Study Suggestion Number 20

[study-suggestion-response@xxxxxxxxx]

Submitted By:
[Redacted for privacy reasons]

Topic:
Timeliness of proxy services in relaying communications to registrants and/or 
revealing the identity of underlying registrants (â??licensees,â?? per RAA 
3.7.7.3).

Hypothesis:
That proxy services are untimely and unreliable conduits of communications to 
registrants, and/or that proxies are in breach of registration terms dictated 
by RAA 3.7.7.3 (requiring that registrant reveal identity of domain licensee 
upon reasonable evidence of actionable harm).

How the hypothesis could be falsified:
Demonstration with statistical significance that proxy services do three things 
at a reliability rate of 95% or higher: (1) relay communications received by 
e-mail, mail, and/or fax to registrants; (2) reveal accurate registrant names, 
physical addresses, and e-mail addresses to requesting users who present 
reasonable evidence of actionable harm; (3) transmit this data within 
forty-eight hours of receipt of request; and (3) give notice to registrants of 
a request for information within twenty-four hours by e-mail to valid e-mail 
addresses.

Utility:
Data collected could inform and quantify the need for additional regulation of 
the responsibilities of proxy services to relay communications and/or to reveal 
registrant contact information upon receiving reasonable evidence of actionable 
harm.

Type of Study Needed:
Survey of proxy registrars, brand owners and law enforcement officials screened 
for experience in requesting data from proxy registrars, and registrants 
through proxy services.  

Alternatively, a pretextually-based study of the reliability and timeliness of 
proxy services (using domains registered for the purpose of the study without 
the registrar or proxy providerâ??s knowledge) could also be considered.

Data that needs to be collected:
Similar data would be collected in two categories:  (1) communication between 
proxies and registrants; and (2) communication between proxies and data 
requesters.  The study would require three separate questionnaires screening 
and targeting each of the three classes of respondent required for the study 
(proxy registrars, proxy registrants, and data requesters with experience of 
requesting data through proxies).

Category 1 (to be sought from proxy registrars and prior / current registrants 
of domains through proxy services)

--      number of instances in 2007 in which proxy service received request for 
name, address, and e-mail address of respondent

--      number of instances in 2007 in which proxy service sent notice to 
registrant about request for name, address, and e-mail address

--      duration of proxy registration in 2007 (registrant respondent only)

--      use of domain for e-commerce, online journalism or blogging, marketing 
a product or service, political commentary, cultural commentary, or other use 
(registrant respondent only)

--      longest time between receipt of inquiry and transmission of notice to 
registrant

--      shortest time between receipt of inquiry and transmission of notice to 
registrant

--      median time between receipt of inquiry and transmission of notice to 
registrant

--      number of instances in which notice from proxy failed to reach 
registrant (bounces to proxy, proxy respondent only)

--      number of instances in which notice from proxy failed to reach 
registrant (UDRP complaint, law suit, or demand received by registrant without 
having received separate notice from proxy, registrant respondent only)

Category 2 (to be sought from data requesters and proxies)

--      number of instances in 2007 in which requester consulted WHOIS data to 
identify registrants

--      number of those instances in which registrant was identified as a proxy

--      number of instances in which proxy failed to respond to request for 
name, address, or e-mail address of registrant

--      longest time between transmission of inquiry and receipt of response

--      shortest time between transmission of inquiry and receipt of response

--      median time between transmission of inquiry and receipt of response

--      number of instances in which response from proxy contained refusal to 
comply with request

--      number of instances in which response from proxy contained inaccurate 
or incomplete data

For the alternative version of the survey, the survey provider would register a 
number of domains similar to the brands of cooperating brand owners (who would 
be transferred the domains upon conclusion of the study). The survey provider 
would be authorized to send a request to the proxy provider alleging 
infringement. The survey provider would track the transmission of the 
communications to and from the proxy provider and actions taken by the proxy 
provider in response to the pretextual allegations of actionable harm.


Population to be surveyed:
Three groups

Group 1:  registrars offering proxy services to domain registrants.

Group 2:  domain registrants who during 2007 used proxy services.

Group 3:  data requesters, consisting of brand owners, law enforcement 
officials, journalists, and consumer-protection organizations.


Sample Size:
Because the pool of available proxy registrars is comparatively small, 
participation near 100% of Group 1 should be the target to permit statistical 
inferences from responses.  For Group 2, a sample large enough to permit 
statistically significant analysis of results by use of registered site (e.g., 
e-commerce), meaning a sample of probably more than 200 respondents.  For Group 
3, likewise, a sample large enough to permit differential analysis of responses 
for each of the types of respondent (e.g., law enforcement, brand owners), 
meaning probably a sample of more than 200 respondents.

Type of Analysis:
Tabulation of raw data, regression of results for times and quality of 
responses according to respondent category.