The landscape of academic library acquisitions has changed tremendously in recent years. Many libraries have faced significant pressure regarding their ability to purchase monographs for the humanities and social sciences. There has been substantial consolidation in the vendor community, with YBP and Coutts being purchased by EBSCO and ProQuest respectively. Some wonder if monographs and other books are experiencing a format transition, while substantial work has been underway to develop open access models for their publication. With this context and the support of the Andrew W. Mellon Foundation, Ithaka S+R launched a research project to gather data on the distribution and acquisition of materials in the academic library sector.

Our initial intention was to gauge how distribution channels were changing—specifically the impact that book purchasing through Amazon was having on library and university press markets. Distribution channels are a key issue for any marketplace, but they are especially important in understanding and predicting the environment in a period of so much change.

We started the project believing that we should survey the library director population, asking them to provide aggregate statistics on their acquisitions through various vendors and from various publishers. Since there are so many ILS offerings, and each has been implemented somewhat differently across institutions, we hoped that we could generate a small number of extremely specific questions that could be reported appropriately from each respondent. We recognized that this was a difficult project, and would involve custom reporting work at many institutions. And we were concerned that the quality and comparability of the data we would generate would be low.

We began to work informally with a group of acquisitions and systems librarians to explore the environment and our options. Through this process we came to realize that a different approach might not only increase the quality of the data we could gather but also provide much more granular data that would allow us to address a far greater set of research questions. The key point of departure was the comparative standardization—some would say inflexibility—in the implementations of the so-called next generation ILS platforms. If the standardization across institutional implementations meant that certain key variables could be tracked in a uniform manner, perhaps a single query or set of queries could be run across dozens of institutions using the same next generation ILS. We ultimately concluded that we would attempt to gather a report listing each purchase that a given library had placed over the past three years, including such information as title, publisher, call number, price, format, vendor, and more. These data could be analyzed in a variety of ways, not only to answer our initial research question but to provide much greater business intelligence for academic libraries and scholarly publishers.

To test this possibility, we shifted our emphasis to investigate the specific reporting capabilities of the ILS instances. We attempted to investigate all the major ILS packages used by academic libraries, working in some cases with the vendors and in other cases with librarians to investigate their capabilities. None of the traditional ILSs had robust cross-institutional comparative reporting as a capability. Among the cloud ILS packages that have come into use among academic libraries, we found the experience to be extremely variable. For example, we worked with several libraries running Innovative’s Sierra to generate identical reports on their acquisitions. We found that, although the reports were identical, some of the specific coding was highly variable from institution to institution in a way that would make broad comparative analysis across dozens of institutions complicated if not impossible.

We undertook a similar exercise with several libraries that run Ex Libris’s Alma package, working through its very helpful product working group. We found that not only were the results from multiple Alma libraries highly standardized in a way that would ease comparability but also that the query itself could be readily shared from library to library from within the Alma system, eliminating the need for it to be recoded every time another library wanted to join our research effort.

These cloud based ILSs certainly aren’t designed to accommodate the type of research we are conducting. Nevertheless, given the nature of the types of collaboration and coordination that ILSs must support, it was interesting to note these differences across systems. In a subsequent blog post, we will share some of the types of data we are beginning to generate from this pilot project. Later in 2016, we hope to gather data from far more institutions that run ILSs with robust comparative reporting capabilities, with the objective of providing systematic analysis on library acquisitions.