News Stay informed about the latest enterprise technology news and product updates.

Gartner: Nostradamus with Vegas odds

Gartner goes the usual Nostradamus routine one better by assigning some of its predictions probability values. They don't just tell you what's likely to happen; they tell you how likely it is to happen with a number.

I recently had the pleasure of attending the Gartner Group's conference Networking Beyond the Enterprise in downtown San Francisco.

Those of you familiar with Gartner and its team of technology analysts are probably already acquainted with the Gartner style - lean, mean, and frequently peppered with interesting predictions concerning future technological developments.

Now, this is hardly unique in the world of technology analysis. After all, the whole point of analysis is to reduce a complex industry in a state of considerable churn to key points so that you can see where things are going.

Gartner, however, goes the usual Nostradamus routine one better by assigning some of its predictions probability values. They don't just tell you what's likely to happen; they tell you how likely it is to happen with a number.

Nostradamus himself, a guy who had a lot to say on a lot of subjects, didn't give us Vegas odds.

And to do it in networking is particularly gutsy. Few technology arenas have gone through an evolution as rapid as the networking industry's. Few are expected to.

How many network pundits in 1998 would have been able to foresee that within two years, the single biggest choke on corporate networks would come from peer-to-peer sharing of pirated music? And that a scant year later, aggressive court rulings would have addressed the issue, dramatically scaling back the effect?

How many would have told us there was an X percent chance all this would happen?

We at searchNetworking applaud Gartner and its determination to reduce the chaos to a probability value (but we admit we're curious about the math involved.)

Let's take a look at a particular case in point.

Voice over IP (VoIP) has been a curious networking subcategory in a number of respects. The strongest VoIP argument has, in the past, been this: Do it because you'll save money. When you transform a phone call into just another collection of packets and deliver it over the public Internet, you free yourself from the fiscal tyranny of the established telephone service carriers.

An appealing argument. However, the VoIP horse hasn't exploded out of the gate with quite the intensity that some expected of it. Why?

Many reasons. There have been quality problems. There have been interoperability problems. Deployment costs have been higher than they might have been. Next-generation services based on the concept of voice as data have remained, for the most part? next-generation.

And in the end, the majority of today's enterprise operations have responded to the VoIP industry by saying "Thanks, but we're happy with our existing phone service right now. But get back to us when things get better."

The VoIP industry, of course, isn't sitting still and is in fact responding to these criticisms and improving its offerings.

But how fast? How soon can we expect VoIP to represent a widely deployed challenge to the entrenched switched-circuit model? Is there any way to know?

At Networking Beyond the Enterprise, Gartner took up the VoIP gauntlet, and, undaunted by the complexity of the issues cited above, predicted this:

Through 2004, successful applications will bridge circuit-switched voice and packet networks instead of forcing VoIP deployment (0.7 probability).

There we have all the elements: a technology statement, a date/deadline, and a probability assignment. It's a complete prediction, bundled up and ready to take to your network planning and budgeting sessions.

But will you?

How accurate do you think the predictions of technology pundits really are, and how much weight do you place on them in making real-world decisions?

Take our latest searchNetworking poll and let us know.

Have more to say about Gartner or about voice over IP? Tell us about it in our discussion forum.

Dig Deeper on Network Infrastructure

Start the conversation

Send me notifications when other members comment.

Please create a username to comment.