TORTOISE vs. HARE

It seems to be the well-worn path in telecom to announce you've got a product or service as soon as you possibly can - even if you've only run one demo of the thing and people won't be seeing it in action for months.

This has been a particularly frequent phenomenon in the world of Internet protocol applications - for example, everyone was talking about voice over IP long before anyone was talking through voice over IP. Maybe all the bandwidth and speed that's necessary to get these applications out is behind this rush to announce the next big thing. You start talking faster when you talk about fast things. And for many applications, maybe this can work. Keep plugging your new service, and little by little it will seep into networks and start getting used.

But for one application in particular - video over IP - rushing to market won't help anyone. Like so many other applications, it has hurdles to overcome: standards, quality of service, getting past corporate firewalls. It has even more basic challenges - most networks just don't have the bandwidth to handle video well yet.

It is also an application that can't be used half way. Customers' networks don't handle it well yet, and quality will be so bad that it will actually detract from a conference or announcement or whatever it is being used for instead of helping it. Currently, businesses are better off not using video over IP at all, rather than trying it and getting terrible results.

"Can you do video over IP? Yes. Are people doing it? No," says Andrew Davis, an analyst for Wainhouse Research, a videoconferencing research group in Brookline, Mass. "You won't see it really take off until the end of next year - not until quality of service is well-established."

He says that within fifteen months, many more people will have high-speed access methods such as DSL, and many companies will build out their network infrastructure more, making video a much more convenient thing to offer.

Which raises another unique point about this app - is it really necessary? How many people need to see each other in order to talk? Until video is an easier, more logical app to offer, it would seem that people could get by just fine without its difficulties.

"Say you've got a sales meeting with twelve people in a few different locations," Davis says. "If you use a simple conferencing system, you've got a low level of difficulty and a high level of quality. But if you add video, the quality is less, it's much more complex and the connection is more difficult. In many cases, the value of having video is not equal to its cost and complexity."

It will be different, Davis believes, when video is easier. In two to three years, it will be as common as having CD-ROM drives in computers, he says. "It will be as easy as picking up the telephone. And it will all be over IP networks."

A key move for video will be when it is integrated into other business applications. "Over time, you will see video-centric applications get woven into business processes," Davis says, adding that this will be due in great part to the rise of application service providers. And of course, one of the advantages of running video over IP is its ability to be integrated with other IP applications.

Until this performance improvement can occur, however, conferencing companies will do what they can to provide video. Companies like MSHOW (see story on page 18) are combining Internet, data, voice and streaming video - all over IP - through one conferencing service. It's a good way to get the ball rolling in this IP video race - almost as if it's an animal somewhere between the tortoise and the hare. That may not sound very attractive, but with this technology that's the point - don't show up for the race until you look like a winner.