Usage Statistics

From MPublishing

Jump to: navigation, search

[edit] Introduction

We have two systems for tracking usage of a publication:

  • Google Analytics: This provides visual and tabular data based on unique URLs. You can filter the data based on substrings of URLs, allowing an easy way to track usage of particular titles or particular system functions.
  • Michigan's homegrown system: This system is used for publications hosted in DLXS on The system is aware of DLXS's URL structure and is able to equate URLs that are functionally equivalent and compile data on different types of searches and different types of browsing

[edit] Google Analytics

A publishing partner's principal contact at Michigan Publishing can get you access to Google Analytics for your publication. Just provide an email address attached to a Google Account which which you'd like access to the data to be shared.

By default we set up GA to track all web pages hosted on on (where Michigan Publishing journals and other digital publications live) and, if applicable, on any pages on a domain name hosted by Michigan Publishing. If a publishing partner maintains a website on their own devoted to the publication, visits to that site are not included in our GA by default but can be added to the tracking by coordinating with your contact at Michigan Publishing.

In the summer of 2013, practicum student Clayton Hayes created a series of guides for our publishing partners who wish to use Google Analytics. These include:

These guides are available under a Creative Commons Attribution license. Please use them freely and share them widely--bearing in mind that they were designed for use by Michigan Publishing platforms and partners.

[edit] The Michigan system

Michigan's system is the successor to a different system used before May 2007. Data from before May 2007 was imported from the old stats system and is unreliable.

There are two interfaces to the Michigan statistics system (both of which have extensive help documentation explaining how to interpret the statistics):

  • – From this page, you can generate:
    • Custom reports showing usage from users anywhere and, if accessed from a known institution, usage from that institution's IP range
    • COUNTER reports documenting usage from users anywhere or, if accessed from a known institution, usage from that institution's IP range
  • – This page is only available to those who are designated as a manager of a publication. A publishing partner can request access to the manager interface for their publication from their contact at Michigan Publishing. From this page, you can generate:
    • Custom reports for any publication for which you are a manager, with statistics broken down by known institution
    • COUNTER reports documenting usage from users anywhere

The way the COUNTER Book Reports 1 and 2, and text views, are calculated was changed on Nov. 15, 2011. As of May 24, 2011, we are now recording section hits for encoded-text items (in addition to page-image items), allowing generation of COUNTER Book Report 2 data for these items. For Michigan Publishing publications, Book Report 2 is best for all monographs and for any publications with page images; use Book Report 1 for others (non-monograph publications without page images).

As of March 21, 2012, the system includes PDF views and downloads as a type of view for most publications. (Previously these views were not included at all, and certain PDF views were erroneously included in the "pg/img" subtype from Jan. 7, 2011 to March 20, 2012.) As of April 6, 2012, clicks on links to download a PDF are included in Book Report 1.

Note that if a user tries to access an item to which they don't have access (because it's available only to subscribers or only to University of Michigan users), they will reach a page that says "Sorry, you aren't permitted access to this item". That page view will still be included in the usage stats because at this time we have no way of removing these visits from usage stats.

However, we do exclude from the statistics visits from known web crawlers, such as those operated by web search engines (like Google) and web archiving services (such as the Internet Archive), that if included would inflate usage data.

Personal tools