MARGINS AND ERRORS

Internet service providers are making some seriously harmful mistakes in pricing their consumer Internet and data services. Their existing pricing structures don't work to capture the elastic properties upon which long-term success will rest; they have instead based their models on unlimited usage at a particular speed — an arrangement that keeps the provider from seeing the additional revenues that come from increased network usage.

As new applications gain traction (like we saw with the meteoric rise and fall of Napster and its bogging down of entire networks), end users will sap more bandwidth with no additional costs. This isn't necessarily a doomsday scenario, as the profit margins on these unlimited usage models can still be extremely high at these stages. Furthermore, it's likely that the network costs of providing these Internet services will continue to drop.

However, a potential problem lies in the fact that because end users are now accustomed to this unlimited usage concept, service providers have likely lost the flexibility to implement structural changes to this scheme. And as higher and higher speeds with unlimited usage become available, if the costs of providing such speeds don't drop at a corresponding rate, the very proposition of providing Internet data services might become an unprofitable one.

Interestingly, service providers have run into similar problems in the wireless sector. Those giant bundles of flat-rate minutes also run counter to the usage-based model that was the foundation of so much success and profit. However, as next generation networks are being rolled out, the smarter operators are building pricing structures that are based on the number of bytes transferred — i.e. usage.

Elasticity is not infinite. There is, after all, a finite amount of funds that end users can spend on network services. But if service providers can learn to leverage bandwidth-intensive applications to their benefit, then we're still far from reaching such a flattening of the curve. Specifically, returning to the model that Napster exploited, we can realize that a pay-for model for the music itself can reallocate the funds being spent at the record store into the network. Service providers must return to a usage-based model in order to actually be paid for delivering such music to the end user.

These problems are exacerbated by the effect that Wi-Fi and access sharing will have on usage because more users can leverage a single connection. In fact, service providers could possibly leverage Wi-Fi as a catalyst for shifting back to usage-based models. They'd better get on it soon, though, or the speed-based models will hurt margins for years to come.