Re: PAB The Green Paper and competing registries

Kent Crispin (kent@songbird.com)
Sun, 22 Feb 1998 15:14:34 -0800


On Sun, Feb 22, 1998 at 10:00:41PM +0000, Jim Dixon wrote:
> On Sun, 22 Feb 1998, Kent Crispin wrote:
>
> > > Competition between TLDs is alive and very well. There is weak
> > > competition between the nTLDs, and reasonably strong competition
> > > between the individual nTLDs and .COM/NET/ORG. This competition is
> > > good for all concerned.
> >
> > It appears to be the case, however, that registry prices in general,
> > and on average, have been pegged near or above the $50/year that
> > InterNIC sets for a relatively long time, despite the fact that their
> > "production costs" are certainly far lower than $50, and they could
> > easily set prices lower. So this "alive and well" competition is
> > actually moving prices *up* from "production costs".
>
> There is a very large world outside North America.
>
> Inside North America * there is no competition *. Everything is
> run by NSI.
>
> Outside North America there IS competition.
>
> So look for your data outside North America.

I was. I was referring to the ccTLDs, just as you were. Last I heard
the average price for a ccTLD, outside North America, was quite high,
and the policies ranged from excellent to nightmare. The last time
this came up someone from Netnames provided some data...perhaps Ivan
or Antony could chime in?

> In the UK registry prices have fallen steadily since Nominet was
> set up.

One shouldn't generalize from one example, didn't I just hear you say?

> > In fact, I think an unbiased examination would reveal that the
> > "registry industry" (to use Jim Flemings term) is so complexly
> > regulated that to say that competition is "alive and well" is absurd.
>
> Especially if you rely upon no evidence at all.

Jim, why do you waste *your* time with gratuitous little insults like
this?

> > Also, neither your example nor Javier's analysis go far enough:
> >
> > Registries actually compete on two fronts -- to gain new customers
> > (market share), and to keep the customers they already have (revenue).
> > We may see examples of competition in the former, but experience with
> > the latter is limited. In fact, I am perfectly willing to concede that
> > registries can compete in the search for new customers, and, in fact,
> > I expect that new registries will compete briskly for market share.
> >
> > But all the non-competitive practices at issue concern
> > re-registrations.
>
> They don't. Out here where we have practical evidence, this is
> not true.

The non-competitive practices at issu involve things like "lockin".
You have practical evidence that lockin concerns times other than at
re-registration? That doesn't even make sense. Or are you just
leaping to contradict a statement without actually reading it? Or
are you saying that new registries don't compete for market share,
which I just conceded?

> Real world issues include price and policy. In my example, which
> is NOT theoretical, people migrated to .UK despite higher prices
> because of better policies.

One can make precisely the same argument about the .us vs
.com/.net/.org situation in the US -- people avoid the .us domains
because they have incoherent policies, even though the registrations
are cheaper. (BTW, I understand from previous discussions that you
are not familiar with this use of the word "incoherent" -- it doesn't
mean "rambling in speech or reasoning" -- it means "not in phase" --
it's a usage I picked up from being around laser physicists, I
guess.)

As I said, I agree that there can be competition in *new*
registrations, and I think the Nominet phenomenon you describe is
almost completely explained by that -- in fact, as you described it,
the primary problem with the name committee was precisely getting
a *new* name approved.

> > In a "competitive registry" environment the obvious
> > evolution is to a pricing structure where the longer you have been a
> > customer, the more you pay. Initial registrations will be nearly free
> > -- we would be inundated with enticing "new domain owner specials!"
> > while long term prices would steadily rise as registries take
> > advantage of lock in effects. Of course, these changes will not be
> > drastic -- they will be gradual, and spread as each registry sees what
> > other registries can get away with.
> >
> > Note that Nominet's competition with .com is a fool's paradise right
> > now. If NSI maintains it's monopoly, and is freed of regulation,
> > .com registrations will mushroom in the UK.
>
> On what vast experience of the UK market do you base this judgement?
> I know I only have experience and know what I am talking about, but
> our real world customers switched to .UK _despite_ the InterNIC and
> .COM having a price advantage.

Are you claiming that most of your your customers *changed* from a
.com to a .uk domain -- that would be news! Or, more likely, are you
saying that *new* registrations are going into .uk (for which there
are all kinds of explanations possible, as I have mentioned.)

> > NSI can certainly afford
> > to undercut even Nominet's prices in that tiny restricted area, gain
> > substantial market share, and relegate Nominet to an unimportant
> > backwater of the domain name space.
>
> Nominet can lower its fees to registrars to the sub-10 pound range
> with no difficulty at all.

Oh -- NSI certainly could as well. In fact, I expect to see that, as
the marketing blitz's begin...

> I know theorizing is lots of fun, but have some respect for the real
> world, please. In the real world, the UK is 25% of the size of the
> US - and Europe is bigger than the US. This market is scarcely
> a "tiny restricted area".

Sorry -- I was teasing. I have more respect for the real world than
you can possibly imagine.

-- 
Kent Crispin, PAB Chair			"No reason to get excited",
kent@songbird.com			the thief he kindly spoke...
PGP fingerprint:   B1 8B 72 ED 55 21 5E 44  61 F4 58 0F 72 10 65 55
http://songbird.com/kent/pgp_key.html