Acquisitions' Next Step

Acquisitions modules are the Rodney Dangerfields of library automation: they get no respect. Conventional wisdom holds that technical services functions don't sell library systems because "they're only for staff." Their flashier public services cousins hog the limelight in presentations. With rare exceptions, functional development for acquisitions modules is a low priority among integrated library system (ILS) vendors. This tyranny of the majority ignores the fact that technical services work is the foundation of good public services.

Meanwhile, those toiling in the back room search, order, cancel, claim, correct, receive, pay for, catalog, and process thousands of books, print and digital—often with systems only partially developed. No wonder so many automated tools for selection and ordering originated with companies other than ILS vendors. Book vendors, online booksellers, bibliographic utilities, and libraries have all invented sophisticated online systems to buy books. (For a complete list and comparison of functionality, see "The Acquisitions Tool Belt" in the netConnect, Summer 2005, p. 2–3, mailed with this issue.)

As advanced as the non-ILS systems are, there are too many of them—each unique, proprietary, and standalone. Selectors and acquisitions staff now work in a virtual Tower of Babel, searching for one workflow that applies to all. That's the challenge today for technical services librarians and their vendors: integrating disparate content streams.

Because their product isn't unique, booksellers compete with discounting, service, and innovation. The most successful strategy is to save staff time—the more support a vendor can offer for selection, payment, cataloging, and shelf preparation, the more business it is likely to attract. Accuracy, speed, and price are all important, but it's more compelling to understand and solve the library's problems.

Since the 1960s, vendors have provided economies of scale, and through approval plans and new title notification services, they organized (and filtered) the publishing universe. In the early 1990s, vendors began to make their title databases available—eventually via the web. Vendor systems quickly morphed into freestanding ordering tools, sometimes becoming extensions of the library's system and work processes. They provided functionality lacking in the ILS, along with vast bibliographic databases (with price and availability).

But early vendor systems were also incomplete—librarians needed to record the transaction in their ILS. This meant double work. In response to librarian demands for collaboration between book and systems vendors, a streamlined selection and ordering process emerged.

In essence, the book vendor's sophisticated systems for selection, record export, duplicate detection, local data management, and reporting were grafted onto the ILS, which managed additional duplicate detection, purchase order creation, and fund accounting and enabled bibliographic overlay and electronic invoicing at receipt. Meticulous data mapping and careful load testing must be done at each site, since there are so many variables. But once the system is implemented, there are enormous workflow benefits. At its best, this model saves so much time and effort that libraries place everything possible with the primary vendor—and the combined system becomes a powerful "order magnet."

Once an integrated workflow had been established, librarians, of course, wanted improvements. Book vendors, faced with lengthy lists of enhancement requests, felt the kind of pain previously reserved for ILS vendors. Most of the suggestions derived from two sources: widespread admiration of Amazon.com and expectations that book vendor systems operate more like ILS acquisitions modules.

Librarians wanted an Amazon-like display: more detailed bibliographic information, tables of contents, cover scans, links to reviews, and more. They wanted variant editions linked, to inform both selection and duplication control: paper/cloth, U.S./UK, and, eventually, print/ebook. They wanted selection lists and a convenient way to determine whether they or a consortial partner already held the title. It would be useful to incorporate out-of-print titles in the database. What about AV? Foreign-language titles? The motivation was easy to understand—once there was an efficient workflow, they wanted as much content as possible to go through that channel.

As for functionality, librarians wanted multiple permissions levels, validation for fund and location codes, extensive support for electronic selection, email alerts for faculty, and peer comparisons, among others. They wanted automatic holdings searching and vendor databases enabled as OpenURL targets and sources. Perhaps most important, they wanted good response time and reliable 24/7 access, since vendor systems had become a primary tool.

Much like their ILS counterparts, book vendors formed user groups, trained, and provided support. They invested in hardware, software, and staff. They built partnerships with ebook companies, out-of-print booksellers, and OCLC. By 2000, a web-based selection and ordering system had become mandatory for any library supplier. The broader the content and the more sophisticated the functionality, the better.

Acquisitions systems multiplied. While they all perform the same basic functions, each is built on a proprietary data structure, a distinctive functional design, and a limited view of transactional history. No standards exist for this kind of system, beyond the lowest common denominator MARC record. For users, this means that each system is unique and must be accessed and learned separately. This level of complexity bespeaks an opportunity.

Most libraries purchase enough English-language monographs to justify learning the systems offered by their primary vendors. But for smaller streams of material, there may not be enough activity to warrant implementation and training. Each requires different user names and passwords, has different search capabilities, and offers different data elements with varying degrees of validity.

A couple of years back, we described a solution to this dilemma: the selector's workstation. "The selector requires a synoptic view of possible selection determinants [and sources], including descriptive metadata, holdings information, fund availability, and a record of comments or decisions by other selectors" ("The Amazon Effect: Virtual Approval Plans and the Changing Nature of Book Selection," Against the Grain, Vol. 14, No. 3, June 2002). A single interface to handle books in all languages, ebooks, AV, and o.p. might make practical sense, yet it appeared back then to be in no vendor's interest to create it.

In 2005, it became in an academic library's interest to confront the problem. Cornell University Library staff designed and built an integrated tool for selection and ordering (ITSO), which collates new title data from several sources into a convenient electronic form for selectors. Unlike vendor systems, ITSO provides bibliographic records from many sources; like them, it enables evaluation, selection, addition of library-specific codes, and export of records to the ILS. It represents a significant conceptual advance in acquisitions tools. (For more on ITSO, see "Many Vendors, One Face," netConnect, Summer 2005, p. 4ff.)

The idea isn't new, of course. The Library of Congress Resource File, Books in Print (BIP), and OCLC all serve some of the same aims. In fact, in their newest incarnations, the BIP web products offer functionality very similar to what we've described: individual selector alerts based on profiles; the ability to select and export bibliographic records and local data in MARC format to an ILS; and a feature called Hooks to Holdings, which automates duplicate searching, at least in small batches. Title Source II, although owned by Baker & Taylor, can also be used in a vendor-neutral fashion.

The most likely locale for viewing integrated streams of content is the ILS itself. Some ILS vendors have included the ability to import weekly files of new title announcements for review and action by selectors, but this has never been widely adopted. ILS vendors have developed other features, such as Dynix's VIP (Vendor Import Protocol) and Innovative's Quick Click and Inventory Express, which interact in real time with vendor databases—but in item-by-item rather than batch mode. And, of course, the ILS is the source of accurate fund balances, another piece of information selectors often need.

ILS vendors routinely create modules designed to be interoperable with other systems (e.g., SFX from ExLibris), so it is no surprise that one ILS vendor has developed a "neutral" module for selection and acquisitions. The Library Corporation's Online Selection & Acquisitions (OSA) provides selectors in public and school libraries with access to multiple vendor databases (Ingram, Brodart, Baker & Taylor, and BWI at present), from which they can create selection lists. Acquisitions can export MARC records and place orders in any ILS—again capturing activity across multiple vendors. (For more on OSA, see "The Paperless Dream," netConnect, Summer 2005, p. 10ff.).

The next level of integration can come in one of two ways: by using the position and strength of the ILS acquisitions module, or via third parties—players that obtain new title data from all sources and present it through a common interface. Right now the four vendor-neutral systems (ITSO, OSA, OCLC's Worldcat Collection Analysis service, and Bowker's BIP) represent partial solutions, but they point the way toward improved productivity.

Demand for full-text ebooks continues to grow, as does rich descriptive metadata. Amazon's "Look Inside the Book," Google's print scanning and indexing initiative, and, increasingly, publishers' XML-based production workflows all point to an opportunity to transform book selection. When a selector (or even a patron) can preview and evaluate an entire book online, acquisitions will change. Here are two possible scenarios: books will continue to be evaluated and purchased individually, as print monographs are now, or books will be packaged and sold as digital collections, as most e-journals are now.

The first could eliminate returns and drastically reduce shipping and transaction costs: in a virtual book room, the selector could browse content, link to reviews, and check the library's and consortial partners' holdings. The selector could then choose the desired format and edition: ebook, print book, or trigger a print-on-demand request and indicate which bib record to export to generate the encumbrance and purchase order. Beyond just the availability of full text for new titles, this also relies on negotiation of "preview" rights. Many of the necessary elements have already been brought together in Amazon, vendor tools, BIP, etc. But the ILS remains a more logical locus, as it would allow all content review and decisions in the same interface. Where are the vendors with this in development?

As for the digital collections model, early experiments in selling subject-based collections proved unsatisfactory to many librarians—but more because of the age and provenance of the titles than the concept itself. Subsequent collections from Safari, Knovel, netLibrary, ebrary, and publishers have fared better—partly because selectors have some latitude in choosing titles. But these collections also work well as databases and are especially valuable if multiple publishers are represented. As the past ten years with e-journals have taught us, patrons use a far broader range of resources than expected—if that range is available. Under this scenario, acquisitions systems may end up more like those evolving to manage electronic resources and use electronic resource management modules, link resolvers, coverage databases, and metasearch tools.

For the next five years, it seems likely that librarians must contend with both futures. Perhaps the distinction between purchasing and subscribing will remain intact in the digital world, though that is increasingly unlikely. Other lines are blurring, such as that between buying and borrowing (acquisitions and interlibrary loan). Full-text availability, especially a short-term rental model, may eventually provide a lower-cost alternative. Increasingly, acquisitions systems need to interact with electronic reserve and course management systems. Streaming video and audio will gradually supplant their physical counterparts. Consortial purchases complicate the identity of the customer. Someday the line between acquisition and access may exist only as an accounting transaction.