Revisions Needed to ICANN gTLD Metrics, Stakeholders Say

ICANN “has not obtained consensus that this initiative is leading to a meaningful outcome,” Verisign said. “While engaging an economist to support this effort may be beneficial, ICANN should allow a proper economic evaluation to be performed without presumptions at the outset as to what a ‘healthy’ marketplace may be. Similarly, an economic analysis should not be limited to only those metrics which are conveniently available. Such limitations will likely yield an unreliable and potentially misleading index. The process of developing this initiative thus far has not led to a community consensus and instead appears to present only ICANN staff’s perspective.” The nonprofit should move away from seeking feedback on specific metrics to developing “a process to lead the community through developing a mutually agreed upon set of goals for a marketplace health index,” Verisign said. “Once these goals are collectively agreed upon, then data to characterize progress towards those mutually agreed upon goals can be collected."

Stakeholders questioned measurement of geographic diversity in terms of the number of registrars and registries in regions and specific jurisdictions. Focusing on the number of suppliers “does not allow for detection of undue market domination,” said the At-Large Advisory Committee (ALAC). The geographic diversity metrics “remain overly simplistic,” said the Registries Stakeholder Group (RySG). “A registrar may provide high-quality service and compete effectively across many jurisdictions beyond the one in which it is based provided that language, legal, payment and other issues particular to that jurisdiction are taken into account.”

The jurisdiction “is even less relevant when applied to registries, because the registry operator is not generally the primary party engaged in customer support, payments, or other interactions that are highly affected by jurisdiction,” RySG said. The registry geographic diversity metric “may be interpreted as a way to indicate where registry operators are able to be successful but it does not effectively measure where registrants and domain users do and do not have choice,” said Verisign. The Generic Names Supporting Organization Business Constituency (BC) said it believes the geographic metrics “are currently not reflective of reality within regions, but provide a good beginning view intra-regionally.”

ICANN’s competition metrics are more helpful, but more details are needed, some said. “True competition in a market is not solely a measure of the market offering but it also revolves around the share of market from the leading competitors,” said ALAC. Each registrar and registry operator family is “counted once, then added to the number of independent gTLD registrars or gTLD registry operators,” the BC said. “It is desired that Competition reveal registrar and registry operators operating independently vs. part of larger families with a corporate parent, the latter of which ICANN’s infographics provide.”

Verisign encouraged ICANN to expand the competition metrics to include the country code TLD market. Donuts said it’s concerned the competition metrics “presume that only growth in these numbers will indicate marketplace health.” The domain name industry could, for example, enter “a period of consolidation, where the absolute number of providers decreases, but products, services and marketplace penetration expand,” Donuts said. “Alternative points of reference in such instances could be useful.”

ICANN’s market stability metrics require further revisions since “it seems that voluntary vs. involuntary de-accreditations will be difficult to separate, as abandonment and failure to pay fees can constitute an intentional opt-out,” the BC said. Definitions for both trust and stability “need to be defined relative to the audience that needs to trust the marketplace and perceive it to be stable,” Verisign said. “Evaluating metrics as they relate to trust and stability without a clear audience defined is not possible and will not yield meaningful or reliable data.” Additional “granularity may be helpful in understanding the data," said the International Trademark Association (INTA). "A summary of the reasons why registrars were de-accredited, including both voluntary and involuntary de-accreditation, would paint a more useful picture.”

Donuts said it’s still concerned about using the accuracy of WHOIS records as a metric because “this statistic is not reliable.” The WHOIS Accuracy Reporting System “is a new technology still being developed and refined -- just recently, an error with the ZIP code accuracy process was discovered,” Donuts said. “The resulting changes yielded significantly different numbers.” ALAC said it “would be interesting to note why registrars are, voluntarily or involuntarily, deaccredited. Was that due to high ICANN fees, noncompliance/legal issues, technical incompetence, lack of interest, etc?” The BC said it supports use of the WHOIS accuracy metric.

Rightside said ICANN should focus on measuring TLD usage as a metric since usage “is a better indicator of long-term viability of a TLD than renewal rates.” Rightside also suggested ICANN measure the concentration of registrants per TLD and the average number of years of registration length per TLD. Donuts urged ICANN not to use pricing “as a metric in any scenario. ICANN is not a pricing authority and should not report on pricing in any format.” INTA said ICANN should add additional metrics, including “domain name renewal rates, new registration velocity, average registration and renewal prices at retail, and the number of accredited registrars per gTLD. It would also be helpful to distinguish between renewals and new registrations; to the extent this distinction is not already captured in the Index.”