Comment From Carlos Raul Gutierrez

[Amy Bivins <amy.bivins@xxxxxxxxx>]

This comment was received by ICANN on 1/22/2016 and is being submitted to the 
public comment forum on behalf of Carlos Raul Gutierrez.

--------------------------------------------------------------------------------------------------------------------------------------------------------------

To Amy Bivins
amy.bivins@xxxxxxx
 From Carlos Raúl Gutiérrez
crg@xxxxxxxxxx
Comments to gTLD Marketplace Health Index Proposal:

1. Framework(s). As far as the CCT work is concerned, the IAG has suggested in 
Appendix 2 of
its report a matrix format, in which 7 categories of data (I. Technical 
metrics, II. Registration
Data, III. Law enforcement/domain abuse, etc. etc.) would be examined on the 
axes of choice,
competition and trust. I think it is worthwhile to use the same framework or 
similar approach for
the health index so as to avoid doble work or overlaps. So far the health index 
seems to be gourd
only along the second axis (competition, choice and trust). Neither is there a 
rational for why the
CCT metrics follow 65 indices and the health index another different and 
smaller group of indexes.
Is the Health index a summary of the CCT effort? Why is it better?

2. Demand side economics. The CCT review is waiting for the second phase of the 
Nielsen surveys
and the Economic study. They both are scheduled to be ready in the first half 
of 2016. So I
don´t know how an overlap with the proposed health index can be avoided. This 
is relevant since
the proposed health index has both a) a strong supply side focus, b) a lack of 
the key relevant
measure for competition, which are market structure and prices (both wholesale 
and retail). It is
very difficult to call a market healthy and diverse if the registries and 
registrars cannot cover its
operating costs (or are making huge profits). As far as supply side and the 
diversity is concerned,
the first KPI Robust and Competitive gTLD Marketplace, the first proposed KPI 
a) Number of
countries with at least one ICANN-accredited registrar, could be as important 
as the number
of registrants from each particular country that choose a (new) gTLD instead of 
a ccTLD.
This information should be readily available from whois without compromising 
the identity of
the registrant (City and Country).


3. Market dynamics. A case for analysis the demand side is particularly 
relevant if the new registration
is just the same person or company moving from an old one, or even more 
important from
the economic perspective, is moving from a ccTLD. Furthermore chances are high 
that this new
registry is being done by the same registrant that is keeping the two old ones 
(gTLD and ccTLD).
While your metrics would look at the number of new domain names, we have no way 
to check if
the same registrant is registering more names without making full new webpages. 
Since the privacy
of the registrants is still an open issue, I suggest that statical samples have 
to be analysed on
a regular basis to check for the level of activity of the domain names (either 
parked with an email
address or redirection to an active website, versus active websites).

4. It is my PROPOSAL that in the segment "Robust and Competitive gTLD 
Marketplace" is necessary
to have a pretty exact idea of the STRUCTURE AND PRICES prevalent in the DNS 
market,
by reporting the overall (new, old and cc) number of gTLDs that have been 
registered, not
only to follow up overall growth in the numbers of gTLDs over the years, but 
also because they
represent the economic basis for budgeting the income side of ICANNs (as well 
as financial basis
of the income side for Registries and Registrars).

5. Furthermore it is necessary to clearly understand how are the being used and 
their level of activity
over specific time periods. In that way we can understand how many registers 
are done for defensive
purposes only, as opposed to real activity. For general statistical 
publications it should be
enough, after the total number of gTLDs, to report the relative level of 
activity over a specific
period of time, as opposed to pure parking of names:
. # of names registered under each gTLD
. % of names with active addresses only under each gTLD
. % of names with active address and webpages under each gTLD


6. How can this be best achieved? Trough regular statistical analysis of 
representative random samples
of gTLDs provided by the Registries, keeping the privacy of the individual 
registrants as
well as numbers per Registrar, but reporting their geographic distribution 
(country of origin of
the registrants) ICANN could efficiently collect the data required to measure 
these additional
concepts. In that way the sample analysis at the beginning and end of the 
period by ICANN staff,
could also help ccTLDs with a point of reference to compare numbers and level 
of activity of
gTLD versus their own ccTLD market. It should be left to each ccTLD to 
volunteer with a similar
sample of their database for purposes of ICANNs reporting.


Friday, 22January, 2016

Amy E. Bivins
Registrar Policy Services Manager
Internet Corporation for Assigned Names and Numbers (ICANN)
amy.bivins@xxxxxxxxx


One World. One Internet.

Direct: +1 (202) 249-7551
Fax:  +1 (202) 789-0104
801 17th Street NW, Suite 400
Washington, DC 20006