0. Phase II Assessment of the Competitive Effects Associated with ICANN’s New gTLD Program
Prepared for: The CCT Review
August 29, 2016

1. Study Goals & Who We Are Study goals:
Understand the competitive effects of ICANN’s New gTLD Program on the marketplace for domain names by analyzing changes in competition metrics.
Our team includes:
Catherine Tucker, Mark Hyman Jr. Career Development Professor and Associate Professor of Management Science, Sloan School of Management, Massachusetts Institute of Technology.
Greg Rafert, Vice President, Analysis Group, a firm specialized in economics, health care analytics, and strategy consulting for Fortune companies, global health care corporations, government agencies and law firms.

2. Our Approach
To assess the extent to which ICANN’s New gTLD Program affected competition, we first published a report approximately one year ago (the Phase I Assessment) that established a baseline description of metrics that would be used to assess, in the future, changes in the competitive conditions in the marketplace for domain names.
Examples of metrics include prices charged by registries to registrars (wholesale prices), prices charged by registrars (retail prices), retail markups, changes in registration quantities and shares of legacy and new gTLDs over time, and registration shares held by registries and registrars.
Now, in our Phase II Assessment, we assess the extent to which the New gTLD Program has affected competition in this marketplace over the past year by analyzing changes in these metrics.

3. Examples of Data Relied On
Historical registration volume data were obtained from monthly transaction reports, covering all legacy TLDs and new gTLDs.
We also obtained WhoIs registration data from DomainTools for specific geographic regions, such as New York City, to analyze the impact of geo- TLD entry on legacy and new gTLD registration activity in these regions.
Historical wholesale price data were provided by registries for the majority of new gTLDs in our sample, while wholesale prices for legacy TLDs were obtained from ICANN.
Registrars did not provide sufficient transaction-level data in response to our data requests (almost no transaction-level data were received). Thus, for retail prices for TLDs in our sample, we relied on publically available list prices for one-year registrations and add-on offerings.

4. Summary of Findings
With respect to changes in wholesale and retail prices, as well as changes in retail markups (over wholesale prices), we find:
Average and median retail prices for new gTLD registrations, as well as new gTLD retail mark-ups over wholesale prices, have declined since Phase I.
For example, median retail markups for new gTLDs declined from 85% to 74% (from Phase I to Phase II).
There has been zero, or at most a very small decline, of wholesale prices for new gTLDs over the last year.
For example, the median wholesale price is equal to $20 in both Phase I and Phase II, while the average wholesale price declines from $ to $21.46 (from Phase I to Phase II).

5. Summary of Findings (Continued)
With respect to changes in the registration shares held by registries and registrars, we observe:
Declines in the share of all registrations (new gTLD and legacy TLD) held by the top 4, top 8, and top 15 registries and registrars between Phase I and Phase II.
For example, the share of registrations held by the top 4 registries declined from 97.9% to 90.7% (from Phase I to Phase II).
Significant declines in the share of gTLD registrations held by the top 4, top 8, and top 15 registries and registrars between Phase I and Phase II.
For example, the share of registrations held by the top 4 registries declined from 66.6% to 31.5% (from Phase I to Phase II).
The largest registrar in Phase I (Network Solutions) dropped out of the top 15 registrars ranked by total new gTLD domain registrations and was replaced by a Chinese registrar whose share increased by ~22%.

6. Summary of Findings (Continued)
With respect to registration activity more generally, we find:
The total number of registrations in both new gTLDs and legacy TLDs continues to increase.
There is no identifiable effect of new gTLD entry or registrations on legacy TLD registration levels or growth.
In analyzing the impact of geographic TLD entry, there is a decline in new gTLD and legacy registrations after the entry of the geographic TLD in the region relevant to that TLD.
For example, after .nyc entered into the marketplace, average monthly registrations in the New York City area for both new gTLDs and legacy TLDs declined by 47% and 15%, respectively.
And finally, we find that new gTLDs continue to target registrants with a variety of interests, and the entry of new gTLDs in a given interest area is often associated with a decline in registration shares of other similar gTLDs.

7. Contacts
Catherine Tucker, Associate Professor of Management Science
Greg Rafert, Vice President