WHOIS Accuracy Rates Improving Under 2013 RAA But At What Cost?

WHOIS Accuracy Rates are slowly improving, as reflected in the latest Preliminary Findings paper [PDF, 668 KB] published by the National Opinion Research Council (NORC) at the University of Chicago.

These findings suggest that there is still more work to be done to bring up accuracy rates. After examining nearly 100,000 WHOIS records, NORC’s Preliminary Findings highlight important observations regarding the Pilot Study:

• Operationally, Registrars under the 2013 RAA have more accuracy for email addresses than Registrars under the 2009 RAA;
• New gTLDs have slightly better operational email accuracy rates than prior gTLDs;
• Prior gTLDs have more operational accuracy on telephone numbers, but the two groups are equal on operational postal address accuracy.

The fact that New gTLDs have better email accuracy really stems from the first point. That is because all New gTLDs only operate under the 2013 RAA from the start as only registrars that have signed the 2013 RAA are allowed to register new domains.

The problem is that this slightly better email accuracy has lead to a lot of domains and websites being suspended. Registrars are not ready to do the required validation and domain owners are punished for that.

The All Things WHOIS Session at ICANN 51 in Los Angeles explored these findings in detail: All Things WHOIS – Now and in the Future – Presentation.

The Preliminary Findings paper [PDF, 668 KB] describes the analysis conducted by NORC, after collaborating with intergovernmental and private sector firms. Relying on expertise from the Universal Postal Union, the UN agency responsible for coordinating global postal policies, and commercial service providers such as StrikeIron, DigiCert and Whibse, NORC applied the latest technology available to validate email addresses, telephone numbers and postal addresses to varying degrees, each at a syntactical and operational level. This expertise is reflected in the methodology adopted by NORC in assigning the different accuracy rates described in the Preliminary Findings paper.

The purpose of the Pilot Study is to test assumptions using real data. The methodology can be adjusted based upon the consultations in Los Angeles, as well as the public comments received. The paper [PDF, 668 KB] is published for the LA Meeting to demonstrate how ICANN’s WHOIS Accuracy Reporting System (ARS) is being designed.

Study Design

ICANN turned to NORC to design, work with validation providers, and conduct the analysis necessary to produce and deliver this analysis. NORC’s past accuracy studies influenced the WHOIS Review Team. The Review Team’s Final Report called for ICANN to publish ongoing statistics on WHOIS accuracy.

The Pilot study examines accuracy levels by applying syntactic validation and operation validation tests to a Registrant’s postal address, email, and telephone numbers listed in a WHOIS record. Although the study did not attempt to apply identity validation techniques, ICANN is exploring the feasibility of including identity validation in subsequent development phases of the ARS.

The sample sizes used in the Pilot Study is described in the chart below:

Data Element	Syntactic Validation	Operational Validation
Postal Address	10,000	1,000
Telephone Number	10,000	1,000
Email Address	98,821	98,821