Will Metro Ethernet make it? - Business Networks

Is metro Ethernet dead or not? Metro Ethernet service was widely and loudly heralded as the salvation of the networking industry. Then Yipes, the most prominent proponent, went bankrupt. Consequently, all the heralds proclaimed metro Ethernet dead. Yipes, however, has re-emerged and many ILECs and IXCs have floated metro Ethernet offerings. Three issues must he resolved before its ultimate fate is known: How can it be used? Who will offer it? Can it be priced profitably?

The early marketing spin was that metro Ethernet was bargain-basement bandwidth--traditional best-effort Ethernet with disruptive pricing. A few start-ups talked about gigabit service priced at $1000 per month, undercutting the incumbents and ushering in a new era in which the cost of bandwidth no longer mattered. Metro Ethernet continues to labor under this market positioning despite the now painfully hollow nature of these claims. Metro Ethernet also is often equated with Gigabit Ethernet--used by blue-chip companies just like OC-48 and wavelengths. Other possible Ethernet roles include a new fast-packet service that competes with frame relay and ATM, the next POTS (the universal vehicle for all voice and data) and faster DSL (true broadband Internet access).

Metro Ethernet's fate also will be determined by how service providers perceive it to fit within their service portfolios. Cogent and Yipes have defined metro Ethernet primarily as Internet access service targeted at small and medium enterprises (SMEs) located in multitenant units. ILEC initial offers seem to be targeted to the blue-chip custom network crowd. AT&T and its competitors have launched two services-metro Ethernet offered by the local business unit as high-speed connections to POPs and long-haul service offered at SONET data rates. ISPs seem to regard Ethernet as just another Internet access service.

The tentative nature of most incumbent carriers' metro Ethernet offerings is due, I believe, to a great deal of fear, uncertainty and doubt about the economic consequences of metro Ethernet service. There is a fear that if metro Ethernet is just discount bandwidth, deployment costs will exceed future revenue streams. Even worse, the service offering may cannibalize existing services, especially the highly profitable T1 and frame relay offerings. However, all financial uncertainty is not on the downside. Metro Ethernet equipment vendors and the Metro Ethernet Forum have offered economic models that show that optical Ethernet products possess lower total cost of ownership than legacy technologies. In addition, many vendors have added QoS to their Ethernet solutions that permits delivery of maximum-bit-rate and guaranteed-bit-rate services. These QoS capabilities support more varied and flexible service offers that, in turn can be used to create sophisticated pricing plans and strategies-an important tool for increasing profit margins in tight markets.

Metro Ethernet has several good chances to survive, even in the current depressed market. First, well-focused strategies of selling Internet access to SMEs should succeed if capital spending follows--not leads--sales development. Cogent and Yipes, in particular, have said they intend to follow this strategy. Sale of Ethernet service to SMEs should also be attractive to the ILECs--they possess strong account control in the SME segment and have much better fiber coverage than anyone else. However, ILEC Internet access offers are a regulatory gray area, and since resolving regulatory issues always takes time, this creates a window of opportunity for Cogent, Yipes and other facilities-based competitors.

The point-to-point Gigabit Ethernet offerings of SBC and others are already successful but there are only a handful of very large enterprises that need such service for a very tiny number of very large offices of data centers. Consequently while gigabit service should be successful, it will not be materially significant to the finances of ILECs or IXCs.

Finally, new deployment of optical Ethernet technology including advanced QoS capabilities and Ethernet rings--whether RPI (Resilient Packet Ring) or another approach--will permit the creation of new Ethernet-based alternatives to TDM private lines, frame relay and ATM. This new technology incorporates DWDM packet switching and true TDM into a single system. This lets the service provider flexibly provision services using a single software system, which eliminates costly manual procedures--especially truck rolls--and improves time to market. Capital costs also are reduced. Unfortunately, I do not believe any ILEC or IXC has the courage or, in some cases, the money to make such a commitment now. Consequently, true service innovation is likely to come slowly. When it does, we should see ILECs making these investments in order to capture IXC's frame relay customers and, in turn, IXCs will begin building out metro networks to capture ILEC access customers.

Michael Kennedy is co-founder and Managing Partner of Network Strategy Partners LLC (NSP)--management consultants to the networking industry ([email protected]).

COPYRIGHT 2002 Horizon House Publications, Inc.
COPYRIGHT 2003 Gale Group