Talking Telecoms: We Know It's the Bandwidth, Genius
March 15, 1999 6:47 AM EST
By Casey Freymuth
Special to worldlyinvestor.com
My column normally appears every Wednesday. However, after reading David Baker's article dealing with the idea of a bandwidth glut, I felt compelled to respond to allegations in the article that indicate that subscribers to the theory of a bandwidth glut are "stupid."
I can tell you with confidence that subscribers to the theory may be many things, but one thing they're not is stupid. Let's set the record straight.
As a telecom veteran, most of my background has been in facilities-based settings. When I first heard of the glut prediction, I disagreed immediately, along with everyone else.
Bandwidth demand was a bottomless pit. But it wasn't until a client, which happened to be headed by one of the country's leading network gurus, wanted me to write a business case that incorporated the idea that I seriously looked into it.
Having done so, I firmly believe that no intelligent person could seriously look into this argument and not come to the conclusion that a glut is, at the very least, a possibility.
But this is the problem with most people who discount the glut theory outright - they disagree conceptually and don't take the time to look in to the foundation or specifics of the theory itself.
It's a shoot-from-the-hip response - "I have to wait forever to connect to my Internet provider during peak times. The number of users continues to grow by leaps and bounds. Video demand is huge. No way a glut is coming." And that is the extent of the investigation.
In this case, Mr. Baker never picked up the telephone to ask me what it was that caused me to believe that a forthcoming glut is a real possibility, and I can tell from his arguments that he did not invest one minute of research into the basic foundation of the argument.
Instead, he made one of the common assumptions among glut naysayers - that subscribers to the theory don't understand price-driven demand or, more specifically, the role technology plays in this dynamic.
In fairness to Mr. Baker with respect to originality, however, he did make an argument that I had not yet heard: That telecom professionals who subscribe to the theory don't realize that video traffic consumes more bandwidth than a byte of data here and there.
This concept is so basic that I doubt you could locate an entry-level employee within the industry that does not know this. The argument is so bizarre that I don't quite know how to respond to it other than to give it an "A" for originality and a resounding "WE KNOW" on behalf of industry folk, laypeople and knuckleheads everywhere.
Now that I've described what hasn't been investigated, and that the fact that video is a bandwidth hog is not a revelation to anyone, let me tell you what all the on-the-record naysayers would have found if they'd taken the time to research the other side of the coin.
The "old" rumor of a glut, which existed for some time, was based on continued infrastructure deployment and modest improvements in technology.
The Internet as we know it today came along and shot the opportunity for a glut under those conditions right out of the water. When charted, in fact, you can see how the model was altered when the Internet took off. Hence, the dearth we're in right now.
It is this experience, which failed to account for exponential growth in demand, that has caused some individuals - particularly those with investments in bandwidth - to dismiss the current theory without taking the time to review the research methodology behind the current prediction.
Unlike the previous model, the current idea behind a bandwidth glut takes into account geometric growth in demand - technology-powered, supply-driven demand.
More importantly, the new model is not based on modest technological improvements. It's based on groundbreaking technologies working in the lab today that, when released into the market (some are already in deployment, in fact), will allow supply to dwarf increases in demand.
These technologies are Dense Wave Division Multiplexing, Optical Switching and Soliton transmissions. Together, they allow significantly more traffic to be packed over existing light streams while amplifying lightwaves and collapsing multiple layers of current technology into a single layer (simplification of network architectures).
The net result is a massive increase in capacity - enough to drive 100-fold growth. Although demand is growing at a staggering pace, it is nowhere near the pace at which supply will increase as these technologies are commercially deployed.
The significance of this development with respect to a glut is that, unlike time-consuming network expansion activities, such as digging up the ground to lay fiber, these technologies can be rapidly deployed into existing networks. This will create a lag while demand, spurred even further by this massive increase in supply, catches up. Thus, the glut theory.
Betting against it is betting against telecommunications technologies that have been proven in the lab and are waiting to hit the market. And the track record of such bets - from those against voice-over-Internet technologies to those against CDMA - is not impressive.
What this oversupply means to investors is that value shifts from bandwidth in and of itself to application of that bandwidth, or service plays. Will the bandwidth glut be eaten up? Yes, but at a much lower price per unit.
Moreover, proponents of the glut theory do not discount bandwidth's value today. The investment portfolios of many proponents are nearly identical to naysayers.
And here's the kicker, which will probably drop the jaws of all the naysayers who haven't bothered to research the issue: The glut model projects that bandwidth will continue to comprise more than 80% of the revenues of service players for the foreseeable future.
Given that bandwidth pricing is plummeting at a staggering pace and will fall even more when the predicted lag in demand creates the glut, this represents enormous increases in bandwidth consumption as bandwidth-consuming applications are developed and deployed.
The message here is that there's been much more thought given to the equation than naysayers from the old school would like to believe.
Whether or not the glut will occur is highly contested, and there are arguments against it that are much more meaningful than those raised in Mr. Baker's article.
Frankly, when there are reasonable questions about whether or not routers can handle the sudden increase in traffic, individuals making arguments about accounting for bandwidth's consumption in video applications haven't scratched the surface of the issue.
If you have questions regarding the glut, I urge you to research both sides of the argument for yourself, then make your own conclusions about the issue.
How sure am I that this will happen? On a scale of one to five, with one being "no way," five being "no doubt about it" and three being "it's possible but could go either way," I would select four.
There's no doubt that the capacity is coming on line but there is a chance that issues like the routers example will delay it enough for demand to keep pace. We'll just have to wait and see.
In the mean time, how do you account for either outcome? Bank on companies making service plays - it all starts with the customer anyway.
Casey Freymuth is president of Group IV, Inc., a Phoenix-based strategy consulting firm to high tech industries that publishes "The Telecommunications Service Provider: How Much is it Worth?" - a comprehensive report on values and drivers in the communications industry. His Talking Telecoms column usually appears every Wednesday.
|
