Local-loop growth hinges on tapping proven networks and services
To compete successfully in local-loop fiber-optic networks during the short term, telephone and cable companies should aggressively pursue the huge demand for existing telecommunications services
probe research inc.
Several telecommunications company executives contend that the financial battleground of the competitive local loop will initially be a contest among telephone and cable companies for delivering voice and video services, not futuristic multimedia-based services. Presently, the only mass-residential telecommunications services with proven and predictable customer demand are voice telephone and basic cable TV.
The architectural implications of this service scenario for structuring fiber-optic networks in the local loop have crystallized. To the extent that service revenues for the remaining years of this decade are to be driven by broadcast applications--whether analog or digital is not important--then installing a low-cost hybrid fiber/coaxial-cable network is recommended. To the extent that service revenues are driven by symmetrical applications that require heavy return traffic, then pushing fiber closer to the customer, coupled with switched digital approaches, merits attention.
Someplace in the middle are service scenarios that envision a mix of broadcast and switched applications without "killer application" revenues for any single service. In these areas, a hybrid fiber/coaxial-cable network makes sense. In this scenario, a fiber-to-the-curb network (16 to 32 homes per optical network unit) for digital services might need to be augmented by a hybrid fiber/coaxial-cable network that can serve a large customer area (2000 homes) for analog services. Industry studies are inconclusive, though, as to whether a deep fiber-to-the-curb network (four homes per optical network unit), combined with an overlay hybrid fiber/coaxial-cable network, is financially sound at light penetration rates (number of homes that have purchased services versus total number of homes passed).
For all local-loop networks, the issue is not raw horsepower--the unlimited ability to provision any service, at any bandwidth, to any location, at any time. The imperative is enough horsepower to handle near-term demands, enough flexibility to upgrade capability as new services emerge, and at low enough cost to avoid bankrupting the enterprise if the demand curve is less robust than projected. On the other hand, the fiber platform must add more intelligence and bandwidth on a scalable basis. "That`s ultimately important, especially when you don`t know what the services will be," warns Larry Stark, vice president for new business development at Ortel Corp., Alhambra, CA.
During the rest of this decade, local-loop markets are expected to be opened to robust competition by deregulation, thus severing the historic pattern in which networks were optimized for a relatively well-defined set of applications. The new rules guiding deployment of broadband fiber-optic architectures will emphasize flexibility, in part, to handle the heterogeneous nature of offered traffic (high, medium and low bandwidths; short and long holding times; moderate to low latency requirements; and mixed media formats).
At the same time, network providers will find themselves on a slippery battleground that is projected to offer scant guidance on the proper mix of network functionality and market demand, investment cost and service capability. Ideally, network providers would like to install an infrastructure that accommodates all anticipated customer demands, thereby generating the highest revenues at the smallest-possible installed first cost and lowest lifecycle cost. Of course, they would like to avoid overinvesting in functionality that does not hold the promise of generating revenue.
But the looming instability caused by local-loop competition destroys most of the usual rules by which network providers craft their systems. In general, network providers simultaneously lower revenue expectations from core businesses and raise infrastructure investment thresholds. Meanwhile, the advent of robust wireless access technologies, such as personal communications services, reduces the "captive customer" element of a wireline carrier`s business, suggesting the possibility of a substantial shift of traffic from wired to wireless modes. All these factors demand rigorous examination of fiber-optic infrastructure assumptions from the standpoint of cost, anticipated customer demand and revenue.
Elements of the multimedia infrastructure require similar judgments. Assume, for example, that a small number of video movie titles--20 to 25--generates 50% to 80% of a company`s service-delivery revenues out of a library of 1000 titles. In this case, a near video-on-demand system offering start times of approximately 10 minutes would cost almost half as much as a true video-on-demand system offering the same 1000 titles. A realistic full-deployment scenario would probably see an increase in buy rates for the true video-on-demand system, but possibly would not result in a net-revenue increase, as customers might substitute video-on-demand spending for other video purchases. Alternatively, true video-on-demand buy rates might be 25% higher than but cost twice as much as near video-on-demand.
For network providers, a growing divergence between cost per home and cost per customer highlights the nature of the service delivery risk. According to David Reed, senior advisor, strategic planning, at Cable Television Laboratories Inc., Louisville, CO, providers should assume a per-home cost of $696 to $1395 for a residential broadband infrastructure.
These figures, of course, represent one set of thresholds--the minimum cost to build a ubiquitous fiber-optic network in an area. What these figures do not capture is the size of the investment to be amortized over the base of customers. In a cable-TV environment, this investment is typically expressed as cost per subscriber, a metric that hints at a profit/loss potential. For example, assume construction of a hybrid fiber/coaxial-cable network that anticipates a 60% penetration (six of every 10 homes passed have purchased services). In this context, network costs would range from $424 to $1684 per customer, Reed argues.
As a starting point for revenue-potential discussions, consider the gross monthly revenues for a small number of voice-telephony services that achieve close to 100% penetration:
Local Access Calls $20 per home
Intra-local Access Transport Area Tolls $11 to $15 (where a call per home originates and terminates in the same area code)
Long-distance Calls $11 per home
On the video side, cable-TV services reach approximately 60% of U.S. households. Ad-supported cable TV, the next most successful network-delivered service (after telephony), is bought by approximately six out of 10 homes at almost $22-per-subscriber level. Such premium cable-TV services as Home Box Office are bought by approximately 20% to 30% of households, and the average premium customer buys 1.8 services. Pro-rated on a per-household basis, this video-service delivery works out to:
Basic Cable TV $13.35 per home
Premium Cable TV $6 per home
Although consumer use of personal-computer-based e-mail services is exploding, interactive revenue streams for most other multimedia-related applications remain uncertain. Equally unfortunate for local-loop stakeholders with nearly 100% share of a protected market, the only direction the incumbent`s customer base can go is down, once competition begins.
Some observers anticipate a near-term network provider financial strategy based primarily on attacking existing telecommunications and cable-TV revenue streams. In this approach, network providers would seek to achieve market share gains at the expense of incumbents. This appraisal, however, does not decrease the importance of cultivating the new markets that will emerge as broadband capability is extended to a critical mass of households.
Consider, for example, a new entrant`s local telecommunications and cable-TV revenue potential, compared to revenue from such new applications as information-on-demand and transactions. In this case, the initial targets are anticipated to come from:
Wireline local access
Intelligent network services
Next, consider a scenario in which an attacking company captures 1% of its primary line business in its first year of operation and 13% by year seven. In addition, assume another 1% penetration of additional lines business, growing to 15% after seven years. Furthermore, presuppose a local customer bill of $21.22 a month, plus $12.50 per line for long distance and telephony features, including call waiting and call forwarding.
If rate increases are held to inflation-only levels, and if the attacker prices are at 70% to 80% of local exchange carrier levels, revenue per subscriber should begin at approximately $180 a year and scale to approximately $202 after seven years. Annual revenue per home passed would begin at $5.58 ($.47 per home per month) and grow to $56.49 ($4.71 per home per month) after seven years.
Compare that revenue stream to the current cable-TV revenue of $7.56 per home, per year, for pay-per-view movies; $11.52 per home, per year, for advertising; and approximately $1.68 per home, annual revenue, for home shopping.
To expand the plan, consider that the same infrastructure used to provide local access at 64 kilobits per second might also be used to provide high-bandwidth data connections for work-at-home, education, entertainment or transactions. Moreover, assume that personal-computer penetration reaches 45% of homes by 1998 or 1999, and that an initial 2% of those homes want high-bandwidth connectivity, growing to 22% after seven years. Assume a $20 monthly access fee and a personal-computer penetration growth to approximately 22% of homes.
To examine the local-loop market from a local exchange carrier perspective, assume the entry by a telephone company into video entertainment similar to that provided by cable-TV companies. Also assume a community where cable-TV penetration is 71% with a monthly basic service bill of $18, premium penetration of 23% and an average $10 per month premium billing. In addition, assume capture of the basic cable-TV business of 12% initially, growing to 37% after approximately seven years.
If the local exchange carrier charges $16 a month for its basic cable offering, per-subscriber revenue is $1.84 per home passed, per month, growing to $5.92 per home, per month. On an annual basis, basic cable might represent an annual revenue of $22 initially, growing to $71 a year at 37% penetration.
The implication for fiber-in-the-loop planning seems clear. Fiber should be driven far enough into the neighborhood to support low- and medium-penetration services as a first step. As multimedia applications emerge, fiber to the curb will make economic sense. u
Gary Kim is senior vice president at Probe Research Inc. in Cedar Knolls, NJ.