How do libraries determine if users are accessing subscribed eresources using Google?

Is there a way to determine the number of users accessing full text of library subscribed database via Google search, or via Google Scholar?

aarontay

• statistics

Comments

• jdscott50: Can users get to library databases via google? I thought that was hidden by the invisible web?
• Ed Summers: s/Googel/Google/
• aarontay: You can google for an article title, it might bring you to a sciencedirect page or jstor page. The you can login using athens or manually add the ezproxy stem etc http://www.lib.umich.edu/mlibrary-labs/proxy-server-bookmarklet
• Peter Murray: Ah, that isn't a 'database' in he sense of an index/abstract service. (That is what I think of as a 'database'.) 'digital-collections' may not be the best tag either because that is commonly used for library-hosted content.
• aarontay: hmm wouldn't databases include full-text databases? I would think for most users the idea that a library database is I&A only is a more alien idea than full-text databases. Anyway yeah, digital collections is not suitable, I couldn't find a better tag and I don't have enough points to add new tags

Answer by Ben Ostrowsky

It looks to me like you can use your own link resolver with Google Scholar's Library Links program:

I have a home grown link resolver. Can I still sign up for the Library Links program?

Yes, but it will take some work. First, edit this file and put it up on your library website. Next, make your link resolver export your electronic holdings in this format and put those files up on your website as well. Finally, email us the URLs for these files. We'll review your configuration and get back to you within a week or two. Once we add you to the program, our robots will periodically visit your URLs and update the links on our website appropriately.

Your link resolver can then track the HTTP referer and other juicy tidbits of information about who's using which databases and how they found them. If your authentication page happens to include some Google Analytics code as well, so much the better.

If you want the same Google Analytics information about on-campus users who are authorized by IP address, you might consider an interstitial page saying "Now forwarding you to (resource)... click here if it doesn't load", and have JavaScript do the forwarding after Google Analytics exits. But tread carefully here, as you'd be inconveniencing your users in order to get better statistics, in order to learn how to better serve your users, when one important way to better serve your users is to — wait for it — avoid inconveniencing your users!

Comments

• aarontay: Yes, we do use the Google analytics code on the authentication page and also yes using link resolvers using the google library links program. We don't allow on-campus users authorized by IP so not sure how that works, all have to use ezproxy to gain access.