Lessons from the current approach in Scotland

The Scottish Higher Education Digital Library (SHEDL) is a closed, all-inclusive consortium of the 18 higher education (HE) institutions of Scotland, plus the National Library and Museum. Scottish HE libraries also utilize the Jisc Collections NESLi opt-in consortium for the purchase of many journal and database bundles.

We can usefully classify information using the structures illustrated in Figure 1. Some information is hearsay – having little value as it is unauthenticated, and may be just rumour, a blog post, a tweet or an anonymous comment. ‘Open’ information is exemplified by web-based material, ideally from a published or recognized source (e.g. NASA, BBC, BioMed Central), and is highly valued as it is both accessible and trusted. ‘Licensed’, ‘local’, ‘distant’ and ‘unlicensed’ material all fall into the domain of libraries. We strive to provide as comprehensive and balanced a set of information for students and researchers as our budgets permit. ‘Confidential’ material is restricted, and ‘unknown’ information is characterized as research, or the results of research, which are known only to the discoverer. Unconsidered, represents information which remains unknown to even the most advanced researcher – information from the future.

Libraries and library consortia are focused on managing information in the ‘open’ through to ‘confidential’ categories. SHEDL has focused on the sustained purchase and management of e-journals and currently includes bundles from ten publishers including Springer, CUP, OUP, IEEE and ACS. Core SHEDL aims are to provide the same resources to all researchers across the HE sector and to maintain pricing as fair and affordable. In expenditure terms, SHEDL covers about 20% of journal spend in Scotland, and, when taken together with the NESLi UK consortium purchases, this represents about 70% of library journal spend. The value of SHEDL journals is judged on download statistics, and the expanded coverage of titles. Pricing broadly follows an historic purchase model. Many new readers obtain material which moves from the ‘distant’ or ‘unlicensed’ classes of information to the ‘licensed’ class.

Recently, SHEDL has extended collective purchasing to include an e-book bundle and the consideration of further e-book opportunities. We face two challenges with e-book bundle purchases: firstly, finding the financial resource (books budgets are defined and carefully allocated) and, secondly, determining value within the many e-book purchase opportunities that exist. Ideally, we could judge value on a cost-per-chapter download (or similar) metric, but for the immediate purchase of e-books, this data is not available. We have developed a model to estimate value within e-book purchase, although it requires further exploration and utilization.

E-book titles are marketed to libraries, or consortia, within a wide variety of purchase schemes, which we can group into either publisher or aggregator derived, and then divided into: title by title, PDA, subject collections or more comprehensive collections. Publisher sourced e-books are generally newer and have less restrictive digital rights management (DRM), while aggregator sourced e-books are more comprehensive in title coverage. From both, e-books can be purchased, leased, or provided on various rental/token approaches.

To estimate the value of an e-book, or a collection of e-books, we have first estimated depreciation rates of academic books, and used this to create a model that permits the value of purchase or lease of e-book collections to be compared. Depreciation rates vary across disciplines, and between titles within disciplines, but reasonable estimates can be made, and these are supported by usage data. Generally the newest material is most highly valued and utilized, and a rapid decline in value occurs in the first three to five years of use, followed by a more gradual decline. Some titles will be replaced with new editions, other titles superseded, representing a fall in value to near zero, while others may increase in value. A general depreciation curve, for a mixed e-book collection, may be proposed as 25% depreciation in the first five years, 15% for the next ten years and then 5% for the remaining period.

Modelling depreciation permits the value of titles to be estimated over purchase or lease time periods, and thus permits a comparison on purchase methods. This requires a fair estimate of e-book value at the time a title is first published (date and list price), and further pricing modifications can be accounted for (e.g. bundle or consortium discounts, lease periods, token schemes). We will soon publish the value model in full, to allow others to use and develop it. An important characteristic of the approach is that it permits the value of different purchase and lease opportunities offered by publishers and aggregators to be compared and purchase decisions justified.

Library Consortia The International Confederation of Library Consortia (ICOLC) represents a useful source of examples of library purchasing consortia. Challenges to sustaining consortial purchase schemes are global. It is important to appreciate the general through to specialist requirements of different institutions. SHEDL has to be flexible to accommodate changes in institutional activity, and SHEDL partners have to appreciate that not every contract will work perfectly for them.

The benefits of engaging or creating a library purchase consortium are clear. There are gains in efficiency, value and sustainability. The publishing industry is well established and resourced. Many publishers are also driven to enhancing business revenue and therefore pricing approaches can be complex, changeable and opportunistic. For publishers, the transition from physical to digital is changing their production cost base and their potential market, and therefore challenging established pricing models. To counter this, library consortia can effectively create a potent collection of shared minds, with different and wide ranging knowledge, experience and expertise. Working collectively can ensure that value and provision are maximized.

Variety is the spice of life: the challenges of managing collections shaped by user choice and publisher bundling

Twenty-first century collection development, governed on the one hand by publisher-defined collections and on the other by patron-driven purchases, could be seen as posing a major challenge for academic librarians. Indeed, some may argue that collection development driven by publishers and patrons leaves precious little for librarians to do. Whilst it is true that the rise of the ‘big deal’ for journal collections and patron-led acquisition (PDA) of e-books have dramatically impacted on the traditional collection development role of the scholar or librarian, this does not mean we have abdicated responsibility. The challenge we face is determining the optimal technique to deliver maximum benefit for our core users.

My earliest experience of acquisitions work was governed by the traditional collection development model. It was characterized by complex decision-making processes, particularly around journal purchases, and funding allocations were rigidly defined by format. Further subtle distinctions existed within formats, for example, between serials and irregular series; monographs and research monographs. Selection decisions were principally at title level and were largely ‘just in case’. When overflowing shelves were reluctantly weeded, it was apparent that much of the material so zealously purchased, or acquired by gift, had never been used. For those institutions with limitless budgets and infinite space, this might not have posed a problem, but few libraries were ever in that luxurious position.

Thus the emergence of the big deal initially seemed like manna from heaven. The age-old, intractable problem of journal prices failing to keep pace with library budgets appeared to have been resolved. For many libraries, the big deal delivered major expansion in journal collections for a relatively small additional cost. This was particularly true for some of the newer or smaller universities, where the big deal enabled a level of collection growth which would have been almost impossible to achieve using traditional collection development models. The move to electronic journals made research collections accessible even when the library was closed or users were working off campus. It also freed up library space. Everyone was happy. However, in the last few years, cracks have begun to show in the big deal model. The mismatch between journal pricing and library budgets is more striking than ever (the big deal only ever having provided temporary relief). Now, the big deal is part of the problem, rather than the solution. Big deal expenditure typically accounts for the lion's share of the library's resources budget. Multi-year deals and the rigidity of packages based on historic print spends have become straightjackets which inhibit our capacity to respond to changing interests. Universities are organic – core subject interests remain constant, whilst others bloom and fade, with new interests emerging all the time. Library collections must evolve to meet the changing needs of their users, or we fail in our core role as an academic service.

The emergence of patron-led, or patron-driven acquisition, particularly for e-books, represents a significant new approach to collection development. This method, more than any other, puts the user firmly in the driving seat. It offers highly personalized, ‘just in time’ provision, which may help to address the perennial challenge posed by student satisfaction surveys: “There aren't enough books!”. After two and a half years of experimenting with patron-led acquisition of e-books, we can confidently report that students love it, especially second and third years, who require wider reading for dissertations. Monograph inter-library loan requests have declined markedly since we introduced the patron-led service, and this year, we noted that six of the top ten schools who use the service had improved NSS scores. Whilst there is currently insufficient evidence to suggest a causal link between improved NSS scores and access to the patron-led e-book service, we believe this merits further investigation. Critics of PDA cite the difficulty of controlling expenditure and the risk that it will lead to the creation of unstructured collections. It can be hard to control and predict the spending, but it is not impossible to manage, and just because we allow users a greater say in collection development does not mean we absolve ourselves of any responsibility for collection development.

So can these very different approaches to collection development co-exist? I believe that they can. It is not always easy to blend them, and the approach that works in one institution will not necessarily translate unchanged to another. In formulating your strategy, it is important to remember your core mission: academic libraries exist to serve their core user community, and collection development is the means by which we do this. It is not an end in itself. Be prepared to experiment. Adapt and survive is a better guiding principle than resist and perish.

Thoughts on the impact of discovery services on the usage and exploitation of collections

Looking back into the mists of time to when we didn't have a Primo or a Summon or an EDS, what made us think we needed one? The obvious reason was the need for something new, something that met the Google challenge. You cannot read up on the subject of resource discovery systems without seeing references to ‘a Google-like experience for users’. That was the Holy Grail. Our users were using systems that were intuitive, whose results made sense, which looked attractive and then, when they switched back to library resources, they got something which looked like it was on a 1980s nostalgia night, which told them the books and journal titles held in the library but not much else. Although we did have federated searching from the early 2000s, the user experience was akin to watching paint dry.

It was clear that we had to come up with something new to try and match Google because if library catalogues had not moved on, libraries and our users had. At the same time, the financial crisis that hit us all in 2008 meant we had to demonstrate the value we provided; we had to ensure all our resources were visible and being used. When you match the need for a search tool to meet user expectations with the need to leverage greater value from library resources, it really was a no brainer to move to a resource discovery tool.

UEA were development partners with Ex Libris on Primo from 2007. When it went live it did so as ‘Broadsearch’ without the Primo central index and as only an alternative to the OPAC offering three search scopes: catalogue only, the catalogue plus MetaLib quicksets and the digital repository.

The game changer was the development of Primo Central Index (PCI), which finally pulled through pre-harvested data, thus raising the visibility of most of our resources. Though unfortunately, due to selective deals within the industry, we do not have data from some of the most significant databases available within PCI. This is a great shame and I hope the decision can be reversed as it is holding back the development of comprehensive discovery services. However, in the main, PCI brought us a step closer to our destination.

PCI was the catalyst for us doing something different. The starting point was to get more direction from our users. We wanted to move beyond librarians talking to technical people talking to librarians. So, in February 2011, we had a focus group with a small group of students from across the university, with the aim to get them talking about how they look for information. They told us that they wanted something that would provide ease of use and access to full-text resources; they didn't want tools that made their lives harder. They wanted something akin to the systems that they were using elsewhere.

It convinced us that we needed to make real changes to our Primo installation, basically reinvent it to better meet the needs of our users. So, over the summer of 2011, we killed off ‘Broadsearch’ and Primo ‘OneSearch’ was born. The new default search scope brings together the Primo Central Index, the library catalogue and our repository. We added an algorithm which would weight our books to ensure they came to the top, a facet for e-books and the bX article recommender tool. In essence, the reinvention was our attempt to provide a ‘UEA Google’.

Have these changes improved usage? Reassuringly, yes, we are seeing greater access to our full-text resources. We have seen a 57% increase in our full-text article requests between 2009/10 (the last year we didn't have PCI) and 2011/12 (the first full year of the new ‘improved’ Primo OneSearch).

However, there do remain further challenges. Relevancy ranking and the complexity of some of the systems remain a challenge that we need to work with suppliers on; it is not yet the smooth searching experience that we would like.

In addition, we have to be conscious of how our users are finding their information and what works best for them, and not be afraid to go to them and embrace new technologies. There are lots of technological changes afoot and how we respond to them will determine how successful we are in giving our users what they need.

As in previous years, the feedback from delegates on the day was overwhelmingly positive and many reported that they would be taking lots of useful facts and questions with them as they returned to their day job. So, the lesson to be learned is – book early for UKSG events in 2013 and check the UKSG website for details of forthcoming conferences, webinars, etc.