Robert,
I'd like to experiment here a bit, and expand on a few other cost elements.
Very often the loop costs, along with the capital outlays for the DSLAMs
and remote modems, are cited as the main drivers behind the overall costs
associated with DSL services. Add to these the normal costs of doing
business, the cost of money, infrastructure, IP administration, overhead,
etc., and then add in a margin for profit. But is that all? There's more.
There are the edge network costs. And the fat pipes that connect to the core
to be considered, as well. In the past, the per-user costs associated with
these upstreams has been a fraction of what they will be going forward with
DSLs and cable modems.
If you price out a T3 internet access arrangement from CERFnet or
any of the other Tier One ISPs or backbone providers, you may very well
find an astounding cost associated with that one port. Greater than $30,000
to 40,000 per month, depending on IP admin charges, etc., if my memory
is correct.
Add the loop cost for the T3 which could be anywhere from 2k to 5k to
this figure, and it could total $40,000/month or more.
[NOTE: All figures used here are order of magnitude for discussion purposes,
only, and I don't make any claims as to their accuracy.]
A T1 port would cost somewhat (but not proportionately) less. Maybe
$1,800/month for a T1 presence on a router to the upstream. Again, then you
must add the loop costs of $500 to $1,000/mo = $2,500 to $3,000/mo.
Descending, there are the lesser forms of access, such as aggregated dial-ups
and dedicated DSLs. The allocation of upstream bandwidth for DSLs has
to be taken into account here, too. But in these instances, historically, these costs
have been hidden in the overall cost to the ISP of some 19.95/month, or whatever.
The more attention the provider pays to this upstream matter, the more expensive
it becomes.
In the case of DSLs and Cable Modems, the upstream allocations could be
more than 10, 50 to 100 times more than they were for their 33.3kb/s or 56k
predecessors. Not surprisingly, these costs must be covered by charges to the
user, and they will be considerably more than those for the dialups.
The flip side to this is... if there is less attention paid to the availability of
upstream bandwidth to a pool of DSL users, once they start to crank up
their usage, then those DSL users may find themselves sucking in a
vacuum. This is a problem that the smaller mom and pops face with users
who merely use V.34 modems at 33 kb/s. They present a choke point
before their surfers even get out of the starter's gate.
Thus far, I've seen only a few tariff-like criteria or SLA information on any of
these offerings with regard to QoS as they would relate to overall
throughput expectations. What kinds of guarantees does a user have in
connection with having a fair shot at a window to bandwidth once they've
satisfactorily achieved connections on a 384kb/s DSL access pipe? Beets
me.
This part of the edge (between the DSLAM router and the upstream
Router) is a huge contention domain. If one service provider makes more
abundant resources available than another provider at that point, and by
the same token provides more bandwidth to the upstream tier, or to the
NAP directly, then they will charge more than the provider who has not,
accordingly. They must charge more, in order to recoup their costs and
make a profit. Having said all of this...
Assume that the DSL provider does not have to worry about the
upstream provisions towards the core, because their partner, such as
an AOL, has their own routers and backbone provisions that they can
leverage against. This, IMO, would make a very substantial difference in
the incremental costs associated with a DSL service to that population of
users who subscribe to the partner's services (in this case AOL).
Perhaps enough of a difference to make the disparity seem entirely off
base, unless these factors are properly taken into account.
And lets not forget the scale of AOL's operations. No one at this point
can compare to them in terms of size or reach. Likewise, no one else
could expect the same level of leverage as AOL at the negotiating table.
If you add these two factors together - AOL's own backbone and router resources,
plus their overall leverage - then I don't think that anyone should be surprised if they
are able to achieve some very low price points with the xLECs.
Regards, Frank_C. |
