Home / Blogs

Resolving the ICANN-Proposed TLDs Debate

The Internet Corporation for Assigned Names and Numbers (ICANN) has recently decided to allow any entity to register a top-level domain name (TLD). The best mechanism for valuing this decision, a mechanism that outperforms crowdsourcing, blogs, and committee decisions, is the legal and easy-to-implement solution known as prediction markets.

In handing down its TLD decision, ICANN never identified the problem it was trying to solve. This gap has resulted in numerous unresolved discussions and speculations.

The discussions have focused on whether:

  1. An open-ended generation of TLDs is needed — that is, would it be value adding compared to the current case-by-case decision?
  2. The open-ended proposal would create unintended hardship for current brand owners, hardship that cannot be internalized. The reasoning is that, to protect their intellectual property, brand owners will not be compensated for any damage and thus will incur the unnecessary additional costs of having to register their brand names under each of the new TLDs.

How should ICANN and our community resolve the above two issues? If ICANN were a for-profit organization, only the first question would apply. But, as a nonprofit and a community-conscious organization, ICANN also has to worry about the decision's negative side effects, including brand dilution. In either case, prediction markets can supply the answers.

The central strength of markets is that they give participants the right incentive to disclose their private information and tastes. In effect, participants are rewarded with money when they get the price right, and they lose money when they are wrong. Theory and evidence demonstrate that other view-aggregation mechanisms are less desirable for handling the ICANN decision.

Let's take statistical crowdsourcing. Despite its popularity, the technique is effective only when a majority rule is used and when it's more likely than not that each participant will answer correctly. Under such conditions, the crowd's decision comes closer to 100% accuracy as the number of participants increases. (An example of failure is when crowds are asked to guess the distance between the earth and the moon, a topic on which each participant is more likely to be wrong than right.) Thus, the approach would not help us if participants had speculative reasons in favor of increasing the TLDs, or if they believed their brands' value would suffer because of the ICANN decision. Under such a scenario, expert opinion is best.

Decisions by committee deliberations fail when there are better coordination mechanisms. For example, the infamous Soviet five-year economic planning system failed in favor of markets. But one should not forget that, in the cases of Wikipedia and open source, the final product involves committee decisions and filters. For Wikipedia, the final editor can make numerous changes. Apache, to take another example, has adopted a system of e-mail voting, with complex rules to ensure consensus within the Apache Group.

Blogs are not effective in aggregating diverse information. I outline their limitations in a forthcoming essay, but my claim is supported by the numerous numbers of posts and comments on the topic on this blog alone. None have resulted in any convincing understanding of the new TLD decision's implications.

Many corporate giants, including Google, HP, and Microsoft, have successfully used prediction markets internally. For example Google used them to answer how many people will use Gmail accounts in the next three months. HP has used the mechanism for sales projections, among other decision issues.

Nevertheless, once our community has experience with one such market, future implementations would be much easier. There is no better time to start than now. Moreover, such a market is not difficult to set up and participants can easily be shown how to take part. For example, each HP participant was given just 15 to 20 minutes of instructions.

There are a number of popular online software vendors already available. In addition, there is available help by experts to set up and make the interface easy to use. Legal concerns are not an issue, since the markets do not place real money at risk and therefore do not fall under the legal definition of securities trading. Instead the rewards can be in virtual currency and gifts, such as free domain registrations, hosting, free UDRP, free trademark application etc.

Last, but not least, who should be entrusted with the process? Is ICANN the best entity to manage the markets and the follow-up steps of designing the allocation mechanism? The answer may be simple, as this announcement gives us a clear idea as to how ICANN works, and to many it is a sad case of déjà vu.

By Alex Tajirian, CEO at DomainMart

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


By Alex Tajirian  –  Feb 17, 2009 1:56 pm PST

Good comment.

Let me clarify using a simple example.

Case 1: Decision facing a for-profit organization

Objective: Maximize shareholders' wealth


Option 1 -
Net Profit from New TLDs to Registrants/Company ($): 125
Net Profit to Brand Owners ($): -30

Option 2 -
Net Profit from New TLDs to Registrants/Company ($): 110
Net Profit to Brand Owners ($): +20

Under typical conditions for registrants/shareholders, option 1 is the best. Companies can then translate the selected option down the organization by using simpler language:

GM might tell employees that their objective is to beat Honda.

Sometimes, a seemingly clear language can be lost in the translation. For example, Jack Welch, former GE's CEO, translated value maximization to mean that GE should be number 1 or 2 in each of its markets. Sounds great, but it was a bad translation as each manager defined their market in a way that best suited them in meeting the objective.

However, under certain conditions, shareholders may prefer options 2. For example, if the company produces pollution as a byproduct (Exxon learned the hard way), or on moral grounds, such as not investing in South Africa during apartheid even when the decision would generate lower value than that of the otherwise best option.

Case 2: A non-profit organization

Objective: Value maximization is not necessarily the main objective. The objective needs to take into account the impact on all stakeholders.

Option Selection: Using the same example,

Option 1 -
Net effect on all stakeholers ($): 95

Option 2 -
Net effect on all stakeholers ($): 130

The company/ICANN should choose alternative 2.

Case 3: Policy Consideration Example: Expanding scope of software patents

Expanding the scope of software patents may help, say, one third of the software industry whose profits depends on the sales of a proprietary finished product. However, the policy can raise the cost without increasing the benefits for the rest of the industry whose profits are based on services. Which option should be adopted?

Add Your Comments

 To post your comments, please login or create an account.



DNS Security

Sponsored byAfilias


Sponsored byWhoisXML API


Sponsored byThreat Intelligence Platform

IP Addressing

Sponsored byIPv4.Global

Brand Protection

Sponsored byAppdetex


Sponsored byVerisign

Domain Names

Sponsored byVerisign

New TLDs

Sponsored byAfilias