The five applications seeking "personal" TLDs are set forth in the following table:
Threshold Review As a threshold matter, the evaluation team applied the first and the last criteria to these applications and compared applications for the same or similar TLD strings on an overall basis as described in Part III.B.1.a. As a result of this review, the evaluation team concluded that the following applications merited further review:
The technical team concluded that the application of The dotNOM Consortium did not demonstrate the technical ability to operate a TLD targeting a large group of potential registrants and end users with high reliability, good performance and strong security. The technical team concluded that the technical plan as presented was inconsistent with and insufficient for the load predicted by the applicant for the TLD, particularly at startup. Further, the technical team also found the description of the security mechanisms to be unsatisfactory when compared with other proposals in this group. The business/financial team felt that from a business perspective this group was the most competitive. Further, it concluded that the application of The dotNOM Consortium was not as strong in significant ways, including its marketing plan and overall assessment of the market, as the other four applications. Comparison Having applied the first and last of the August 15 Criteria to these applications as a threshold matter, the evaluation team evaluated how each of the August 15 Criteria should be applied to proposals for these TLDs in light of the diversity in purpose and targeted markets reflected in the categorization. The evaluation team concluded that the August 15 Criteria apply as follows:
Enhancement of competition, proof of concept, meeting unmet needs and enhancement of diversity seem particularly relevant to analysis of these TLDs.
The proposals suggest different formats for personal names. CORE Internet Council of Registrars proposes names in the form john.smith.nom, with all second level domain names reserved and created automatically as necessary. Global Name Registry proposes to reserve all second level names for family names, creating them as necessary. Its form of name would be john.smith.name. JVTeam proposes names of the form johnsmith.per. Sarnoff proposes several different forms of names at different prices. It proposes to issue numeric names in the form 0123456789.xx.i for free, allowing the user to choose the numbers and two letters. Sarnoff proposes to offer standard names in the form john-j-smith.0001.i for $2.00, with the unique number automatically generated for the user. It proposes to offer most other names as "custom" names for $4.00, such as johnsmith.i, subject to specified restrictions. Evaluation of whether these proposals enhance competition for registration services encompasses several inquiries, including whether the proposed TLD and registry are likely to be attractive to a significant market segment in which they can compete effectively, whether distinctive services are being proposed that will meet the needs of those not being served adequately by existing services, the adequacy of marketing and promotion plans, the competitiveness of proposed pricing and service levels with other TLDs, and operators having significant market share and restrictions on accredited registrars. In order to be attractive to a significant market segment in which it has a realistic prospect of competing effectively and to provide an effective proof of concept, a general purpose TLD applicant must realistically assess the business, financial, technical, operational, and marketing requirements for implementing the proposal and procure firm commitments for necessary resources. Some of the significant factors in evaluating whether these TLDs will likely be successful additions to the DNS are summarized in the following table:
As the table indicates, wide variations exist among these applicants. Expected demand10 in year 4 ranges from 5.3 million to 76.0 million registrants. (Of the 76.0 million registrants estimated by Sarnoff, 30 million registrants are numeric names which Sarnoff is offering at no charge to the registrants.) In comparison, at the end of September, 2000, .com had approximately 20 million registrations. However, even if those applicants forecasting a smaller demand for a personal TLD are correct, the projected market demand would constitute a significant level of potential demand. The initial equity investment also varies widely. Some of the variations may be explained in part because some applicants propose leveraging existing infrastructure, including outsource partners, while others must build the infrastructure. Even accounting for the stated reasons for the variations, the table indicates that JVTeam, Global Name Registry, and Sarnoff appear to be willing to devote significantly more resources in order be attractive to a significant sub-market and to compete effectively for the demand generated by that segment. In addition, Global Name Registry's application is similar to the business of its parent, NamePlanet.com. CORE has not provided budget information for marketing efforts and delegates most of the marketing activities to the registrars. Another method of being attractive to a significant sub-market in which it has a realistic prospect of competing effectively, meeting unmet needs, and providing an effective proof of concept, is enhanced service content, particularly with respect to the registry interface protocol as discussed in Part III.B.1.a. A summary of the service offerings of these applicants with respect to the registry interface protocol is set forth in the following table:
Sarnoff believes that its unique feature of granting 30 million free numeric names, along with standard names for a modest price of $2.00 per year, will result in large numbers of people having registered names. Large numbers of people with registered names, Sarnoff asserts, will enable many new applications for the Internet since it will create the infrastructure required for mass adoption. Examples Sarnoff cites include providing a generally accessible repository for public keys, aliasing for people with multiple e-mail addresses, better support for roaming, and support for private networks of Internet-enabled appliances. In order for any applicant in the personal group to be attractive and therefore possibly be commercially successful, meet unmet needs and provide an effective proof of concept, proposed pricing and service levels must be competitive with other TLDs and operators serving that market segment today. The following table summarizes the applicants' proposed pricing and service levels:
All of the applicants propose a price of $5.30 per year or less (some significantly less) for registrations, except as indicated above. The proposed pricing by all four applicants is less than Verisign's current price of $6.00 per year. In addition, Sarnoff's application proposes free subscriptions for registrations of numeric names. All applicants provide an acceptable level of availability for the SRS services. "Time to Reliance" varies widely based on the underlying update design philosophy. Update times in minutes are probably more useful than the half-day average times for CORE and Sarnoff. On the other hand, one must consider whether today the personal name space needs to be held to the same high standard of availability and performance as the general group. Transaction times available shown in the table are appropriate, although the CORE number (1 tps) seems much too slow. Another competitive issue considered in evaluating the proposals is whether the proposals restrict the ability of accredited registrars to offer registration services within the TLD. None of the applicants impose restrictions on accredited registrars.
Enhancement of diversity is another of the August 15 Criteria that seems particularly relevant to analysis of these TLDs. Evaluation of whether these TLDs enhance diversity encompasses several inquiries, including diversity in business models and of geographic locations. In addition, some of the factors identified in the August 15 Criteria in connection with effective proof of concept largely overlap the diversity evaluation. Some of the significant factors in evaluating whether these proposed TLDs are likely to enhance diversity are summarized in the following table:
Two of the applicants in the personal group are based in the United States. CORE is based in Geneva, Switzerland and The Global Name Registry is based in London, England, either of which would enhance diversity of geographic location of operators of large TLDs. Diversity of geography in ownership is also a relevant inquiry. Core is an international consortium of 72 member registrars in 20 countries on four continents. JVTeam is owned by NeuStar and Melbourne IT, which is an Australian company. Two of the applicants in this group have interests in other applications in other categories and groups. CORE is the proposed registry operator for the World Health Organization (.health) and the Museum Domain Management Association (.museum). JVTeam has submitted an application for .biz, is the proposed registry operator for SRI International's application for .geo and is listed as the proposed registry operator for Blueberry Hill Communication, Inc.'s application for .kids (although JVTeam has not yet committed to do so). Sarnoff is a wholly-owned subsidiary of SRI International. All of the applicants have a subscription-based revenue model, although JVTeam offers volume discounts and Sarnoff offers free subscriptions for registrations of a numeric name.
Enhancement of the utility of the DNS is another relevant August 15 Criteria. These applications appear to sensibly add to existing DNS hierarchy, do not appear to create or add confusion to the existing DNS hierarchy, and are semantically far enough from existing TLDs to avoid confusion. Sarnoff's application requests a one character TLD - .i. In answer to a question about acceptable lengths for TLDs, ICANN responded prior to the end of the application period in FAQ #47 that ordinarily TLD labels proposed for the new TLD program should be between three and sixty-three characters long, inclusive. The response emphasized that, under the current practice of the Internet Assigned Numbers Authority, one-letter labels are generally reserved from assignment to allow for future DNS extensibility. Although explicitly acknowledging in its application the reservation of one-letter codes, Sarnoff requested .i. However, Sarnoff said in its application that it would consider adopting an alternative string, such as ".idi", ".iii" or ".one", although Sarnoff believes that such a string would be less distinctive and therefore pose a greater possibility of being confused with a traditional TLD. Sarnoff's claim to have an "untraditional" TLD, however, appears to be significantly overstated; the evaluation team believes that the specific plan of Sarnoff calls for a relatively mainstream personal-use TLD.
The four applicants request a name space that will be reserved for personal usage. All applicants require that registrants declare that they are reasonably related to the name being requested although none proposes any technical mechanism to verify this. Beyond this, each has different mechanisms for enforcing such usage.
This applicant allows non-commercial use as well as limited commercial use (e.g. selling your house) as long as it is effectively personal and does not conflict with a true commercial activity (e.g. Mr. McDonald may not offer hamburgers for sale). Trademark holders may request exclusion of their names from registration. Furthermore, the registry will require registration of names as two level entities (john.smith.nom). It is expected that this restriction will greatly reduce non-personal registrations.
Trademark holders may register their names on a notification list. They will be alerted if a registration attempt is made on that name. JVTeam also proposes to extend the UDRP to allow for challenges to improper use of the name space. Thus others could object to the commercial use of a name.
The registry will prevent the registration of trademark names and dictionary words. Names in the space will be non-transferable (you may only release your registration) and once a registration is released the name will be returned to the pool of available names only after a random period of time has passed. Registrants will be restricted to owning a limited number of names.
Name registration will only be allowed at the third level since the registry reserves all SLDs as family names. Trademark holders will be allowed to provide their names for inclusion on a watch list and will be notified if a registration attempt is made.
As with all applicants in the general purpose category, protection of the rights of others is another of the August 15 Criteria relevant to analysis of these proposals. As a practical matter, proposals in the personal group must propose a well-thought-out plan for the allocation of domain names and provide adequate protections to stakeholders and third parties. Though they tend to restrict the registrant's commercial use of the domain, the personal group also presents a risk to intellectual property rights of others. As a whole, these proposals rely on the registration processes and restrictions to minimize the potential for registration abuses. All proposals provide basic methods for protecting and enforcing infringed rights (i.e. status quo) and offer limited extra protections. If one or more of the applicants in this group is accepted, the evaluation team recommends that further clarification and direction be required. All applicants in this group require registrants to declare they are reasonably related to the name requested although none of the applicants propose technical mechanisms to verify the declarations. The proposals provide some differing approaches for the protection of the rights of others, summarized as follows:
No sunrise registration period but during queue building process allows registered trademark and service mark holders to file exclusion requests. To restrict abusive registrations, registry requires second and third level registration and restricts transferability of the registration. During the initial period, at least, registrations will be controlled by a random selection process. Whois service is maintained by the registry but will only publish minimal data. Third party Whois requests are required to show a legitimate interest in the data to receive access. Dispute resolution will be handled through a modified UDRP conformed to meet the needs of .nom registrants. Registrants are allowed limited commercial use of the domain as long as no conflict exists with a true commercial activity.
No sunrise registration period but a fee-based intellectual property notification service allows parties to register their marks with the registry and receive notification if an attempt to register the mark is made. Registrations are first-come, first-served and dispute resolution is controlled by a modified UDRP which allows for challenges to improper use of names. Whois service is publicly available for limited registrant information. The TLD is restricted to personal use and requires registrant certification of use.
Proposes a 90-day daybreak period for registered mark holders and a lottery period for the initial rush of registrations. Under the daybreak plan, only valid trademark or service mark holders are allowed to participate and have two options: (1) pre-empt registration by other parties; or (2) pre-register. After the daybreak period, a lottery period during the initial registration rush will operate to allocate user-selected domain names in a lottery fashion. A single central publicly-queried Whois database is available at the registry level with registrant name and number of registrations. To limit abusive registration practices, dictionary words and geographic terms are restricted from registration, domain names are non-transferable, a "takedown" provision to limit continued registration by bad faith registrants will be implemented and released names will be returned to the available pool after a random period of time has passed. The UDRP will be enforced and registrants will be restricted from commercial use.
Proposes a sunrise period for registrations which is not well defined but appears to conform to its overall attitude to maintain a hands-off policy with regard to the registration and use of domain names. There will be no pre-screening of domains: all domains are registered on a first-come, first-served basis. Trademark and service mark holders will be allowed to provide their registered marks for inclusion on a watch list and will be notified if a registration attempt is made. Whois service will be searchable for domain names, but not registrant names. Reliance on the UDRP and release of only third level domains is stated to greatly reduce the potential for abusive registration. There are no provisions for restricting the TLD to personal use only. Recommendations This group includes several strong applications, justifying selection of a limited number of them. The selection(s) should be made from among the applications that underwent further review, based on discussion above. Footnotes: 10. The applicants in this group were asked about their assumptions on expected demand. Global Name Registry, the only applicant requesting more than one string, was asked about its assumptions on expected demand. Its estimate is based on being granted .name. In addition, the applicants were asked about their assumptions regarding other potential new TLDs. CORE responded that it assumed that ICANN would launch between six and 10 new TLDs, at least one of which would be a new generic TLD along with .nom granted to CORE. JVTeam responded that it assumed the introduction of additional personal TLDs as a result a successful proof of concept. Sarnoff responded that it made no specific assumption about the number of other personal TLDs introduced in the future. 11. During the first 6 month start-up period, Sarnoff anticipates a rush on custom names and will charge on a cost recovery basis, which Sarnoff anticipates to be $10.00. Comments concerning the layout, construction and functionality of this site should be sent to webmaster@icann.org. (c) 2000 The Internet Corporation for Assigned Names and Numbers. All rights reserved. |
