At-Large Phase II Assessment of the Competitive Effects Associated with the New gTLD Program Workspace

AL-ALAC-ST-1216-01-01-EN.pdf

ALAC Statement on the Phase II Assessment of the Competitive Effects Associated with the New gTLD Program[1]

The genesis of this Assessment comes the Affirmation of Commitment that required ICANN ‘promote competition, consumer trust and consumer choice in the DNS marketplace’. With the introduction of new gTLDs, ICANN commissioned the Analysis Group to assess the impact of the new gTLDS program’s impact on competition in the domain name marketplace.  Phase I of their assessment was published in September 2015, a year after the introduction of new gTLDs.  This is Phase II of their assessment and measures outcomes against the baseline findings it made a year before.

‘Competition’ is only one of the three areas of impact that has been assessed with the introduction of new gTLDs. But based on this Assessment and its attempt to measure ‘competition’, there is no clear case that end users have benefitted from the expanded DNS marketplace.

The Report is unequivocal that the introduction of new gTLDs have not impacted either price or non-price competition. The first issue raised by the Assessment is what they mean by ‘competition’.  Competition in a market place is generally taken to mean a marketplace in which goods that are substitutable for each other competing against each other.  Clearly, that classic understanding of the term does not apply in this situation.  What the Assessment does instead is look for changes in the DNS marketplace in both price and non-price terms.

Arguably, an expansion in the DNS market could be seen to benefit end users of the Internet. Indeed, the Assessment points to its possible benefits: the introduction of new registries, new registrars, a decrease in prices and greater choice of domain names.

However, the outcomes of the Assessment are, as best, equivocal. As it concludes:

While we are unable to draw conclusions about whether the New gTLD Program has caused a change in competition in the domain name marketplace, some of these changes in the past year are consistent with what one would expect to see in a market place with increased competition.[2]  

The most useful information from the Assessment is about the changes in registry numbers and their share of registrations. Based on the sample of new gTLDs studied [3] the greatest change was in the set of entities included in the largest 15 registries and registrars, and a decline in the share of registrations held by the top 15 registries and registrars between Phase I and Phase II. The largest percent in registry operators was in the Asia Pacific and European regions. Specifically, the top registrar, by share of new gTLD registrations in Phase II went from 2.9% to 24.8% between Phases I and II, with the next largest jump from .0% to 6.3%. [4]  Another potentially positive outcome of the study is that new gTLDs target registrants with a variety of interests.[5]

The Assessment’s conclusion on the impact of new gTLDs on the market was that there was worldwide impact of new gLDs on legacy gTLDs: registration of legacy TLDs ‘followed the same pattern before and after the beginning of the new gTLD program. It did observe, however, a decline in both new and legacy gTLDs after the entry of regional TLDs.  On those issues, it concludes:

We however do not have sufficient data to fully analyze the substitutability of new gTLDs for the legacy TLDs.[6]

A recent study on the impact of new gTLDs was far less positive about the introduction of new gTLDs. Their research found that:

…only 15% of domains in the new TLDs show characteristics consistent with primary registrations, while the rest are promotional, speculative, or defensive in nature; indeed, 16% of domains with DNS records do not even resolve yet, and 32% are parked.[7]

The definition they use of ‘parked’ domains is ones that are ‘owned by an ad network or are for sale by their owners and typically return Web pages dominated by ads.[8] Their conclusion: the new gTLDs have ‘yet to provide value to the Internet community in the same way as legacy space’.[9]

If the outcomes on non-price ‘competition’ are questionable, outcomes on price are more so. As the Assessment report notes, requests for current and historical pricing data were sent to all registrars. Only 14 percent of registration volume of the new gTLD sampled was provided in both Phase I and II.  In Phase II, the study had to rely on publicly available retail price data but, as the report notes, because the researchers had to rely on publicly available data (covering 39 of the 54 registrars sampled) the analysis of retail pricing ‘may not be representative’.[10] The charts on pricing, therefore, are necessarily on “Collected” information.[11]

Clearly, this Assessment was carefully done, with as much comparison with Phase I outcomes as could be reasonably be drawn. What the Phase II outcomes do not do, however, is to make a clear case for significant benefits arising from increased ‘competition’ with the introduction of new gTLDs.  

As the Assessment itself concludes, they were

… unable to draw conclusions about whether the new gTLD Program has caused a change in competition.[12]

At the introduction of new gTLDs, the ALAC argued for a range of metrics to be used in assessing the success (or otherwise) of the introduction of new gTLDs under all three headings: competition, choice and consumer trust.  At that time, the ALAC proposed metrics for consumer trust and choice that could be used in assessing their achievement:

• End user confusion
• Growth in the use of domain based and not domain based alternatives for Internet resource access
• Complaints to, and action taken in relation to new gTLDs
• Transparency of contact information
• Accuracy of promotion of new gTLDs to end users
• Technical issues encountered.

We are now aware that in relation to end user complaints pertaining to DNS abuse and action taken in respect of gTLDs in general and new gTLDs in particular, there are significant differences between incidents and their impact reported by ICANN compared to other reputable independent agents.

The ALAC awaits the testing of the expanded DNS marketplace against metrics for consumer trust and choice. This is particularly true for developing and under developed economies that lack the basic infrastructure and know-how to take advantage of such opportunities.

In conclusion, the outcomes of Assessment are, at best, equivocal. While there has been some expansion in registry numbers and new market entrants, only 15% of the new domains have the characteristics of primary registration. From an end user perspective, most of the resultant new registrations are speculative, defensive, unused or parked – adding little of value to end users. And from an industry, there is no clear evidence of lower prices or more choice. Based on this Assessment, there is little evidence of benefit to end users with the introduction of new gTLDs.

The ALAC has argued for a range of metrics to be used in assessing the success (or otherwise) of the introduction of new gTLDs under three headings: competition, choice and consumer trust.  Prior to the introduction of new gTLDs, the ALAC proposed metrics that should be used in assessing the achievement of consumer trust, as follows:

• End user confusion
• Growth in the use of domain based and not domain based alternatives for Internet resource access
• Complaints to, and action taken in relation to new gTLDs
• Transparency of contact information
• Accuracy of promotion of new gTLDs to end users
• Technical issues encountered.

The ALAC was disappointed that additional metrics were not adopted to fully determine the impact of new gTLDs on one of the three metrics - consumer trust. 

Nevertheless, increased competition that leads to a decrease in price and greater choice could be seen to benefit end users. Phase I, undertaken in 2014 with the introduction of the new gTLDs was undertaken to provided a benchmark against which impacts of the new gTLDs can be measured. The Phase II Assessment outcomes, therefore, provide useful information in determining the extent to which new gTLDs have resulted in overall benefits to end users.  

The outcomes of Assessment II however are, as best, equivocal.  As that Assessment concludes:

While we are unable to draw conclusions about whether the New gTLD Program has caused a change in competition in the domain name marketplace, some of these changes in the past year are consistent with what one would expect to see in a market place with increased competition (P. 2).

The most useful information from Assessment II is about the changes in registry numbers and their share of registrations. Based on the sample of new gTLDs studied [1] the greatest change was in the set of entities included in the largest 15 registries and registrars, and a declide in the share of registrations held by the top 15 registries and registrars between Phase I and Phase II. The largest percent in registry operators was in the Asia Pacific and European regions. Specifically, the top registrar, by share of new gRLD registrations in Phase II went from 2.9% to 24.8% between Phases I and II, with the next largest jump from .0% to 6.3%. (see Table 3D, p 28).  Another potentially positive outcome of the study is that new gTLDs target registrants with a variety of interests.(p. 2)

The Study’s conclusions on the impact of new gTLDs on the market was that there was worldwide impact of new gLDs on legacy gTLDs: registration of legacy TLDs ‘followed the same pattern before and after the beginning of the new gTLD program. (P. 53) The studydid observe, however, a decline in both new and legacy gTLDs after the entry of regional TLDs. (p. 53) On on those issues, the study concludes:

We however do not have sufficient data to fully analyze the substitutability of new gTLDs for the legacy TLDs (p. 53)

It is more difficult to draw any conclusions on the impacts of new gTLDs on price. As the Assessment II report notes, requests for current and historical pricing data were sent to all  registrars.  Only 14 per cent of registration volume of the new gTLD sampled was provided in both Phase I and II.  In Phase   II, the study had to rely on publicly available retail price data, but, as the report notes, because the researchers had to rely on publicly available data (covering 39 or the 54 registrars) the analysis of retail pricing ‘may not be representative’.(p. 10) The charts on pricing, therefore, are necessarily on “Collected” information. (see charts pp 12-15.)

Clearly, Phase II of this program was carefully done, with as much comparison with Phase I outcomes as could be reasonably be drawn. What the Phase II outcomes do not do, however, is make a clear case for significant benefits arising from increased competition with the introduction of new gTLDs.   As the Study itself concludes, they were

… unable to draw conclusions about whether the new gTLD Program has caused a change in competition. (P. 2)

The ALAC, therefore, believes that, until clear evidence of increased choice and consumer trust can be established as arising from the introduction of new gTLDs,  the case for introducing more new gTLDs has not been established.