Throw it On the Wall and See if it Sticks

Though it isn't the way service providers are used to doing things, allowing customers to discover services they like is about as rational as anything else a provider might try. Especially when everybody is in discovery mode anyways. Mobile providers "discovered" the popularity of short message service, rather than being able to predict such success. And if it doesn't die of its own weight, IMS is supposed to allow such discovery and trial processes. The Web is the other way discovery processes happen every day.

In 2005, Forester Research estimates, just 2.6% of Western European broadband consumers used VoIP for almost all their fixed-line calls from home. But matters will change dramatically. Forrester Research expects three in four European telco network broadband subscribers to use VoIP within 10 years. But the average incumbent telco will not get more than €63.58 (a bit less than $7 a month) in net annual VoIP revenues per broadband user in year 10. That, plus a dramatic shrinkage in revenue-beating access lines, as indicated by Technology Futures Inc. forecasters, highlights the urgency and strategic importance of new revenue streams. There simply isn't going to be all that much consumer voice revenue to get.

Churn: Watch What :People Do, Not What They Say They Will Do

As wireless number portability comes to the Canadian market, consumers seem to be telling The Yankee Group there is a significant chance they will use the opportunity to switch providers. But that's what people told researchers when number portability came in the U.S. market as well. While one has to be careful about inferring too much from experience in other markets, there is a fair to good chance not much incremental churn is going to happen. People often say they will do this or that. Sometimes there is a huge discrepancy between what they say they will do, and what they actually do. This is probably one of those cases.

50 Mbps Symmetrical Access From SureWest

SureWest Comms. soon will offer a 50 Mbps Internet access service (up and down!) for residential customers with access to the company's FTTH network. The company has been selling 10 Mbps and 20 Mbps symmetrical services for some time. SureWest is bundling the 50 Mbps offering in its newly-created “Ultimate” quadruple play package, which includes the 50 Mbps access service; a 250-channel digital TV service; National Unlimited local and long distance telephone; and Unlimited wireless with 1,000 travel minutes each month. The package costs $415.18 a month. If it were offered on a stand-alone basis, SureWest says the 50 Mbps service would be valued at $259.95 per month.

So think about it this way: A DS3 service should cost less than $300! And we can only hope that Verizon is able to stick to, or increase, its own FTTH deployment. Not that all that many residential customers need symmetrical 50 Mbps access today, at least not for any purpose within the scope of "acceptable use" policies. The problem is that at such rates, about the only application that really benefits is hosting of video servers inside the home. And that typically falls outside a service provider's notion of acceptable use. Still, it's an awesome advance in access bandwidth.

Many Internets, Many Models

Whatever positions one takes on the "future of the Internet," or the "future of the telecom business," one thing seems clear enough: The Internet is going to change, because it is fragmenting into multiple Internets, private Internets, regional Internets, application-specific Internets (as odd as the concept seems). We aren't going to live in a world with a single Internet, but rather a world in which there are many types of IP networks, some more open than others.

Mike Volpi Cisco SVP, argues that service providers (telco and cable, for example) can reshape at least parts of the legacy Internet. For starters, transport and access providers probably will be able to charge differential rates for varying levels of quality and higher levels of policy management, especially for real time services. Providers probably also will find they can charge different amounts of money for different amounts of upload bandwidth as well.

Stepping back from the inevitable policy debates about how much intelligence "needs" to be in the network, and what such intelligence might mean for new gatekeeper roles for transport networks, it seems clear enough that services such as high definition video and audio do require policies such as enterprises normally apply for their own traffic, to prioritize packets and bandwidth access. This is less a matter of "controlling the Internet" and more a functional requirement, though the danger of abuse cannot be discounted.

And while the notion of an "application-optimized" IP network seems counterintuitive, it makes lots of sense, if one assumes there are large, sustainable markets for applications of various types that benefit from network tuning. Video and voice provide the most obvious widespread examples, but there are lots of vertical market apps that also would benefit from network tuning. Gaming networks, security networks, hospitality and medical segments come to mind. Advertising, marketing, video post-production and other specialized news feed apps are obvious as well. Some apps just require network tuning. And it probably will turn out that these new apps are the ones with clear revenue models attached to them.

VoIP Mostly Works

Recent Mean Opinion Scores of VoIP traffic by Minacom show that VoIP audio quality pretty much works. Quality isn't uniformly high, because of the unmanaged nature of access bandwidth and the general state of networks some places in the world. But it works well enough to be useful. This obviously raises a question.

At some point, when the technology underpinning voice is nearly 100 percent IP, there may yet be ways to differentiate services based on levels of assured audio quality.

Managed networks probably still will be able to provide higher MOS scores on a consistent basis, compared to unmanaged networks, even though performance on unmanaged networks also will improve.

Of course, the other quality metrics should be capable of differentiation as well. Session integrity is the other current example of varying quality. Even when a VoIP call "sounds good," the integrity of the session might not be as good as a PSTN call. Voice VPNs will help, of course. So the issue is the degree to which unmanaged connections can be made more reliable by addition of VPN capabilities.

Simplicity Wins

U.K. small and medium enterprises surveyed by the Bathwick Group say they prefer to limit the number of suppliers they must deal with. Some companies are large enough to aggregate large amounts of value and can satisfy more wants and needs with one relationship. The other approach, better suited to smaller or specialized suppliers, is to provide a managed solution that removes complexity from the premises. Either approach will work.