Home / Blogs

Switching on the Light: Expression of Interest for New TLDs

They say late converts are the most passionate believers. Until now I haven’t supported the Expression of Interest (EOI) for new TLDs, the proposed mechanism to measure the number and type of likely applications. Not because it won’t work (I think it’ll work fine) but because I didn’t think it was necessary. I’ve changed my mind. Here’s why.

The EOI was conceived to counter claims from the anti-TLD crowd that tens of thousands of TLDs will ‘flood the root infrastructure’ and ‘dozens’ of TLDs will create morality, trademark and geographic problems at the top level. Having spent the better part of three years talking with potential TLD applicants, I know these claims are untrue. There are two main reasons we won’t see huge volumes of TLDs or a lot of ‘problem’ strings: (1) obtaining and operating a TLD is very expensive; and (2) there are carefully constructed procedures to eliminate problematic applications.

Factoring in all the costs, it takes an applicant at least $500,000 to get a TLD ready for root entry. Given this cost, only a fruitcake would apply for a string that faces legitimate trademark, morality, community, confusing similarity or geographic concerns. And there aren’t that many fruitcakes with $500,000. Even someone wanting to make a political statement has much faster and cheaper ways to do it than applying for a TLD. Will some applications generate problems? Perhaps a few. Most strings will be very carefully chosen by their applicants to move through the process without objections. Talk to anyone who’s dealt with a lot of applicants and you’ll hear the same thing.

Will thousands of new strings ‘flood the root infrastructure’? Not from what I see. It’s expensive, and there simply aren’t that many viable applicants out there. Also, many applications will be duplicates for the same string (e.g. something popular like .BLOG might see a half dozen EOIs) so the number of EOI applications will be greater than the number of actual strings entering the root. If more than 300 unique strings emerge from the EOI process I’ll buy everyone a drink at the ICANN Latin America meeting in December.

Why did I become an EOI supporter? First, some people don’t seem to believe what I explained in the previous paragraphs. I continue to see exaggerated claims of TLD volumes/ problems, and I continue to hear this alarmism repeated back to me when I visit political offices in Washington DC. It’s as if someone’s actively spreading disinformation there. Second, there’s no reason not to have an EOI. As long as it’s mandatory it will produce precise planning data, and there are no harmful effects from it.

I don’t think anyone can object to the EOI’s principles. It’s about broader communication and more accurate information, and those can only be good things. If you believe there’ll be rampant trademark infringement at the top level, why oppose an EOI that measures this? If you think there’ll be widespread morality concerns, why oppose an EOI that provides data on this? If you worry there’ll be thousands of TLDs flooding the root infrastructure, why oppose the precise measurement of actual TLD load?

As the EOI can’t be opposed on principle, the arguments against it tend to be peripheral, and weak. Here are the ones I’ve seen:

‘It Will Favor ICANN Insiders’

First, as others have pointed out, the majority of ICANN insiders don’t want new TLDs. Insiders have been around a long time and have typically found comfortable niches with the status quo. New TLDs upset their world. The majority of TLD applicants are very new to ICANN—I wouldn’t call them insiders.

Also, there’ll be an extensive communications campaign before the EOI window opens and this campaign will educate a far wider audience than current ICANN followers. It will be essentially the same campaign that would have preceded the opening of the full application window (had there not been an EOI). The campaign will alert new people that TLDs are happening and tell them how the application process works. Given that it’s the same campaign (the same size and duration) why would having it in conjunction with an EOI favor ‘insiders’ any more than having it in conjunction with the full application window? This criticism isn’t logical. To maximize community awareness and involvement, and to get timely EOI data, we should start the communications campaign as soon as possible and run it for as long as necessary. The planning for it is already completed. It’s ready to go.

‘EOI Will Create a Market in TLD Slots’

This doesn’t make much sense. Here’s why:

a. The Rules Are Still the Rules. Even if an EOI position is sold the applying entity cannot change. The new owner of the entity must continue meet all DAG obligations (including the ‘Applicant Background’ requirements at evaluation question 11). So we’re not talking bait and switch. Evaluation of the application (and any objections to it based on its ownership) don’t start with the EOI, they start when the full application is submitted. Given this, why do we care if there’s a new owner for the EOI entity? If George forms Acme Registry Company, submits an EOI, and then sells Acme Registry Company to Susan, why do we care if Susan is the new owner, as long as Susan meets all the evaluation criteria? If the concern is that Susan didn’t know about the EOI process and bought Bob’s company for an inflated price let’s solve that problem by launching a strong communications campaign as soon as possible. That will educate Susan and others to form their own companies and submit their own EOIs.

b. It Won’t Happen Any More Than it Would Without EOIs. The EOI doesn’t increase the likelihood of TLD applicants being bought or sold. The cost of a ‘place at the table’ has always been $55K. Under the pre-EOI model you submitted $185K but you could quickly get $130K back on refund—so it’s always cost $55K. Having an EOI won’t increase any potential speculation.

c. Speculation is Very Risky. Submitting an EOI purely for speculative sale is extremely risky. If it’s a popular string, there will be multiple applicants and you may find yourself with no string and no $55K. If it’s a less popular string, and you’re the only applicant, you may not find a buyer for it. If you’re trying to sell the string to someone with community, geographic or trademark rights, they don’t have to buy it from you, they can just Object to you (which costs them much less than buying it). The net result—speculation is very risky, and the payoff is low, so it’s unlikely to happen with any frequency.

‘EOI is Too Expensive for Some Applicants’

As a self-employed person I sympathize with cost concerns, but this isn’t a criticism of the EOI. It’s a criticism of the cost of the TLD program. The EOI hasn’t added to applicant cost. Previously, the fees for getting a TLD were $185K. With EOI, the fees are $55K plus another $130K if a full application is submitted. The total in both cases is $185K. Applicants will actually be better off with the EOI because it involves fee payment in two phases, rather than all up-front.

So this criticism isn’t about the EOI. It’s a call for ICANN to start subsidizing some applicants (or somehow guess which applications will be cheaper to process). Subsidization is a very separate issue from the EOI, but let me make one important observation about it.

Running a new gTLD registry will be very expensive and the $55K EOI fee is the tip of the cost iceberg for applicants. The DAG has been loaded up with extensive new requirements that add to operating cost (e.g. DNSSEC, financial continuity instrument, numerous trademark and other rights protection mechanisms, malicious activity mechanisms, audits, data escrow, compliance reports, etc.). If a decision is to made to subsidize someone because they can’t afford the $55K EOI fee then whoever is subsidizing that applicant should also plan to subsidize their ongoing registry operations. For example, this same applicant will almost certainly need to be subsidized for the Continuity Instrument at DAG Evaluation Question 50 (this instrument requires up-front funding for three years of registry operation). EOI cost is insignificant compared to overall registry cost.

In summary, I think the benefits of the EOI significantly outweigh the criticisms of it. I encourage the Board to vote ‘Yes’ in Nairobi and start the communication campaign immediately.

The Communications Campaign Should Begin Immediately

Approving an EOI without committing to the communications campaign is meaningless. Here’s why:


The Sooner We Get EOI Results The More Useful They’ll Be

Operational readiness and root scaling planning is going on right now, but it’s based on assumptions (e.g. how many contention sets will there be? how many morality objections should we plan for? what size hamsters should be on the wheel powering the XYZ root server?). The sooner we get EOI data the more useful this data will be as an input to planning—and we can’t get full EOI data till we’ve performed the communications campaign.


Greater Public Awareness is a Good Thing

The communications campaign will educate new people that TLDs are happening and how they’re happening. Let’s involve those people now so we get maximum participation in the process and maximum involvement in the outstanding issues. Why not involve new people now?


Starting the Communications Campaign Isn’t a Policy Issue

Whether or not we should have a Communications Campaign is policy—but that policy has been decided in the affirmative. ‘When’ the campaign starts is an implementation decision. Let’s stop emasculating the staff by making every implementation step a policy matter.


Starting the Communication Campaign Doesn’t Set an EOI Closing Date

Opening the communications campaign doesn’t mean we’re tied to dates for the EOI window. We can always extend the communications period or we can delay closing the EOI window. Also, we could have the EOI window open during (and close after) the communications campaign. This would give us EOI data sooner, and would add a useful sense of urgency to the communications message.

As the staff report indicates, not every issue has to be resolved before accepting EOIs. Let’s take the thorniest remaining issue - vertical integration. Let’s base the EOI and DAG on compromise language developed in 2009 (the 100K name limit) or on language from existing contracts such as MOBI and TEL. Either of these approaches are reasonable compromises. If the PDP ends up changing things we’ll adjust.

I’m hoping the Board approves the EOI in Nairobi and starts the Communications Campaign immediately. We’ve been hearing about the TLD boogieman for years now. Let’s switch on the EOI lights and see if he’s real.

Filed Under


Commenting is not available in this channel entry.
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

Brand Protection

Sponsored byCSC

Domain Names

Sponsored byVerisign

Threat Intelligence

Sponsored byWhoisXML API


Sponsored byDNIB.com


Sponsored byVerisign

New TLDs

Sponsored byRadix