision. The company's CoreBuilder is far from a best-of-breed product, for example. It made a big splash when it appeared on the market, but 3Com failed to deliver on long-term promises for feature and functionality upgrades.

The second reason, Claflin said, is the poor return on investment. The high-end LAN and WAN switching business sapped 60 percent of 3Com's development resources while providing only 20 percent of its revenue and none of its profits. That's not a bad reason to get out of the market either, but it leaves many customers stuck with a large investment in 3Com infrastructure.

Gary Habermann, Widener University's director of technical resources, said he's very disappointed with the company's actions. And while he appreciates 3Com's Extreme Networks migration program, he has doubts about the viability of a 400-person company delivering products for his network core.

3Com hasn't totally abandoned its customers but has put a significant time and budgetary crunch on large IT shops, forcing them to cancel and postpone other rollouts in order to re-evaluate their 3Com infrastructures. For some customers considering jumping ship anyway, this is the kick in the pants they needed. 3Com customers have until June 30 to get their final orders in for the discontinued lines. 3Com will support the hardware platforms for up to five years, until contracts are fulfilled by attrition. Likewise, software development for the hardware platforms will continue until June 30, 2001, to ensure that outstanding customer issues are properly put to bed before the development teams move on to other projects. 3Com maintains most of its engineering expertise and all the intellectual property for the CoreBuilder and NetBuilder products, as this technology is being recycled into some of its midrange enterprise offerings.

Competing vendors are champing at the bit to seize a chunk of 3Com's market share--18.3 percent of all ports shipped globally, representing 10.8 percent of global market revenue for Ethernet switches, according to fourth-quarter 1999 statistics from the Dell'Oro Group, a market research firm. Cabletron, Foundry Networks and Lucent were quick to offer to buy your old 3Com gear and give some juicy discounts on their own stuff. And 3Com has worked its own deal with Extreme Networks, sending 200 sales and marketing staff to Extreme, and cutting a good migration deal to 3Com customers willing to take the purple Extreme road.

3Com will exit other lines, too. Its analog modem business is being sold off wholesale to a new company being formed by Accton and NatSteel. 3Com will have a 20 percent minority investment in the new venture. This sell-off includes all the U.S. Robotics products.

Also on the way out is the high-end Transcend network-management platform. 3Com will retain the Transcend Supervisor element manager software, which is merely a single switch management agent for Wintel-based small-to-midsize shops. All the high-end offerings (node management, network design, planning, discovery, VLAN management--essentially all the enterprise-caliber tools) are being taken off the market. The RMON (Remote Monitoring) line, which includes Traffix and LAN Sentry, are staying, however, and could be good sell-off targets, as so few RMON vendors are left.

The new 3Com will focus on four key markets: Web-based solutions, LAN switching/Internet telephony, wireless access products and broadband devices. At least, that's the company's story. Web-based solutions fall under a partnership with a company called March First, which will be the consulting services arm of 3Com's new corporate model.

In the enterprise space, 3Com will continue to sell its NICs, SuperStack hubs and switches, OfficeConnect products, and other products that don't fall under the CoreBuilder umbrella. 3Com has no intention of exiting the LAN switching business. However, if aggressive vendors capitalize on the fear, uncertainty and doubt 3Com has created--and who wouldn't?--the company is likely to lose significant market share.

3Com's drastic move shocked the industry. It surprised investors, too; however, Zacks analyst Mansoor believes 3Com's poor market response was clouded by the Palm Computing spin-off, which happened almost simultaneously with the enterprise divestiture. Arbitrage traders had a significant impact on 3Com's stock following the Palm spin-off.

But 3Com is really putting all its cards on the table. It has told the world what it's going to do, and rather than dragging the CoreBuilder products and customers through a long, slow death, the company is attempting to come clean now.

Moving forward, the company sees new product lines arriving on the market much faster and with a lower price tag. Claflin speculated that because 3Com is no longer supporting non-revenue-generating products, it will be able to offer more competitive pricing in its markets. The R&D budget will shrink by about 1 percent to 2 percent, but the overall R&D effort will focus on a smaller range of products. Thus, the overall effect is a significant improvement for new products.

Where will 3Com be in two years? The company views itself as showing expansion in new markets, with total company growth in excess of 20 percent and operating returns of more than 14 percent. "We expect ourselves to be absolute leaders in these markets," Claflin said. 3Com is trying to become a broadband company. With 3Com's leadership position in the cable modem, DSL and CDMA wireless markets, we understand why. As for the enterprise market, 3Com will continue to try to penetrate the wiring closet with high-function, high-density edge switching. But frankly, its real strength will show in the SOHO (small office/home office) and small business markets, where the word "Cisco" doesn't carry the brand recognition it does in the enterprise.

The Cabletron Schism

On Feb. 10, Cabletron announced it would break up its core enterprise into four separate businesses, and the organization formerly known as Cabletron would be a holding company for all four businesses: Enterasys Networks, Riverstone Networks, Aprisma Management Technologies and Global Network Technology Services.

Fiallo said Cabletron's assets were undervalued, and the move would unlock the company's value for its shareholders. Cabletron's stock sunk as low as $7 in 1999 but was recovering when the announcement was made. With the announcement comes the opportunity for four IPOs (initial public offerings) on the stock market. Of the three companies covered in this article, only Cabletron has adopted this strategy. Will it work? The jury is out, but at press time the stock appears to be sinking again.

The initial plan is to spin two of these companies into separate start-ups by year end. The other two would follow in the first half of 2001. Cabletron has not disclosed which companies will go first. However, it is interesting to reflect on mid-April's market fluctuations and wonder just how successful this plan will be. Cabletron seems to think its new vision and management team will belt a home run on the stock market. Still, the company will hold out for favorable market conditions to deliver on its four IPOs.

The Cabletron announcement generated a lot of confusion in the market, which resulted in a negative response on the stock market, according to Mansoor. Specifically, arbitragers may have played some role in the poor performance, and more acutely, analysts had a difficult time revising their estimates because of confusion over the new Cabletron's growth strategy going forward. This speaks volumes for what the investment community thinks of Cabletron's split up; the unanswered question is, what makes the new companies bigger and better than the sum of their parts?

Industry insiders will tell you the Cabletron name carries with it love-hate associations. Cabletron customers are often on the love side of the fence. But many people cite Cabletron for its aggressive, almost strong-arm tactics on the sales battlefield. Those folks tend to hate the Cabletron name and everything associated with it.

Once the new companies become established, will customers and investors forget that these firms are really just parts of the old Cabletron? Will they be able to attract talented management to guide these new enterprises? Will they be more effective in executing on strategic initiatives, a long perceived weakness of Cabletron.

Splitting up into four companies does let each company focus on a specific set of products and technologies. If a company uses that new focus to drive better products and better returns, then the IPOs may indeed have some merit. If the changes at Cabletron improve the morale, work ethic and attractiveness as a company to work for, the companies may see significant profit. And, as a customer, you may see a better product. But there are a lot of ifs in there.

Enterasys received the lion's share of Cabletron's networking assets. Enterasys will sell enterprise-class networking solutions and resell Spectrum network management from the Aprisma group. Enterasys will sell most of Cabletron's existing product offerings--specifically, the SmartSwitch 2000, 6000 and 9000 modular switches; SmartSwitch Router 2000, 2100, 8000 and 8600; SmartStack and the new Matrix E7 switches; and the RoamAbout wireless access products. All VPN and VoIP (voice over IP) products also fall into Enterasys' domain.

Enterasys also inherited Cabletron's entire service and customer support division. For now, the company is leasing first- and second-level support services to its sibling companies, until these companies either build out or outsource their own support structures. Enterasys also inherited roughly 500 engineers from the core groups of the product lines it is keeping.

Riverstone will sell the Internet Appliance line of load-balancers and resell the SmartSwitch router to carrier-class network providers. Aprisma will sell Cabletron's Spectrum management software suite. Global Network Technology Services is purely a customer consulting and service agency, much like the Digital Networks services group of old.

The list of Cabletron products that will go away is short but will strongly affect some of the company's inherited customer base. Specifically, the Digital Networks group is being sold to a third party in the next 90 days or so. FlowPoint (DSL technologies) was sold to Efficient Networks. The NetVantage line of stackables (certain SmartStack models) is also being sold off. Five-mirror (an ISDN SNA remote backup solution) and Network Express/Avail (RAS modem banks and NAS) are being discontinued.

The new Cabletron companies will focus marketing, sales and development on six technology areas: hardware switching, hardware routing, VPN and security, wireless and RF, next-generation management, and voice over IP/convergence technologies. This puts the Cabletron companies head to head with Cisco in the enterprise space, 3Com in the e-business space and a slew of vendors in the convergence space. There will be room for growth, but with other companies--such as 3Com--focusing more tightly on these markets, it will be a tough battle and one that Cabletron has not historically dominated.

Cabletron says the new management team is readier and more willing to set the pace and make the changes necessary to make Cabletron a faster-moving, more focused company. The management speaks about the four companies as an ecosystem; we think there is a lot of interdependency among these companies. We asked how the new companies will operate independently and were told that for the near term the corporations would do a lot of sharing. A software "bank" has been set up that lets the companies share new software code between groups.

Cabletron had spent 12 percent to 14 percent of revenue on R&D. After the schism, Cabletron executives suggested their R&D budget as a percentage of revenue would be slightly higher than that.

Cabletron is trying to win new midrange-enterprise accounts. Enterasys managers believe only Cabletron and Cisco are positioned to meet the enterprise customer's needs now that 3Com has left that market. Nortel Networks Corp. and Lucent aren't key players in the enterprise infrastructure market, they say. However, Nortel says it has 9.6 percent market share by revenue and 10.9 percent market share in worldwide Ethernet port shipments, according to fourth-quarter 1999 reports from the Dell'Oro Group. Cabletron is aggressively targeting 3Com customers looking to migrate. But if you were a 3Com customer who's just been burned, would you go with Cabletron?

The Cabletron schism is complete, but the IPOs and distribution of old Cabletron stock are not. Poor market response could drag out the Cabletron split for longer than the company would like, and no dates are set yet. If the stock market takes another big dive, it could spell bad news for Cabletron's child companies. Cabletron doesn't see anything stopping the eventual stock offerings, but if market conditions continue to sour, customers might be left wondering where to turn. For now, Cabletron customers will be turning to Enterasys for their enterprise needs, or perhaps they'll start looking elsewhere.

Lucent Behavior

In November, Lucent Technologies underwent a massive internal restructuring to help better align its business units with consumer goals. For the end user, that means a leaner, meaner Lucent with more purchasing power, quicker time to market and more strategic alliances with e-business organizations looking to deliver custom solutions to end customers.

On March 1, Lucent announced its intention to spin off the Lucent enterprise-computing group. This new company, not yet formally named, is being referred to as the new Enterprise Networks Group of Lucent Technologies. Lucent has set a target date of Sept. 30 to complete the split.

Of all the companies doing the corporate shuffle, the changes at Lucent will probably be the most transparent to the customer. After all, the company restructured in November, and this spin-off is the culmination of changes that took place internally months ago.

We asked Lucent why the company chose to take the split-up route. Karyn Mashima, vice president of strategy and CTO of the group, said the new group will be able to focus on product development without worrying about its impact on the larger Lucent entity. Before the breakup, Lucent had to be very careful about how many dollars it allocated to the enterprise group, based on the impact it would have on Lucent's service-provider business. The separation will allow the new group to move more quickly on partnerships and acquisitions--two things that had previously been entangled in red tape. Lucent management also felt that market conditions were favorable for spinning off a new division that has the potential for higher earnings than the core company. Lucent is not spinning off the unit in an IPO, however; stock will be distributed to current stockholders when the separation finally takes place.

The biggest downfall for Lucent and its customers will be if the company doesn't follow through in managing its customer relations. Although the spin-off is still five months away, Lucent hasn't been overly communicative with its customers, according to some who spoke with us.

Unlike 3Com and Cabletron, which have significant stakes in the enterprise LAN market, Lucent's announcement will have a more significant impact on enterprise telecom managers. Lucent's Enterprise Networks Group comprises a wide range of product families: the Definity voice products line, Lucent's extensive CRM (customer relationship management) software suite, its entire line of Europe/Middle East/

Africa offerings (voice and CRM products sold specifically and only in those markets), the messaging solutions group (including Audix and Octel voice messaging), small-business communications, IP Exchange, conferencing and collaboration tools, the GuestWorks hospitality solution, and the Cajun campus family of LAN and ATM switches. Also part of the new group is the Systimax cabling line. Lucent has approximately 151,000 employees. About 34,000 employees will join the Enterprise Networks Group. In terms of revenue (for fiscal year 1999, ending in September 1999), Lucent is a $38 billion company, and the Enterprise Networks Group represents $8 billion (21 percent) of that.

In addition, the Enterprise Networks Group will license a number of products from the "old" Lucent, including Pipeline access routers and the SuperPipe multiservice access router and Lucent's Access Point Internet access routers. Both companies will split engineering resources from the RealNet Rules development team and will sell separate products to address their respective marketplaces. The Xedia VPN device will also remain in the old Lucent's hands, while the Enterprise Networks Group will resell and license the technology from the old Lucent. Finally, the QIP address-management software will remain part of the old Lucent. We find that surprising, considering the enterprise nature of that software. However, the Enterprise Networks Group will license that software and resell it through its own channels.

Easing Support Concerns

In terms of R&D and future resources for the company, more than 3,000 engineers have been moved to the new group, which is expected to be self-sufficient in terms of R&D and financial resources. Customers have expressed the most concern about whether Lucent will continue to support call-center applications, CRM, messaging and convergence technologies. Lucent has responded that the same account representatives and service groups will support existing accounts, since most of that staff is coming along with the product lines.

And what of the old Lucent? It will focus primarily on the service provider and microelectronics space. There, customer relationships have been developed mostly through exclusive, one-to-one sales and support teams. Those resources will stay with the old Lucent. Lucent will maintain its own service and support group for NetworkCare Services (comprised largely of resources from the INS acquisition). The new company plans to use the consulting and services arm of Lucent extensively in the future.

Bruce Boardman contributed to this story.

Send your comments on this article to Joel Conover at jconover@nwc.com.

Executive Summary Infrastructure Trends Cabletron Systems', Lucent Technologies' and 3Com Corp.'s decisions to reorganize, refocus and spin off their businesses have sent shock waves through their customers' lives. 3Com is dropping dozens of WAN- and enterprise-focused products--notably, its CoreBuilder, PathBuilder, NetBuilder, AccessBuilder and Transcend lines. Instead, the company will target small and midsized businesses and network service providers, with IP telephony, wireless and broadband access devices, and Web-based solutions. Cabletron's changes involve the creation of four new companies: Enterasys Networks, Riverstone Networks, Aprisma Management Technologies and Global Network Technology Services. Enterasys, which appears to be the largest, will sell the company's enterprise-class products. Lucent, meanwhile, is putting the finishing touches on a massive restructuring that began in November. The company is spinning off a good chunk of its telecom business into a new business that has yet to be named but is being called the Enterprise Networks Group for now. These spin-offs have one thing in common. Managers from all three companies said they are attempting to unlock the value of their organizations' stock. Cabletron intends to do this by offering each of the four companies as IPOs, 3Com is shedding unprofitable lines and Lucent is trying to free up dollars for its enterprise business. Prudent, perhaps. But some customers complain of being left in the dark as products disappear and sales relationships get muddled.