This proposal outlines the creation of a new market for the efficient disposal of domain names that are about to expire and how Registrants are best included as an integral part of this process. This document does not make any proposals regarding the existing primary or secondary markets for domain names.

Further, this proposal is not intended to criticise, condemn or make any declarations about the appropriateness of any particular primary or secondary market business models. As an underlying principle, this document assumes that the market itself is in a far better position to make judgments of this nature.

This proposal first describes the current state of the market and then recommends a fundamental structural change that will allow the market to function in a healthier, productive and more efficient manner.

Like the internet, the domain name market continues to evolve and is not in its final state. The current state lends more to the imbalance of information in the market than to any compelling business forces. Specifically, Registrants have little knowledge of the value of the domain names they let expire. A market that permits registrants access to that information would result in them benefiting from typical supply and demand forces with which particants in other, more efficient markets, benefit.

Furthermore, this arrangement create tensions in the market that would incent all parties to engage in rational, sustainable and efficient behaviors over the long term - a feature that does not exist under current arrangements.

Names in the Expiry Process…

There is an extremely active trade in domain names that have been previously registered and are allowed to expire by the Registrant. These names are bought and sold by a small group of businesses who take advantage of two elements of value. The first is the latent traffic that all “expired” domain names possess. This is website traffic that continue to come to the domain name, either through historical inbound links or errant navigation (misspellings or semantics) that can be used to create “clicks” on contextual links that are aggregated and sold to Overture, Google and other cost-per-click (“CPC”) marketplaces. The second element is the potential value of the domain name on the secondary market.

Estimates indicate that 2-5% of expiring names have a value reasonably in excess of what it would cost to renew it. Anecdotal evidence indicates that at least one name per month, often more, sells for over $10,000.

Importantly, these uses are other than what were viewed as “expected use” for a domain name (such as in connection with a website or email) when the industry structure was initially set forth.

Supply dynamics…

Approximately four larger suppliers and a number of smaller ones service this market. Let’s call them Expiry Service Providers or “ESPs”. Their role is twofold. First, they aggregate demand via auctions and similar mechanisms. Second, they regularly flood the registry with a continuous barrage of registrations requests over the shortest period of time possible just as desirable domain names become available for re-registration. The results of these auctions determine the intensity of the registration requests dedicated to each name.

The volume of registration requests an ESP can dedicate to a particular transaction is limited by the level of resources that each registry provides to individual registrars. Each registry provides each registrar with a specific level of access to the registry resources. Each registrar is afforded the same level of access with no differentiation made on the basis of transaction volume, sales revenue, resource usage or other factor. In ICANN-speak, this arrangement is called “equivalent access to registry resources”.

ESPs work around equivalent access limitations by aggregating the resources of multiple accreditations (which are borrowed or otherwise accumulated through a multitude of different schemes). The more registrar accreditations an ESP has access to, the more resources it can bring to bear on the registry.

This behavior was institutionalized in 2001 with the creation of the first gTLD registry “batch” pool. Registration requests for “about to expire” domain names (then call “add-storms”) had become fierce enough to negatively effect the regular course of business for normal registry and registrar operations. The solution was to implement a batch pool of processing resources as a temporary fix to the service degradation and resource contention problems caused by the add-storm within the registry’s primary production environment. This allowed those wishing to re-register expiring domain names to do so without effecting the normal course of business of registrars engaged in regular registration and administrative activities.

Accumulating, maximizing and optimizing access to the batch pool quickly became an arms race. ESPs began paying for use of a registrar’s regulated access to the batch pool as a normal cost of business. This started with registrar’s that were inactive and quickly moved to include registrars of all types. Monthly fees paid for use of an accreditation or “cred” peaked earlier in 2004 in the range of $20,000/month.

As the value of a cred grew, the number of registrar accreditation applications to ICANN grew exponentially. It was widely reported at the recent ICANN meeting in Kuala Lumpur that one party had recently submitted and received approval for a rumored 100-plus accreditations that have since been devoted to the war on the batch pool. ESPs have also gone to the significant expense of locating specialized data centers in Virginia to minimize network latency by being as “internet-close” as possible to the registry as possible in an attempt to optimize their request efficiency to yet another degree.

It is important to note that all of this activity exists simply to deal with imperfect information and regulatory limitations; “first-come-first-serve” registrations, equivalent access to registry resources and mandatory flat pricing of domain name registrations at the registry level. These are all functions of an evolving market and yesterday’s rules relating to today’s problems.

Demand dynamics…

Best estimates indicate that this is a $4m/month market and that significant growth will continue for the foreseeable future. A few large players have secured the lion’s share of the market, but this does not seem to have deterred the hundreds of smaller players who continue compete in this space.

There area number of secular drivers that will ensure the growth of this market for the foreseeable future; growth in CPC rates, the economics within CPC itself and an increase in the number of participants in this market. The market will also benefit from the growth in Internet usage in general.

The Problems…

First the good news. There is real value in the aggregation of demand that has taken place. There have been efficiencies created in information, in aggregating traffic and in monetizing it. Most importantly, a floor has been set on the value of a domain name.

But there are significant problems with this state of affairs.

First, these arrangements threaten to fundamentally, and inappropriately change the concept of what it means to be a “registrar” in the ICANN context. Accreditations are now being put to a uses almost diametrically opposed to those accounted for at the onset of competition. This has implications in terms of regulating registry access, in terms of fees, costs and budget contribution and also in terms of policy development at the GNSO level (who does the registrar constituency actually represent?).

The extent of the distortions to the concept of “registrar” are best illustrated simply by noting that it is currently much more profitable to simply “rent” an accreditation to an ESP than to actually operate a business selling domain names to the public.

Second, the registry is subject to significant and costly load that it must devote considerable time and investment to sustaining. Unable to recover even basic costs, the registry must continue to make increased allocations for no additional benefit. The issue of whether the registry “should” or “should not” provide such services or whether they are well compensated for doing so is besides the point. The real issue is that the registry is expending much in the way of time, money and people to a market other than that it was originally contemplated that they would serve.

Third, and most importantly, the market exists because registrants are being kept in the dark. This can be demonstrated conclusively with one simple question; “Would any current registrant with perfect knowledge that the value of their expiring domain name was in excess of $50 let the name expire for nothing? At $30?” It may be reasonably argued at what level this line should be drawn, but certainly not that it should be drawn.

The market evolved…

The current market exists for the benefit of three groups. ESPs, registrars renting access to the batch pool and acquirers of large pools of names. This is an unsustainable situation. Better information or more efficient market alignment would inevitably work to the detriment of these three groups and to the benefit of registrants.

The most helpful analog is the online auction phenomenon. Before online auctions there were whole classes of goods that gathered dust in various closets, garages and attics. With the advent and adoption of new auction technology, efficiencies were realized and entire markets opened up. The market for expired names currently operates as if some small group of people had access to those goods and sold them solely to another small group. The price that the buyers pay is less than what might otherwise be generated and the money all goes to those with access to the goods. Put another way, this is regulatory arbitrage.

This is not intended as an aspersion. To the contrary, this underscores the creativity and level of innovation that occurs when monopolies are opened to competition. However, in the interests of continued innovation and increased competition, the market must evolve in order to remain viable. It must become a marketplace open to all.

It should also be reiterated that the current situation is part of a natural market evolution, not some failure on the part of ICANN or the registry. Markets always evolve in unforeseen ways and we cannot expect those charged with administering those markets to know the unknowable.

When inefficiencies exist in a market there are two responses. They can be exploited in the short term, usually for significant economic benefit. The alternative is to introduce efficiencies that tend to be of lower absolute economic benefit but have much greater longevity. The market for expiring domain names is now at a stage where it is ready for the second, healthier, response.

In sum, both ICANN and Verisign registry are currently forced to spend significant, unsustainable resources in terms of people, time and money dealing with the current inefficiencies. This is all for the benefit of a small group that exists primarily because of imperfect information available to current registrants.

Essentially the demand side of the ESP market has evolved to create a useful efficiency. The supply side, however, remains woefully inefficient. This must change. The catalyst that will allow the market to move towards greater efficiency and a healthier market is to increase the flow of information to current registrants and accommodate the corresponding effect of increasing the flow of benefits to them.

Elliot Noss & Ross Wm. Rader, August 31, 2004. This post originally appeared at “Random Bytes”

This work is licensed under a CreativeCommons License.