You are here

WHOIS ARS Phase 2 Cycle 1 Overall Results Summary

This page provides high level results and sample design summary of the Phase 2 Cycle 1 Accuracy report. For the full report, see here. For additional result summaries, see here

Phase 2 Cycle 1 Findings

Phase 2 Cycle 1 examined both syntax and operability accuracy of WHOIS contact information over several dimensions, focusing on rates of conformance by contact mode (Email, Telephone or Post) to the requirements of RAAs (2009 RAA or 2013 RAA). For over 75 percent of domains, the contact information in the registrant, administrative, and technical contacts is identical for all three contact modes, revealing why accuracy rates among the three contact types are all similar.

 

Table 1. Syntax Accuracy

 

E-mail

Telephone

Postal Address

ALL 3 Accurate

All 3 Contacts Accurate

99.1% ± 0.2%

83.3% ± 0.7%

79.4% ± 0.8%

67.2% ± 0.9%

 

Our analysis finds that approximately 99 percent of email addresses, 83 percent of telephone numbers, and 79 percent of postal addresses met all of the baseline syntax requirements of the 2009 RAA, as shown in the table above. For syntax accuracy, there was a drop in telephone number accuracy from Phase 1. This change cannot be attributed to effects of Phase 1, however, since Phases 1 and 2 overlapped (i.e., Phase 2 Cycle 1 data was pulled before the conclusion of Phase 1). Other findings include:

  • The syntax reasons for error had very similar distributions to those in Phase 1.
  • The drop in telephone number accuracy seems to be due to an increase in missing country codes among the telephone numbers sampled for Phase 2.
  • For postal addresses, the vast majority of errors in both Phase 1 and Phase 2 were due to missing fields that were required such as city, state/province, postal code, or street.

 

Table 2. Operability Accuracy

 

E-mail

Telephone

Postal Address

ALL 3 Accurate

All 3 Contacts Accurate

87.1% ± 0.7%

74.0% ± 0.9%

98.0% ± 0.3%

64.7% ± 0.9%

 

In terms of operability accuracy, our analysis finds that approximately 87 percent of email addresses, 74 percent of telephone numbers, and 98 percent of postal addresses met all of the baseline operability requirements of the 2009 RAA, as shown in the table above. Sixty-five percent of domains passed all operability tests for all contact types (registrant, administrative, technical) and contact modes (email address, telephone number, postal address). Unlike for syntax accuracy, the contact mode with the highest rate of passing all operability tests was postal address. The mode with the lowest rate of passing all operability tests was telephone numbers. Other findings include:

  • Of those email addresses that failed operability, the majority (10 percent) bounced while very few were simply missing.
  • Of the telephone numbers that were present, but failed operability, there were roughly equal numbers that were disconnected, invalid, or that simply did not connect.
  • For the small numbers of postal addresses that failed operability testing, almost half did not have an identifiable or easily deduced country.

Phase 2 Cycle 1 Sampling Design

A two-stage sampling method (initial –> analyzed subsample) was designed to provide a large enough sample to reliably estimate subgroups of interest, such as ICANN region, New gTLD or Prior gTLD, and RAA type. The initial sample contained nearly 150,000 records and the analyzed subsample contained 10,000 records representing all 678 active gTLDs at the time of the report. The results from the subsample testing are used to estimate the results for the entire gTLD population or the particular subgroup of interest. Based on sampling error, there is a 95 percent chance that the true parameter (e.g., the percentage of all administrative telephone numbers with a North American registrant that pass all syntax accuracy tests) is within the 95 percent confidence interval.