RE: Unique sites

Thank you Aleecia!

171 sites PER MONTH seems much more reasonable (than per day as conveyed on the call).  :)

- Shane

From: Aleecia M. McDonald [mailto:aleecia@aleecia.com]
Sent: Wednesday, January 04, 2012 1:06 PM
To: Tracking Protection Working Group WG
Subject: Unique sites

On the call today we talked about average number of unique sites visited per day, and Shane asked me to follow up on that. The answer was users visit about 171 unique sites per month, based on a 2008 Nielsen study where they recorded the websites users visited at work and at home. Here's a summary of data on unique sites:

Nielsen Online reported the average number of unique websites that United States Internet users visited at home and at work during March, 2008 as 66 unique sites from work and 119 from home.[43] The overall average number of unique sites visited per person for the same time period was 171.[44] The overall figure is lower than the sum of sites visited from work and home because there is duplication. For example, imagine someone who visits Google both at work and at home. Google would appear once in the count of unique sites visited at work, plus once in the count of the unique sites visited at home, yet only be one unique site overall. As depicted in Figure 4, on average Internet users visit 52 different sites exclusively at work, 105 different sites exclusively at home, and 14 sites at both work and home.

[43] Nielsen/Net Ratings, Internet Audience Metrics, United States http://www.netratings.com/resources.jsp?section=pr_netv&nav=1 (accessed February 26, 2009) (site now updated with data reflecting the present time period).

[44] Nielsen/Net Ratings, "Nielsen Online Reports Topline U.S. Data for March 2008," news release, April 14, 2008, http://www.nielsen-online.com/pr/pr_080414.pdf.

The Nielsen data is all that's relevant, but for completeness, the summary text above is from McDonald and Cranor, "The Cost of Reading Privacy Policies," I/S: A Journal of Law and Policy for the Information Society, 2008 Privacy Year in Review issue. Pedantic note: in the original there was a typo where I wrote that the overall is count 105, when clearly that is instead just the count from home. Copy paste error, published. Sigh. I fixed it here so the text makes sense.

If you're really astonishingly bored, the full paper contains further discussion of estimating the unique sites per year, and a 2008 study of an authors' acquaintances (with all that implies for issues with generalizable data) finding an average of 390 unique sites over an average observation time of 105 days. In the end, I used an estimated range of about 1350 to 1500 unique sites per person per year. Of note: these estimates are for first party sites. Since that's what we're talking about, that should be fine.

I'm trying to understand the objection to branding or other means of user communication. That does not seem like a high hurdle, and could be as simple as adding a logo to a website template. It seems a much lower burden to ask DNT implementing companies to take on once, rather than all users for all time for all sites trying to know who owns what. User expectations is a massive improvement for companies over the single site technical approach we've seen to date. Can anyone share specific use cases you have in mind where a "discoverable" approach is necessary? I don't think I'm understanding this yet.

            Aleecia