Survey of Usage of the .US TLD
[ Introduction - Methodology
& Results - Conclusions & Future Work ]
Abstract: Recent policy changes allow registrations in .US with few restrictions. The author collects data about all known .US registrations, analyzing their registration patterns and usage. Certain registrants are found to register more than 2,000 domains each; these registrants may be gathering domains for commercial applications requiring many domains or for future sale, and large registrants (with ten or more .US domains) jointly hold a total of 46.4% of .US registrations to date. Non-Americans are found to register 7.0% of domains, and some of these registrations may violate .US registration restrictions that require nexus in the United States. The overwhelming majority of .US registrations as yet provide no original web content; working .US web sites are found to be clustered with certain registrars, while certain other registrars tend to register domains that offer no web content and domains offered for resale.
In 1985, Jon Postel created a series of top-level domains for use by interested countries; among these so-called country-code top level domains ("ccTLDs") domain was .US, bearing the two-letter country code ordinarily associated with the United States. For more than a decade thereafter, .US registrations were generally permitted only within a strict hierarchy reflecting both geographic and organizational categorizations (i.e. www.k12.wa.us for the public schools in Washington State). However, the National Telecommunications and Information Administration of the US Department of Commerce in 1998 began a consultation process to consider a liberalization of the .US registration hierarchy, and in the spring of 2002 .US was opened for second-level registrations (i.e. cars.us) via newly-selected registry operator Neustar and competitive registrars.
Four months after the opening of .US to public registration, the author seeks to investigate usage to date of the .US space. Such investigations in part follow the model of the author's previous studies of domain names, quantifying top registrants, registrar market shares, warehousing, defensive registrations, and cybersquatting. Analysis further considers .US domains that may not comply with .US registration restrictions.
To analyze domain registrations and use, the author began with a full listing of all registered .US domain names. Many TLDs provide such a listing upon request, via a so-called "zone transfer" often accompanied by a license agreement; however, Neustar told the author that zone file is currently available only to .US registrars but not to the general public (email communications of May 21, 2002 and August 12, 2002). The author sincerely thanks a .US registrar, who prefers to remain anonymous, for providing the zone file so central to subsequent analysis.
The author used automated systems to collect data about each registered .US domain. From publicly-available WHOIS data, the author collected registrant name and organization, registrar, and date of registration, as well as the country of registrant and of administrative, billing, and technical contacts. The author further collected the title of each domain's default web page (when available).
Analysis uses the .US zone file of August 13, 2002, which includes a total of 307,788 distinct .US domains.
Results include the following five sections:
Registration Patterns of Top .US Registrants
Registrations by Non-US Registrants
Rate of Registration
Registrar Market Share
Domain Usage & Registrar Specializations
Registration Patterns of Top .US Registrants
Certain registrants were found to register a large number of .US names. For example, Bryon Uding of the American Spirit registered 2,494 .US domains, Bradley Norrish of Internet Registrations Worldwide registered 1,746, and Sanda Yackolow of Marblehead Consulting registered 1,500. The tables linked below summarize and detail .US registrations by registrants with ten or more .US domains.
.US Registrations by Top Registrants
.US Registrations by Top Registrants, with domain listings
Page 1, 2, 3, 4, 5, 6, 7, 8
Note that some "top registrant" listings reflect registrations in the name of a registrar. The author contacted representatives of certain registrars and was in most instances told that these registrations will in due course be modified to reflect registration by registrars' existing customers.
Inspection of the registrations of top registrants shows five notable categories of .US domains. The first three categories listed below are specific to the "us" string specifically, while the final two are consistent with open TLDs generally.
Registrations by Non-US Registrants
Certain .US domains were found to provide (in their WHOIS registrant and contact data) addresses outside the US. Under the .US Nexus Requirements (PDF), .US domains may be registered only by 1) US citizens or residents, 2) US entities or organizations, and 3) foreign entities or organizations with a bona fide presence in the US. To determine whether a foreign entity has such a presence, registration requirements specify consideration both of a registrant's ordinary lawful activities within the United States and of a registrant's offices or other facilities in the US.
The author knows of no automated means of confirming that a registrant in fact complies with nexus requirements; indeed, Neustar's Nexus Requirements document contemplates only occasional scans and spot checks. However, when a non-US registrant holds many .US names, the registrant may be particularly likely not to comply with stated nexus requirements; a non-US registrant holding dozens or hundreds of .US names, each pointing to an "under construction" site or an error message, might be thought more likely to be a domain warehouser or reseller than a foreign entity with a bona fide presence in the US. Accordingly, the author below reports those registrants who have registered 5 or more .US domains that each provide addresses outside the US for registrant address and for administrative, billing, and technical contacts. Of course, many of these registrations no doubt comply with Nexus Requirements, but at least some may reflect registration by a non-US individual or by a non-US organization without the required US nexus.
Of .US registrants with all contacts outside the US, some registrants registered many .US domains. Silver Back Corp (of Antigua and Barbuda) registered 834 domains including all-animalsex.us, alyssa-milano-gallery.us, alyssa-milano-naked.us, and alyssa-milano-nude.us; Global DNS Services (of the Netherlands) registered 474 domains including 11b.us, aanbieding.us, aanbiedingen.us, and adult-toy.us; B.Stone of the Netherlands registered 226 including afterparty.us, ahold.us, americangigolo.us, and americanpornstar.us. A total of 613 distinct registrants registered 5 or more domains with all contacts outside the US, and a total of 21,639 domains were registered with all contacts outside the US (7.0% of all .US registrations to date).
As detailed in the .US WHOIS FAQ, each .US domain includes a designation of its nexus code. Possible codes include C11 (US citizen), C12 (permanent resident of US), C21 (US organization), C31 (foreign entity or organization with bona fide US presence, regularly engaging in lawful activities in the US), and C32 (office or other facility in the US). As among .US registrations with all contacts outside the US, the table below reports the number of registrants providing each of these nexus codes:
Number of .US Registrations |
Proportion of .US Registrations | |
(among .US registrations with all contacts
outside the US)
|
||
US Citizen |
3392
|
15.6%
|
Permanent Resident of US |
751
|
3.5%
|
US Organization |
1373
|
6.3%
|
Foreign Entity or Organization with Bona Fide US Presence, Regularly Engaging in Lawful Activities in US |
12766
|
58.7%
|
Office or Other Facility in the US |
2848
|
13.1%
|
No nexus data available in WHOIS |
609
|
2.8%
|
Impermissible .US registrations (by registrants without the required nexus) may tend to take place within certain purported nexus codes. However, research to date has not identified nexus codes disproportionately used for this purpose.
WHOIS "Domain Registration Date" data provides information about the date of registration of each registered .US domain.
.US Registrations Per Day - chart and table
This data reflects that more than 56% of .US name were registered on or before April 30, 2002 -- in the .US sunrise process. Since that time, approximately 1,000 .US domains have been added per day; this rate has remained roughly constant since June 1, 2002. Clear weekly trends reflect fewer registrations on weekends than on weekdays.
With approximately 308,000 domains registered through August 13, 2002 and a growth rate of 1,000 domains per week, extrapolation suggests a total of approximately 332,000 domains on January 1, 2003.
Interested registrants obtain .US names through accredited .US registrars. The chart and table linked below summarize registrar market share to date.
.US Registrar Market Share - chart and table
This data reflects that leading .US registrars are Go Daddy (55,687.US names, 18.09% of registrations to date), Register.com (42,645, 13.86%), Verisign (38,578, 12.54%), Enom (27,437, 8.92%), and Directnic (16,740, 5.44%). Together, these five registrars sponsor 58.84% of .US registrations, while a total of 65 other registrars sponsor the remaining 126,661 .US domains.
Among the twenty largest .US registrars, 12 registered between 40% and 80% of their .US names during the .US sunrise. Certain other registrars did not participate in the sunrise; Wild West Domains registered its first domain on July 16, and Stargate Communications registered its first on July 27. Other registrars have registered only minimal names since the sunrise; such registrars include "Official US Domains" (6,245 sunrise registrations and 494 subsequently), Encirca (5,008 and 940), and Namescout (5,537 and 388). Additional details are available in the chart and table linked above.
Domain Usage & Registrar Specializations
The author attempted to obtain the default web page from each registered .US domain name, and when those pages were available, the author categorized available content into the groupings described below. Review and grouping of HTML page titles provided automated categorizations of the majority of tested web sites, while manual review was used for certain additional domains that could not be classified based on their ambiguous or omitted HTML page title.
Proportion of all .US domains | |
Fails to provide a valid HTTP response ("cannot connect to server") |
30.2%
|
HTML body is blank, provides a redirect, or includes "under construction," "coming soon," or similar |
50.6%
|
HTML title or body contains an offer of sale |
2.8%
|
Error message |
1.1%
|
Uncategorized (includes domains with actual content) |
15.3%
|
The large number of .US domains without default web pages, with blank pages, and with "under construction" or similar pages is consistent with the author's prior study of other top-level domains including .BIZ and the open country-code top-level domains of .CC, .TV, and .WS.
With knowledge of each .US domain's registrar, the author tabulated domain usage by registrar. As detailed in the table linked below, registrars vary greatly in their customers' usage of .US domains. Large .US registrars with relatively high estimated rates of provision of web content include Tucows (29.6%), Bulkregister (24.9%), Enom (23.5%), Go Daddy (17.9%), and Verisign (17.1%). Large registrars with substantially less frequent provision of web content include Itsyourdomain (5.0%), Dotregistrar (6.2%), and Directnic (7.7%). Note, however, that certain large registrants can sway these estimates dramatically; for example, a single Bulkregister registrant provides substantially the same content on all of its 1,333 .US domains, but since these domains do provide actual web content, they count towards registrar Bulkregister's total and increase its "actual web content" proportion by 11.5%.
Registrants of certain registrars chose "under construction" pages for the overwhelming majority of their domains; for example, fully 83.1% of domains by registrar "Official US Domains" provided "construction" pages or were blank or redirects. However, registrations by certain other registrars disproportionately tended not to provide valid HTTP responses, perhaps because these registrars do not provide "under construction" pages or because their customers prefer not to use such pages; such registrars include R&K Global Business Services (87.3% of registered domains fail to provide a valid HTTP response), Encirca (88.5%), and Emarkmonitor (96.0%).
Domains with offers of sale tended to be clustered with registrars Domain Discover (42.5% of registered domains included an offer of sale on their default web pages), Directnic (11.0%), and CORE (18.9%). This result also primarily reflects clumping of registrants -- that these registrars each have one or several large registrants with many domains offered for sale.
According to Neustar's Director of Policy and Business Development, the .US registry is and ought to be "a national public resource" (cite, PDF). In this context, evaluation of .US's registrations to date may properly examine registration trends with an especially detailed level of scrutiny. Registrations like those of Silver Back Corp of Antigua and Barbuda (834 .US domains including all-animalsex.us and alyssa-milano-naked.us) may come into question for disputed compliance with .US nexus requirements. In addition, while the resale of domains is permissible under .US registration rules, those who register hundreds or thousands of names for the purpose of resale may also find their actions controversial.
The author knows of no proactive enforcement of .US registration restrictions, and it is therefore perhaps not surprising to find many thousands of domains that may be inconsistent with registration restrictions. Consistent with the author's prior investigations of .NAME and .BIZ, compliance with stated registration restrictions seems to require an active and proactive enforcement mechanism; merely demanding that registrants certify compliance with stated rules may not suffice to ensure compliance. Of course, effective enforcement may be difficult and costly; indeed, it may be sufficiently difficult and costly that the Neustar registry and the US Department of Commerce on balance decide against such enforcement. Nonetheless, if current registration restrictions reflect an explicit policy decision as to proper usage of the .US TLD, current enforcement systems may be ineffective at carrying out this policy. Were the DoC and Neustar to remain committed to the current registration restrictions, they might put into place special checks for those registrants outside the US. For example, a first-time non-US .US registrant could be required to fill out a web form detailing its US activities and/or US offices or facilities, the factors considered under Neustar's current statement of Nexus Requirements.
Future work might consider the following questions:
Thanks to Bret Fausett for suggesting this project and to Tim Hewitt of myOstrich Internet for information on .US sunrise registration restrictions. Thanks also to an anonymous registrar that provided a current .US zone file.