‘ Just in case ’ vs . ‘ Just in time ’ : e-book purchasing models

The academic e-book market is experiencing a period of significant growth and development. Libraries now have choices to make not only regarding title, format and supplier, but also purchasing model. And the range of purchasing models available for e-books is growing! E-book purchasing constitutes around 8-9% of UK higher education (HE) library spend. Figures for other countries vary. Many librarians feel that percentage would be higher if more content were available. However, publishers report that e-book sales are growing steadily. The Bookseller reported in March 2010 that:


Introduction
The academic e-book market is experiencing a period of significant growth and development.Libraries now have choices to make not only regarding title, format and supplier, but also purchasing model.And the range of purchasing models available for e-books is growing!E-book purchasing constitutes around 8-9% of UK higher education (HE) library spend.Figures for other countries vary.Many librarians feel that percentage would be higher if more content were available.However, publishers report that e-book sales are growing steadily.The Bookseller reported in March 2010 that: "In total 68.6% of publishers have seen an increase in e-book revenue over the last two years.For those publishers with e-book programmes growth had been extraordinary, with one publisher recording e-book growth of 44,000%.Even without this figure included, publishers recorded growth in e-book sales of more than 200%, with commercial publishers seeing a rise of 345% and non-profit publishers growth of 108%.
While e-book sales still account for a relatively small amount of total book sales, the survey found that for commercial academic publishers they now represented almost 8% of business, while for non-profit publishers it was more than 10%.For 'very small' publishers it was above 17%, while for 'large' publishers it was close to 13% of book sales.Overall digital accounted for 9.4% of total book revenue, a three-fold increase on previous estimates." 1 Leeds University, e-book purchases represented around 6.6% of book expenditure in 2007/08, and individual e-book purchases accounted for 6.7% in 2008/09.At York, the figures are slightly lower, with around 5% of 'information' spend being on ebooks in 2008/09.In both cases, these figures are well below the reported national average of 8-10%.
Libraries need to consider who they hope to serve with electronic provision, and what they hope to achieve in terms of service delivery or improvement.Is single-user or multi-user access required?Is unlimited usage needed or will some kind of 'ceiling' on usage be sufficient?Does the library wish to own the e-book outright or should it be 'leased' (an important consideration if a work is likely to run to newer editions)?Is there value in buying a collection or is it more efficient to purchase on a title-by-title basis?

202
There is probably no one-size-fits-all answer.Responses will vary from institution to institution, from publisher to publisher, from subject area to subject area, and maybe even from title to title.

Title-by-title purchasing
In the UK, title-by-title purchasing has become the most popular method of e-book purchasing in higher education libraries, but it still accounts for less than package/collection purchasing in purely monetary terms.Which suppliers offer this model?Dawson Books through their DawsonERA platform, Ebook Library (EBL), MyiLibrary (via a range of partner companies such as Blackwells, Swets, etc.), and, increasingly, publishers themselves -though this is often aimed at end users rather than libraries.
This method of purchasing has the obvious advantage of allowing libraries to tailor their collection to meet local needs, while allowing the integration of e-books and print ordering workflows, often through mainstream library suppliers.There are potentially advantages for library customers too as purchasing through one supplier means that they only have to become familiar with one interface.Individual purchases also generate individual records in library OPACs, meaning that customers can find e-books easily alongside print books, and can use federated searching, especially of research-level material.
However, at present, not all titles are available from one single aggregator, resulting in various interfaces for customers to interact with, each of which may have its own terms and conditions or digital rights management (DRM) restrictions, e.g.DawsonERA e-books have a usage cap of either 325 or 400 usages per annum but unlimited concurrent users, while MyiLibrary offers options to limit concurrent users and varying levels of annual usage.The model also mirrors older print purchasing models which were speculative, i.e. 'just in case', and offer no guarantee that purchased titles will be used.It is often reported that over half the printed material purchased by academic libraries in the UK is never used.Reports from the United States indicate that the figure there could be as high as 70%.Is this really a model we should be mimicking?
Figures from Leeds and York certainly indicate that e-book usage is growing.At Leeds, 538 DawsonERA e-books were purchased between August 2009 and March 2010, all had been used at least once, and many had been heavily used.Only 77% (419) of the titles had been used in print.Of the 281 MyiLibrary e-books purchased in the same period at Leeds, all had been used, though only 40% (111) of the titles had been used in print.
At York, of the 17,980 print items added to stock in 2009, only 11,730 were borrowed in 2009.This represents 65% usage of print copies, which is significantly higher than figures that come out of US studies.But, is 65% high enough in times of financial cutbacks, especially when compared to ebook usage?During 2009, 97% of the e-book titles purchased from MyiLibrary in 2008 and 2009 were used.
It is not just the number of e-book titles used that is increasing, but also the volume of usage.What is not clear with this model is exactly what customers are using the e-books for.We certainly know from student feedback that many simply use the electronic version as a quick reference for viewing TOCS, etc., and then make use of the printed version.

Package/collection purchasing
Package/collection purchasing constitutes a larger percentage of HE library spend than individual purchase, though it is unclear whether this is because libraries purchase packages as a way of building 'critical mass' or because they are looking for economies by bulk purchasing.A number of these packages also offer add-on benefits, such as ebrary's DASH! software which allows customers to upload locally-created content to the platform.E-book packages are available from aggregators such as ebrary and EBL, and publishers including Taylor & Francis (T&F), Cambridge University Press and Wiley-Blackwell.These range from small subject collections to larger packages, such as ebrary's Academic Complete collection of over 47,000 titles.T&F, for example, offer around 30 annual subscription subject collections, alongside a number of 'Best-sellers by Subject' packages of 100 e-books.
The University of York library began by subscribing to two ebrary subject collections (Social & Behavioural Sciences in April 2005 and Education in July 2007), which together gave access to over 14,000 e-book titles.In 2009, 41% of the available titles were viewed, giving an average cost per title of £0.45 or an average cost of £0.015 per section request.This certainly reflects good value for money, despite the fact that less than half of the available titles had been used.
York also purchased an e-book package from a major society publisher.This was a one-off purchase of content published in the years 1966-2009.In 2009, 32% of the available titles were viewed, resulting in a cost-per-section request of £6.13 overall.Though initially seeming expensive, this cost-per-section request will reduce over time as usage increases.
It can be seen from these two examples that there can be great variance in the cost-per-use between different packages, so it falls to each individual library to decide on a level of usage which justifies making an outright purchase rather than subscribing to content.
UK higher education libraries are not unique in purchasing e-books.In Spain, for example, 36 of the 50 universities have at least one e-book collection, with the two main inter-disciplinary collections being E-libro and NetLibrary.

"This is a sector that is definitely taking off in Spain and will require progressive acceptance on the part of the academic community. Librarians have the responsibility to aid their users in understanding the growing complexity of the information market and the increasing range of resources available for research … when it is borne in mind that e-books will have a crucial role in the new model for education advocated by the European Higher Education Area (EHEA)." 3
The University of Bergen, realizing that many Norwegian publishers had not yet embraced e-book technologies, began by purchasing collections from NetLibrary and ebrary, alongside more specialized content such as Gale Virtual Reference Library, Safari and Encyclopaedia Britannica.They, too, found that e-books are most often used for literature searching and quick reference, with average usage time being between 10 and 20 minutes.Having tried individual title purchasing via NetLibrary, and finding it too expensive, they now participate in consortial purchasing through the Norwegian Archive, Library and Museum Authority (ABM). 4

Reconsidering purchasing models
So, with a range of purchasing models available to them, why are academic libraries in the UK reconsidering their purchasing options?Firstly, it is a developing market and new models are becoming available.Secondly, inflation on materials continues to rise, and with major cuts in budgets and efficiency exercises, libraries are looking to make their budgets work even harder.Increases in student numbers and raised expectations of electronic provision mean that there is an increasing demand for e-provision, which is subject to VAT.
Libraries are therefore looking to improve methods of selecting material, responding more rapidly to customer demands, and increasing the likelihood that library materials will be used.Though library staff may be best placed to decide which purchasing model to adopt, libraries are starting to look more to their customers to inform decisions on what content to buy.

Patron-driven acquisition
Patron-driven (or user-driven) acquisition is being explored both in the UK and the United States.In effect, control of content selection moves away from library staff and into the hands of the library customer.Working in partnership with a publisher or aggregator, the library loads bibliographic records for an agreed collection of e-books into its OPAC, allowing customers access to any of the titles.After an agreed number of accesses, usage of the e-book converts to become a library purchase of that title.Suppliers such as ebrary, MyiLibrary and EBL, and publishers such as Springer and, most recently, Elsevier, offer patron-driven acquisition.As Leslie Lees, Vice President (Content Development) for ebrary, explains: "As usage plays a key role in determining the value of electronic products and services, patron driven acquisition is quickly evolving as a model of choice …" For the customer, access is seamless and instantaneous, i.e. 'just in time', with the purchase being administered after the fact by the library.The disadvantages for the library are that the model is more difficult to map onto existing budgetary structures and can make monitoring spend more difficult, with purchasing patterns being less easy to predict.There is scope for a small number of customers to monopolize purchasing (unless the library sets some local limits), and there are difficulties in knowing what is the most appropriate purchase threshold.With some suppliers, the library has no choice in what that threshold is; other suppliers, like EBL, do offer some flexibility.
The University of York first investigated patrondriven acquisition between May 2009 and February 2010 with MyiLibrary when a 'windfall' pot of money became available to put down as a deposit.The aim was to create a collection of around five times the value of the initial deposit, to ensure a wide choice of content.However, it proved more difficult than anticipated to limit the collection using the subject profiles, and eventually an initial collection of just over 3,000 titles was selected (around 7.7 times the value of the deposit), in six subject areas: Language & Literature, Health & Nursing, Social Policy, Political Science, Physics, Chemistry and Electronics.The collection grew slightly as new titles matching the profiles were released each month.Purchases were triggered after two usages.No publicity was done, and by the end of February 2010, 433 titles had been purchased.This represented 14% of the total number of titles made available.A further 684 titles (23%) were viewed only once, and therefore did not trigger a purchase, so these were essentially 'free' views.
York also buys e-books from MyiLibrary via the title-by-title selection route, so has been able to compare the average costs and page views for both purchasing models, albeit over slightly different time periods.Books acquired via 'Patron Select' (between May 2009 and February 2010) received an average 87 page views per purchased title, with an average cost per page view of £0.61.The average cost of each per title purchase was £66, with the average cost per title viewed (both purchased and non-purchased) being £26.The e-books acquired from MyiLibrary via the title-bytitle selection route were usually ordered alongside print copies because they appeared on reading lists or high-demand reports; they were therefore likely to be core material for students, and did attract more usage than the Patron Select titles.However, their average cost was slightly higher at £72, and the average cost per page view during the calendar year 2009 was £0.41.
York was also keen to trial other patron-driven models, and embarked on a pilot with Springer in December 2009 to investigate whether patrondriven would work for e-book collections.(See Sarah Thompson's article on this pilot project 5 .) The concern for librarians is that allowing automatic purchasing of e-books may result in less balanced collections.Price and McDonald 6 presented a paper at the 2009 Charleston Conference, entitled 'Beguiled by bananas: a retrospective study of the usage and breadth of patron vs. library acquired e-book collections', outlining some findings from a survey carried out at Claremont University Libraries in the US.Interestingly, two of their key findings were that patron-selected e-books were used on average 2.5 times more often than library-selected e-books, and that they are used by two to three times more unique users.
It would appear from these findings that customers know better what they want to use than librarians do.So, should we librarians be afraid for our jobs?Or just adapt our purchasing methods to meet these changing user needs?Some publishers are already pushing the patron-driven concept further.Elsevier recently announced their Evidence-Based Purchasing Model, which may prove more attractive to librarians concerned about losing control of their budgets.A collection is agreed with the library and a downpayment of between 10% and 50% (depending on size and publication year) is paid.Unlimited access to the collection is made available via ScienceDirect for a period of 12 months.After this period, the library is provided with usage statistics, which can be used to inform purchasing decisions.Unlike other patron-driven purchasing models, no title is automatically purchased, and the library is free to choose to buy in perpetuity only titles that have proven themselves in terms of usage over the year.
But, there is still more to do.As Mitch Ratcliffe 7 points out: "The entire world is moving to a market ideal of getting people what they want or need when they want or need it.Publishing is only one of many industries battling the complex strategic challenge of just-in-time composition of information or products for delivery to an empowered individual customer." Libraries need to look for purchasing models that give them the freedom to deliver content to their 'empowered individual customer' in a way that is meaningful, efficient and, above all, practical for that customer.This is a point that was picked up in a recent article by Dan d'Agostino: "Instead of focussing on books downloadable to e-readers or smart phones, academic libraries have created enormous databases of e-books that students and faculty members can read only on computer screens.The result … is that these collections are used almost exclusively for searching for information -scanning rather than reading." 8e example of a library that is proactively looking at its purchasing methods very closely is the University of Utah's.Rick Anderson, Associate Director for Scholarly Resources & Collections, believes that the financial crisis and attendant budget cuts have only strengthened the feeling that we must move away from the model that has librarians trying to guess what patrons are going to want.The library's book budget is format-neutral, so e-book purchases are funded in the same way and out of the same pot of money as print-book purchases, and there are already plans for moving towards making approval plans e-first.The library is also running patron-driven purchasing with both MyiLibrary and NetLibrary, and has even purchased an Espresso Book Machine to enable print on demand.Rick went on to say that this combination of factors will allow the library to make millions of records for digital books visible to its patrons without either loading those records directly into the catalog at Utah or purchasing copies of any of the books ahead of time; their patrons will have the option of printing them up and buying them, or (in many cases) printing them on demand and then borrowing them, after which they will add them to the collection.

Conclusion
A significant mind-shift appears to be taking place in libraries regarding e-book purchasing, which arguably reflects changes in higher education and wider society: a willingness to relax control and empower the customer, giving the customer more choice and minimizing the time taken to provide the end product.Whether the patron-driven model will overtake more traditional purchasing models remains to be seen, but it is certainly provoking a great deal of interest from libraries, publishers and vendors alike.

Case studies from the Universities of Leeds and York, together with the experiences of some libraries in the United States and Europe, are used to help demonstrate the range of purchasing options currently available to libraries. The advantages and disadvan- tages of routine title-by-title purchasing of e-books 'just in case' are examined, the emerging patron-driven 'just in time' model is explored, and the value of large publisher and aggregator e-book packages in com- parison to individual purchases is assessed.
201 'Just in case' vs. 'Just in time': e-book purchasing modelsBased on a breakout session held at the 33rd UKSG Conference, Edinburgh, April 2010

Serials -23(3), November 2010
At the University of Leeds, MyiLibrary titles were accessed 9,955 times in 2008, and this figure rose to 20,540 in 2009.In the first quarter of 2010 the figure was already 7,583.Usage of DawsonERA titles rose from 1,055 in 2008, to 5,191 in 2009, and 4,651 in the first quarter of 2010, according to COUNTER Book Report 6. 2