The Predictability Problem of the New gTLD Process (Part 1)
Updated: Sep 14
Note: Jeff Neuman is one of the two Co-Chairs of the New gTLD Subsequent Procedures Working Group within ICANN’s GNSO. However, the opinions expressed herein are my personal opinions and may not necessarily reflect the views of the Working Group, its leadership and/or Members. The Draft Final Report of the Working Group is currently out for public comment at : https://gnso.icann.org/en/drafts/draft-final-report-new-gtld-subsequent-20aug20-en.pdf.
The number one complaint we heard about the ICANN’s process to introduce new gTLDs in 2012 was that the process was unpredictable. Though years had been spent coming up with the policies and rules governing the introduction of new gTLDs, there was a perception that ICANN frequently changed the rules to serve its own needs without relying on a consistent process involving impacted stakeholders in their decisions. In addition, there was a perception that consistent lobbying of the Board on various subjects by the community, including from those that never wanted new gTLDs in the first place, caused substantial unreasonable delays especially because the ICANN Board had no mechanism (or frankly the expertise) to deal with the onslaught of pressure they were facing.
This pressure mixed in with some unforeseen issues not only resulted in changes to the new gTLD Program which many of the new gTLD applicants believed were fundamentally unfair, but, in some cases, also forced applicants to change their business models in ways from which they could not recover. Despite predictability being one of the overarching goals of the new gTLD program, such predictability was never achieved. As more and more issues were introduced and with the passage of time and natural events, the new gTLD program became less predictable than the two previous introductions of new gTLDs in 2000 and 2004. And given that those processes were based on mostly subjective criteria, that is saying a lot.
Just some (of the many) examples of changes introduced into the program after applications are submitted included:
introducing a new baseline registry agreement (more than 1 year after applications were submitted);
adding new terms and conditions for accessing testing environments and rights protection mechanisms;
adding additional legal terms (public interest commitments) on certain sensitive TLD strings including a new dispute resolution process to handle alleged violations of those commitments;
the initial prevention of registering certain second-level domains in a particular TLD because of a “possible name collision risk”; followed by a completely new mechanism being required causing delays in utilizing a TLD as a result of new name collision analysis;
the substitution of a new mechanism to place applications in a queue for evaluation;
the shutdown of the application system due to security breaches;
the introduction of new rights protection mechanisms;
changing to testing protocols to satisfy vendor providing testing services to ICANN;
Changing of back-end portals and customer support systems;
the introduction of new requirements for registries to deal with domain name abuse; and
preventing applicants that applied for a string considered to be generic from being used in a closed manner whereby registrations were limited to the applicant and its affiliates.
At the end of the day whether you believe these changes were positive or negative for the program that is not really the issue. The fact is that these changes were all introduced in an ad hoc manner, without following a standard process that ensured feedback from all impacted parties. This resulted in substantial delays for applicants to complete the evaluation process and launch their TLDs. So much so that some applicants lost faith in the program, the value of new gTLDs, and ICANN itself. It is the number one reason why many brands that applied for TLDs were forced to withdraw their applications or ultimately terminate their registry agreements.
How so? Without mentioning any specific names, I personally worked with a number of brand applicants that began preparation for their applications in 2010. Some of these projects were sponsored internally by executives in large corporate marketing departments. After nearly 2 years of delays in opening up the application window, as well as security breaches in the application system, the earliest many of these applications could start being evaluated by ICANN was not until well into 2013. But then, under a new CEO, ICANN introduced completely new registry agreements, public interest commitments, new name collision requirements and ultimately disallowing brands from using a generic term in a closed environment. Each of these changes came well after applicants paid their $185,000 application fee. Not only had ICANN already taken applicants money, but they now were changing fundamental the terms and conditions of how that asset could be used.
In parallel, as early as 2012, brands had been asking for revisions to the registry agreement that truly reflected the nature of their closed TLDs. These were vital changes needed to recognize that brand TLDs were not operating in the same manner as those TLDs offering registrations at the second level to third parties. For .brands, the sole registrant is the Registry itself.
ICANN staff and its board had no process to deal with these requests, and spent years trying to figure out from a process perspective how it could reasonably accommodate the brands (which I believe it wanted to do). Without a predictable process in place, it took until mid-2015 until .brand registries got these accommodations and were able to sign fit for purpose agreements. It is also worthwhile to note that for most of that time, applicants were not allowed to communicate directly with ICANN staff or the Board about their issues. Applicants were forced to rely on ICANN staff and Board members to figure out the issues (from many letters exchanged on these issues). All decisions were made behind closed doors based on Board Papers drafted by ICANN's legal department, most of which continue to be substantially redacted as confidential. Sure, ICANN had put out some of the changes for “public comment”, but neither applicants nor the community even knew what role the comments played in the ultimate decisions made by the Board or frankly whether those comments were ever considered.
By the time ICANN was able to figure out an ad hoc process to accommodate the brands (five years after these applicants planned to apply for and use a new gTLD), many large corporations lost interest in their applications and doubted whether they would ever be be able to get their TLD, much less use that TLD in the manner for which they planned. They lost faith in ICANN as an organization. In their eyes, ICANN lost all credibility.
Internal corporate brand TLD champions were disgraced in front of their executive teams to whom they had to fight hard to convince that spending $185,000 for a TLD was a worthwhile investment. Budgets for the use of their new gTLDs within these companies eventually disappeared and those that initially championed the TLDs internally were laid off or were otherwise no longer at those companies to continue to champion the use of their TLDs.
With the turnover of those champions, responsibility for TLD applications internally often were moved around like a hot potato within these corporations with no one and wanting responsibility for monitoring, let alone using, a TLD. There was no certainty around the timing of when these TLDs would be available or even around the rules under which these TLDs could be used. The new owners within their respective companies for the TLD applications had seen what had happened to those before them and the political hits that their predecessors took for believing in new gTLDs and for championing their value. No one wanted to put themselves out there within these companies to support ICANN or their applications. The predecessors had been badly burned and they did not want to suffer the same fate.
.Brand applicants also knew that signing the Registry Agreement was just the first of many steps towards using the TLD. Brands that signed their Registry Agreements in 2015 still had to go through testing (which had a number of months back-up) before delegation. And once delegated, these TLDs had to wait another quarter of a year before it could start having names resolve (as a result of fears over something called “name collision”).
In reality, therefore, it really wasn’t until 2016 that most .brands could actually start using their TLDs. This was at least 6 years after beginning to plan for obtaining and using a TLD. 6 years is many lifetimes in corporate speak. Think about it. Six years ago, there was no apple watch, “the cloud” was still a good idea looking to replace private data centers, Tiktok was not yet released, there was no Amazon Alexa, etc.
All of this illustrates how important it is that the next time we introduce new gTLDs, we get it right. The process MUST be predictable. Changes to the program should not be made lightly midstream. Changes should also only be made when absolutely necessary AND must involve getting the appropriate feedback from stakeholders that will be most impacted by the changes. I believe ICANN has one more shot at this. If the process is predictable and operates as planned, then the program will gain its needed legitimacy. ICANN will also earn back substantial credibility. If not, I am not sure what fate lies ahead for new gTLDs and ICANN as an organization.
But there is hope!
Stay tuned for Part 2 where I go into detail on how the Working Group on Subsequent Procedures for Introducing New gTLDs attempts to address the Predictability Problem. If you want to read it for yourself, go to https://gnso.icann.org/en/drafts/draft-final-report-new-gtld-subsequent-20aug20-en.pdf, Topic 2.