Vonage Hits ISP Resistance

Steve Gibbard scg at gibbard.org
Thu Mar 31 19:28:57 UTC 2005

On Thu, 31 Mar 2005, Steve Sobol wrote:

> Bill Nash <billn at billn.net> wrote:
>> I find this to be entertaining, since as a VOIP consumer, I'm reimbursing
>> my ISP for the cost of the traffic as part of my monthly tithe.
> Not proportional to the potential cost of providing the service.
> I have no idea what my cable company pays for their bandwidth, but I am
> certain it's more than the $40 per month I pay for my 3Mbps down/256 Mbps
> up... and I am able to actually *get* 3Mbps on many occasions, and I average
> between 1 and 2 (on HTTP/FTP transfers, fwiw).
> Yes, I know the connectivity cost is shared between several thousand 
> customers in this area, but what happens if large numbers of customers 
> start using VOiP on a regular basis?

I'm replying to stuff on NANOG too much.  I should stop...

That said:

This is really a matter of adjusting business models as the costs of 
providing various types of services change.

As others have pointed out, all end user telecommunications networks that 
I'm aware of are based on some amount of oversubscription.  This is not 
necessarily in the sense that pipes are full and users are getting poor 
service, but that it's assumed the users won't all use it at once.

When doing flat rate, unlimited use, billing, the goal is generally to set 
the price such that each user covers the cost of serving the average user. 
Assuming the prices aren't excessively high, the ISP or phone company (or 
all you can eat restaurant, for that matter) probably loses money on its 
heaviest users, does quite well on the users who barely use the service at 
all, and in the end it all balances out.

But if the average user starts using the service more, the cost model 
breaks.  At that point the ISP or phone company needs to either find a way 
to lower the cost of providing the service, discourage people from using 
it, or raise prices.  This isn't unique to VOIP or PtP; it's a general 
issue with flat rate business models.

Another approach is to bill based on usage, in which case if you're not 
losing money on every bit, you've got an incentive to encourage your 
customers to use your product more.  Even there, you've got 
oversubscription issues to contend with:  If you're billing your customers 
per minute, or per megabyte, you still need to hope that they're not all 
going to use the same minute, or all send their megabyte at the same time. 
If they are, the model breaks and needs to be fixed somehow.

What I generally see as I look at this industry around the world, is that 
pricing models adapt to fit local conditions, and continue to adapt as 
those conditions change.

In the US, broadband providers tend to do flat rate billing because it's 
easy to administer and it works.  Colo providers tend to do usage based 
billing, because the spread between the cost of hosting somebody whose 
website gets occasional hits versus the cost of somebody who is constantly 
saturating a 100 Mb/s pipe is just too big.

Elsewhere, things sometimes work differently.  Suresh was saying earlier 
that Korea Telecom is switching to usage based billing for broadband, 
presumably because they hope that will be a better fit for their market 
than flat rate.  In Nepal, New Zealand, and Western Australia, all places 
where long distance capacity is very expensive, I've seen pricing 
differentiation between local and long distance Internet use.  In Nepal 
and Western Australia, it's been flat rate billing for local use, and per 
bit billing for long distance, while in New Zealand there's at least talk 
of providing New Zealand only connectivity.  In the US that sounds 
horrendously complicated, but where the wholesale monthly cost of 
international bandwidth is $5,000 per Mb/s and the monthly cost of handing 
traffic off to other local ISPs at the local exchange point is around $50 
per Mb/s (Kathmandu), it makes a lot of sense.

So, I don't know if VOIP use will measurably change the costs of broadband 
providers in the US.  If it's only a few users, I suspect it won't.  If 
it's a lot of users, and there's big market demand for it, I suspect the 
ISPs that survive will find a working billing model.


More information about the NANOG mailing list