GX clouds the Verizon
By Bret Swanson
February 9, 2001
Verizon's announcement on February 7 that it will enter the international business telecom space is good news for Telecosm companies Global Crossing, 360networks, Metromedia Fiber -- and even WorldCom.
The move further validates Crossing's and 360's focus on subsea links and cross-border customers, as does the fact that WorldCom's fastest growing division is International.
But Verizon is seriously behind in this game. One, it has not targeted this market before. Two, it mostly will not have its own facilities but will be using those of FLAG and MFN, for example. Three, other non-facilities based service providers (or quasi-facilities based providers) are already up and running...and losing business to GX, WCOM, and soon to 360. How Verizon plans to succeed where others are currently being disrupted is a real unknown.
But many are wondering…
…Whether or not to be facilities-based. That is the frequently asked question. And surprise, surprise...it depends.
In general, I think service providers domestically will succeed if they outsource substantial portions of their networks, that is, if they are not facilities based. And international providers will win if they own the network (the facilities) end-to-end. But there are many exceptions and ifs and buts involved, too.
Take Yipes, the new metro EtherSP, for instance. They have built their own metro networks by leasing fiber from MFN, Level 3, and others, and connected customers using Extreme switches and Juniper routers. But they outsource their long-haul transport and Internet backbone. Cogent, on the other hand, has built both metro webs and a long-haul network of its own using one nationwide fiber from Williams. The optical nature of Cogent's network, powered by the Avanex OADM, is attractive, but its choice to build a long-haul backbone of its own is curious. Numerous robust backbones already exist. Why not use one of them?
Even SBC, clearly a large telecom player, has outsourced its network to Williams who handles all its long distance and data traffic. SBC concentrates on marketing and sales and customer service, and Williams concentrates on driving down the cost per bit and cost per lambda by deploying new optical systems and expanding co-location facilities, for example. Price and provisioning competition here will be fierce. It makes sense for SBC to be non-facilities based (at least in the long-haul) and for Williams to follow a pure, horizontal business model and not worry about selling DSL services to Joe Smith on Walnut Lane.
The international arena is different, I think. First, international telecom is mostly commercial, not consumer. Second, cross-border patch networks are complicated by tariffs, laws, protocol mixes, margin addition for each link, etc., etc. Third, there just aren’t many seamless global nets. GX is really the first. 360 will be the second (and will remove yet another layer of complexity by completely banishing Sonet). WCOM is a hybrid. And Level 3 is also in the mix. So GX or 360 or WCOM or LVLT can be largely facilities based but go the vertical route and legitimately target commercial customers, rather than just provide bandwidth to the AT&Ts and Deutche Telekoms of the world (which they do as well). This is why FLAG or TyCom will not be as successful because they will sell low-margin bandwidth under the seas but will not be able to take advantage of any high-margin opportunities on land -- they don’t have terrestrial networks. GX and 360 can do both.
Crossing’s $300 million plus deal with Swift last week confirms the seamless, global, broadband model. Swift connects 7,000 banks in hundreds of countries and channels over $5 trillion and 5 million messages each day over an outdated narrowband quilt network. Add new era scalability to Swift’s previous requirements of security and reliability and you begin to understand its need for one, and only one, world wide web.
Verizon’s is not a bad idea, but Gary Winnick, Greg Maffei, and Bernie Ebbers are way ahead. And if by slim chance Verizon, Sprint, and other new service providers do succeed, it will largely be because they used GX, 360, and MFN broadband pipes. |
