As noted, academic libraries make substantial investments in subscriptions to electronic resources. The proportions vary, but it is not unusual for a library associated with a major university to allocate more than 90 percent of its collection budget to subscriptions for electronic content products. The annual increases in subscription fees places even more pressure on collection budgets, often resulting in further diminishment in spending for print monographs or other types of materials. Not all academic libraries devote such high percentages of funds to electronic resources. In some global regions, academic libraries continue to spend more on print resources. At a recent conference I attended in India, for example, many librarians mentioned that they spend less than half of their budgets on electronic resources and continue to purchase print books and journals.
Any electronic resource management strategy needs to be based on comprehensive and accurate use statistics and analytics. While it may not be possible to gather data under all the different use patterns, it is important to gain as much data as possible describing the use of electronic resources at the most granular level possible.
The most reliable measure of use of electronic resources comes from the publisher's content delivery platform. Only the publisher can reliably count the number of access requests for restricted content items made from researchers associated with any given subscribing institution. To gain access to restricted subscribed resources, researchers must authenticate themselves and their institutional affiliation passed to their content delivery platforms. Publishers are then able to provide statistics to each subscribing institution regarding the number of searches and document downloads, preferably using the formats and delivery mechanisms specified by Project COUNTER. The SUSHI protocol for harvesting COUNTER statistics can be implemented in an electronic resource management system to automate the otherwise tedious process of gathering statistics from all of the library's suppliers of electronic content.
COUNTER continually enhances the formats for reporting use statistics. Release 5 of the COUNTER Code of Practice has recently been issued (https://www.projectcounter.org/ friendly-guide-release-5-librarians/). With each new release of the Code of Practice, publishers and libraries need to update their statistical platforms to accommodate changes. To facilitate the adoption of this new release, COUNTER issued a Request for Proposals for validation and harvesting tools in March 2018 that can then be shared with the broader publishing and library communities (see: https://www.projectcounter. org/request-proposals-tools-support-counter-community/).
Once the raw statistics have been collected from all the content providers, a library can begin producing reports and analysis that describe patterns of use. The analysis will reveal content packages that may not be as heavily used as expected or other scenarios that may warrant attention or action. A library often enters a contract with a publisher that gives access to a very large portfolio of materials at a set cost, rather than subscribing to titles individually. These “Big Deals” will naturally include some titles of high interest as well as others that may not attract high usage statistics. One of the analytical exercises that a library may want to periodically perform would assess whether the cost of the comprehensive package compares well to individual subscriptions of the high-interest titles. Other forms of analysis can determine the “cost per use” of any given title, which may be important information to inform renewal or cancellation decisions.
Armed with a complete assortment of usage data, reports, and analytical tools, an electronic resources librarian can begin taking measures to improve the performance of collection items. Strategies to improve the use of resources include adjusting the selections of content components to identify and eliminate those of lower interest, ensuring that the profile of disciplines covered aligns well with the research and teaching disciplines of the institution, and optimizing the library's discovery environment.
Usage statistics and performance analytics can help inform the procurement decisions that ultimately determine the shape of the collection. Since library budget increases rarely keep up with annual increases in subscription fees, libraries must continually pare items of lesser interest from their collections. These cancellations also make room for new subscriptions to resources of stronger potential interest or those that have been specifically requested by researchers. Each cycle of cancellations, renewals, and new acquisitions can potentially result in strengthening the performance of the overall collection when informed by sufficient data of prior use and predictive analytics. Regardless of these efforts, however, use patterns across the collection will likely be uneven and require continuous adjustments.
These techniques may not apply as well to the gradually increasing realm of open access materials. These resources are available without restriction, and therefore it may not be possible to comprehensively track all use by members of a given institution. Usage statistics of these materials can at least be partially captured by other means, but the overall strategies for assessing collection performance will increasingly need to take into consideration the impact of open access resources.
The discovery services employed by the library can also make an impact on the use of electronic resources. It is important to note that only a portion of access to the electronic resources a library provides take place through the discovery tools it provides and controls. A large portion of uses will gain access to resources within the library's body of subscriptions through global services, such as Google Scholar, or may connect directly with publisher content platforms, both of which bypass library-provided discovery tools. That said, the library's discovery service stands as one of the important layers used to gain access to electronic resources. Any gaps in coverage in a discovery services central index can make an impact on usage levels. When statistical analysis reveals underperforming content, whether individual titles or across the broad holdings of a given publisher, the library should test the availability of these materials through its discovery service. If lapses are revealed, possible corrective measures would include validation that the items are activated in the discovery service or issues with the coverage in its central index. Some issues may need to be addressed by the provider of the discovery service. While coverage of resources in the major discovery products can be considered quite broad, gaps remain. Libraries might consider urging the developer of their discovery product and all of their content providers to follow the Recommended Practice of the Open Discovery Initiative (http://www.niso.org/standardscommittees/odi) to foster more complete representation of resources through library-provided discovery services.
These techniques, based on leveraging data and sophisticated analytical and resource management tools represent one aspect of monitoring and optimizing the performance of a library's collection of electronic resources. These technical approaches will not work in isolation of other aspects of a broader strategy. Conversations between librarians and researchers or instructors in their respective disciplines, outreach activities that heighten awareness of key resources, and other professional activities can also lead to improvements in collection usage patterns. As with almost any aspect of library work, positive results depend both on adequate technologies and on the expertise and efforts of library personnel.