Frank, Re your position, " ... each of the service providers platform models which have been publicized (that I have heard about, in any event) will find themselves becoming "tapped out" where their abilities to "scale" is concerned. They all are susceptible to a place where their heads hit the ceiling." Probably true, with the significant exception of a true PON architecture FTTH topology. Here, the problem is a bit different -- it's not an inherent capability ceiling, it's the transition strategy, the motivation to invest in network capabilities far beyond current demand levels.
Let's suppose, for discussion sake, that XYZ Telecom (I won't mention their ticker symbol) has started shipping an amazing new passive optical network that exploits passive DWDM and very high powered lasers permitting unprecedented splits. The system can deliver hundreds of broadcast liner video digital streams, maybe several hundred megahertz of analog NTSC for settop-less TV connection, 4 to 6 lines of traditional telephony, and, say, 100 - 200 Mb/s of data, shall we say symmetric data, per customer. Not my personal specifcation for a residential PON, but clearly, beyond most concepts today of what demand for residential bandwidth may be in 10 to 15 years. If you think the future demand is even higher, substitute your specification -- that is not the issue here. Your engineers feel that with a concerted effort, up to 2 million lines a year could be built in your 20 million residential line ILEC. These guys have in the past missed several of their estimates of construction capability, but the way.
Finally, the real golden fleece -- the system capital cost is, say, $200 per home passed, plus say, $500 per home served. If you were an ILEC CEO and your engineers brought such a system to your attention, what is the appropriate response? Not in general, but in terms of a real action plan.
You have a ADSL rollout in progress right now. It is needed to check the competition from cable modems, and you are right now deciding how best to expand your ADSL coverage with multi-billion dollar, multi-year alternatives. Right now, you are close to a technical decision, but you are waiting to pull the trigger until you have a better grasp of the regulatory regime in which you will spend the next few billion dollars to broaden the coverage of broadband service. Basically, though, there is no turning back on the ADSL expansion commitment, it is a tactical necessity, along with the capital cost it entails. Once that technology is deployed, you have multi-megabit asymmetric data capability along with multi-line voice into every home within the footprint of your upgrade activity. Your engineers say it will take fours to complete. You made a mental estimate of 6 years and crossed your fingers when you heard their guess. You also wonder if they will soon start pitching an extra few billion dollars to convert deep-fiber-fed ADSL to a bit deeper fiber fed VDSL. Who knows? All you are sure of is that ADSL, or ADSL upgraded to VDSL, suffers the capability ceiling exactly as Frank C. has noted. At least you have this PON alternative. What to do with it?
For "only" $40 million a year, we can start seeding PON facilities that pass 2 million households a year. [remember, this is a hypothetical, fantasy cost, please.] Of course, to use any of that investment, say, to get to 10% penetration each year, it will cost another $10 million -- what a bargain. Does the system look economic with 10% penetration? Well, no, I need 25%. So let's "plan" 25% penetration each year. Still only $25 million a year. Great!
Now, where do I start building? The marketing department wants us to start with the best demographic areas, and make sure we cover every home passed in the region before we move on. How else can they market their service suite effectively? How else can they even think of offering video services? Well, the engineers want to knock off all of the new growth and most of the rehabilatation each year first, and spread the rest where they can achieve the lowest cost build -- generally where there is high residential density and aerial plant. Unfortuanately, that puts their build activity predominately in the inner city for several years -- and the marketing department warns that their revenue commitment is null and void if they get primarily that demographic. The engineers counter with the observation that the marketing department's key early target already has or soon will have good coverage with ADSL capability; has marketing figured out yet how they will exhaust that architecture with HDR service offerings? Hardly looks like it, the ADSL sales are pretty anemic.
You are the CEO, so you want to take the long look (not quarter to quarter, indulge me). Shall we optimize for combined broadband access capability in 5 years, while we also push the tactical response to competitive broadband offerings? Why not a 10 year horizon for optimum integrated access capability? It would take at least that long to do all the PON construction. But if the ADSL upgrade and expansion is going to be largely complete by 2002, about the time the PON construction juggernaut is really getting traction, then what am I getting for my PON expenditures? No one in the company has come forward with a credible residential service requirement exceeding a few megabits of data rate, except linear video. And I'm not sure I really want to be entering the linear video business in 2003 or 2004; isn't the paradigm changing away from liner video? My people aren't sure, and so neither am I.
Maybe the PON deployment is a great bargain, until I consider the extent of the enhancement in real service delivery capability, the delta from near ubiquitous ADSL capabilities. I can't come up with many incremental revenue streams that the PON will service. Maybe if I talk myself into massive operational savings, I can justify the deployment even without much new revenue. Ah, but I remember all those other times we tried that logic. Those considerable savings that look so logical in the business case seem so elusive when we start trying to trip operations budgets. Any chance of that fiasco happening again here? I hate to risk it again. The analysts won't be kind to us if we let ourselves be duped yet again by that trap.
Well, where shall we start building PON? Maybe if we wait a few years until the rationale for deployment is a bit more compelling, the costs will be even lower. We will certainly have a better handle on the level and location of broadband demand in a few more years. And we hate to get stuck with a network dinosaur; what if the PON we build begins to look obsolescent even before we get much built or any really fully operational? I know the engineers told me we can always change out the electronics and the glass will always be good, but those electronics are pretty expensive, too.
I am almost sorry we already have all the capabilities we derive from our existing deeply penetrated loop fiber and ADSL technology. Those remote DSLAM and NGDLC plans just make the analysis worse -- such a slender delta to to after. If only we were starting from scratch, we could exploit all of the capabilities of the PON, not just enjoy raising the capabilities ceiling for some time in the future.
Maybe we should build the PONs, but outside our service area. Hey, my engineers tell me all I need is CLEC status to get access to rights of way. That is pretty easy, even though it has to be done in every state where I might want to build. And we are obviously a lot smarter than our ILEC brethern, why, it would be like taking candy from a baby to build PONs and eat their lunch. Yeah, maybe that is what we should do. We still have to decide if we should build PONs in region, but hey, why not wait for a bit more regulatory certainty, and see what kind of results we get out of region? Shouldn't take more than a few years to see what's what. And in the mean time, we have our ADSL expansion as the tactial response we need to in-region competition. Sounds like a plan.
Does any of this little hypothetical business development tale sound plausible? What do you think it might take for this hypothetical ILEC to "raise the ceiling" on residential access services capabilities? Does it look like a vendor cost issue? But is that were most of the attention seem to be focused to catch the trigger? Hard to say for sure. It's so difficult to make real planning decisions from hypothetical considerations. Even more difficult than simple arithmetic for some CEOs. |
