Comments on new gTLD Draft Applicant Guidebook
Thank you for the opportunity to submit comments on the gTLD Draft Applicant Guidebook. These comments are a redacted version of my blog post of November 20, 2008, found at http://www.namesatwork.com/blog/2008/11/20/the-4-biggest-flaws-in-icanns-new-tld-process-and-how-to-fix-them .
First, let me congratulate the ICANN Board and staff for putting together a process, that, while still flawed, begins to fulfill the promise made at ICANN's founding to promote competition and diversity in the gTLD space. It is long past due, and badly needed. There are those who would argue for no new top-level domains, but although vocal they are voices of entrenched interests. Do we need new top-level domains? Anyone who has tried to register a domain name recently (and found that not only is Choice #1 not available, but also choices #2 - #100 are not available), will answer with a resounding "YES."
That said, the Draft Guide contains some major and minor flaws. Below are what I consider to be the four major flaws, with proposals on how to fix them. I call them major flaws because unless they are fixed it will be very difficult for most applicants, even those with bona fide business plans, to attract investment or to afford ongoing costs.
MAJOR FLAW #1. ICANN can change the contract whenever it wantsICANN’S PROPOSAL: The contract between a new registry and ICANN can be changed by ICANN, in any way, at any time, for any reason, and the only appeal is pro-active action by 2/3 of the affected registries, and even that can be overridden by the Board on the elastic grounds of “security and stability” of the Internet.
SITUATION: ICANN wants to find a way to administer all the new contracts all at once, instead of having to renegotiate each one. That seems perfectly valid, although it does leave one wondering why they need $75K a year minimum from each registry if they’re not even going to administer the contracts individually. More on that below.
MY TAKE: The idea that you would sign a contract that the other party can change at will is insane. This is not a serious proposal. Instead, I sense a negotiating position. If the entire new TLD effort is to be cost-neutral (as is claimed in justification for the $185K application fee), then the ongoing cost of administering contracts has to be covered. So if people object to this provision, ICANN can claim that it then needs the other thing that everyone wants to be rid of: the $75K annual fee.
THE FIX: Let ICANN make changes to all contracts at once, but the changes have to be approved by 2/3 of the affected registries. It still takes 2/3 of the registries to make a change, but it shifts the burden and will make ICANN an advocate instead of a dictator.
MAJOR FLAW #2. The $75K/5% annual feeICANN’S PROPOSAL: New registries will pay $75,000 per year, or 5% of their domain revenue, whichever is greater, to be paid quarterly.
SITUATION: This fee has united everyone from small non-commercial users to large domain companies — in opposition. Unlike the staggeringly long calculus used to defend the $185K application fee, the Guide provides absolutely no justification for this punitive and fundamentally unfair fee.
MY TAKE: A source who wishes to keep his job told me that a figure of $20K/year was being discussed just days before someone came up with $75K. I can only imagine that this is another negotiating position. If community outrage reduces it to $50K, then that’s still a lot of money to ICANN. The problem with this provision is that it effectively scuttles all applications which don’t have big-business ambitions. A perfectly valid TLD could be started for a small linguistic or ethnic community, for instance an American Indian tribe, or speakers of Aramaic. This TLD might never aspire to more than 5,000 registrations. They might be able to come up with the $185K initial fee, but the ongoing fee would doom them. This fee is an abomination and would prevent much of what ICANN says it wishes to accomplish, namely diversity and competition.
THE FIX: A (lower) percentage of revenue, or (better) a low per-domain fee, with no minimum.
MAJOR FLAW #3. Demonstration of Financial CapabilityICANN’S PROPOSAL: “The criteria and evaluation should be as objective as possible” but “purely objective criteria such as a requirement for a certain amount of cash on hand will not provide for the flexibility to consider different business models.” In other words, the core of the application is not objective at all. (See Attachment to Module 2: Evaluation Questions and Criteria)
SITUATION: Unknown panelist(s) of unknown domain name experience and unknown business experience; who may or may not understand how business works in your country; who may or may not have any experience with startups; who may or may not come from a place where every executive needs ten assistants; who may or may not think that everyone should work from a physical office; and who may or may not “get” your business plan — this unknown person or persons will decide the fate of your application.
Furthermore, you cannot communicate with them, and, if they decide they want to communicate with you (to ask a question, for instance), they can only do so once, and only through the as-yet-untested online workflow application (called TAS). I quote: “Evaluators are entitled, but not obliged, to request further information or evidence from an applicant, and any such request will be made solely through TAS, rather than by direct means such as phone, letter, email, or other similar means. Only one exchange of information between the applicant and the evaluators may take place within the Initial Evaluation period.”
Finally, in the event that you are rejected, you have to pay for an extended evaluation, where exactly the same panel will once again decide your fate. You may or may not be told why you failed, and if they do tell why, the reasons may not be specific, leaving you without any way of knowing how to improve the application. In this extended round, again there is only one round of communication, and in the event you are failed a second time — again, without any promise that you’ll learn why — you’re finished.
MY TAKE: This procedure is a failure of imagination — or common sense — on ICANN’s part. My $185K ought to at least give me an assurance that my application will be reviewed by someone competent and knowledgeable about my field and about business conditions in my region. Simple fairness dictates that if rejected, I should know that I’m going to get a reasonably specific justification. The “one-time, one-way” communication rule defies all common sense, and flies in the face of all accepted practices of judging business soundness. Finally, the appeals process (extended evaluation), which ought to be a stop- gap in case this ill-conceived procedure yields bad results, suffers from the same structural flaws as the initial evaluation.
In addition, the procedure encourages gaming. It is clearly in the applicant’s interest to purposely downplay the expected success of the business, and instead produce a business plan that is the mirror opposite of the one shown to investors. Instead of large volumes of registrations, you would want to show very few. Instead of rapid growth, you would want to show minimal increases. You will want to appear as overfunded as possible, and as a result the panelists are going to see a lot of, ahem, conservative applications.
Anyone who has watched decisions that come out of the UDRP (Universal Dispute Resolution Process) knows that the exact same fact pattern can yield opposite results (see for instance Michael Geist’s paper on UDRP decisions at http://aix1.uottawa.ca/~geist/geistudrp.pdf, page 20). Like this application evaluation, the UDRP is a panel, and the panelists are supposed experts. And yet the history of the UDRP is littered with really crappy decisions. Is there any reason to expect that this process will be different, or better?
THE FIX: If (as ICANN admits) the evaluation is subjective, then make it more like a “fair” subjective evaluation, instead of this ghastly hybrid of “objective” technology (TAS) and procedure (only one back- and-forth). If panelists are going to evaluate, let applicants have some insight into who their panelist are, so they can have a clue about how they will decide, and therefore how to make their application as strong as it can be. An online or face-to-face meeting with the evaluators would be better than what is being proposed. Any investor in TLD must have grave concerns about the arbitrariness of this process. Finally, we need an appeals process that is different and fairer than the original evaluation — different people, different processes, and a higher level of scrutiny. People will spend years putting their applications and funding together, and they deserve a fair hearing.
MAJOR FLAW #4. Registrant Protection ProceduresICANN’S PROPOSAL: You must have a financial instrument (or actual cash) equivalent to 3 - 5 years of operational costs, in order to protect registrants in case of your failure.
SITUATION: ICANN wants to protect the registrant who registers a name in good faith in a given TLD. Agreed. Worthy goal. ICANN is worth nothing if it doesn’t do this.
MY TAKE: The proposed solution is lunacy. Does ICANN have 5 years of operating costs stashed away in case it goes broke? (Maybe I shouldn’t say that, I can see fees being raised again — and you could make a good argument that an ICANN failure wouldn’t affect registrants at all.) But consider: if it costs $1 million per annum to run a given TLD, that means I need $3 - $5M in the bank, or in a bond. If I go to an investor to raise $5M, the investor will very likely want to get something like $25M back in a few years. If I go to a bank and ask for a loan, that will be very expensive as well - and did anyone notice that banks are a little shy of lending these days? None of these sources of capital are going to be available to a new company, unless the principal is very well-heeled — and who wants to tie up $5M because ICANN has a case of the shivers? This approach is like asking a renter for 5 years of rent beyond the lease termination, in case the landlord can’t find another tenant.
THE FIX: ICANN mentions that it is “considering the option of a pooled registry continuity plan where the proposed registry pays a fixed fee in addition to the application fee to an outsourced entity to provide name resolution services.” Obviously this approach, though still needlessly expensive, is much more sensible. Even better would be to let registries make arrangements with other registries deemed “safe” by ICANN (e.g., VeriSign, NeuStar, Afilias, etc. etc.) to take over resolution activities in case of failure. The longer-term fix, by the way, is to accredit registry operators, and simply move domain names from one to the other in case of failure, as happens now with failed registrars.
If these problems are fixed, I think this process is workable. But these are major areas of concern.
My thanks once again to ICANN for making it possible for me and others to comment.
Antony Van Couvering Names@Work