Nudging the envelope : the hard road to mainstreaming UCD Library e-book provision

Results of a local survey in late 2008 show that e-books have not yet reached a tipping point at University College Dublin (UCD) Library, with only about a quarter of those surveyed saying that they have used any. There is therefore everything to play for in this developing area.We describe briefly the stages of e-book development at UCD Library, and then focus upon the main themes that have emerged and our experience in each case. These themes are: collection development and acquisition; resource discovery; statistics and promotion; then future challenges and plans. There is discussion of the real challenges that we have seen over the three years that we have had a specific focus on e-books, in terms of the inter-team nature of the enterprise, the rapid transition to new working methods, and the threat presented to future e-book developments by the current financial situation.


Introduction
E-books encompass many forms, from incunabula through to student texts, and present a complex library management scenario. As well as being available through aggregated packages, numerous titles can also be obtained individually from a variety of publisher platforms. Purchase models are equally diverse, from outright purchase to annual subscription. In a comparatively short space of time, the commercial production, sale and distribution of e-books has challenged libraries to creatively adapt both vision and practice to encompass the format.
Three years ago the Library at University College Dublin (UCD) initiated a special focus on e-books as a distinctive product, looking to develop best practice guidelines: principles for purchase decisions; choice of delivery platform; administrative workflows; promotion as a part of our digital collection. In an Irish context, all the universities are grappling with this new medium, with some taking a more co-ordinated approach while others are at the experimental stage of dealing with the resultant complexities. UCD 1 is the largest university in Ireland, and includes five Colleges encompassing 35 Schools, plus a range of research institutes and centres. The subject range taught and researched is extremely diverse, including the only veterinary medicine and library and information science courses in Ireland. There are approximately 25,000 students, of which around a quarter are postgraduates. The University and Library have both gone through major re-structuring and the University is heavily committed to becoming a quality research-intensive university, with a new emphasis on the postgraduate fourth level emerging, having historically been more focused upon teaching. The Library service is on a scale to match, with five library buildings and around 150 FTE library staff when at full strength.
While it may sound easy enough in theory to succeed with the uptake of e-books, with much guidance to hand 2 , it is sobering after three years of effort to conclude that despite progress, until we more successfully address the entire context in S13 which we develop the electronic library -engagement with users, communication and marketing, users' awareness of the library brand and offerings -take-up of the transformed e-books collection described below may continue to fall well below our aspiration as information providers.

Summary of this case-study
We use both a chronological and a thematic approach. We first list briefly the principal UCD developments in e-books from 2006-2009, and discuss some major themes in more depth. We look at the approaches we have taken in the areas of resource discovery; purchase; workflow models; and promotion. We also look at how UCD library moved to raise awareness of the value of e-books among both library and academic staff; the approaches we are now using to develop our e-book collection; and those we rejected. We reflect critically on progress made to date. Finally, we draw attention to current and future challenges in this area, especially those specific to an Irish context. The contribution of all involved is acknowledged, though it is not possible to detail every area of work here.

What is an e-book?
There is no fixed understanding of what the term e-book encompasses, neither for content or mediating channel. For a brief summary of these issues, a recent primer from Jane Lee of the California Digital Library is useful. 3 We have to date focused upon individual e-book titles and collections that are accessible and delivered to users online through a web browser. The bulk of this article refers to web-mediated e-books but reference will be made in the closing remarks to the e-book reader and its implications.

Chronology of key developments -taking stock, general aggregators, subject collections, individual titles
In 2006 a small working group was set up in UCD Library to progress e-book development, particularly for undergraduate learning. The initial work included: pulling together for the first time a summary of exactly what e-books we had available; providing some website information about this collection; a literature review to establish best practice; and discussions around the major issues. These key themes included: reforming our view of what makes a library collection; raising awareness among library staff on availability and benefits of e-books; implementing novel workflows; addressing resource discovery and uploading MARC records to the library catalogue; and promotion and marketing to our users.
The main activity in 2007 was a pilot project to look at all these issues and to work with academic colleagues to test our assumptions. We were aiming at this time to pick a single aggregator to provide the bulk of our e-books. Much was learned in the pilot, though ultimately the aspiration to acquire the bulk of e-book titles on one platform proved to be unworkable.
There were two main factors that led to abandoning that model: our checking revealed there were not enough titles available in any single package in relation to what academics wanted students to read; and academic units putting forward requests to purchase subject packages, such as PsycBOOKS 4 , which had been escalating in availability since 2004.
In addition to this shift to subject-specific packages, a second pilot was set up in 2008 to identify and mainstream the processes involved in purchasing individual e-book titles, from where to check for e-content availability, how to store the acquisitions record, how to provide resource discovery, to identifying what usage statistics are required.
In the 2009/2010 academic year, with major constrictions in our budget, all book purchasesboth print and electronic -are severely constrained. It is not now possible to commit funds to recurrent subscription packages of e-books, and we have therefore, at the time of writing, returned focus to a title-by-title approach to e-book acquisition, using a perpetual purchase model.

Collection development and acquisition
Three quite clear strategic approaches have emerged as regards acquisition and collection development of e-books, and these are explained below. Our acquisitions department has been challenged to accommodate a new purchase model which is more complex but also of necessity more flexible than has been the norm to date. Our collection development strategy is similarly varied and defies uniformity, illustrating the somewhat 'messy' nature of e-book provision.

The aggregator approach
At the start of our e-book focus in 2006 we had a mix of single titles and some subject packages but collection development was ad-hoc and we wanted to move beyond that stage. Our first pilot project was set up both to look at the potential of e-book aggregator products and to determine if e-books would prove popular with our students.
In the selection process we invited suppliers to give demonstrations of their products. We then looked at the content in ebrary, MyiLibrary (Coutts), and NetLibrary to determine which best reflected the titles that were on UCD reading lists. As Dawsons were in transition to a new platform at that time they were not included 5 . Beyond mere title availability, we also looked at features such as: ease of use; off-site access; multiple simultaneous uses; print and download options.
We determined that NetLibrary most closely matched the content we were looking for, though it should be noted overlap did not exceed 17% of required titles in any cases (see Table 1). It must be stressed, however, that coverage is likely to have increased in the interim.
The Schools of Politics, Psychology and Philosophy agreed to participate on the pilot -we had hoped to include Nursing and Veterinary Medicine but title availability was negligible. Overall, 84 titles were purchased, on a perpetual access model.
The pilot highlighted some interesting issues, one being the entirely different development strategies that can be followed. Because of the large classes, the School of Psychology pursued a strategy of complementary printed book (p-book) copies, choosing e-books also available in the print collection. In contrast, the School of Philosophy believed they had enough copies of many texts already and so they selected textbooks not already available to their students, promoting them to students as e-only primary reading material.
Substantial work was carried out promoting the e-titles to users, including awareness raising at library training sessions, promotion by the academics directly to users, and follow-up promotion within our virtual learning environment (VLE), Blackboard. Although we had an expectation that the availability of core material in electronic format would result in overall higher usage statistics of material in demand, we discerned no overall increase. Perhaps this more accurately reflects the key point that users do not always read the mandatory or 'essential' reading material for their courses -a practice that is not related to the medium, but to student choice. Usage of e-books remained relatively low, with the exception of a few very popular titles. Usage of print titles, where available, was also very patchy. Title-level statistics were available for the e-books, although they include total access counts only, with no detail about duration of reading sessions. Sample statistics for the calendar year 2008 show: ■ potential usage of the 84 titles at 2,661 desired sessions (successful access 1,753 times; turnaways 908) ■ the top title had 332 accesses ■ Six titles were not used at all ■ 50% of titles had ten or fewer accesses in the year ■ philosophy titles are the least used, perhaps a result of not being core texts on reading list.
The problems with the lack of unlimited simultaneous users in the NetLibrary resulted in high turnaways of the most popular titles -with 880 of the 908 turnaways being for the top two titles.
Other factors that may have contributed to poor usage include: the limited print options that NetLibrary impose in order to obtain a good list of titles; and some issues with the NetLibrary interface, such as titles being presented in two different formats, which can cause confusion to users.
In the questionnaire that ended the pilot, over half of the respondents assessed their knowledge of our e-book collections as poor. Half had not used any e-books in the last academic year or were not sure, despite being the target group of the pilot project promotion. This was sobering, in spite of certain mitigating factors, such as the seamless linkage from PsycInfo into PsycBOOKS perhaps resulting in students accessing an e-book without realizing it. Half of the respondents had found and accessed the e-books via a catalogue record, thus highlighting the importance of this route for resource discovery. The views on whether e-books were liked in comparison to print books were completely spread.

The subject e-book package
Throughout the period, there was an increased range of e-book subject packages coming onto the market, and our academics increasingly asked us to buy these subject-specific collections from their funds, often on subscription models. We accepted that our e-book policy had to encompass these multiple platforms. Of course, a greater variety of platforms inevitably brings a greater variety of problems for our users, but also for staff in terms of administration, licensing, etc. As a user-focused service we therefore had to abandon the single platform approach in the light of this development.

Mainstreaming purchase of individual titles
As well as subject packages, the working group recommended that the book ordering process for individual titles be altered to include information on both e-book and p-book options. This was a significant step forward for UCD Library, and included altering the order form, the related business processes and the workflow as required.
A second pilot, influenced by reviews available at the time 6,7 , was set up in 2008. It explored the purchase of individual titles which were specifically required via reading lists to support both the teaching of large classes and users based off campus.This pilot focused on the whole process of checking multiple platforms for content and ordering individual titles in the arts and human sciences area. A number of notable things emerged from this: ■ the amount of material that is not available in eversions anywhere ■ the considerable time needed to access imprint information and review costs closely ■ the expense of e-versions compared to print when available -compounded in the Republic of Ireland by the punitive Value Added Tax (VAT) of over 20% charged on electronic material, whereas print material incurs none ■ the number of sources of information to be checked for availability of e-content, combined with the need to develop a priority list of these platforms to use when titles are available on more than one ■ the complexity of costings including platform charges and variations in prices depending on choice of purchase or subscription.
Total mainstreaming of e-books into our order process is becoming a reality, and during the purchasing of reading list materials to support the 2009/2010 academic year, e-book titles were identified for purchase in the selection process: 40 titles in the social sciences; 38 titles in the humanities; 20 titles in science, technology and medicine (STM). The high numbers of e-books being purchased for those Schools involved in the original pilot suggests that advocacy and promotion of the e-book option with the academic units is extremely important. Throughout all the phases of our e-book purchasing, our Liaison Librarians have been heavily involved, working with academic colleagues both to promote e-books as outlined below but also to ensure titles can be purchased swiftly. E-books have a big advantage in that they frequently can be accessed by a user almost immediately after a purchase request. To leverage this advantage further, the UCD Library Information Resources Development Policy allows individual purchases of e-book titles or packages by Liaison Librarians with no further approval process where material cost €400 or less.

Resource discovery
It is a consistent truism for librarians that the best purchases are effectively worthless if they cannot be found, and in this, e-books are no exception. This has been a concern from the very first, that whilst putting e-book collection descriptions onto website listings was necessary, it certainly was not sufficient. Literature at the time suggested a correlation between title-level catalogue records being made available and increases in e-book usage statistic 8 . The group had no hesitation in recommending that MARC record sets be uploaded for all existing e-book packages as soon as the Library had MARC21 in place; that further sets be acquired/purchased for all our e-book packages including new purchases; and finally, it recommended the retrospective uploading of MARC record sets for our current collection.
With the library's move to MARC21 in summer 2007 we were now free to add these titles to the catalogue. From the point of view of resource description we had to look both at the minutiaefor example, just how rigidly should we be applying certain cataloguing rules -but also at the broader picture: what effect would the addition of hundreds of thousands of MARC records have on our catalogue? There was considerable debate as to whether the catalogue should contain such records at all and, interestingly, this debate was not limited to staff involved directly in cataloguing and metadata. There was also much discussion around the advantages/disadvantages of including free e-books on the catalogue -some of which are born digital, and others which are effectively free online surrogates of a print item in our collection. While we have achieved some resolution in this area, the very fact that e-books are proliferating will undoubtedly require us to revisit it again.
In terms of the practicalities of incorporating such large quantities of MARC records into the catalogue, we determined that, where possible, we would use vendor-or publisher-created records, with a level of quality assurance applied which should pick up errors but not delay our work unnecessarily.
In terms of challenges faced, and solutions used to surmount them, the experiences of UCD are not wholly uncommon 9 . While we were reasonably confident that the ends (higher usage) would justify the means (changing workflows; a degree of compromise on quality; duplicate entries for print and online copies, inevitable when bulk loading records), enacting these changes was a challenge. This has proven to be a period of intense transition, and it would be remiss to pretend that it has been easy or even 100% successful. In both acquisitions and cataloguing, we have had considerable, if not complete buy-in from staff and the reaction has been, generally speaking, positive. Our new purchasing and cataloguing processes are increasingly mainstreamed and have over the last year been rigorously stress tested.

Statistics and promotion
Usage statistics of e-books, available to the 'e-books group' in 2006, were very patchy. They showed a very mixed use of the products by users including low usage of some expensive items. This was not surprising as users would have had to discover the e-book content within subject databases -Source OECD , for example. While this was addressed to some extent by uploading MARC records into the catalogue, as discussed above, there was a clear need evidenced to sort out a cohesive and regular production of e-book statistics, and also to use them as one element in an advocacy campaign.

Promotion
In 2006 we developed two web pages on e-books, the first being a general advocacy page, the second a listing of all the UCD e-book collections. This was the first time we had systematically gathered all holdings together. The second web page listed the purchased material (a number of these products being financed by the national IReL initiative rather than UCD Library locally 10 ), profiled links to major free e-books and also pointed to the potential of search engines to locate free e-books.
To further promote e-books in UCD Library, a series of 12 printed and online guides to various key e-book packages were produced -sadly, these were not a success, with little take-up from users, and have not been maintained. We also provided general promotion to the academic community and researchers via the Library Newsletter (See Figure 1).
We now actively promote our current and new subscriptions through profiling on the 'news section' of our web pages, via Web 2.0 channels and by sending targeted e-mails to academics drawing their attention to usage statistics and new e-book products. Attention is also drawn to our ebooks collection in bibliographic instruction at the start of every academic year, and academics are encouraged to link from Blackboard. We are aware that promotion of our e-books holdings must be carried out on a more ongoing basis beyond mere announcement of a new product. It is worth noting that some of the usage figures show that where active promotion took place in 2006-2007, usage was higher than in the subsequent year (see Table 2).

Statistics
There are still a few e-book providers giving intermittent or no statistics of use but in the main we now get a full and regular usage report, which we aggregate. This is useful when considering S17 Figure 1. Promoting e-books via the Library Newsletter cancellations and also in targetting promotional activity. There is clearly a great range of usage levels but without benchmark data from other institutions, useful interpretation is limited. Table 2 serves to indicate the main e-book packages and titles we now have, and shows the range of usage across them.

Challenges ahead for 2010
The challenges ahead can be grouped into four main categories: ■ finance and recession ■ new ways of working ■ user awareness of the e-book collection ■ resource discovery.

Finance and recession
Due to current financial constraints, and the 'long tail' pattern of usage of materials, we are looking critically at what we can afford to purchase. This perhaps highlights the need to go back to the basics of book selection, regardless of the format of the item, and to look again critically at the number of copies we purchase. As with other institutions, the selection of material recommended for reading by academics and the uptake of usage by students is sometimes starkly divergent. Much of the e-book content available to UCD is supplemented by the resources available via the Irish Research eLibrary (IReL). This is a nationally funded electronic research library, established on a consortium basis and available to the research community throughout the sector. It clearly differs from UK activity in that the individual institutional library does not pay the costs locally. With regard to financing individual e-books ourselves, they usually have a higher basic cost per copy than print, they often incur additional platform charges, and in the Republic of Ireland context have a heavy VAT commitment of over 20%. This contributes to a slow uptake of e-book material as there is no VAT on print books.
Publishers started offering e-books on their own platforms from 2004 11 . As already outlined, these are very popular with some of our academic units. Many of these packages use a subscription model and, whilst not a problem in the past, this presents challenges in a time of constricting budgets. The notion that book content may 'disappear' in the event of budgetary constraints is not an attractive one, but 'permanence' and 'books' may be two concepts that have to split apart in the future.

New ways of working
In reviewing the work of the e-books group, it is notable that although the key issues listed below were very well understood right from the start, three years on it has taken longer than expected to progress some of them. As with any new medium, it is not always easy to amend our working practices to exploit them fully. Initially at least, the feasibility of mainstreaming e-books into library workflow at all levels was compromised to a degree by the feeling among some staff (in all areas) that there may have been an element of faddishness. In this we suffered from leading the way in this area -at an Irish level. However, it is now apparent to all that e-books are both invaluable and here to stay, and this has allowed us to tailor what we do to accommodate this new medium.

User awareness of the e-book collection
A detailed survey was carried out in UCD Library late in 2008 as part of a consultancy process. The good response rate of 1,689 people gives us some pointers as to where we are with e-books. The picture is somewhat mixed, however, as this summary from the report indicates: "E-books are not yet widely used, with only 26% saying they had used one from UCD. A not insignificant 15% said they were not sure..." and "There is little difference in e-book usage between members of the different colleges, with Arts and Celtic Studies being a little above average, with 33% having used an e-book". One question on e-books usage achieved over 871 free text comments, with a very wide range of views and suggestions for improving usage. This, the consultants noted: "…indicated low awareness and understanding of what e-books have to offer", and they recommended further awareness raising to promote e-books.
Here we provide a few samples of the 870 free text comments on e-books given in the survey: A key theme at the 'e-Books and e-Content' event held at UCL in May 2008 was that e-books should form part of an overall e-learning and e-content environment. Taking this into account, we therefore need to pursue in more detail with Schools, when they request a new subject package of e-books, how they are going to build learning content round the titles. We also need to encourage linking to whole titles or chapters within Blackboard, as well as continuing to investigate other ways to promote this material to students.

Resource discovery
In terms of resource discovery, the library is implementing Ex Libris's SFX/Metalib product in summer 2009 12 For now, a clear resource discovery strategy has been agreed for e-books whereby individual titles will be found in the catalogue through MARC records at title level -we have currently loaded some 180,000 records for nine major packages. Subject-search of the content of e-books will be achieved via individual or federated search of e-book collections in Metalib. We will be extending the number of free e-book packages catalogued in Metalib in subsequent phases of the project. Our hope for the future is to implement an overlay solution like Primo to search across both, but questions now surround the financing of that next step.

The longer term: UCD Library, RIP?
E-books could utterly change the Library's nature and purpose if they become widespread. But will they?-that is the question. Should we expect a gradual evolution towards e-books or a more disruptive transformation of the university library as we know it?
Currently, we are very much steering away from polarized views and high emotions that any threat to the printed book tends to produce 13 . We are not committed to a long-term e-only future as yet, and are following a middle way, supporting Jane Lee in her view that: "Digital resources, including e-books, are here to stay and will continue to experience accelerating growth for the foreseeable future. Whatever happens to print shouldn't distract us from what we will need to offer our patrons in terms of electronic resources" 14 .
It is unlikely that we will be able to ignore, over the next years, a number of very major developments including both offline e-book readers 15 and Google Books. There are others, too, such as direct marketing of e-books with a complementary online study environment to teaching staff and students, as is already being pursued by some major textbook publishers. 16 In early 2006 the e-books group noted that "ebooks that can be downloaded to PDAs, phones, laptops or iPods is a strand of development that we recommend is tracked. These developments are something that may need to be fed into our e-book strategy in a later phase as they could be very attractive to our users". We do not yet own or license e-books through hand-held devices, such as Kindle or Sony Reader, but we did purchase one of the latter in 2008 to familiarize library staff with such a device. Perhaps the e-book reader has reached tipping point. The British Library had models on show until July 2009 for users to try; various groups of university libraries in the USA are experimenting with loans of preloaded Kindles to users.
An alternative point of view highlights the preference users may have for the reading of ebooks on a multi-function mobile device, and the availability of Stanza and various other book readers in the Apple iPhone application library has received a great deal of attention in the media. It is clear to us that this is going to have some significance for a university library over the next planning period, but we are not quite clear as yet what that is, or how we need to respond. Some pilot projects may well be the next step in this area.
Much the same lack of clarity surrounds Google Books, launched in 2004 17 . As stated, we are not currently assuming that e-books will make p-books obsolete and are assuming co-existence. But though the vision of a totally free online book collection has somewhat receded since the class action settlement in the USA, we know from experience how rapidly some developments can lead to a complete paradigm shift in library service, and that we may be off the mark. This is another fruitful area for a pilot project: to explore just how much of our print collection, how much on the reading lists, and how many of the book orders for research monographs can currently be sourced in Google Books.
The disruptive effect that these major developments may have was well summarized by the CIBER team: "The expected popularity of e-books will come with as yet an unknown price for university libraries. Fewer people will visit the library. Academic libraries tend to occupy an enormous amount of premium space and this space is largely filled by books and students. Provide access in students' dorms, bars and recreational space to the relatively small number of books they need and who would bet against a huge drop in library visits? This, of course, will then lead to questions about whether the library needs all the space it occupies. While the advent of e-books in numbers will mean libraries will become more remote from their users, it will also mean that publishers will become ever closer, because they will have all the knowledge of how the user behaves -the users' footfalls now take place in their virtual space. Furthermore, they can offer the products directly to the user. Indeed, with e-textbooks, emonographs, e-journals and e-reference works being bundled together in some publishers' offerings it might be the publisher who will provide the e-library experience" 18 .

Conclusion
Significant progress with e-books has been made at UCD Library since 2006, and many issues have been tackled over that time. With a new University Librarian arriving and much greater awareness of e-books at the popular as well as scholarly level, there is everything to play for, but within the new financial reality.
We hope that our new e-resource discovery system and a renewed focus on marketing and promotion will improve our users' awareness of our e-book collection. At UCD we intend to consider and apply the findings of the major JISC national e-books observatory project to inform our immediate next actions 19 and, at the time of writing, eagerly await the final report from that project. We are heartened by the preliminary finding in the UK, from deep log analysis and surveys, that there has been an increase in use of ebooks over the last 12 months (between January 2008 and January 2009) approaching 10%; there is more promotion of e-books by teaching staff in most subjects; dissatisfaction with the printed textbook collection of university libraries has declined somewhat, and: "More information needs are being satisfied through library-provided e-books. It is [not] known if this is due to increased availability, or simply through more intensive use of what was there already. But university libraries are now the provider of first choice for the majority". 20