new gtlds

George Kirikos: “How can *objectively* argue that sunrise program has been a success, when the data says otherwise?”

George Kirikos has been arguing about the New gTLD sunrise success with the ICANN community and the GNSO Working Group.

In my opinion his views are 100% correct and his suggested procedure could save all a lot of time and money, and stop the sunrise exploitation by some trademark sharks. Sunrise has been a waste of time and money even for the large companies that is supposed to be protecting.

Here are 3 of the comments that George Kirikos left:

“Hi Claudio,

Changing the topic to one that is more appropriate, the numbers speak for themselves, namely a greater than 99% reduction in sunrise utilization rates per TLD.

Trying to reframe the issue, to look at things in aggregate can’t hide that startling truth. That’s the same kind of “bad math” that suggests that the new gTLD program is a “success” because there’s 20+ million registrations in aggregate (despite the fact that’s spread over many, many new TLDs, and that nearly all, if not all of them, individually are considered weak, with stats propped up with high counts of sub-$1 registrations, or even domains that are stuffed into the accounts of registrants of other TLDs (e.g. .xxx with NSI, or .kiwi).

If you wanted to start looking at things in aggregate, consider a company like Microsoft, registrant of approximately 80,000 domain names according to DomainTools:

or Google with 20,000:

and so on.

i.e. the fraction of domains they’re acquiring via sunrise is still a tiny proportion of their overall holdings. All that would happen, when sunrises are eliminated, is that they would shift their spending to the landrush period instead. Easy-peasy.

Why not aggregate the number of unique users of the sunrise, even when across TLD? If Microsoft or Apple or Dell or Google register 2 or 3 or 10 marks each in sunrise, across most/all TLDs, that starts to look like a very narrow group of stakeholders who would be affected by its elimination (and the extent that they are affected is small, given they would just shift their demand to landrush), compared to a situation where it’s many different sunrise users in different TLDs.

Suppose I own an ice cream shop had sells a cone generating $X in sales per day, and I decide to launch a new flavour of that ice cream cone. That new flavour generated less than 1% of $X (i.e. a 99%+ reduction). I would call that a failure. I would call that a disaster. I would not be “declaring success” and launching 1000 more flavours each generating less than 1% of $X.

So, the new talking point is the 99%+ reduction in sunrise utilization rates per TLD (no need to even mention 130 domains anymore, when one can simply talk about a 99%+ decline).

As for DPML, that’s essentially an attempt to privately replicate and monetize the poorly received “Globally Protected Marks List” proposal.
Can’t see how folks like paying for domains they can’t use (i.e. it’s a scaremongering purchase, buy this or someone else might get it, and
once again using TMs to jump the queue), and also preventing good faith purchasers who might desire it. That’s even worse than sunrise,
in my opinion.


George Kirikos


How can *objectively* argue that sunrise program has been a success (I can see how one can argue the political angle, but we’re not here to
argue politics, we’re here to look at facts and evidence), when the data says otherwise? We know that on average a mere 130 registrations
occur per TLD in sunrises, which means that the benefits are small, and one must compare those with the costs.

Let’s try to put 130 per TLD in perspective. I was looking for stats on the .eu sunrises, and perhaps others have better sources, but according to a Google search for “.eu sunrise period registrations total” one of the hits was to the book “Information Technology Law” (Diane Rowland et al), it stated there were 346,218 applications filed for 245,908 different domain names. Those numbers don’t provide a citation, but they seem consistent with a Eurid report:

Click to access quarterly_2006_q2.pdf

which states (page 9) that the validation agent had validated (while sunrise validations were still in progress) 140,000 applications.

I couldn’t find the .info stats (although I did find that there were more than 15,000 *challenges* to sunrise registrations, see, so the aggregate total must have been much higher ),  but I did find the .asia ones, where there were 30,780 sunrise domain applications:

.Asia Sunrise Completed with Over 30,000 Domain Applications provided their own stats:

.xxx sunrise: 80,000 blocks (2011)
.co sunrise: 11,000 domains (2010)
.asia sunrise: 32,000 domains (2008)
.mobi sunrise: 15,000 domains (2006)
.eu sunrise: 140,000 domains (2006)
.biz IP claims: 80,000 (2001)

If new gTLDs had anywhere close to those sunrise statistics, it would be clear there were substantial benefits, and there would be no argument from me. If that was the data, anyone would be laughed at for trying to seriously suggest the benefits were small, given the large
uptake. I would be on the other side, arguing that the benefits were obviously high.

But, that *isn’t* the data. We know that the numbers are very small. So, let’s face the facts, the sunrises were a complete disaster in terms of uptake. That speaks directly to the “benefits” part of the equation.

And we know what the costs were, I won’t go into them again.

So, again, I ask anyone to objectively attempt to argue that the new gTLD’s sunrise policy was a success, given those disastrous figures compared to .eu, .xxx, .co, .asia, etc. (perhaps someone else can add the complete .info stats with citations, so that we have a full picture).

Instead, the only “basis” for perpetuation of the failed policy is “let’s not rock the boat”, or “GAC members might get upset” essentially, rather than calling a spade a spade — it’s been an obvious failure.

Let’s do our job, look at the evidence objectively and fairly, and use evidence-based policymaking.


George Kirikos


On Thu, Aug 10, 2017 at 11:21 PM, Greg Shatan <> wrote:

I don’t see the math that created your “talking point” of a “99%+ reduction
in sunrise.” Can you show your work please?

The post at:

showed numerous sunrise statistics, ranging from 15,000 on the low end
for .mobi (.co was slightly lower, although that’s a ccTLD, not a TLD
that ICANN is involved with in any way), 32,000 for .asia, 80,000 for
.biz/.xxx, and who knows what it was for .info?

Even taking the lowest of those (15,000) as the base, 130 (average
new gTLD sunrise from The Analysis Group report) divided by 15,000 =
0.0087 = 0.87%, which is less than 1%, i.e. a 99%+ reduction. Of
course, if one chose a higher base (.asia, .xxx, .biz, .etc.), or an
average of those other sunrises, the reduction is even greater than if
one had used the lowest sunrise (from .mobi).

As for your other statement:

We can’t expect Sunrise registrations to outperform the New gTLD Program generally.”

While the new gTLD program has been a disaster, it hasn’t been an
underperformance of 99%+ of expectations (perhaps more like 80% to 90%
underperformance). Thus, while it’s obvious that both have been
failures, sunrise usage is an even greater failure than new gTLDs
overall. So, even on that relative scale, the sunrise period should be

Since I know you’ll ask “George, why do you say there’s been an 80% or
90% underpeformance for new gTLDs?” let me answer that now to save
time. I’ll use as my reference (besides the obvious general
observations of most informed observers) ICANN’s own stats:

where the numbers came in at just 18% of ICANN’s original 2014
expectations. For the math-challenged, 100% – 18% = 82% as the level
of underperformance.


George Kirikos


About Konstantinos Zournas

I studied Computer Engineering and Computer Science in London, UK and I am now living in Athens, Greece. I went online in 1995, started coding in 1996 and began buying domain names and creating websites in 2000. I started the blog in 2012.


  1. Just to be clear, in the first email to that mailing list quoted above, I had made a typo in the section:

    “stats propped up with high counts of sub-$1 registrations, or even domains that are stuffed into the accounts of registrants of other TLDs (e.g. .xxx with NSI, or .kiwi).”

    “.xxx” should have been “.xyz”, of course. I corrected that in a followup comment a few minutes later:

    One could read the entire archives of the latest month at:

    although this PDP has been going on for over a year now:

    It’s never too late to join, see the instructions at:

    Expect 2 or 3 hours per week of time spent (90 minute phone calls once a week, plus time spent reading the mailing list, etc.).

    Other folks already participating can be seen at:

    As you can tell, it’s dominated by trademark lawyers trying to maintain their unfair advantages over other domain name interests. Many in the trademark camp want to even extend those advantages, giving trademark owners even more benefits, making an unbalanced system even more biased in favour of trademark interests.

    • Most of these lawyers don’t want to protect their customers as they say. They only want to charge them more for “protection”.

      • I agree that there’s definitely a conflict of interest that exists. If the sunrise period and/or the trademark clearinghouse was eliminated, some lawyers would have fewer services to sell. Instead, their brand clients would have to compete in landrush or GA just like everyone else on a level playing field.

        If you keep reading that entire mailing list, there’s great pushback from the trademark interests to any idea that would upset the status quo (status quo which favours them!), and things have gotten pretty heated at times. More voices are need to help argue the case for the elimination of sunrises, the TMCH, as well as changes to the UDRP/URS that would create a less biased system.

      • I am not talking about sunrise only. There is also an industry of useless UDRPs. (Some are even requesting cancelation of domains, etc.)
        By the way the UDRP needs to have some strict rules to protect owners: e.g. tm registration after domain registration or owner change? Denied automatically. Domain registered 10+ years ago and no complaint? Then all complaints denied automatically. Etc.

  2. Yes, changes to the UDRP are long overdue. This is the first time that the policy will be reviewed since its inception in the 1990s! And, we’re not there yet (dealing with sunrises, TMCH, etc. first). UDRP might be next year.

    A “statute of limitations” to prevent a UDRP from being filed after a certain number of years is one of the things I’d push for — although, 10+ years would be too long. Maybe 2 years, as it is for .cn domains and their dispute policy. And of course the example you gave about the TM being created after the domain name is “low hanging fruit” that should be dealt with too. There should also be other changes, too many to list here, but including proper due process for domain name registrants, including longer time periods to reply (perhaps based on domai name age). It’s unfair that a complainant can take years to craft a complaint, but a domain name owner (who might be on vacation, etc.) has only a few weeks to respond.

    The UDRP providers need to also be held more accountable, through contracts with ICANN and better oversight to get rid of some of the rogue panelists. We know NAF was forced out of consumer credit arbitration after the Minnesota government filed a lawsuit:

    What measures are in place to protect domain name registrants in UDRP cases, compared to those consumer credit arbitrations?

    • 30 to 45 days at least should be the time for a respondent to reply.

      Also emailing the postmaster@domain.etc is a joke. The providers should do what they can to make sure the respodent has received the complaint. Why not also call and/or send a simple letter. I used to get the complete complaint by courier. Now they send nothing but an email.

      • In a real court case, one has to “serve” the lawsuit on someone. That’s when the clock starts ticking, when they’ve been properly served. The UDRP doesn’t do that.

        There are so many things that need fixing, but unless domain name owners show up and fight for their rights, the other side might block those necessary changes. Phil Corwin of the ICA is a co-chair of the PDP, and there are a few others like myself who have the interests of domain name registrants at heart, but just take a look at the participants list, and the responses on the mailing list to see that we’re outnumbered.

    • Re: “statute of limitations”

      You can’t have any limitations in case of cybersquatting, but in the case of generic dictionary words, this is a must.

      And if a mark holder doesn’t apply/register during sunrise or even a couple of years later, shouldn’t it be assumed they don’t wish to protect their mark in the said tld?

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.