Karl, that was an interesting, if not familiar collection of thoughts on the subject of bandwidth commoditization. While I'm not in disagreement with what you've stated, I'd like to add my obervations and my views on this subject, as it has been at the root of much discussion and debate recently.
I have more than one client in banking and brokerage who, at the upper management levels, has made corporate-wide proclamations to mid-management and line staff that bandwidth is nothing more than a commodity, calling for sourcing of the most aggressive types, beating the carriers over the proverbial head for the best price. A dangerous tactic, I advise them, if they are using price as the sole determinant, and not looking over the horizon with respect to their choice of carrier.
There is nothing that a carrier can do in their eyes to dispel these notions, they often say. Water is water, air is air, and bandwidth is bandwidth. While there is some underlying truth to this, flags go up nevertheless, immediately, when I hear this.
The bidding parties result in the shaking of heads in disbelief amongst the competing carriers followed by not just a little commiserationg over beers and shots between them, afterwards.
Perhaps these clients are correct when taking into account the "core" bandwidth availability that exists, especially on very-high- density routes between major metropolitan area centers.
But supplying and parsing [in short, administering] this b/w at the denominations needed to satisfy less-than-gargantuan flows for discrete corporate applications, however, is an altogether different story.
This should be no surprise, for it forces the customer to buy very large pipes, and leaves the splitting and routing of virtual channels to the carriers, instead of allowing the customer to order up chunks of the stuff as needed to satisfy onesy twosy applications, ad hoc. But that is by far not the only reason.
The major carriers are now doing what they euphemistically are calling "bandwidth engineering reviews" to assess demand over the next six to twelve months. QWhat was a no brainer for them six or nine months ago, suddenly is something that now requires a one to two week analysis before they can commit to availability.
They would have you believe that they are sizing demand that would result in the provisioning of a just-in-time quantity of bandwidth to meet customers' demands, in other words.
But the bandwidth almost invariably always does exist in terms of available "spectrum" in the backbone.... what they are actually assessing is the need for additional boxes that can harness this bandwidth in slices that are useful for merchandising purposes. In this respect, its a matter of packaging, shelf space arrangements, pricing, perquisite determinations and promotional considerations.
Translation: The Carriers are busy determining how many DWDM devices and ports are needed, how many SONET OC-48 and OC-12, and -3 ports are needed, how many digital cross connect frames and ports are needed, and how many router and mux channel ports are needed.
In effect, they are attempting to determine how many PORTS, in general, are needed to capture and distribute all of this bandwidth that they have been plowing into the ground. As long as the xCLEC ports on the ends of their links are wanting, they will tend to keep the cost of the intermediate sections high as well. No sense getting hopes up if you are going to be the provider of an end to end service, if you cant get timely delivery of the last mile. In some cases, as availability of the end office provisions increases with respect to port count, they will lower the overall cost of the longer pipes as well. In some ways, this is like the interstates riding the tails of the locals where pricing is concerned, and this relationship is nothing new.
This becomes only so painfully clear when the top three carriers bid aggressively to capture business, and make T3s and OC-3s available with three month lead times, only to very-often miss their appointment dates due to the last device in the chain lacking a necessary circuit pack that contains the necessary denomination's port size.
Getting back to our new age corporate managers who have been overly inspired by what they read, those who proclaim bandwidth a commodity are in many cases the very same ones who are reduced to terror tantrums when their commodities are delayed for an additional two to three months or more, because the inter-state carrier's intermediate and terminus DACS locations, or those of the hand-off ILEC's is out of steam, or when the last mile of fiber capacity has been exhausted due to arcane planning and forecasting methods, forcing the xLEC to "prematurely" [read: at a premium] pull new silica in the streets after much grief and not just a little penance, and the calling in of long-forgotten markers .
The problem of bandwidth availability lies therefore in the availability of switching-location, router-location, and end-office-location- PORTS. And to suggest that a macro-fiber availability is the way to offset this deficiency at this stage of technological preparedness, is folly. Although, if one can sustain the requisite delays I've cited above, then the bandwidth is indeed there for the taking.
The carriers are not about to split off bandwidth at the lambda level right now for distribution to customers using mirrors and prisms, when it is going to cost them ten times the amount that it would if they can deliver capacity at a fraction of the cost using established and reusable T1/T3 grooming and distribution techniques.
The problems that I am citing here are experienced daily by the nation's largest firms in each of their respective sectors, so this is not just a matter of buying clout. By inference, however, one could justly conclude that conditions are in fact much worse for those with less clout, and proportionately so, for those with much less clout.
A major factor behind these conditions, but by no means the only one, is the lack of coordination and synchronization of incentives and goodwill between the IECs and the local carriers to properly forecast and meet demand. Not to mention the fact that oversizing the end office environment could potentially adversely affect the supply-demand dynamic. Also, it would violate inventory controls, since ports are very expensive "commodities" to have lying around spare, doing nothing, waiting to be used.
The introduction of CLECs in many sectors in recent years has reduced this type of supply side restriction, to a degree, if the CLEC is permitted to operate as an independent. But look what has happened to the largest CLECs. TCG, MFS, Brookes, and others have all been gobbled up, rendering them the last mile extensions of larger IXCs.
This being the case, then, why is it that the Top-Two IXC carriers have such a difficult time fitting their chosen CLEC into a service design in a timely deployment schedule, and are forced to revert to ordering local loops from the incumbent BOCs? It once again gets back to the shortage of boxes, ports, and fiber builds, even when they own the fiber capacity into targeted locales.
Some of these problems are often overshadowed by all of the hype surrounding Moores Law and the explosion of wavelength availability in the long haul, nowadays. But they exist, undeniably, at the operating levels of these carriers, regardless of whether the IXC is a Top Three, or a QWST/ LVLT or WMB, or whether it's a tier-three reseller.
>>The key to this dilema regarding pricing and loyalty/retention is how to properly incentivize consumers with an ongoing recognized value proposition...Fortunately, there are marketing methodologies to accomplish those objectives.<<
Good points, if all other factors were equal. But the main key lies in the availability of [at the risk of being repetitive and annoying at this point] switching-, routing-, collection- and distribution- ports in hardware. Without them all of the smoke and mirrors dreamt up by marketing departments, and all of the intermediate route macaroni, falls flat on its face.
I would be interested in seeing what your recommedations for the carriers would be concerning the existence of "marketing methodologies to accomplish those objectives." |
