Home / Blogs

A Framework for Selecting New TLDs

Your corporate domain names send implicit messages (signals) through their Top-Level Domains (TLDs) and their second-level words. Shape your domain names so to send the right messages and to avoid sending unintentionally confusing messages. The post focuses on a framework to help bidders determine which TLDs send messages that are potentially profit generating.

As many readers are aware, the Internet Corporation for Assigned Names and Numbers (ICANN) has recently approved relaxing the rules for the introduction of new top-level domains (TLDs). Soon TLDs such as “.car,” “.cars,” “.green,” and “.eco” will be available to any qualified body whose request is favored by the allocation system. The system being discussed is a combination of beauty contests and auctions.

Once new generic TLDs are established, companies and the public decide which second-level domain names to register under the approved extensions. The value of the signal depends on its desirability, which is driven by the demand for the associated second-level domains. The value of branded TLDs depends on the final registration restrictions adopted by ICANN.

Designations and Branding

The value of “.com” is determined by the branding message it transmits, which has evolved from the TLD’s original designation for commercial use. Two of the original TLDs, “.net” and “.org,” were intended to signal an Internet company and a nonprofit organization, respectively. In practice, bodies that wanted a “.com” but couldn’t get one have settled on “.net” and “.org” as second best. True, the TLDs’ original signals have not faded from the minds of Internet users: consider the indirect marketing by Microsoft’s “.Net” project and the use by the recently created BushClintonKatrinaFund of “.org” for advertising and branding even though the group also owns the “.com” extension. But “.net” is often used for name-servers (”.net” represents 58% of worldwide hosts and 30% of the world’s nameservers according to VeriSign’s article “There’s More to .Net Than Meets the Eye”) or as the TLD for employee email addresses; Forbes, for example, uses “.com” as its main TLD branding and “.net” for email.

There are two widely held theories as to why additional extensions such as “.biz” and “.info” are being introduced. One is that they are intended to relieve the strain placed on the original TLDs by the tremendous growth of the Internet, with “.biz” signaling an online business and “.info” an online information site. The other view is that they were intended to break the monopoly of “.com.”

The “.us,” although technically a country-code top-level domain (ccTLD) for the United States, has been used by U.S. companies as an alternative TLD and as a way of signaling an American business presence.

Message Types

Below I outline three kinds of primary message and one kind of secondary message. As noted earlier, a message can be sent through either the TLD and/or the second-level domain name. To decide at which level to signal, you need to be clear on what you want to signal and how your message can get across.

The three primary message types are:

  1. Location. A company can send a message about the presence of its office in a specific location. For example, “.SF” can signal San Francisco. Thus, the Italian carmaker Fiat can signal its presence in San Francisco by using “Fiat.SF”.
  2. Brand positioning. Suppose that Honda wants to expand from cars into the aerospace market—it registers “Honda.aero.” Or Exxon signals expansion into green energy, or merely the birth of an environmental conscience, by using “Exxon.green.”
  3. Branding. For trademark owners, the decision to register defensive TLDs depends on the final version of the ICANN-approved trademark registration regime. If registering a brand TLD adds value, i.e., an offensive registration, then it should be registered irrespective of whatever regime ICANN finally adopts for domain-name intellectual property.

A text-based URL can never look as snazzy as an elaborate graphic image or animation, but a domain name must still have a visual personality. For example, “.BMW” has less visual personality than “BMW” or “BMW.com.” Also “BMW.cars,” if “.cars” were approved, has no branding personality; thus, if the eventually adopted trademark regime restricts brand-based extension ownership to brand owners, BMW should not register the “.BMW” TLD. At the second level, capitalizing key words, as in a headline, can make a name into a memorable statement instead of a jumble of letters. Compare thisisourwebsite.com and ThisIsOurWebsite.com.

New proposed TLDs that use geographical designations can provide personality as well as information. For example, compare “SFBMW.com” and “BMW.SF.” In online games, for example, the use of the “.SF” TLD would signal the presence of a BMW dealership/store in San Francisco. Of course, and this is especially the case with superlatives and geography, a personality needs to be authentic if it’s going to be effective.

Demand for a new TLD can be estimated using prediction markets and statistical modeling. The demand for second-level domains has to be estimated over a period beyond the initial speculative period. Perceived failures, such as the underwhelming number of registrations for the recently released TLDs (”.info,” “.biz,” “.cat,” and others), do not imply failure for future TLDs. Nevertheless, there can be valuable lessons to learn from the failures. For example GM’s decision to scrap the production of electric cars gave Toyota Prius a lifeline. Also, Apple’s failure with Newton did not mean the death of PDAs or keep Apple from pursuing what turned out to be the phenomenal success of the iPhone. One potential reason for the limited success of certain TLDs is the lack of a roadmap such as a signaling framework. A second is the lack of rigor in estimating demand for individual TLDs.

Concluding Remarks

When considering registering a TLD, you need to analyze the implied message and personality.

Failure of previous TLDs does not imply the failure of new ones. Nevertheless, demand should be estimated using scientific methods such as prediction markets and statistical techniques.

Related References
Roles of Corporate Domain Names
Branding Strategy: The TLD Dimension
Carlton and Kende’s Narrow Understanding of Corporate Domain Registrations

By Alex Tajirian, CEO at DomainMart

Filed Under


Thanks Alex -- perhaps you could provide Antony Van Couvering  –  Jul 25, 2009 12:49 AM

Thanks Alex—perhaps you could provide more detail on which statistical techniques you espouse?  Also, as I understand it, prediction markets work if and when there are enough people (and, importantly, different kinds of people) to produce a valid sample.  How can this be achieved?

Small corrections: .SF, as a two-letter TLD, would not be allowed under ICANN rules, as two-letter TLDs are reserved for country codes.  Also, .AERO could not be put to the uses you suggest, since it’s already delegated to SITA and reserved for parts of the airline industry.

Anthony, good to hear from you.Each applicant Alex Tajirian  –  Jul 27, 2009 6:31 PM

Anthony, good to hear from you. Each applicant will have his or her proprietary demand/revenue prediction model. Although there is no one right model, there are lots of wrong ones such as “Because brand owners have not registered under all current TLDs, they don’t need to worry about new TLD infringements.” Obviously, the models will not result in low margin-of-error forecasts, but can provide actionable information. I am sure there will be “experts” who claim to provide “accurate” value forecasts, which is a pure garbage concept. I am sure you know the difference between a market solution and one based on crowdsourcing that requires a large number of participants. For price efficiency in arbitrageable markets, all you need is one smart (and rich) investor and all the rest can be crazy. Markets with few participants will not be liquid, but can be informationaly efficient although with systematically a higher return to risk ratio. The challenge in prediction markets is the design of incentives, as it is illegal to use real money. Nevertheless, the approach has been implemented in small corporate business units of major companies such as Google, Microsoft, and HP. On the other hand, with a crowdsourcing approach, you are correct in that they are useless when participants are biased. That’s why the average of predictions by a large number of participants about the distance between earth and the moon or guessing the number of red jelly beans in a jar are not reliable. My philosophy, in general, is that a task that is not challenging is not worth pursuing, with few exceptions such as in some domain name ventures. I had a feeling that there were restrictions on two-character TLDs because I haven’t heard any complain about new TLDs being ccTLD typos, and that New York wants .NYC instead of .NY. Thanks for pointing out their misleading use in my examples. Well, now that I have a lemon, can we make lemonade? What prevents ICANN, which likes beauty contests, from changing the restrictions? A more interesting question may be, do we need the restrictions?

Comment Title:

  Notify me of follow-up comments

We encourage you to post comments and engage in discussions that advance this post through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can report it using the link at the end of each comment. Views expressed in the comments do not represent those of CircleID. For more information on our comment policy, see Codes of Conduct.

CircleID Newsletter The Weekly Wrap

More and more professionals are choosing to publish critical posts on CircleID from all corners of the Internet industry. If you find it hard to keep up daily, consider subscribing to our weekly digest. We will provide you a convenient summary report once a week sent directly to your inbox. It's a quick and easy read.

I make a point of reading CircleID. There is no getting around the utility of knowing what thoughtful people are thinking and saying about our industry.

Co-designer of the TCP/IP Protocols & the Architecture of the Internet



IPv4 Markets

Sponsored byIPv4.Global

New TLDs

Sponsored byRadix


Sponsored byVerisign

Brand Protection

Sponsored byCSC

Domain Names

Sponsored byVerisign

Threat Intelligence

Sponsored byWhoisXML API