During the American Library Association Conference in Washington, D.C. in June, I spoke on a panel with Oliver Pesch from EBSCO and Bob McQuillan from Innovative Interfaces Inc. about The Three “S”s of Electronic Resource Management (ERM): Standards, Systems, and Subscriptions. The meeting attracted more than 150 attendees and was one sign of the challenges faced by librarians who manage e-resources. Each year, the percentage of acquisitions budgets directed towards digital resources is increasing by several percentage points and has done so for most of the past decade. The majority of libraries currently dedicate more than 50% of their acquisitions budgets to digital content. A few librarians have even indicated that they are likely to move to an acquisition strategy of 100% digital in the coming years, one of many indicators of the growing importance of e-resources to both librarians and patrons.
Relatively new systems have been developed to store and curate the information necessary to order, process, and monitor electronic products, and a variety of standards and best practice projects addressing ERM have evolved. However, managing these resources continues to be problematic for a variety of reasons. In part, this is due to the complexity of digital products and the way they are packaged for sale. The rapid pace of transition away from print and toward electronic resources has not been matched in many libraries with an equivalent transition of the human resources and skill sets necessary to effectively manage these products. Also, the development, deployment, and population of management systems naturally lags behind changes in practice given their costs and complexity, both on the vendor and library sides. However, the majority of attendees to the ALA session (granted a very un-scientific study, although similar research supports this) had either implemented or planned to implement an ERM system.
What is it about digital resources that make them more complicated to manage than their print counterparts? While the item management lifecycle for a print product is linear and moves from selection through ordering to receipt, cataloging, circulation, and eventually de-acquisition, the lifecycle for digital resources is quite different. The electronic resource lifecycle is circular and iterative and contains many additional steps not relevant in the print world. Product selection can require both trial use and technical evaluation, because e-resources are often encompassed in their own information system. Many e-resources come bundled in packages that have to be evaluated as a whole as well as for their individual resources. E-resources are usually licensed, not sold like their print counterparts, so along with price consideration one must negotiate a license that matches the intended use, population to be served, and other terms. Providing access is no longer a matter of simply cataloging and then placing the resources on shelves. Electronic access includes IP address management, A to Z list management, authentication setups on both the library and publisher sides, possibly user ID setups, possibly OpenURL knowledgebase management, and whatever setup or policies are needed to ensure license compliance. Not to mention all the issues of ongoing support, such as troubleshooting, downtime and other problem management, usage monitoring, user training, etc. And this entire process begins again at renewal time. The availability of titles within an electronic collection can change — even midsubscription — requiring a re-evaluation of the whole product. And the previous year’s usage may necessitate license renegotiation, a process that usually can’t be relegated to a third party such as a subscription service agency.
The development of standards in this area largely began with a joint NISO and Digital Library Federation (DLF) workshop on Standards for Electronic Resource Management in Chicago in May of 2002. Out of that meeting and with additional work undertaken by the DLF, a report of the Electronic Resources Management Initiative was issued in 2004 that specified the requirements of an ERM system. That laid the foundation for the development of many of the ERM systems on the market today. It also led to a variety of other standards initiatives, often building on each other.
Several of those initiatives involved licensing, such as ONIX for Publications Licenses (ONIX-PL) that created encoded exchange of licensing terms, NISO’s License Expression Working Group that mapped the license syntax between ERMI and ONIX, and NISO’s Shared E-Resources Understanding (SERU) that provided guidelines for those who want to forego negotiated licenses.
A major advantage of the new electronic content systems was the ability to track usage. This led to the development of the COUNTER Codes of Practice to standardize what was counted and how. The success of COUNTER resulted in NISO’s Standardized Usage Statistics Harvesting Initiative (SUSHI), a protocol to automate the harvesting of COUNTER data.
A significant new capability with e-resources was OpenURL linking. Another success story, OpenURL evolved into a formal standard (ANSI/ NISO Z39.88) and generated another project, the NISO/UKSG Knowledge Base and Related Tools (KBART) initiative. They issued the first of their recommended practices earlier this year to improve the quality of OpenURL knowledge bases and their metadata. Another NISO project, Improving OpenURLs Through Analytics (IOTA) is looking at how to measure this metadata quality.
Systems-related standards efforts for ERM include NISO’s Cost of Resource Exchange (CORE) project to develop a protocol for exchanging financial information between an ILS and an ERM, and a project to develop best practices for Single-Sign-On Authentication so users don’t have to log in over and over.
Some of these projects have had tremendous success and are being rapidly adopted in the community. SUSHI and SERU are two examples whose success points to the underlying reasons why standards are adopted generally. An inefficient business process causes “pain” in the form of wasted time, money, or resources. In the case of SUSHI, it was the gathering of usage data from several dozen to as many as a few hundred content suppliers. For SERU it was the effort to negotiate licenses, which becomes completely unscalable when the number of licenses reaches a few dozen. With libraries unable in this economic environment to add staff directly dedicated to these tasks, a different approach was necessary.
The “pain” relieved by using SUSHI (ANSI/ NISO Z39.93) to systematically automate gathering usage data is not trivial. Before SUSHI, some libraries reported having at least one year-round FTE staff person dedicated to e-resources usage data gathering and consolidation. Release 3 of COUNTER’s Code of Practice included SUSHI compliance as a requirement. The wide market acceptance of COUNTER and the fact that the provision of COUNTER-compliant usage data is included in many content licenses have led to a rapid adoption of SUSHI. As of May 2010, there were more than 110 publishers who were compliant with COUNTER Release 3 — and therefore SUSHI-compliant. Most major ERM vendors are incorporating SUSHI compliance into their systems to enable usage data to be easily imported. By reducing the costs of gathering and managing usage data, the SUSHI standard has proven a direct and quantifiable business value that has supported its adoption.
License negotiation is another point of significant “pain” for both libraries and publishers. Often, the license negotiation process can take longer than the agreement for the business terms of the sale, which not only adds to the total acquisition cost but also delays making the e-resource available to end users. Various approaches to streamlining negotiations have been tried, including the distribution of model licenses and the development by some libraries of their own standard license. The Shared Electronic Resource Understanding (SERU) project took a different approach, envisioning an environment of shared understanding and good faith. The SERU recommended practice (NISO RP-7-2008), released in the spring of 2008, articulates well-established and widely accepted common expectations between libraries and publishers and can be referenced in a purchase order in lieu of negotiating a license. The SERU registry of parties willing to use the guidelines with some or all of their e-resources lists more than 130 libraries, eight consortia, and 44 publishers and content providers. In addition, there is talk of “internationalizing” the document — since it is based solely on U.S. Copyright law — so that it can be applied in other countries where the underlying intellectual property protections are different.
Some standards may be of great interest but actual adoption is slow. EDItEUR’s ONIX for Publication License (ONIX-PL) is an XML communication structure for making licensing terms machine-readable. The terms can then be added to an ERM system and delivered to end-users in real-time and contextual with the e-resource being used. While the standard was published in 2008, it has seen little adoption — despite its significant potential — due partly to the complexity of turning a legal document into structured formats and terminology such as “Permitted,” “Not Permitted,” “Silent,” or “Interpreted.” The actual encoding, which required some knowledge of XML, has recently been simplified with the availability of the open source ONIX-PL Editor tool. Who actually does the encoding — the publisher who delivers it with the resource or the library after acquiring the resource — is also a debated issue due both to resource constraints and to license interpretation. The use of third-party encoding has encountered push-back due to possible liability and indemnity issues. There are also cases where ambiguity with their license terms is preferred versus the clarity provided by an XML-encoded structure. Organizational needs differ in terms of the level of detail needed. However, some recent ERM projects, such as that of the Statewide California Electronic Library Consortium (SCELC) and the JISC Collections Group in the UK, are experimenting with the use of ONIX-PL and may set the stage for additional uptake. We are still in the process of determining whether the pain threshold of managing licenses badly — using paper in file folders — is less than the system costs of using and encoding the licenses.
There is also a chicken and egg problem about the creation of communication protocol standards, such as the ONIX-PL or the Cost or Resource Exchange (CORE) standards. In order to be effective, a communication standard requires not just one implementer; it requires two. Like any conversation, talking with oneself isn’t terribly productive. These communication protocols require multiple implementations to be successful. However, not all companies have the same business goals, development priorities or system models. While there is no one right or wrong approach, it makes coordinating development schedules difficult, which delays adoption. Again, the questions of whether the old “painful” way of addressing the problem is worth the investment in systems to overcome the problem is a balancing act that system suppliers need to weigh carefully.
Each of these ERM-related standards addresses a piece of the total ERM puzzle. As yet, there is no overall framework of standards for ERM in the way that libraries have become accustomed with their ILS. And there are still gaps in the e-resources cycle where no standardization has yet occurred.
Looking forward, NISO has chartered a working group to conduct a gap analysis of ERM-related data, standards, and best practices. The findings and recommendations of the working group, led by Ivy Anderson at the California Digital Library and Tim Jewell at the University of Washington, will set the stage for the next phase of standards work in this important area. They are scheduled to release a report of their work by year’s end and will be discussing their work-to-date at several fall meetings, including the Charleston Conference, the LITA National Forum, and NISO’s Electronic Resource Management Forum in Chicago in October.
One thing that I stressed during the ALA presentation, and at many other times during the ALA conference, is that content providers and systems suppliers are very responsive to customer concerns and needs. If enough librarians demand systems and products that use license encoding and license transfer protocols, suppliers will adopt ONIX-PL. If determining cost-per-use calculations is taking far too long and requiring too much data entry and manipulation, ERM vendors could implement CORE to address their customers’ problem. This process has worked well with COUNTER, SUSHI, and SERU adoption; the library customers were demanding the standards and system suppliers saw the value of implementing them. As the old saying goes, the squeaky wheel does get the grease, which is just as true with libraries and vendors as it is with your car and the mechanic.