Hi Martin,
"Just thought some may be interested in these numbers."
Thanks for that post. I'm finding traffic and revenue metrics in this new paradigm rather interesting to analyze.
Earlier, someone (I think it was Raleigh B.) made reference to Jeff Pulver's projection that VoIP may account for some $60 Billion in market share in some segment by some date. Beware... How many times are the same calls being counted in that projection? There is no consistent answer to that question because it depends on the modality and the configuration of the substituting call arrangement... whether they are end-to-end, or PC to Phone, PC-to-PC, etc. In the latter case (PC-to-PC) there is virtually no charge to speak of that is noteworthy, so I wouldn't get too excited over the prospects of that group from a 'carrier services' revenue generating perspective. From a hardware sales standpoint? That's another story.
Assuming $60 Billion "is" the current valuation of the traffic to be commandeered, at a range of 10cpm to 25cpm today, say, what would it be worth at 3 or 4 cpm, or lower, in the future. The long haul rate structures will likely be entirely turned inside out by that time. Even free in some instances, to large corporates who have very large data comm (VPN) appetites.
In your example, you note:
"I will divide by two as always since we need an inny and outy to make a call."
The innies and outies are required at 'both' ends of the call path, aren't they? See what I mean? It follows then that the 96K revenue would be amortized across four ports, not two. Hence, each end would generate 48k gross annually. Correct me if I'm mistaken about this, since there may be something I'm not taking into account here.
Let's take it a bit further and examine some other costs. Shall we start by accounting for colocation costs (which happens to be the FCC-recognized spelling of the term - which I had to learn the hard way about six years ago when we prepared handouts for a colocation conference presentation in Washington, and we had it spelled: collocation!) and facilities costs, billing services & customer care desks, maintenance, T-line$ to the local carriers for LEC hand offs, and rentals of ISP ports, peering charges, and so on?
There is a lot of money to be made in this game, no doubt. Tons. Critical mass is important, however, and where it cannot easily be reached, partnering, outsourcing and other means of leveraging are the name of the game. In short, I think that if we start to post gross revenue figures, we then ought to examine the overhead charges as well.
Such a breakdown could lead to an analysis concerning which is the more profitable business to be in at this time and going into the future: Hardware sales or carrier services? They take different business models, each based on their own sets of independent assumptions. Interesting, and thanks again for bringing it up.
BTW, your port averages are fairly consistent with what we are finding to be the case in some CLEC studies done recently, with the First T-1 in a mixed voice-usage (bus and res) trunk group accounting for between 260,000 to 280,000 minutes of use (MOUs) per month, and subsequent T-1s in the same trunk group generating about 15% to 20% less, or roughly 225,000 MOUs. Keeping in mind, of course, that there are usually 22 to 24 DS0 traffic bearing channels in a T-1.
FWIW, Frank |
