RE: [gnso-vi-feb10] SRSU

[Jeff Eckhaus <eckhaus@xxxxxxxxxxxxxxx>]

Sorry , maybe I assumed that the term reseller was clear. A reseller is another 
company that buys the domains/services from a registrar and resells them. Your 
term "captive reseller" is an entity owned by the Registrar,  so it is no 
longer a reseller, it is part of the Registrar. 


________________________________________
From: Alan Greenberg [alan.greenberg@xxxxxxxxx]
Sent: Saturday, July 03, 2010 1:52 PM
To: Jeff Eckhaus; Gnso-vi-feb10@xxxxxxxxx
Subject: RE: [gnso-vi-feb10] SRSU

Jeff, your description of resellers is once possibility, but there
are others. A "captive reseller" is my term for a reseller
substantially owned by a registrar or both owned by the same parent.
For an example of what I was referring to, see scenario b) in
http://forum.icann.org/lists/gnso-vi-feb10/msg02222.html.

I do not at this time favour any level of integration above the
nominal 15%, but that message proposed one possible set of
constraints (not yet fully fleshed out) that  should be mandatory if
integration is ultimately allowed.

Alan

At 03/07/2010 03:23 PM, Jeff Eckhaus wrote:

>I cannot seem to find the captive reseller example, but think it may
>be worth explaining to the group how the reseller model works so
>that there is clarity on the issue.
>
>A domain reseller is an entity such as Google, Yahoo, Intuit, or
>maybe a mom and pop hosting company that has an existing business
>but would like to sell domains as a complementary part of their
>business. They do not want to become ICANN accredited and do not
>want to build a registrar system since that is not their core
>competency. Some offer free domains as a loss leader for their main
>business line and that is their business choice and allows consumers
>a choice on how they acquire their domain name.
>Any domain that a reseller registers lists the Registrar in the
>whois record. So if XYZ hosting is a reseller of Network Solutions
>than Netsol will come up in the whois record.
>
>If in the proposed JN2 model a Registry owns Registrar X and applies
>for .TLD, then Registrar X or any of its resellers cannot offer .TLD
>through Registrar X. This is easy to monitor as I mentioned
>previously since Registrar X would show up as the registrar of
>record in the whois.
>
>Now I know everyone can think real hard and think of an edge case
>where there could be some secret agreements, but, do these
>agreements have anything to do with cross ownership? If the co-owned
>registrar and its affiliated registrars (if any) are restricted from
>offering that TLD and can be verified in the whois, wouldn't any
>harms from co-ownership be mitigated and be put into the same
>situation if there was no co-ownership?
>
>Right now it seems to me that we are willing to block out new
>competition, market entrants and possible innovation even though we
>have mitigated harms due to the minor possibility that an entity can
>possibly think of a way around the normal process. Is this the best
>decision for new TLDs? For consumers? That is what we really need to decide
>
>Have a great weekend
>
>
>Jeff Eckhaus
>
>
>
>
>________________________________________
>From: Alan Greenberg [alan.greenberg@xxxxxxxxx]
>Sent: Friday, July 02, 2010 1:54 PM
>To: Jeff Eckhaus; Gnso-vi-feb10@xxxxxxxxx
>Subject: RE: [gnso-vi-feb10] SRSU
>
>See comments embedded.  Alan
>
>At 02/07/2010 02:34 PM, Jeff Eckhaus wrote:
> >I also like the idea of exploring exceptions including the SRSU
> >model as described below, but I have a more fundamental question on
> >these exceptions.
> >
> >Why is there no question or discussion on compliance abilities with
> >regard to SRSU or other exceptions but arms start flying when other
> >types of co-ownership are brought up
>
>I thought that Eric's message timestamped 2 Jul 2010 06:15:20 -0400
>didjust raised just that issue.
>
> >?  When I look at the idea of a Registry being able to own a
> >Registrar but not be able to sell the TLD it owns it is actually
> >simple to monitor, since the Registrar and affiliates could not be
> >accredited in that TLD. If it is not accredited it cannot register
> >any names. With mandatory thick whois, the Registrar of record is
> >displayed.  All very easy to monitor.
>
>And easy to circumvent. See my earlier example of how through the use
>a captive resellers.
>
> >The SRSU model (which I said is worth exploring) has an incredible
> >number of moving parts that need to be monitored and by many
> >estimates there are expected to be over 200 .brand TLDs, yet the
> >compliance issues and harms are not brought up.
> >What is it about .brand SRSU TLDs that make it easier to monitor and
> >protect than another TLD that allows cross-ownership?
> >
> >
> >Thanks
> >
> >Jeff Eckhaus