A little like Goldilocks

As carriers jump on the DSL bandwagon, a key will be finding the magic price point - not too high, not too low, but just right

Most service providers today are enhancing their business plans to include a DSL service offering. With more than 500,000 DSL links deployed through 1999, DSL rapidly has become a requirement for the competitive portfolio. Unfortunately, many service providers are struggling with DSL deployment, starting with the pricing strategy.

A recent sampling of 15 DSL packages offering 1.5 Mb/s/385 kb/s asymmetrical DSL (ADSL) reveals monthly rates of $79 to $389. This broad range suggests that in the rush to deploy, pricing strategy fundamentals may have been inadvertently abandoned. Of course, in major markets where DSL competition is fierce, new entrants default to going rate market pricing; a de facto ceiling already has been established. But in new, raw markets - or for new advanced DSL applications - there is a clear opportunity to strategically price the service and understand, ahead of time, the effect that pricing will have on demand.

Figure 1 shows current price points for ADSL service offerings.

However, as historical price data for a service as new as DSL is largely non-existent, several considerations and strategies must be examined to price DSL within the context of fundamental economic principles. Price elasticity, penetration pricing and price skimming are three principles that can help managers price DSL.

Is it elastic?

Price elasticity of demand is a measure used by economists and managers to determine how the demand for a good or service will respond to a given price. A service can be considered inelastic or elastic. An inelastic service is basically unresponsive to changes in price. Whether the price goes up - or down - demand remains basically unchanged. If a service is elastic, however, then a change in price will result in a change in demand for that service.

T-1 exemplifies an elastic service. Tables 1 and 2 illustrate the price and deployment of T-1 services between 1994 and 1999. Although price has decreased by only 18%, the number of T-1s deployed (quantity demanded) has increased by 177%. This indicates a very elastic service where a small change in price equals a big change in demand. Of course, other factors have contributed to the rapid deployment of T-1, but if T-1 service were not price elastic, the other effects would have had much less of an impact.

Price elasticity is typically represented by a numerical value derived from the following formula:

Price elasticity (Ep) =% change in quantity (sales) % change in price

If the result is less than 1, demand is inelastic; if greater than one, demand is elastic. How much greater (or less) than 1 indicates the level of elasticity (inelasticity). Using this formula, we can determine that in the T-1 example, price elasticity would be 9.8 (177% (18%). The result is much greater than one, or very elastic.

Typically, determining price elasticity requires extensive pricing and demand data. Unlike T-1, however, there is insufficient historical data from which to evaluate the elasticity of demand for DSL services. Yet even when historical data is not available, price elasticity principles, if not necessarily the formula, still can be applied. If empirical information is not available, or too unreliable, then service providers canevaluate two key determinants of price elasticity to estimate a logical level of pricing. The primary determinant is the availability of substitute services; the secondary factor is degree of product versatility.

Services that have many substitutes (such as answering services in lieu of voice mail, 56 kb/s frame relay instead of dial-up 56 kb/s modems or PCS rather than wireless) tend to be more elastic. That is, if a customer has numerous acceptable alternatives to a currently used service, then an increase in the price of that service will result in their moving to the alternative. Demand and revenues for the current service decrease, while the substitute service gains customers and an increased revenue stream. By corollary, the fewer substitutes available, the more flexibility the company has in setting a higher price.

Evaluating substitutes obviously applies to pricing a new DSL service. The key is to accurately determine what constitutes a viable substitute product. For example, dial-up modems are not a substitute for DSL; the many advantages of DSL place it in an entirely different class of service. Nevertheless, there are numerous Internet access technologies available today, including fractional T-1, ISDN and 56 kb/s frame relay, which can be considered acceptable substitutes.

Another factor of price elasticity that is especially applicable to DSL is product versatility. That is, if a service can be used for two applications instead of one, it's more versatile. The more versatile a product is, the less sensitive it will be to a price change because it can do more, and it can't be as easily substituted. This principle is important to understand because DSL has been touted as a technology that supports multiple applications. A market educated to understand the value of versatility will bear a higher price, at least initially. So, appropriately marketing DSL beyond high-speed Internet access will allow service providers to charge more.

For instance, if a 384 kb/s fractional T-1 pipe to the Internet costs $200 per month, a 384 kb/s DSL pipe may be competitive at $220 per month. Why? Because the versatility of DSL makes it less easily substituted and more elastic; a higher price will have less of an impact on customer churn. The customer is paying for a 384 kb/s DSL connection today knowing that later he can non-intrusively increase bandwidth to 768 kb/s, add a virtual private network (VPN) circuit or add voice services without the typical delays associated with adding more lines.

A perfect world

Simple models often can be helpful in visualizing the effects that pricing will have on demand for a good or service. Figure 2 is the typical representation of price elasticity of demand.

Service Provider A has no DSL competitors in his market. Service Provider A has acquired 1000 customers on demand curve D1 at a price of $79.95 per month. Service Provider B enters the market and defaults to status quo pricing. At the going rate of $79.95, Service Provider B acquires only 500 subscribers. Realizing that it provides a viable substitute, Service Provider B lowers its price to $59.95. The result, because demand is elastic, is the demand curve shifts. This represents the price-sensitive customers who move to Service Provider B because of the change in price. The result is that Service Provider A's customer base decreases to 800 users while B's increases to 700. Service Provider B's 25% change in price is offset by an even greater increase - 40% - in demand. Monthly revenues grow by close to $2000, or approximately 5%.

This is an effective tool that should be considered by those developing their DSL product offering. It can help managers visualize the impact that DSL pricing will have on the take rate for service and customer growth rate.

All about pricing

Assessing versatility and availability of substitute products can assist service providers in determining how to price their service. Equally important is integrating the economic principles of penetration pricing and price skimming into the carrier's strategy. Both pertain specifically to new service pricing, making them very applicable to DSL.

Penetration pricing is the process of setting a relatively low initial price for a service. This type of pricing ensures a broad penetration of the market in the current period and a large market share in subsequent periods. This strategy is generally adopted in two instances: when the company needs early acceptance and when the company wants to maximize long-run profits. Because of the high initial cost of deploying telecom services, this strategy is not used frequently. One of the RBOCs, however, used penetration pricing for DSL to great advantage. Its initial pricing was much lower than prices offered by its DSL competitors. In a very immature market, this is unusual, yet was highly successful.

Price skimming is not the nefarious pricing strategy that the name implies. It is simply a decision to set a high initial price for a new good or service. This strategy is adopted to maximize short-run profits. That is, charge more until the market will no longer bear it. That's what a prominent data competitive local exchange carrier (CLEC) did. As the first to market, in late 1997, this data CLEC took advantage of its position by setting relatively high prices. When competitors trickled in, the carrier reduced its rates accordingly. As the data CLEC with the largest number of lines deployed, almost 60,000 by the end of 1999, this seems to have been the correct strategy.

Setting a strategy

Service providers should use penetration pricing or price skimming for new service deployments. The overall state of the market they're trying to enter suggests which pricing strategy should be applied. Figure 3 illustrates the process of determining whether a penetration-pricing or a price-skimming strategy should be adopted.

The graphic highlights four key variables that help determine the overall market state: barriers to entry, availability of substitute products, general market perception and likelihood of market acceptance. Careful consideration of these factors will help determine which strategy to chose. Of course, it's too much to expect a completely unambiguous assessment. The model is simply a tool. The relevance, or weighting, of each variable also must be factored in. For example, in some markets the perception of DSL is swaying between premium service and commodity access. The former suggests price skimming, the latter penetration pricing. Fortunately, DSL is quite versatile. A service provider can adopt a penetration pricing strategy for commoditized DSL, while simultaneously implementing a price skimming strategy for its advanced DSL services, such as VPN.

At this point, the service provider will know what profit expectations it should set - short run or long run. If that objective is not acceptable, then another market should be selected or operations costs should be adjusted to increase profitability. Price should not be altered. If price is set too low in a price-skimming strategy, profit will be sacrificed. If it is set too high in a penetration-pricing strategy, revenue will be foregone.

The concept of elasticity, including substitution and versatility, is one of the most important aspects of demand analysis. The relation of product price to sales volume is the basis for pricing strategy. Yet busy professionals sometimes forget the fundamentals when developing new services. A logical framework can simplify the process and help managers avoid pricing pitfalls. And because no strategy lasts forever, these basic pricing principles should be regularly reviewed to ensure that profit is maximized.