The Lippis Report

During our Dec 4th ?¢‚Ǩ?ìIP Telephony Economics: Does It Add Up To Savings and Productivity?" webinar, it became obviously clear that IP Telephony is moving into prime time. There are a few data points that bring me to that conclusion. First, the sheer number of attendees (as mentioned above) indicates wide spread interest. Second is the large number of questions, over 100, and the type of questions asked during the webinar. Third, and the strongest indicator of all, is the excellent large scale deployment experiences the panel members shared with the audience. In fact, every speaker presented over an IP phone!

But to gauge the market´s acceptance level, it is the questions that provide the most insight. They indicate that Chief Network Architects (CNAs) are thinking through developing their implementation plans. As deployments follow education, one thing is very clear: there is a large demand for IP Telephony information. Only when that demand is satisfied will CNAs move toward implementation.

Here is a sampling of the type of questions asked and education needed to move IP Telephony into prime time.

Justin Stevens, CNA International Monetary Fund or IMF: ?¢‚Ǩ?ìWere phone features a high factor in your vendor selection either in parallel to your existing pbx or to your greenfield installations? We are looking at this also and are finding phone features vary among the different vendor offerings. Was this an issue with your organization?"

Brian Walker, CNA, SunTrust: ?¢‚Ǩ?ìHow were the savings achieved, specifics please."

Leon Schuman, CNA, Sedgwick CMS: ?¢‚Ǩ?ìWhat monitoring and diagnostic tools do you use?"

There were many questions on cost justification through PBX or Centrex displacement, if and how to factor in IP Telephony applications into the cost justification, quality of service requirements, feature comparison between vendors, meeting e911 requirements, security concerns, plus probing on IP Telephony reliability etc. We were able to go through a lot of these questions during the webinar. Rather than re-hash them here, you can view the webinar on demand at www.en2004.com. What follows are a few insights gained from the session.

IP Telephony: Does Cost Justify!

If you´re not exploring how to take advantage of IP Telephony with its cost savings and corporate productivity increases, you´re delinquent in your duties as a chief network architect. A few examples: Lehman is saving $1M/year alone in the cost reduction of moves, adds and changes (MACs). MAC is a huge and relatively easy cost justification metric for IP telephony as is the cost avoidance of running IP voice and data over a single RJ45 Ethernet connection. A client of mine estimates their MAC cost at some $6.5M/yr, most of which can be eliminated with IP telephony´s mobility feature. Outside of MAC savings are toll bypass, especially for international links and employee-to-employee voice calls over the existing wide area IP network. IP telephony does provide productivity increases measured in increased revenue per employee. Yet no CNA factored this into their cost justification. In short new business process drives new applications, which is driving IP telephony deployments.

IP Centrex´s Success Doubtful

There is huge interest in IP Centrex offerings by AT&T, MCI, BT, Cable & Wireless, SBC, Verizon, Quest, Level(3), et al which all describe versions of IP Centrex in various stages of development. But these service providers may get a chilly reception when selling their services to enterprises. For example, David Greenfield, International editor for Network Magazine asked John Manville of Lehman Brothers, ?¬¢‚Äö√ᬮ?√¨Did you ever consider the hosted VoIP/IP Centrex solutions? And why didn’t you pursue them?" John Manville and all the panel members answered the question in the same way: IP Telephony is a new technology with inherent risks. None of the panel members trusted putting this technology and their enterprise voice service in the hands of service providers.

These chief network architects wanted control of their architecture, including its design, its deployment and operations, and wanted to use it as a platform to wrap applications around business process. They all agreed that it was the right choice not to use IP Centrex, as an IP Telephony implementation requires tender loving care to keep it running smoothly. They also considered their converged network an ?¢‚Ǩ?ìapplication development platform", not just an IP phone that rides on top of an IP network. Their
thinking is, IP Telephony is a platform on top of which they are building applications. Most importantly, these applications increase the productivity of their organization and create a competitive advantage. The panel members also agreed that there is no killer application for IP Telephony. Rather it´s the culmination of small applications that incrementally provide real, tangible productivity value.

This thinking doesn´t bode well for AT&T, MCI and all the others offering IP Centrex service. The other knock against IP Centex is that most if not all of the regional service providers insist on bundling their own transport with IP Centrex, thus driving up the cost. I believe that the traditional service providers are entering the hollowing out process as enterprises move their voice service to private IP Telephony implementation. In fact, this may bode well for different service providers such as Dimension Data www.didata.com, IBM www.ibm.com or wholesalers such as Level (3) www.level3.com, who do not force companies to buy transport from them. If this model becomes reality then the traditional service providers may be relegated to transport providers while high value services such as IP Telephony will move to companies such as IBM Global Services, Dimension Data, Level (3) et al.

It also became clear that companies such as Vonage, Net2Phone, VoicePulse, SBC, Packet8 and Skype who offer basic voice services at zero to near zero cost may also receive a chilly reception in the large enterprise market as IP Telephony is much more than low cost dial tone. While there is always a large market for free things, corporate phone service is not one of them. Corporations will always expect reliability, quality and service, especially from their enterprise voice service.

Wither MPLS?

One topic that surprised many was the high quality voice service the panelists are experiencing on both their local and wide area network with limited QoS deployed. On the local area, by keeping voice and data in separate VLANs and just tagging packets, quality was reported as being excellent. In the wide area, tagging voice packets and establishing prioritized queues was all that was needed to deliver excellent voice quality. Also, the cost of wide area bandwidth over the past three years declined along with the telecom crash. MPLS based services have not fallen nearly as far as private lines and frame relay, making MPLS relatively expensive. Couple low cost bandwidth with the tagging of packets and queue prioritization and what do you get? Private lines and frame relay, not MPLS for VoIP transport.

VoWiFi Will Be A Big Deal This Year

A surprising number of questions focused on VoIP over Wi-Fi. VoIP over Wi-Fi is the new cordless phone for corporate networks. There is a lot of industry activity within the vendor community such as Avaya, 3Com, Mitel, Nortel, Cisco, Symbol, et al offering Wi-Fi phones. While there are a few Wi-Fi phones on the market today, I expect to see an explosion of them during 2004. The Wi-Fi or WLAN switch vendors need to beef up their voice support on access points as only 5 to 7 IP voice calls can be accommodated on an access point today. But all the questions on this topic were in search of experience and guidance as the market
ponders Wi-Fi corporate IP phones.