--

Abigail H. Goben, MLS

Associate Professor

Information Services and Liaison Librarian

Library of the Health Sciences

University of Illinois at Chicago

1750 W. Polk (MC 763)

Chicago, IL 60612

[log in to unmask]

Hi Eliza, 

Thanks for reaching out to the ToS Listserv about this issue.  

I'm sorry that you're in a crisis-like situation—so stressful! 

I can share some of my experience, which hopefully can help bring new ideas into discussions with your administrators.

• Do you have any experience using alternatives to GA for your library web analytics? What positive or negative results have you observed or experienced?

• My library installed a pilot implementation of Piwik a few years ago, but did not move forward due to the massive server load required to operate Piwik locally. Our systems team reported that we would need a major upgrade to our web server infrastructure to locally host Piwik. (Frankly, this provided a newfound respect for the processing power of Google Analytics, especially their real-time stats.) In response to this, Piwik/Matomo now offers a cloud-hosted version, but for a fee. This is the great advantage of Google Analytics—it's attractive to administrators because it looks free in terms of technical infrastructure and subscription costs. Of course, we know that the tool trades on user data as its currency. 

• Do you have any experience making the case to administrators that your library should avoid GA? How did it go?

• I have spoken with my dean about this, and I would say that there are three main reasons keeping us from moving away from GA:

• Status quo: Our library has been using Google Analytics for several years, and its implementation predates my arrival. Keeping things as they are is, of course, a powerful motivator.
• University tie-in: Our library website is also connected to our university website via Google Analytics, so that keeps us tied to the service and would have to be accounted for if we moved to Matomo or Open Web Analytics. 
• Tool ignorance: Not many people know about alternatives to GA, or about the privacy implications of GA. If you can show that Piwik/Matomo offers the same data and analysis that you would need from GA, I think that could be compelling when combined with the privacy-positive argument. 

• Any (article-length or shorter) readings you would particularly recommend to librarians and/or administrators who want to understand the issues at stake? (See below for our current collection.)

• The two resources that Andy shared (thanks Andy!) could definitely help your conversations. I'm happy to talk more about anything you read there, especially the key point that under certain conditions the Google Analytics tracker participates in the vast network of third-party trackers that are capable of leaking user data across the web.
• Other resources that might be helpful:

• Data Analytics and Patron Privacy in Libraries: A Balancing Act, Becky Yoose
• Balancing Privacy and Strategic Planning Needs: A Case Study in De-Identification of Patron Data, Becky Yoose
• Which Students Get to Have Privacy?, danah boyd 
• Library Values & Privacy Report, Michael Zimmer and Bonnie Tijerina 
• Librarians, Act Now to Protect Your Users (Before It’s Too Late), EFF

• Any other advice or insight you’d like to offer?

• I would ask your administrators to articulate what Google Analytics offers that Piwik/Matomo does not, in terms of user data and web traffic analysis. If they don't know that exactly, I would then ask which metrics are specifically important to them. Essentially—what do they want to measure and why? Based on their response, you might be able to demonstrate that Piwik/Matomo is a perfectly capable tool. 

Good luck with your meeting next week! Let me know if any follow-up questions or thoughts come up before then, and keep in touch with the ToS Listserv as the situation develops. 

All best,