[registrars] Comments on Final Report of the Transfer Task Force on the WLS proposal

[Bruce Tonkin <Bruce.Tonkin@melbourneit.com.au>]

Hello All,

I have submitted the following comments on the Transfer Task Force
recommendations on the WLS proposal.

They are fairly long, but I have submitted them to the list in the hope it
may spark some discussion.  I am sure many of you will disagree with me.

I would like to see a formal vote taken by the constituency on the task
force report, so that the names council reps can vote according to the
wishes of whole constituency.  The vote should either accept or reject the
current recommendations of the task force.

If there is no vote by the constituency, I will abstain from voting on this
issue at the next Names Council.  I am happy to provide my proxy to either
Ken Stubbs or Philip Grabensee if they choose to vote for or against the
proposal.  I just feel personally that I do not know the wishes of the
constituency.

Regards,
Bruce Tonkin


-----Original Message-----
From: Bruce Tonkin 
Sent: Wednesday, July 17, 2002 9:05 PM
To: 'comments-deletes@dnso.org'
Subject: Comments on Final Report of the Transfer Task Force on the WLS
proposal


Hello,

Thanks you for the opportunity to comment on the Final report of the
Transfer Task Force.

My comments are on behalf of Melbourne IT, and do not represent the views of
the registrar constituency.

Recommendation 1: To deny the WLS
=================================

I disagree with this recommendation.
I believe that ICANN needs to be very careful not to stifle innovation at
the registry level.  Any new service will have an affect on the market.  

I agree with the following comments in the Task Force's powerpoint
presentation:
"Competition should always be viewed as to the effect on the eventual
consumer - this is the framework of consideration the TF has taken".

"While individual registrars are acknowledged to have existing vested
interests in the status quo - maintaining today's competition is NOT about
protecting particular businesses, rather it is retaining an open market".

I disagree with the conclusions of the next slide.

The present market of registrars compete against each other to register
names using the "add" domain command.  This competition can occur at the
time when a name is first created (most prominent when a new registry comes
into business), and when a name first becomes available again for
registration (e.g after the name is deleted from the registry).  This
competition is meant to be on the basis of equal access to connections to
the registry so that each registrar competes against each other fairly.
Recently there has been a degree of market collusion where some registrars
team together to combine their connections to the registry to gain an
advantage in registering certain domains over other registrars, and there
has also been instances where companies have become accredited registrars
for the sole purpose of trading in the names that they register for their
own use.

I do not believe that the addition of the WLS service will significantly
impact the competition in using the "add" command.  There will continue to
be a range of business models in the use of this command.

The WLS service in fact will have similar characteristics to the core domain
name service.
There will be competition to add a WLS entry at the time when a WLS is first
created (we could see a similar effect here to when a new gtld domain name
registry starts operation), and also when a WLS is removed from a name (this
will be similar to the current competition for deleted names).  A range of
business models (including most of the existing models for the core domain
name service) will also exist for WLS.  For example, different methods to
determine the optimum time to place a WLS on a name to maximise the chance
of obtaining the name in a deletion (based on knowledge of whether the
domain name is likely to be deleted).

I believe that WLS will not damage competition amongst registrars in
servicing their customers, and it may in fact create more competitive
opportunities in terms of new business models to take the best advantage of
WLS as a registry service.

In terms of consumer acceptance of WLS, this will be a matter for the market
to determine.  I personally think it is an easy service to explain to
consumers.  I think most of the issues surrounding WLS are in fact related
to the processes for deletions.  Consumers do not want their names to be
handed over to the holder of a WLS subscription accidentally.  Provided a
consumer definitely decides not to renew a domain name, there should be no
problem with an approach to take out a first option on the name when it
becomes available.

My main personal advice, is that I don't think there is a strong enough case
put forward from a regulatory point of view to deny the service.  I do not
see how the consumer will be adversely affected from any changes in the
competitive marketplace as a result of introducing the WLS.   It will create
a dangerous precedent that may either limit registry innovation, or provide
ammunition for organisations that would like to see the end of ICANN.  It
could create an example of where ICANN has exceeded its authority, for no
strong gain for the consumer.


Recommendation 1A
==================

"The ICANN board move with all haste to implement and actively enforce the
proposed Redemptions Grace Period for Deleted Names policy and practice."

I strongly agree with this recommendation.


Recommendation 1B
=================

"The ICANN Board reject Verisign's request to amend its agreement to enable
it to introduce its proposed WLS."

I strongly disagree with this recommendation.  See comments under 1 above.

Recommendation 1C
=================

"The ICANN Board reject Verisign's request to trial the WLS for 12 months."

I find Verisign's use of the term "trial" interesting.  A trial implies that
there is something measurable at the end of the period to decide on whether
to proceed.  I suspect that Verisign is talking about a normal commercial
"market trial".  If the service makes money it will continue, if it loses
money it will cease.  This is normal business practice, and is irrelevant to
ICANN.

Another use of the term "trial" could be for ICANN to trial the introduction
of this new service and measure the effect on competition in the domain name
market.  In this sense I can support a trial from an ICANN point of view.
The ICANN Board may set a condition that if the introduction of WLS reduces
the competition in the domain name market (measured in terms of number of
registrars, or concentration of market power by registrars).

So I reject the recommendation as it stands.  I have no interest in
Verisign's market trial.
I would advise the Board however to consider a measurement for the level of
competition before the introduction of the WLS, and after 12 months, and use
this as a condition for allowing WLS from ICANN's point of view to continue.
I expect that this is unnecessary for this service, but it may be an
approach for conditional approval of future registry services.


Recommendation II A
===================

"The introduction of WLS is dependent on the implementation and proven (for
not less than three months) practice envisaged in the proposed Redemption
Grace Period for Deleted Names policy and practice and the establishment of
a standard deletion period."

I agree with this recommendation.  I think if the problem of deletes is
solved first to the satisfaction of consumers, the resistance to WLS will be
minimal from a consumer perspective.

The Redemption Grace Period is one step in this solution.

A standard deletes process is still missing, and I would advise that ICANN
consider setting up a similar if not the same group as for the Redemption
Grace Period, to propose a solution to this problem.


Recommendation IIB
==================

"Verisign has proposed an interim Grace Period. The TF recommends that any
interim Grace Period have all the characteristics and conditions of the
Redemption Grace Period now in implementation." 

I strongly agree with this recommendation.


Recommendation IIC
==================

"Several Constituencies remain concerned that a standard deletion period be
established and implemented. Some TF members believe that this could be
considered separately from WLS. "

I recommend that the standard deletion period be established before the
introduction of the WLS.  There is a significant potential for gaming of the
system.  A registrar may hold onto a domain name without deleting it, until
they can sell a WLS subscription for the name, or if the WLS already exists
on the name, they may hold onto a domain name to sell via a secondary market
or even to the holder of the WLS at a higher price.  The registrar must not
be able to hold onto a name beyond the normal deletion grace period, unless
the original registrant has requested that the name be renewed.  The
registry should consider an auto-delete process to ensure uniformity.  


Recommendation IID
==================

"The WLS include a requirement that notice be provided by the Registry
(through the registrar) to the existing registrant of a domain name when a
WLS option is taken out against that registrant's domain name."

I don't think this is necessary to issue a notice to a registrant.   It is
more likely to be confusing to the registrant.   However I do believe that a
registrant should be able to find out, via a publicly available service such
as WHOIS, whether a WLS exists on the domain name, and the name of the
registrar that placed the WLS on the name.  I don't believe for privacy
reasons that the name and address details of the WLS holder need to be made
public.

The simplest solution would merely be a flag to indicate publicly that a WLS
exists on the name.

Note that a registrar will always be able to find out for a registrant if a
WLS exists, by doing a check or attempting to add a WLS entry.


Recommendation IIE
==================

"The WLS include a requirement for full transparency as to who has placed a
WLS option on a domain name and the registrar that actions the option."

Comments as for IID.  I don't think for privacy reasons that you need to
display the details of an individual or company that has placed a WLS on the
name.  It would be sufficient to identify the name of the registrar.  This
ensures that any gaming of the system by some registrars will become more
obvious.


Recommendation IIF
==================

"Based on the above two points (notice and transparency), the price for the
WLS be set at the same amount as the current registry fee for a registration
- the cost of the WLS function being no more, an probably less, than a
registration - plus any additional costs to "notice and transparency', based
on Verisign's provision of such validating information on such costs to the
Board/Staff."

This is a tricky area.  

With regard to the arguments in the powerpoint presentation, I agree that:
"Registrants are captive to the domain of their registration - the switching
cost is usually too high to contemplate changing domains"
Thus the price to renew a ".com" name is definitely akin to a monopoly
service.  The existing ".com" registrant has no choice.

With regard to WLS, the argument is more complex.
For a person without a domain name, the choices are:
- create a new .com name
- put a WLS order on a .com. name
- choose a name in another gtld or cctld namespace
At this level there is clearly choice and competition.  WLS could be subject
to market forces.

For a person with a domain name, the question is:
- will the existing ".com" registrant feel they must purchase a WLS on their
name for insurance purposes (ie to protect their name)?
If the answer to this question is YES, then I think WLS is a monopoly
service and should be subject to price control.  I would hope that the
problems of deletes and the introduction of the redemption period, would
reduce the need for an existing registrant to take out a WLS on their own
name.

On balance, I think that WLS should not be subject to price control provided
that the majority of registrants feel no need to place a WLS on their own
name.

If the Board thinks otherwise, then price control may be necessary.
I see two approaches to price control:
(a) based on cost plus a reasonable margin
(b) based on the price of a similar service

Given that Verisign has shown no inclination to provide cost details to
ICANN or anyone else, I expect that (a) will be difficult to pursue.

In regards to (b), the difficulty comes down to making an assessment of the
capital costs  and marginal costs, of WLS compared to a normal registration
service.  If WLS had a market demand of the level of a domain name
registration in .biz, .info, etc - then this would help define the price
level.  The problem for ICANN will be in determining in advance what the
level of market demand there will be for WLS.

If possible, it will be better if ICANN does not need to enter into price
control approaches for registry services.  The aim should be to create an
environment (e.g via fixing the deletes issues in advance) that negates the
need to provide price control.

Where a new service is introduced where price control is necessary, it maybe
better to create an environment where competitive tendering is used to
determine an appropriate price for the service.  This was effectively the
approach used for determining the registry prices for domain names in .biz,
.info etc.  For example WLS could be operated by a third party to Verisign
if necessary.


I hope my comments will be of some assistance to the task force.  I urge the
task force to consider the precedents that may be set in this decision
making process.

Regards,
Bruce Tonkin