[ NNSquad ] Re: Metrics of Service

[Barry Gold <bgold@matrix-consultants.com>]

James S. Huggins wrote:
> I don't mind reasonable metrics of service.
>
> What I mind is their approach.
>
> If more than "X" gigabytes is an issue then, more than "X" should cost more.
> And less than "X" should cost less. 
>
> Yes, some limited step functions might make sense ... maybe ... but the
> proposed step seems too much to me.
>
> The Comcast approach seems wrong. People who use less don't get a reduced
> price. People who use more can't. 
>
> For me, if you want variable usage pricing, then go there. 
>
> Comcast seems to be trying to have the best of both worlds.

Actually, it looks to me like Comcast is having the _worst_ of both worlds.

I wrote earlier that corporations are just as capable of insanity as 
people.  This is an excellent example.

Comcast proposes to give heavy users 1 warning and then "goodbye".  Not, 
"You can continue your heavy use if you pay more," or "We'll slow down 
your traffic to let other people get a fair chance at the net."  Just, 
"you exceeded our limit, now you're not a customer any more."

If current predictions are right (and I _strongly_ suspect they are), 
then those caps will either need to increase over the next few years, or 
Comcast is going to start losing customers.

Remember, it isn't your _median_ customer that counts.  It's where that 
"long tail" starts eating into your customer base.  Cut off the top 
0.1%, and it won't have much effect on your bottom line.  Cut off the 
top 1%, well... yeah, we can lose the "most expensive" customers.  Cut 
off the top 10% and, "what happened to our growth?"

That's what's going to happen to Comcast (and Qwest and others) if they 
stick with this policy and don't raise their caps to match increasing 
usage.  People will start switching to DSL, where you pay for bandwidth 
and can use that bandwidth 24/7/360 if you want to.  That will happen 
especially in dense urban areas (where DSL reliability is at its peak, 
but also where provisioning cable service is cheapest per household). 
And the cablecos will end up with the least dense suburban and exurban 
regions, where long distances increase the cost of creating and 
maintaining their cable farm, and hence profitability is lower.

Yes, some people in those sub- and exurbs will lose out for lack of an 
alternative to cable.  But think of the opportunity for the next 
generation's Brett Glass to make money competing with these dinosaurs...

 [ Note that in some areas outside the U.S., DSL is also already
   capped.  And here in the U.S., AT&T -- which until *very*
   recently was publicly saying that they saw no need for caps
   "since their networks didn't have the same limitations as cable"
   -- suddenly is talking about caps also -- and they can't blame
   DOCSIS.  So other factors are at work.  I assert that one of
   these is likely the understandable (from the ISP profit center
   standpoint) desire to promote ISP-generated content (e.g.
   U-verse video) over that of external Internet video services.
   But given the public lack of hard data about ISP deployments
   (which ISPs frequently tag as "proprietary") we're all really in
   the dark when it comes to the details.  And the big ISPs like it
   that way.

      -- Lauren Weinstein
         NNSquad Moderator ]