The two main target markets for broadband are the small and medium-sized enterprises and the consumer markets. Analysts at Morgan Stanley Dean Witter estimate that the SME market will reach $52 billion by 2005. SME services include Internet access, voice services, application hosting, virtual private networking, and Web hosting. The same analysts estimate that the consumer market will grow to $108 billion by 2005, up from $92.5 billion in 1999. Morgan Stanley Dean Witter analysts expect that by the end of 2000, DSL technology will be the access method of five percent of all consumer Internet users, growing to 28 percent of all Internet users by 2005.

Analysts are split regarding which technology will be the winner, and thus the forecasts vary greatly. However, one clear trend emerges: until all service providers concentrate more on customer service and solving back-office issues, DSL will continue to lag behind cable. Poor customer service is the Achilles' heel of DSL providers. The lack of attention to customer service is one of the most troubling issues facing the broadband marketplace today. Customer service is a critical but overlooked ingredient of any e-commerce or broadband access provider. The success of the DSL service provider hinges almost entirely on the ability to obtain and retain customers. Retaining customers is directly related to customer satisfaction.

Forecasting in general is an inaccurate science, and thus analysts' estimates tend to be all over the board. However, that being said, we can make intelligent guesses about DSL subscribership and the growth of the industry. Both Moore's Law, which states that processing speed doubles every year, and Metcalf's Law, which states that network impact increases exponentially, can be used to project future demand for DSL. Internet traffic is doubling almost every year and switching and fiber capacity from wave division multiplexing (WDM) and terabit routers have allowed the network to rapidly carry more traffic than was thought possible only a few years ago. Technology growth will continue to drive down costs, making available more types of applications and value-added services. Internet video—the ability to watch full-motion movies through your PC—is but one example. Providers are building fiber deeper into the networks and moving intelligence to the edge of the networks. The technical problem of delivering video is well on the way to being solved through the use of caching and by bypassing the Internet backbone with satellite or fiber to deliver programming to a server at the edge of a network. The second bottleneck is inadequately provisioned networks. This too is being solved through the use of a new network architecture similar to the designs that SBC is implementing as part of Project Pronto.

Cable continues to have an early lead in deploying high-speed services, but its lead is primarily in residential services, rather than in the more lucrative business services. This lead will likely continue for the next few years as DSL continues to have technical and provisioning problems in getting enough lines deployed on a timely basis. These problems likely will ease significantly with the widespread implementation of line sharing and the work of the DSL Forum on autoconfiguration and in getting to plug-and-play.