The Office of the Federal Register, in cooperation with the Government Printing Office, provided bulk access to the source code of the Federal Register and the Code of Federal Regulations. This source code in XML was then used by three volunteers in California who created a vastly better web site for the Federal Register and made their software available as open source. The Office of the Federal Register wanted to make their own web site better and rather than reinvent the wheel, worked with those volunteers and adopted the open source platform.
The result is federalregister.gov, the daily newspaper of the Federal Government. The web site is not only vastly better than before, any other organization that wishes to replicate that web site and modify it is free to do so. And, the bulk data is available for anybody who wants to create a new kind of web site using this same information.
Agencies can observe the number of users on the web site and run standard web analytics. Better-designed web sites mean more visitors and the visitors that come tend to stay longer.
Another metric are the number of outside organizations working with the data. In addition to the GovPulse.US team which worked with OFR, there are many other groups that have now begun working with this bulk data. For example, the Cornell LII web site has started creating a new web site devoted to the Code of Federal Regulations.