Melbourne IT comments on the new gTLD process

["Bruce Tonkin" <Bruce.Tonkin@xxxxxxxxxxxxxxxxxx>]

Hello All,

Below are a series of comments from Melbourne IT which is based on
face-to-face discussions with many of its corporate domain name
customers in Europe, USA, and Australia.


(1) General feedback from corporate clients
===========================================

- most see that the cost for protecting their brands will increase

- given the high costs of applying for a new gTLD - many will take a
"wait-and-see" approach

- those that feel they need to register for "brand protection" reasons,
don't see the need for more new gTLDs except in limited circumstances
(e.g for under-served language or cultural groups)

- a few see new opportunities - especially around new product launches
and possibly new business initiatives

- most clients feel that ICANN has taken reasonable steps at the top
level to protect the namespace, but have serious concerns about the need
to spend significant resources protecting their core brands and
misspellings of their brands at the second level within new gTLDs.


(1A) Distinguish between "open" and "closed" gTLDs
==================================================

Melbourne IT recommends that ICANN distinguish between "open" and
"closed" gTLDs

"Open" or "Community" gTLD - is open for registration at the second
level by third parties that are not affiliated with the applicant.
Third-party registrants take full responsibility for use of the domain
name at the second level.

"Closed" gTLD - only allows registrations at the second level by the
applicant, and the applicant takes full responsibility for use of the
domain name at the second level.   This type of gTLD could be used by
corporates wanting to create their own hierarchy of domain names under a
new top level domain.


(2) Cost
=========

- Corporates are concerned about the high application fee of $185,000
and the $75,000 annual fee, when they are registered names purely for
brand protection purposes

- there is also a high cost for those wanting to apply for their brand
in multiple languages or for multiple brands in the same industry sector
(e.g brands of cars)

Recommendations:
===============

- for "closed" gTLD - offer a lower annual fee for when a trademarked
name has been approved, but not yet requested to be in the DNS - e.g
small holding fee $5000 per year

- allow a lower application fee for a "closed" gTLD where the applicant
has a trademark on the gTLD string, as costs for reviewing and risk is
likely lower.   Recommend $90,000.

- for "closed" gTLDs - offer a variable annual fee based on the number
of approved applicants.  E.g $75,000 per year if less that 100 new
gTLDs, $50,000 per year (100-200), $40,000 (200-500), $30,000 (greater
than 500).   This is because ICANN will obtain economies of scale with
more gTLDs, and the costs of managing compliance for "closed" gTLDs will
be lower (as there are no third party registrations at the second
level).


- for "closed" gTLDs - for applicants that apply for multiple names that
are trademarks for the same legal entity (e.g product names), provide a
10% discount on application and annual fees for each additional name, up
to a maximum of 50% of all fees.    This can be justified as ICANN staff
can have a single interface with the applicant for multiple names, and
the cost of reviewing multiple applications will be lower as most of the
content of each application will be the same except for the string being
applied for.

- allow a discount for multiple applications for names that have the
same meaning in different languages (e.g travel in Arabic or Chinese).
This can be justified on the basis that the technical and business
aspects of the proposals would be the same, and only the strings need to
be checked.





(3) Technical criteria
======================

- it is likely that many proposals will use an existing gTLD or ccTLD
registry operator

Recommendation
==============

- ICANN should consider accrediting organisations to act as back-end
registry operators to allow applicants to select one of these operators.
This would reduce ICANN costs of carrying out technical evaluations on
an application by application basis and possibly lower the application
fees for applicants.  In terms of potential legal risks, ICANN could
make clear that it has reviewed the operators as passing a technical
standard based on their application, but the applicants should perform
their own due diligence on such operators.  This should be a similar
risk as for the process ICANN uses to accredit registrars.


(4) Rights protection
=====================

- one of the main concerns of brand holders is the potential costs of
brand protection at the second level of new gTLDs.

Recommendation
==============

- ICANN fund a central database of authenticated trademarks that
registry operators will be able to use in their sunrise processes - this
will save costs for brand owners from having to pay the costs of
separate authentication at each registry operator

- use the same principles of legal protection designed for the top level
at the second level

-- ie second level registrations are posted for evaluation prior to
going live at the second level

-- existing brand holders have the right to raise an objection under
either confusingly similar objection to existing name (applicant would
need to show rights in a name that could be considering confusingly
similar) or legal rights objection - if the applicant challenges the
objection they pay a filing fee and an upfront dispute resolution fee

-- the winner gets a refund

-- removes the incentive to make money from advertising while a UDRP
dispute is underway, and also balances the costs to ensure that the
applicant at the second level pays the dispute fees if they lose, should
they proceed to the dispute resolution process



(5) Open versus community based applications
============================================

- there is uncertainty over the prospect of a community based
application getting precedence over an open application.  This will tend
to encourage applicants to propose they represent a community to try to
gain a commercial advantage at the time of contention.

- there is potential for gaming

Recommendation
==============
- if there is contention amongst applicants for the same name and one of
the applicants claims they represent a community - allow any of the
other applicants to provide information on community support before
using a comparative evaluation process to resolve contention.


(6) Options for resolving contention amongst parties
======================================================

- the current rule states that the only option for reaching agreement
amongst multiple parties is for some of the parties to withdraw and
allow one of the parties to proceed

- this does not allow the forming of new joint ventures to "share" a new
gTLD string and limits the options for parties to reach a mutual
agreement

Recommendation
==============
- allow some flexibility for multiple applicants for the same string to
agree on new legal structures to support a particular string


(5) Timeframe
=============

- there is uncertainty over when a second round would occur, and this
may artificially put pressure on brand holders to register in the first
round

Recommendation
==============
- announce the second round as starting 6 months later than the date of
publication of the strings applied for in the first round.  

- allows potential applicants to review the type of names being applied
for, and provides sufficient time to prepare an application



(6) single organisation - registry/registrar separation
=======================================================

- Melbourne IT supports allowing a single organisation operating a
"closed" gTLD to operate both the registry and registrar functions

Recommendation
==============
- to avoid gaming - limit this situation to "closed" gTLDs, when the
single organisation is the "registrant" for all second level domains -
ie second level domains are not able to be licensed to third parties


(7) hybrid model - proposal to allow a registry to own a registrar as
long as not for its gTLD
========================================================================
==============

- there are no real clear benefits in this proposal for new start-up
registries and there is room for abuse, where a registry can have access
to domain name check availability searches and contact information to
give it an unfair advantage in other gTLDs.

- the rules for registry-registrar separation are also not clear when a
registrar provides back-end registry services for an applicant (and
hence may not be designated as the "registry operator")

- small start-up registries will want the option to operate their own
registrars to help with the direct marketing and sales of names to get
the TLD started


Recommendation
==============
- where a registry offers registrations to third parties, allow the
registry to operate its own registrar as well as allowing other ICANN
registrars to offer names on the same commercial terms up to a cap of
50,000 names in total.  Once that cap is hit the registry would no
longer be able to operate registrar services for new registrations but
could support their existing base.   A registry would not be able to
operate a registrar for other gTLDs in this model.