Local cost related to running a TLD

[Jorge Amodio]

Agreed,

but I really don't have any solid estimates, coming out with numbers
out of thin air does not provide what Avri and the NCSG is looking
for. The biggest offset I can think of about running a gTLD in a
developing country is cost of labor.

Cheap servers won't do it, I find hard to believe that anybody trying
to apply to become a registry and gTLD operator will pass the
technical evaluation without a solid understanding and plan to deploy
reliable and up to the task infrastructure, it's not just a pair of
linux boxes running bind.

I agree with Avri that probably the best approach is to put together a
working group and do some study and number crunching and perhaps put
together a paper that helps to develop some solid foundation.

I should start by reviewing the DAGv3 and other documentation to
clearly identify, first one time costs like the initial application
fee $185K (that by the way is not the grand total if you happen to
have a string that will require further evaluation and/or dispute
resolution, that may add another $120K to the bill) and then recurring
fees based on the standard registry contract terms, for example a
registry will have to pay ICANN and annual fixed fee of U$S 6,250 plus
a variable fee based on the number of transactions performed by the
registry.

We can guesstimate the equipment and services costs for what kind of
infrastructure is required to operate the name servers and back end
systems for a given size registry.

IMHO the best way to approach this and come out with reasonable and
arguable numbers is to put together a draft business plan, an
associated budget and financial projections.

There are many things that are being left in the current discussion
about what I refer to run a "real" operation, for example ICANN
requires that the applicant and gTLD operator is a corporation,
institution or organization of some type, no individuals or sole
ownership, you must take in account insurance, a legal reserve fund
and many other little things that affect the bottom line.

Have to go now, I don't have the time right now to do the entire
exercise by myself but I may have some bandwidth available to
contribute in a working group.

Regards
Jorge


On Thu, Mar 18, 2010 at 3:09 PM, David Cake <dave at difference.com.au> wrote:
> At 11:31 PM -0400 17/3/10, Avri Doria wrote:
>>
>> Hi,
>>
>> The issue is that when one complains about this cost, the reply is that
>> the 185,000 fee is nothing when compared to the cost of running a registry.
>> So when one responds that in developing areas, the cost of running a new
>> registry, especially a small  one with a slow ramp up will be much lower, we
>> are asked to prove it.  And to date have not had the figures to do so.
>> I..e when I make the argument that it cost less to do a proper regisrty in
>> Cambodia or Kenya, I am just handwaving, because I do not have the financial
>> facts.
>
>        Even in expensive developed countries,  even taking into account
> Jorges estimates of what it takes to run a registry, the estimates of what
> it takes to run a registry are all basically nonsense.
>        Why?
>        Because the cost of running a registry operation may be large, but
> what is the marginal cost of running another domain for someone who already
> runs a registry, and already has the hardware, support and contracts in
> place?
>        The reality is the 'oh, it takes a lot of money to run a registry'
> line is one of those lines that gets waved around a lot by people who
> already have a profitable business - but the reality is that for those
> folks, the question of how much is costs to run a registry is very much a
> flexible question. But given their relatively low marginal cost for running
> an additional registry, it is in their interest to keep the barrier to entry
> to the business quite high.
>        Regards
>                David
>