Just back from a week of ICANN meetings in Dakar. Is it just me, or is the new top-level domain program starting to feel like a TLD triathlon, where everyone’s now jockeying for position in the final stage—a grueling marathon?

When ICANN’s board approved the new gTLD plan in Singapore, it came with the promise of small but substantive changes to improve the program. I hope that wasn’t empty rhetoric, because as exciting as the program may be, it must be improved if it is going to fulfill its promise to global Internet users and their governments.

I had the opportunity in Dakar to address ICANN’s Board on behalf of the Business Constituency, which, like many stakeholders, sees much to gain from a well-executed introduction of new gTLDs, but also sees much to lose if the new program is poorly executed. To that end, I listed four promises ICANN has to keep in its new gTLD program, at the risk of losing the multi-stakeholder model and ICANN’s future:

1) Enforceability of Registry Restrictions. Perhaps the most important promise ICANN made to wary governments is to let them object to proposed TLDs that may offend cultural, religious or national sensibilities. My concern is that the current process won’t empower ICANN to deliver on that promise.

While we are asking governments and other stakeholders to base their reaction to proposed strings on the proposed terms in their application, those terms won’t actually be enforceable unless they’re part of the formal Registry Agreement. This opens the risk that for some applicants, promised restrictions on registrants and uses of domain names could be ignored after their applications are approved. That would leave ICANN with little leverage to hold TLDs to the restrictions that were relied upon to satisfy governments and other objectors. This loophole should be closed before the first applications are accepted, or ICANN risks breaking a critical promise made to skeptical governments.

2) Serving the Next Billion Internet Users. Internationalized domain names (IDNs) are the major benefit ICANN has promised to global Internet users. Of all the gTLDs that will be created in the new gTLD process, IDNs have the greatest potential to reach the next billion, most of whom don’t use the Latin alphabet as their primary script. But based on early activity in the new gTLD community, it seems like IDN applications will represent only a small fraction of the total applicant pool. It would be missed opportunity if IDN users did not see recognizable TLDs coming out of this new gTLD program. ICANN has fielded a range of proposals to increase IDNs in this round (including targeted discounts and bundling of applications) but has thus far shown no interest. Even a little bit of progress on this front could pay huge dividends to underserved Internet users, and to the governments that represent them at the UN.

3) Protecting Rights, Preventing Fraud. The issue of rights protection and fraud prevention has been a centerpiece of the new gTLD debate, and still dominates discussion about the program outside of the ICANN community. The good news is that the ICANN community developed a suite of rights protections mechanisms that should go a long way toward minimizing fraud and abuse on new gTLDs. But one of the most valuable of these mechanisms—trademark claims notices—is only required during the sunrise period for new gTLDs. I told the board it would be a real shame if TM claims proves to work well, only to have it disappear once a TLD is opened for business. ICANN should find a way to encourage the continuation of tools like TM Claims.

4) Raising the Bar for Registrars. With all the focus on registry agreements, the new gTLD program somehow missed the opportunity strengthen ICANN’s contractual agreements with registrars who will sell and manage names in new TLDs. When millions of new registrants enter the market, it is the registrars—not registries—they will be dealing with. The new TLDs are just as big for registrars as for registries, especially now that cross-ownership and vertical integration are allowed. With the benefit of hindsight, the ICANN community probably should have pushed harder for an improved registrar agreement as a requirement to sell names in new TLDs.

I realize that the ICANN community, staff and board are all fatigued with the new gTLD process. I know I’m already tired, and the marathon stage has only just begun.

But now is not the time to stand back and let the TLD chips fall where they may. Governments, who are inherently skeptical of ICANN’s stakeholder model, will hold ICANN to its promises to maximize global benefits and minimize harms to registrants and users.

As we saw in Dakar, it won’t suffice to tell those governments that ICANN enforcement is constrained to its Guidebook or constrained by its Policy Development Process. Back in June 2010, I described it this way on CircleID:

There’s a common-sense saying, “Don’t moon the giant.” Alienating nongovernmental stakeholders is one thing, but ICANN is risking its very existence if influential governments feel alienated or ignored. That’s because governments have options other than ICANN. They can turn to other venues where their voice is law, such as the United Nations and ITU.

Generally, speaking, promises gain greater weight when your critics and rivals are just waiting to hang those promises around your neck. When it comes to meeting the promises of new gTLDs, nothing less that the future of ICANN and our multi-stakeholder model hangs in the balance.