by The Editors of Business Communications Review

MPLS IS MANDATORY FOR NETWORKED VOIP PER GMACCH
“Please—don’t try this on your old networks,” said keynoter Niraj Patel, executive VP and CIO of GMAC Commercial Holding Corp. Patel operates an MPLS network from service provider Masergy, and uses four classes of service—voice, video and two data classes—to link 109 locations worldwide. “You want your users to have good quality of service, so they will enjoy using all the new features.”

About 500 of GMACCH’s 3,600 users have IP telephones, but all inter-site traffic is IP. Patel said he has saved plenty of money by converging his huge network, driving his expenses down to less than half the financial industry average of 6 percent. Other savings stats:

• An eight-fold increase in backbone capacity at no extra cost with MPLS compared with frame relay.
• International VOIP calls saving $163,000 per year.
• IP videoconferencing (H.323) saving $100,000 versus ISDN.

These are all hard savings because, as Patel put it, “quantifying soft savings is hard.” Nonetheless, he avers, it’s the business value of convergence that sold GMACCH. “There is business value in rolling out all these great tools—or we wouldn’t even be looking at them.”

Patel specifically cited unified messaging, speech-enabled calling, videoconferencing and the flexibility to work from anywhere without sacrificing any features. Convergence proved in easily, said Patel, while he could not make the business case for CRM.

Will you get the same savings? You could—if you are as big as GMACCH, and as willing to partner with vendors on “the emerging side” as Patel has been. My guess is those partnerships are the key source of his incredible cost savings.
—Sandra Borthick, senior editor, Business Communications Review

“JUST BECAUSE YOU CAN DOESN’T MEAN YOU SHOULD”
Phil Edholm, Nortel CTO, got a big laugh with that tag line in the VoiceCon 2004 Great Debate. Edholm was comparing the advisability of a 100-percent IP telephony deployment to Janet Jackson’s Super Bowl flash. Edholm’s main argument was that business value had to drive IP telephony deployment, and he offered a formula for calculating the viability of the decision to deploy:

Business Benefit
Risk x Cost

This formula rightly relates costs and risks, he said, and allows businesses to weight risk according to their tolerance. Edholm said Nortel studies have shown that IP telephony implementers are recovering the costs of Web agents, telework, and consolidating WAN bandwidth in about six months, with other advantages taking from one to two years to recover costs.

Edholm’s debate opponent, Cisco voice technology VP and director, Michael Frendo, argued that IP telephony is fully ready for 100 percent deployments, having overcome five factors he called “myths” that had been holding it back. Contrary to these myths, Frendo said:

• Key standards are in place.
• IP telephony has proven ROI (return on investment) and reduced TCO (total cost of ownership.
• New applications are emerging. .
• IP networks can be made secure and reliable. .
• Migration is possible—no flash cuts are necessary.

The ensuing Q&A was lively and surprising at times. For example, Nortel’s Edholm panned IP phones with their tiny displays, extolling the advantages of his PC as a softphone, while Cisco’s Frendo stressed the advantages of purpose-built IP phones, pointing out that not everyone travels with a laptop, and 50 percent of all office desktops still don’t even have PCs.

Both retreated to their more traditional vantage points, however, on a slightly more gruesome point, with Edholm saying he wouldn’t want anything but a fully reliable traditional phone in the hospital operating room if he were going under the knife, while Frendo said a POTS connection wouldn’t be good enough for him if his surgeon were being remotely advised by a specialist—he’d want imaging and other technologies that can be supported by IP.

What emerged from the dialogue was a shared view of IP telephony more as a change agent and enabler of transformation than as an either-or, all-or-nothing technical proposition.
—Sandy Borthick

MORE ON THE GREAT DEBATE
One reason the Great Debate is a VoiceCon highlight is because it inspires the combatants to heights of rhetoric and metaphor that might seem grandiose in your typical breakout session. Maybe it’s the spotlights, which seem brighter than ever this year, both literally and figuratively. Or maybe it’s the stakes, which get higher each year for the vendors, as more users buy and deploy VOIP.

Among the thrusts and parries in today’s debate:

• Michael Frendo built his introductory remarks around the metaphor of the transition from horse-drawn to motor-driven vehicles a century ago. He likened today’s debate topic, “How Much IP Telephony Is Enough?” to a hypothetical debate 100 years ago: “How Much Internal Combustion Is Enough?” In the new networking environment, Frendo said, having a “hybrid” IP-TDM network is akin to someone saying, in the early 20th century, “I’m going to have one of those horseless carriages, but I’m going to have a horse running behind it.”

Attempting to turn the metaphor against Frendo, Phil Edholm proclaimed that the “field is littered with dead horses behind you”—by which he meant the high-profile stories about important users like Merrill Lynch pulling out Cisco IP-PBXs in favor of the “hybrid” solution that Nortel, Avaya and other legacy vendors advocate.

Frendo shot back that Cisco’s competitors, in “desperation,” were “willing to give away a solution”—though he didn’t say which specific customers might have been on the receiving end of those vendors’ generosity. Nor was he specific when he said the customers who dropped Cisco IP-PBXs did so because they were disappointed when mere “like for like” replacements of TDM systems didn’t live up to the promise of cutting-edge new VOIP applications.

Not to be deterred, Edholm replied, “You could say they took it out because they didn’t have an application and they were silly”—though in fact Frendo never did use the word “silly”—“I’m not sure,” Edholm said, “[that] the people at Merrill Lynch would appreciate being called silly.”

• Frendo also called upon more recent, and telecom-specific history, by bringing up the last big migration in PBXs, from analog to digital starting in the 1970s. Frendo noted that in that migration, one equipment vendor without much market share in the legacy technology began urging companies to abandon the old technology and embrace the new because of the vast improvements in functionality that the new technology promised. That company, Frendo barely needed to say, was Nortel.

• Finally, Frendo slipped up a few times and referred to TDM PBXs as “mainframes.” At least I think it was a slip-up.
—Eric Krapf, Editor, Business Communications Review

VOICECON SPECIAL FOCUS: TWO SIP PANELS
Back-to-back SIP panels this afternoon gave attendees a solid grounding in what SIP is and how it operates, as well as offering several vendors’ opinions on how it could change the enterprise voice business model. In the first panel, “Is SIP Ready for Prime Time?” Lance Lockhart (MCI senior VOIP engineering architect), joined David Sokolic (Microsoft lead technical strategies manager) and Cary FitzGerald (Cisco senior director of engineering), to address the topic. The second panel, “Does SIP Break the Model for Enterprise Telephony?” featured Jack Jachner (senior director, Alcatel), Jim Davies (MCTO, Mitel Networks) and Martin Steinmann (senior marketing VP, Pingtel)

The first panel’s presenters stressed that SIP is not a product, not a feature set, not a buzzword—it’s a signaling and control protocol. As such, it enables new forms of control over communications, and new models for delivering functionality. But it doesn’t tell you what to do with them. “It’s a methods protocol, not a functions protocol,” said Cisco’s FitzGerald, contrasting SIP to older voice protocols that actually operate features and functions.

High praise goes to Henry Sinnreich and Alan Johnston of MCI, who put together a terrific—but lengthy—set of slides on the current state of SIP development, which MCI’s Lockhart delivered. Although he managed the 32 slides in just over 15 minutes (!), attendees were urged to work through them more carefully again.

In the second panel, the presenters basically agreed that SIP evolves the enterprise model, but doesn’t break it. Pingtel’s Steinmann asserted that SIP plus the open source model (which Pingtel has recently embraced) could make a bigger difference, if enterprises adopt it. Mitel’s Jim Davies said his customers are most interested in what he called the “pre-built” model, with which “strong integrators create families of solutions by combining best of breed components.”

Other highlights:

• Asked when Cisco would deliver a SIP phone in addition to or to replace its proprietary Skinny Call Control Protocol (SCCP) phone, FitzGerald said that any Cisco IP phone can download the Call Manager 4.0 client today, which has SIP.
• Alcatel’s Jachner said the company was not really worried about softphones displacing handset revenues, or Microsoft displacing IP-PBX software. If need be, he said, Alcatel will license its software, too.
• Microsoft’s Sokolic urged attendees to ask for SIP from their PBX vendors, because so far there has been more “lip service” than real SIP implementation. He said Microsoft’s next release of Windows, code-named Longhorn, will support the SIP RFC 3261, and the plan is to enable third-party call control from the PC to whatever phone is on the desk.

—Eric Krapf

Posted in Management, Architecture |

You can follow any responses to this entry through the RSS 2.0 feed. Both comments and pings are currently closed.