Enjoy the ride

Did you ever understand those Nissan ads? You remember: After some bizarre vignette they ended with a speechless, elderly, bespectacled gentleman staring at us from a visually disruptive perch, followed by the written tag line, "Enjoy the ride!"

The ads were supposedly a big hit. Unfortunately, they were a big hit only with those who give out advertising awards. Evidently, a lot of people like me didn't get it, and Nissan's U.S. fortunes faltered.

For the same reasons, the coming push by the multiservices hardware and service provider businesses on quality of service (QOS) for Internet protocol (IP) networks, at the fall trade shows and in the press, ultimately will be much ado about nothing.

That's right. I want to be on record: In less than three years, differentiation on the basis of QOS will be meaningless.

Now that I have your attention, I have an admission. This argument has been tried on over a dozen vendors to test its shock value. Based on experience, I now add a caveat in the next breath: For the next two to three years, QOS as a differentiation in multiservice offerings will be tantalizingly and agonizingly important.

The reasoning goes as follows:

* This season's "best-effort routing" for things like IP VPNs will give way to next season's "better-effort routing." And, just in time for the millennium, will appear, "always-perfect routing"-every time, all the time.

* Service provider A will give customers a variety of QOS offers with attendant service level agreements (SLAs) at prices X, one-third X and two-thirds X. (Think of this as first, business and economy class with "seat farming," in each category.)

* Service provider B will offer customers a variety of QOS offerings with SLAs to match those of service provider A, but with first class pricing starting at two-thirds X.

* Service providers C, D, E and F will magically appear, all willing to beat service provider A and service provider B, and willing to dispose of long-term contracts.

Discerning customers will act rationally in this environment. Talented IT shops at places such as company Z will set up what I call, "best-performance routing" capabilities.

Think of this as the multiservices version of least-cost routing. It comes with the enhancement that the customer will call the shots on a long list of parameters, including QOS, on a per networking event basis. The end result will be only the "best" will do, and "best" will become a commodity.

The reason this scenario will take a few years to develop is because the technology is not quite there yet. It will be here sooner rather than later. A lot of "best efforts" will make it happen.

Why QOS is important in the short term is because going forward, IP rules! As such, packets rule and communications of different packet sessions and types need to be prioritized if they are to arrive on time and in order. In addition, LAN connectivity is important, and leveraging the Internet for extraneting is becoming critical.

Add to the above the fact that the business case for voice over IP, including over the Internet itself for personal use, extranets, virtual and multimedia call centers, has become so compelling that 1999 surely will be the year of network transformation for end users as well as carriers. If tariff arbitrage was not a good enough reason to move voice onto IP, then network usage benefits, ease of migration through service intermediation via the addition of SS7 support, high-fidelity voice quality, simplified network management and "future services awareness" should close the deal.

The big question is not whether large-and even not-so-large-customers will move toward virtual convergence (multiservices) networking. Rather, will the solutions chosen be primarily CPE-based or are we entering a golden age of networking where service providers end up being outsourced more and more responsibility? And because perfect won't exist, finding the best effort will be important until technology and competition commoditize things.

In the future, corporate survival will depend upon 100% network-and networking vendor-reliability. And as the gateway oligopoly crumbles, customers will demand one class of services, the "best."