Implementing E-Book Demand Driven Acquisition (DDA) at Asbury Theological Seminary

Theological education, in general, but particularly for graduate students, is continuing to include or transition to online education. Our students may live all over the country, or even around the world. A physical collection, while extremely important, simply cannot meet the time-critical needs of these educational programs. In the context of Asbury Theological Seminary, we experience the complex demands of a quality research library that serves both residential and online students. These two constituencies, while sharing some significant similarities in their content needs, have drastically different mode of delivery needs. Even our residential students tend to prefer digital materials more often.

In this environment, it is increasingly likely that the faculty will request and/or expect that their supporting library have access to (and likely own) highly specialized books in their general areas of expertise. While this is encouraged, it also can be a detriment to the overall stewardship of the library budget. Many of these highly specialized texts may very well never be used, they are increasingly more expensive, and the maintenance or storage of such resources continues to be costly as well. These titles are important for research purposes and for maintaining the research quality of a seminary library. One of the ways we have sought to address this concern is through the use of a Demand Driven Acquisition (DDA) system for these select titles. DDA has been touted by vendors as a potential way to solve understaffing in acquisitions and cataloging. The following will address what we have learned about this style of acquisition, both the good and the bad, and how we see it playing a vital role in our acquisition processes going forward.

As we explored the potential for DDA, we quickly realized everything that had been promised was not as it appeared. It is true that vendors are agreeable to partnering with libraries because they appreciate an opportunity to further market their holdings and ultimately increase sales. It is also true that the vendors provide MARC records of the materials outside the collection for inclusion in the library catalog in accordance with the established profile. However, we found these records to be extremely low in quality. It is true that the user (in our context, limited to faculty, staff, and students) can easily select and initiate a purchase, avoiding review by the subject librarian at the time of purchase.

The first step, once you have a vendor, is to establish a selection profile that requires the subject librarian to work with a vendor to identify a predetermined set of specifications (e.g., subject, format, language, cost, etc.) by which corresponding titles will be provided. In our case, the subject librarian selected title-by-title to create our profile (DDA collection). Through this process, our librarians were able to predetermine e-books available for DDA to retain control and ensure the library’s collection development policy was followed. With our vendor, EBSCOhost, it is possible to establish more than one selection profile with a predetermined expenditure amount for each profile. While many may see the benefits of immediate access to the desired information and increased ability to control cost, obstacles still remain that may deter librarians from implementing an e-book DDA system.

Our initial experiment with EBSCOhost DDA began with 119 titles derived from European publishing companies that focus on Biblical Studies for our Ph.D. program. We limited the budget to $10,000, and if funds remain at the end of the budget, we may purchase all or none of the 119 titles as we determine.

Here are some highlights:

  1. When the full list was compiled, our Director of Strategic Collections exported it as a spreadsheet from the EBSCOhost Collection Manager (ECM) so it could be used to find the MARC records in OCLC. What he found was that more than 50% of the titles were not yet associated with an OCLC record, that is, there was no OCLC number in the MARC record. He waited more than a month to see if this would change, but there were no changes that he could detect in the export list.
  2. After that waiting period, he began making his way through OCLC to find appropriate e-book records, title by title. The Export List did have a full complement of print ISBNs and e-book ISBNs. This was very helpful. His first step was to add the e-book ISBN to the corresponding print book record. This was done because the print records are consistently cataloged correctly, by a human, while the e-book records are consistently created using “vendor supplied metadata”, which were routinely of the poorest quality. By adding the e-book ISBN to the good print record, he was able to retrieve the good print records when doing an eISBN search. He could then manually copy the good data and use it to correct the bad/incomplete e-book records.
  3. This process is time-consuming, on the order of 20 hours of cataloger time to make things right for 120 titles. That’s about 10 minutes per title, working at a brisk and determined pace.
  4. While doing the eISBN searches, he found that EBSCO had attached our holdings symbol to most of the e-book MARC records on the list. The quality of those records ranged from “all capital letter, placeholder” records to the very occasional Library of Congress record. At least half of the time the records with our holdings symbol did not have the 856 URL field with the EBSCO link, so he added them.
  5. Also, we discovered that when a title is purchased, an invoice is sent as an email attachment: one a PDF and one an XLS. We were dumbfounded to find that the invoice was not itemized. Basically, it just says, “Pay us and the amount.” The Title Number is another mystery since it doesn’t correspond to the title ID of either of the titles purchased. This is potentially a big problem. At the time a purchase transaction occurs, we need to create a Purchase Order that accounts for each title purchased. We cannot pay the invoice until we know exactly what we are paying for. At this time, it appears that the only way to find the title information for purchased DDA titles is to go to ECM > e-book Manager > Purchase Type > DDA > Purchase DDA > Apply. In the “expanded view”, the results screen does indicate “date added” but it does not display the purchased price.
  6. At this time all the DDA MARC records have been downloaded from OCLC. They are now sitting in our catalog as if we owned all of them. There are at least half a dozen that our Director of Strategic Collections would expect we would want to own. It will be interesting to see whether or how quickly these titles are discovered.

As a result of this work, we are already seeing benefits. Having the records in our catalog has allowed our constituents to more easily discover these resources. Discovering these resources in our collection increases the confidence in our collection development procedures, even if we have not purchased these titles yet. As we noted, some titles have already seen use and have triggered purchases. While we recognize that it would be possible for our funds to be depleted quickly by a “click happy” patron, the limits in place in the system provide an adequate protection. Our initial appraisal of the process and system is cautiously optimistic.

However, our efforts have revealed some significant issues with catalog record quality. There is a significant need for librarians to continue to diligently produce high-quality title records for our catalogs. A database system, like a library catalog, is only ever as good as the data entered into it. Incomplete and/or badly formatted records like those discovered through this process can greatly impact the usefulness of a library discovery system and produce barriers for access. More feedback and expectations of publishers or vendors to produce at least basic quality title records is essential to ongoing patron research efforts.

These records are the most time-consuming portion of the process. The process could be streamlined to make all of this easier. A library must not simply accept low-quality records in the face of understaffing. This is unacceptable and will greatly damage our guild in coming years, not to mention its impact on scholarship. Rather, vendors and publishers must embrace the need for higher quality modern records that employ greater functional description standards. The pace of change in education and academia simply requires more from libraries than it ever has and high-quality records are a significant portion of those requirements.

The blended concerns and preferences of residential and online educational programs are pushing libraries to continue to try new things, and while DDA is not terribly new, it might be for graduate theological and religious libraries. As we look to the future of theological education and the future of our library, we project that DDA efforts will prove a useful tool for our acquisition procedures. While useful, it is best to view DDA as a tool among several to augment a library’s collection development policy. The way we implemented the e-book DDA allows us to refine a comprehensive collection for future scholarship and address ongoing subject trends while controlling costs. This is a cost-effective and customer-service-oriented process.