Response to the Expressions of Interest new gTLD model

[Kieren McCarthy <kierenmccarthy@xxxxxxxxx>]

In general, I think the community-supplied solution of allowing for
potential new gTLD applicants to put forward “expressions of interest”
before the Applicant Guidebook has been finalised is a useful compromise.



However, the mechanism outlined in the model published for public comment
contains a number of flaws that may end up creating more problems further
down the line.



If the model is going to work effectively as a measure of interest in new
Internet extensions, it needs to flexible and nimble. As it currently stands
however, the model risks introducing bureaucratic strain and further delay.



In particular:



* The model appears to be an effort to reduce the risk of developing rules
for new Internet extensions to as near zero as possible. We can all
understand the desire to reduce risk but in a situation where the core goal
is to increase competition in a market, this is not only an unrealistic goal
but may end up undermining what the model is trying to achieve. ICANN staff
and Board need to become more conversant with effective risk management
techniques and write them into the process rather than try to devise ways to
avoid risk.



* ICANN needs to recognise that there is a far greater community of Internet
users out there that will have a strong interest in new gTLDs but are
currently unaware of these preparatory steps. As such, making it compulsory
for someone to file an Expression of Interest at this stage is an invitation
for further delay and tension in the whole process. The logic of making it
compulsory for people to sign up to a process before that process’ own rules
have been finalized is also questionable. Expressions of Interest should not
be compulsory for the first round of new gTLDs.



* The application fee of $55,000 is high and will only work if expressions
of interest are compulsory for first-round consideration. However, this
combination of early compulsory application and high fee is effectively a
self-selection exercise within the existing ICANN community. ICANN has a
duty to look beyond the few thousand individuals, companies and
organisations it frequently interacts with and serve the broader Internet.
The application fee should be reduced to a figure that will not trigger
formal approval processes within organisations so that the expressions of
interest model itself does not contribute to further delay and tension. A
figure of $10,000 to $25,000 would be more effective.



* ICANN should supply a number of benefits to those that provide Expressions
of Interest, rather than use such expressions solely as a cut-off to the
process. For example, regular applicant-only update sessions by staff; a
promise that questions asked by applicants will be answered directly;
advance notice of papers and/or meetings; exclusive access to the
application software and algorithms, and so on. If ICANN were to supply a
high level of customer service to those that submit a formal expression of
interest (and fee), it may overcome most potential applicants’ reticence to
making their plans known as well as avoid the problems inherent in a
compulsory model.



* The details of the “communications campaign” are dangerously sparse
considering its vital importance in making the Expressions of Interest model
legitimate and effective. The timetable outlined in the model foresees Board
approval of the campaign in February, giving far too little time to develop
it effectively. Even with a fully equipped communications team (and there is
currently an open position for vice president of communications), this would
be a tall order. There are also no metrics attached to the campaign to
ascertain whether it has been sufficiently successful to allow the
Expressions of Interest model to proceed effectively. Much greater attention
needs to be paid to the communications campaign.



* The model misses a vital opportunity to gather useful data that could be
made anonymous, analysed and published to give everyone a better overall
perspective of the new gTLD process. For example, asking applicants to say
how many registrations they expect for the first three years; how much they
expect to spend on registration systems, registry systems, marketing, and so
on; how they perceive their applied-for gTLD (mass market, niche, community,
etc); what their greatest concerns are with the process (uncertainty; cost;
name conflict); what they see as the greatest opportunities that the process
represents. And so on. All this information could be compiled – removing any
identifiable material - and published, giving everyone a better idea of what
the nascent market for new Internet extensions looks like.





Those are my broad observations. Below is some counter-argument to
assumptions included in production of the current model in an effort to
demonstrate why the suggested changes above are not only achievable but also
an improvement on the current approach.







Compulsory nature of the model



The paper represents the Expressions of Interest model as being either a
“firm” (compulsory) or “soft” (non-compulsory) approach. The paper argues
for a firm, i.e. compulsory process.



The “soft” option is represented as “containing minimal data” and having
only the  “perception of movement”. The approach, it is argued, “will
provide no certain data and will consume additional resources (time and
money) on the part of both ICANN and potential participants, with no real
value added.” Finally the non-compulsory approach “could also result in a
perceived bias toward EOI participants in the eventual application process.”



By contrast, the firm, compulsory model is represented as providing “data to

resolve, with certainty, demand, operational, and possibly other issues”. It
“will be the most time‐efficient and allow the most flexibility”.



It strikes me that this is a very one-sided approach to deciding the
fundamental issue of whether this model is made a pre-requisite to a process
that has been going for years. I would be much more comfortable seeing a
list of pros and cons and then a reasoned explanation as to why ICANN staff
felt one approach was preferable to the others.



As it stands, the “firm” approach has several significant cons, not least of
which is that this is a brand new approach that was introduced at the
last-minute at the Seoul meeting and yet it will have a controlling impact
over the entire gTLD process.



ICANN appears not to have learnt the lesson of the overarching issues
introduction after the first version of the Application Guidebook: that it
cannot and should not assume that important actors are following the
organisation’s every move.



The firm approach will certainly gather data but in so doing it attempts to
bring down a shutter on applications a year before the full details of the
process are agreed or published. It is not hard to see how large numbers of
organizations across the world will be frustrated and infuriated when they
hear about the possibility to apply for new Internet extensions only to be
told that they had to pre-apply a year before.



I would strongly question the assumption that the firm approach provides
“the most flexibility” when it clearly offers the least flexibility through
its compulsory nature. A soft approach would be far more flexible. And no
model at all is supremely flexible as it has no rules at all.



The suggestion that the firm approach is also the “most time-efficient” is
also questionable. If expressions of interests are compulsory, it seems all
too likely that the remaining issues to be decided will be held off until
all the information is available. In effect, ICANN would be introducing a
whole new part to the process rather than running it in parallel to ongoing
discussions. In that respect, it is in fact the *least* time-efficient.





The soft approach



The assumptions applied to the soft approach also do not stand up to
scrutiny. This approach would provide “minimal data” the paper argues. But
that assumption itself makes the assumption that the majority of those that
already have plans for new gTLDs would not provide an expression of interest
unless it was compulsory. There is no evidence to back up that underlying
assumption and, in fact, it was the potential gTLD applicants themselves
that put forward the idea of expressions of interest to ICANN.



It seems strange to assume that a process requested by applicants themselves
would be ignored by those same people, especially when the approach has
already been seen to work in other areas of ICANN’s work (registrar
transfers; IDN Fast Track; and others).



The soft approach, the paper argues, would add “no real value”. Again, this
presumes that potential applicants will not provide an expression of
interest unless compelled to do so.



And lastly, a non-compulsory approach “could also result in a perceived bias
toward EOI participants in the eventual application process,” the paper
argues. It is easy to turn this argument around the other way and point out
that there would be advantages for those that send in expressions of
interest – and then clearly list what those advantages are. That way, ICANN
provides a useful additional service, rather than attempting to push the
global Internet community into a pre-formed box.



The advantages to the soft approach are also overlooked. First, and most
significantly, ICANN as an organisation has learned time and again that any
effort to impose strict rules at short notice and without the full buy-in of
the broader community invariably end in failure (or years of temporary
fixes).



The soft approach has the very distinct advantage that it allows for the
accumulation of data - which should help in working through the remaining
overarching issues and in developing the application process – without
imposing a rigid rule.



The important thing to focus on would be to provide incentives to potential
applicants to apply. And that is where a public comment period could prove
extremely useful – asking potential applicants themselves what would be most
useful to them. This is an approach that worked for example in asking
registrars about changes to the Registrar Accredition Agreement. There is no
reason to believe it wouldn’t work for gTLD applicants.



Broadly, I think that ICANN’s staff has been seduced by the procedural
advantages that a compulsory model would have, but without looking at it in
the broader context of the new gTLD process.



It is understandable that having fixed and known quantities would make
things easier from a planning perspective but it ignores the wider reality
that the gTLD process is not something that affects only those that closely
follow ICANN.



Of course, ICANN may still wish to continue down this path, but if it does
so, it would be well advised to do a much better, and less biased, job of
explaining the rationale behind that choice.







Risks and risk management



There is a section in the paper that covers “Risks considered” and then runs
through a number of different scenarios in a similar way to companies
outlines risks to their business models in an Initial Public Offering.



It strikes me that the paper – and the model itself - is confused between
identifying risks and risk management. The paper highlights possible issues
that will arise as a result of the process and then seeks to promote future
solutions for each.



What this approach fundamentally fails to recognise however is that there
will always be uncontrollable risks in this process, since it is an
expansion of an unknown market. If ICANN seeks to introduce a different
procedure for every aspect of risk that appears, or attempts to produce a
new rule for each new issue, it will end up with a heavily delayed and
hugely over-complex process that stifles the very competition it seeks to
introduce.



Rather than provide hundreds of rules, a more efficient and effective
solution is to introduce systems for managing risk. In essence, ICANN needs
to recognise it cannot know the unknowable.



The October 2009 edition of Harvard Business Review advertised itself as the
“Spotlight on Risk” issue. In one of its articles on risk titled “The Six
Mistakes Executives Make in Risk Management”, the authors noted in their
introduction:



“Risk management, we believe, should be about lessening the impact of what
we don’t understand – not a futile attempt to develop sophisticated
techniques and stories that perpetuate our illusions of being able to
understand and predict the social and economic environment.”



The compulsory model proposed in this paper, and the risk-identification
rather than risk-management approach taken, could very easily be seen as one
of these futile attempts to “understand and predict the social and economic
environment”.



The new gTLD process lives within an unpredictable social and economic
environment and ICANN would do well to acknowledge that fact and design
models accordingly.





I hope these thoughts and reflections are helpful going forward. I would
like to highlight that the very nature of this public comment process means
that my response is pre-determined to take a critical approach. As such, I
would like to acknowledge the hard work and constant efforts made to find
workable compromise solutions to a complex and difficult task both by the
staff and the community.



Thanks,







Kieren McCarthy