The networks of yesterday were dominated by voice and low-bandwidth data services, such as e-mail and basic Internet connectivity. They have been planned in the same way for the last one hundred years, with the marketing organization and the planning organization working together in a time-honored dance.

The Old Model - Incremental Planning Built on Historical Trends

The first step in determining the capital required to build a network that was mapped to available budget and expected service demand, was simple in structure, as shown in Figure 1.

Network usage data was traditionally gathered by the planning organization from the as-built network and analyzed using trending methods, then modified by forecast information from marketing to reflect that organization’s changing demands.

The planning organization would use this information to decide which new equipment would be necessary to meet demands, often eschewing the marketing input (i.e. “They incorrectly predicted a large uptake of ISDN once, why should we believe them now?”), and effectively over-built the networks to consume the available budget.

This practice cost more, but assured that resources were available to meet customer demand and, even if the marketing organization was right and the new service was a hit, would require but a small incremental increase in capacity to get by.

Such high-cost planning methods were tenable when:

- The services provided changed little year to year
- The rate of service adoption was relatively stable
- Consumers had little choice in substitute technologies

Of course, even with this relatively simple method, a Tier 1 carrier could find itself with as many as 15,000 separate construction projects per year as capacity was added where needed and reduced where not needed, and new technologies were introduced into the network.

The New Model: Market-Driven Planning

Fast-forward to the present day. Carriers’ marketing, technology and business departments must work together more effectively in the planning stages to ensure success when introducing new technologies, business models, and services.

Standard forecasting and data trending was sufficient for planning yesterday’s networks, but is no longer adequate for introducing next generation network architectures, or new bandwidth-hungry services, as there is no historical data to use for trending purposes. Overall market take rates are unknown, and very difficult to break down across multiple serving territories. And with the large bandwidth demands of video customers, it is not economically feasible to simply “over-provision” the network - yet if a mistake is made and the network is under-built, the quality of service will suffer severely, crippling the market introduction of the new service.

This lack of data means the carrier must derive its master plan proactively from market information, as opposed to historical network information, as shown in Figure 2. Here, the marketing organization must determine the anticipated demand for services and make it available to the network planning organization to create a strategic network plan. Then, since the market demands are probabilistic, they must be checked occasionally - much more often than yearly, as was done in the past - and the plans changed as more information is accumulated about actual market demands, values and locations. The trending information, thus, takes a back seat to the market demand data in driving the plans - trending becomes a “check” on the plan, and, in some cases, a fine tuning of the network plan.