Amazon, Uniregistry, Verisign… here’s who won the new gTLDs lottery

Kevin Murphy, December 18, 2012, 22:32:44 (UTC), Domain Services

Amazon, Uniregistry and Verisign were among the luckiest companies competing in yesterday’s New gTLD Prioritization Draw, our preliminary analysis indicates.

ICANN spent nine and a half hours last night pulling lottery tickets from a drum in order to determine the order in which it will evaluate, negotiate and delegate new gTLD bids.

Applicants representing 1,766 applications bought tickets, a 92% turnout. Internationalized domain names were given special priority, but all other participants were treated equally.

A few hundred people — including Santa Claus, there to represent Uniregistry’s .christmas bid — showed up, with many more participating remotely.

Not many people stayed the course, however. In introductory remarks, ICANN CEO Fadi Chehade promised that the Draw would be “as boring as possible”, and it kept to his promise.

“I think it went really well,” ICANN’s new new gTLD program manager, Christine Willett, told DI today.

“I think people were really pleased and pleasantly surprised with how transparent it was,” she said. “We could have done it much faster electronically, but it wouldn’t have been as transparent.”

I’ve spent much of today drilling into the results of The Draw, using the DI PRO New gTLD Application Tracker, and here are some of my findings.

Uniregistry won the most contention sets.

Uniregistry, the portfolio applicant owned by domainer Frank Schilling, won more contention sets, in percentage terms, than other volume applicants.

This table shows the performance of the top 10 applicants (as measured by the number of contention sets they’re in).

ApplicantSetsWon%TLDs Won
Donuts1585736.08.college .tires .memorial .trading .doctor .mobile .phone .data .city .casino .school .realty .auto .coach .theater .broadway .realestate .film .eco .fyi .site .help .coupons .ski .search .salon .studio .rent .red .racing .radio .plus .photography .money .team .loans .gifts .flowers .vote .juegos .jewelry .wine .娱乐 .guide .garden .free .football .energy .discount .construction .church .charity .boats .beauty .baseball .band .audio
Google581220.69.diy .family .docs .lol .fun .pet .show .vip .talk .drive .mom .moto
Top Level Domain Holdings561628.57.style .book .home .llc .store .love .lawyer .law .dog .deals .casa .science .review .网址 .dds .country
Famous Four Media501530.00.tennis .golf .basketball .poker .run .taxi .hockey .diet .delivery .cricket .chat .bingo .cam .fit .sport
Uniregistry451737.78.inc .sale .art .tech .shopping .restaurant .property .hosting .yoga .gratis .video .furniture .fashion .cars .auction .save .gift
Amazon351028.57.buy .cloud .play .you .group .yun .wow .dev .deal .coupon
Directi28828.57.online .movie .hotel .blog .ping .website .world .click
Afilias Limited16212.50.app .mls
NU DOT CO LLC13215.38.design .web
Demand Media10330.00.rip .map .fishing

Getting the best draw number in a contention set is of course not indicative of any skill or of the quality of the applications, it just means the applicant got lucky.

Neither is it an indication of whether the applicant is likely to ultimately win their contention set; myriad other factors are in play.

There may even be some advantages to poorer draw numbers. We’ll get to that later.

Amazon is the luckiest portfolio applicant.

Amazon was the most successful applicant in the Draw of any company applying for 20 or more gTLDs, as measured by average prioritization numbers.

ApplicantAverage Priority
Amazon777
Afilias878
TLDH909
United TLD (Demand Media)910
Donuts946
Uniregistry961
Directi (Radix)994
Google (Charleston Road Registry)1,050
Famous Four Media1,222

The average for each applicant is of course affected positively by the number of IDN applications it filed, and negatively by the number of applications for which it opted out by not buying a ticket.

Amazon applied for 11 IDNs, increasing its average score, while Google did not buy tickets for 24 of its applications, substantially reducing its portfolio’s mean priority.

Likewise, Famous Four Media did not buy tickets for 12 of its applications.

Dot-brands fared less well, on average, than open gTLDs.

Single-registrant TLDs (which includes dot-brands and generic strings with single-registrant models, such as Google’s .blog application) had an average priority of 983, compared to 921 for TLDs we’ve identified as having “open” registration policies.

Verisign’s clients did better than most other registry back-ends.

Of the registry back-end providers named in more than 20 applications, China’s KNET fared best, with an average draw number of 328, according to our data. That’s to be expected of course, due to the inherent bias in the process towards IDN applications.

Of the rest, Verisign topped the list at 913 (to be expected again, given its own dozen IDN gTLD applications), followed closely by KSRegistry at 915. Minds + Machines got 930, Demand Media 942, Internet Systems Consortium 947 and Neustar 953.

OpenRegistry was unluckiest, with an average of 1,207, preceded by Google with 1,050 and GMO Registry with 1,027. CORE scored 1,000, ARI Registry Services 1,007, CentralNic 983 and Afilias 994.

Tagged: , , ,

Comments (9)

  1. Tom G says:

    Donuts scored well with uncontested TLDs in the first couple hundred, including some IDNS.

  2. Tom G says:

    killer analysis though, I can see why the DI PRO subscription could pay off.

  3. Domain Dan says:

    What’s all the fighting about? It’s like rearranging chair on the desk of the Titanic if you ask me. Clearly, all of the gtld’s will fail miserably anyways so who carez?

    Like someone is gonna really want to use .restaurant! (Sorry Frank!)

  4. Go Figure says:

    Overall, looks like the geos drew the longer straws. Not unexpected.

    • Go Figure says:

      Of the 66 geos, 30 are over 1000. In other words, about half the geos won’t see the light of day until the latter half of 2014 thereafter. Very unfortunate for this group since they have the greatest potential to succeed. No surprise given the history of this initiative.

  5. fred krueger says:

    i would like to see the reverse analysis — of those uncontested non-IDN apps, which ones ended up, for example, with priority <700.

    That's really the metric that counts. If I have an uncontested app, I definitely want to go early. If I have a contested one, you can argue I may want to go later, because I would benefit from more time to potentially get a bigger refund.

Add Your Comment