Sometimes identifying what content is to be purchased is easy, for example, if you encounter

• new or updated course reading lists
• requests resulting from specific research funding
• requests from patron-driven acquisitions (i.e., “Order a copy.”)

However, other requests are more complicated, particularly when it is discovered that there is an electronic equivalent to a print version. This is especially true if your collection policy is to purchase an e-format, when it is available, in preference to print.¹ You may also want to investigate the history of interlibrary loan (ILL) requests and associated costs. In the case of abstracts and indexes (A&Is), full-text databases, or nontextual resources, you may need to determine what platforms host the given resource and which works best in your local environment—you may already use and prefer a particular platform (e.g., EBCSOhost, ProQuest, Ovid, etc.). Lastly, you may become aware of a new product or service and want to check it out to see if it would be a fit for your institution.

It is important to set out the criteria you wish to fulfill and map them to your collection management and development policy. Is the primary use for undergraduate teaching or postgraduate research? Are you purchasing within the existing budget or are additional funds available? How sustainable is this budget? Are multiyear deals a possibility? Remember that electronic resources need at least two years to become embedded into teaching and research. That is, resources usually need a full academic year to appear on reading lists, etc., and only then will your usage statistics start to make sense. Very often, the first year of usage of a resource can be fairly meaningless because the resource is not yet embedded.

For a single order, your criteria are usually short and based on local needs, such as your collection management and development policy regarding format choice, what platforms are preferred by users, and which ones work best in the local environment, for example:

• Undergraduate/postgraduate bias: Some resources cross over, but others are not always appropriate to certain levels.
• Intuitive interface: Is the resource as easy to use as Google? If not, users might go elsewhere.
• Hosting: Many electronic resources are available on more than one platform. The subject coverage of other resources on a given platform will influence any new subscriptions. Subject librarians or faculty will have favorite platforms.
• Shibboleth authentication, EZProxy access as standard: Any resource that relies on individual usernames and passwords for access is creating a barrier to use.
• Unrestricted access: Resources that restrict access by number of simultaneous users often lead to dramatic drops in usage over a period of time as users become frustrated by turn-away messages. In addition, restrictions by location (e.g., campus use, overseas) also result in potential low usage. Unrestricted access rules!
• COUNTER-compliant usage data: You need accurate usage data to show value for money—COUNTER sets the standard.
• Ability to use within a federated or harvested search system: Resources that cannot be added to the federated or harvested search are effectively making themselves invisible to today’s user, who expects a “just in time” approach to resource discovery.

Following on from the last point, if the resource is not making full text available to the various resource discovery systems out there—then why? And does the supplier have a plan to do so in the future?

For larger projects, these criteria become more involved. An example here would be choosing an e-book platform or e-book provider; this would require more expansion about what is desired and what will be provided via the platform selected. Appendix 2.1 shows an example of a supplier platform review document for patron-driven access delivery.

Understanding your institutional needs is essential before you look at the market and talk to suppliers. What do you require the resource to achieve? What are the essential criteria? These criteria should include requirements for interface, content, and administration. The Knowledge Base + (KB+) project in the United Kingdom is developing a community-based shared service for electronic resource management (ERM). As part of the preparation, a small group looked at essentials to check on a license. The crowdsourced results show some essential criteria for any new resource:

• concurrent users
• remote access
• walk-in access
• multisite access
• partner organization access
• alumni access
• interlibrary loan
• course packs
• VLEs
• postcancellation access
• notice period

Knowledge Base +

www.jisc-collections.ac.uk/knowledgebaseplus

KB+ plans to use traffic-light icons to indicate Yes, No, or Conditional for a range of key definitions and clauses. This will make it far easier for the e-resources manager to see criteria at a glance.²

If you are considering the purchase of a full-text or A&I database, use the list below in addition to the points above to help set out your specification.³

• Full-text coverage/full-text linking: Is the full text cover-to-cover, and if not, does the A&I database link or provide the means through OpenURL linking to access external full text?
• Sustainability: If there is full text, is it sustainable? Can savings be made to journal subscriptions? Imperial College London developed a toolkit back in the 1990s to identify just such savings.⁴ For more information on sustainability, see below.
• Cover-to-cover indexing: Many A&I databases list a large number of journal titles in their coverage. Further inspection often reveals that these titles are split between core content (cover-to-cover indexing); secondary content (where more than 50 percent of the material is indexed); and tertiary content (where less than 50 percent of the material is indexed).
• Date coverage: Unless the database specifically covers an archive period, check the ratio of current to ceased titles—you may be surprised.
• Geographical coverage: Is it important where the data comes from? Is US, European, Far Eastern, or other coverage needed?
• Publisher coverage: If an aggregated platform does not have a good spread of publishers, then searching the resource is little better than searching a publisher’s platform, where the functionality could be better.

It is important to expand upon the point about sustainability made above. Sustainability is increasingly important for the collection management and development of e-journals in particular. In the good old days of print, we knew what we subscribed to, and as long as we kept a print copy, we knew we had access to it! Those days are gone for many libraries. Now we essentially rent access to content, so what happens when we stop the subscription? What if we want to ditch the print? What does postcancellation access really mean? All these questions need to be asked at the time of selection, not at the time of cancellation.

Using the Imperial College model for sustainability, there are three rules against which all e-journals can be assessed. E-journals are classed as sustainable when at least one of the following applies:

• There are perpetual access rights to the content via the Web. Perpetual access rights include access via the publisher’s website or via services such as Portico, LOCKSS (Lots of Copies Keep Stuff Safe), and CLOCKSS (Controlled LOCKSS).
• The journal is permanently open-access for all years or certain years. Hybrid open-access journals are not included in this category. For these purposes, we are not interested in sustainability at the article level.
• The content is available in a trusted service such as• Jisc Journal Archives (a community-driven archive of nationally procured journal archive collections in the United Kingdom), JSTOR, etc.⁵

A journal will be considered unsustainable if it fails the above criteria. For example, aggregated services such as ProQuest’s ABI Plus Text or EBSCO’s Business Source Complete do not fulfill any of the sustainability criteria, and therefore titles within these resources would be recorded as unsustainable. In order for e-journals to be sustainable under the rules above, you need subscriptions to a number of “insurance” services, such as Portico, LOCKSS, or CLOCKSS, or archival subscriptions such as JSTOR or Jisc Journal Archives. This practice is very much a part of ERM—call it a modern-day disaster plan. Consideration of subscriptions to these services can also use the TERMS cycle.

Portico

www.portico.org/digital-preservation

LOCKSS

www.lockss.org

CLOCKSS

www.clockss.org/clockss/Home

Jisc Journal Archives

www.jiscjournalarchives.ac.uk

JSTOR

www.jstor.org

In the case of a single e-resource, getting the right team may be as simple as consulting with faculty on whether an e-version will be sufficient. However, even with one-off purchases, there may still be a review panel that approves all new subscriptions.⁶

With larger-scale databases and more complex resources, however, the purchasing decision may require consultation with information technology infrastructure support personnel, or even with university purchasing officers if the cost of the resource means that an Invitation to Tender (ITT) or request for purchase is required. In some cases, the team would be an ad hoc group consisting of

• the e-resources manager
• the subject/liaison team
• the budget holder
• faculty

For large-scale projects, such as resource discovery systems, etc., use the Jisc project template.⁷ For other projects, the members of the team will be dependent on what is being selected, which does not mean that everyone should not be kept in the loop.

You may want to consider running update sessions for your subject or liaison teams to discuss new e-subscriptions. This forum is useful in getting buy-in from other subject teams that may otherwise be overlooked. For example, many resources, such as marketing resources, are used by more than one faculty (e.g., business, art and design, or engineering [for product development courses]). Of course, many institutions make these decisions via a standing committee. If this is the case, make sure that the committee members are as well informed as you can make them.

It is not unheard of for academics to request resources that have already been purchased or that are partly available in another subscription. Often academics will request a resource that they are already familiar with. While acknowledging that content is king, the electronic resource manager has a fiscal responsibility to consider the options before making a purchase. To this extent, it is important to check whether the request for new resources can be satisfied by existing subscriptions or whether alternatives are available. In times of economic austerity, we can no longer afford to subscribe to multiple resources that have a large overlap. A desktop review of the market and the literature may require more investment for larger collections of electronic resource materials or when a particular resource is available on more than one vendor platform. Multiple platforms for the same resource tends to happen more with electronic A&I and aggregated services, where the reviews and trialing of various versions may be critical to the selection of one resource.

Before talking to suppliers, have a look at what is out there. To check coverage and duplication of content between resources—both potential and current subscriptions—use commercial tools such as

• 360 Core Overlap Analysis Tool (Serials Solutions)
• WorldCat Collection Analysis (OCLC)
• SFX Advanced Collection Tool (Ex Libris)
• EBSCO A-to-Z Overlap Analysis tool

Alternatively, use the free Jisc Academic Database Assessment Tool (ADAT) from the United Kingdom or the CUFTS Open Source Serials Management system from Canada. Simple manipulation of title lists (available on most vendor websites) in Excel can also pay dividends, especially when looking at duplication of titles across a range of products. The Charleston Advisor is a great resource for finding product reviews and comparison studies of various content platforms.

Serials Solutions 360 Core Overlap Analysis Tool

www.serialssolutions.com/360-core

OCLC WorldCat Collection Analysis

www.oclc.org/collectionanalysis

Ex Libris SFX Advanced Collection Tool

www.exlibrisgroup.com/?catid=%7B449D6DF8-2061-4E18-9AB0-62B3DF94EE57%7D

EBSCO A-to-Z Overlap Analysis Tool

www2.ebsco.com/en-us/ProductsServices/atoz/Pages/atozinfo.aspx

Jisc Academic Database Assessment Tool (ADAT)

www.jisc-adat.com

CUFTS: Open Source Serials Management

http://researcher.sfu.ca/cufts

The Charleston Advisor

www.charlestonco.com

Once you have done this, use your specification document to further narrow the field. Look at trialing your short list. It is very important to get the timing of your trial correct; it can be frustrating when faculty get in touch on the last day of the trial! Use your faculty contacts to confirm the best time of year; publicize the trial on your blog, wiki, or web pages; and make sure you get usage statistics for the trial. Put a comments sheet together to solicit feedback. The length of the trial is also very important. Some suppliers still offer only a one- to two-week trial, and a few offer only twenty-four hours! Try to get a trial for as long as possible—one month minimum. Some suppliers will negotiate a “sponsored trial” where, for a small admin fee, the trial can be extended for up to six months. This allows you to get a real feel for the usage and is particularly useful for larger subscriptions. Use budget cuts to your advantage—suppliers may be prepared to negotiate the price down.

Make sure that when you disseminate information about the trial that you have feedback mechanisms in place; record any comments and feedback you get, and from whom they come. This will allow you to justify any decisions made and to collate feedback to the suppliers or vendors.

Talking to providers will follow readily from the desktop review and consideration of trial set-up. Be aware that some resources are available through different suppliers. In addition, there may be national or regional consortial agreements in place with preferential prices and licenses. However, some suppliers may have exclusive deals in a given region or territory, meaning that your choice may be limited. Always make sure to let a supplier or vendor know when you are looking at more than one provider for a resource as this may result in your learning a bit more about the product as they try to prove why their version would be best. Make sure that you fully understand all of the contracting and fees associated with any given resource to avoid surprises at the point of acquisition.

Try to get a good representation of your team when talking to suppliers—preferably the same people present for each meeting—and have your specification document to hand to remain focused on task. You may learn things along the way; go back to previous suppliers to verify anything you pick up. This may seem like a lot of preparatory work, but remember that some deals can add up to well over $100,000 over the course of a three-year deal, for example.

Finally, you need to score the resources and suppliers against your specification document, using any weighting you wish based on your priorities (e.g., cost, ease of use, coverage, etc.).

After this review, which may take only a few hours given a single resource or a few months if purchasing a large collection of content, document in your ERM system (if you do not have an ERM system, a spreadsheet will do) any relevant points that went into the purchase decision. These could be as simple as the fact that you subscribe to other journals on the same platform and that the platform functionality works well, or any relevant comparison information you have gathered. If you decided not to take a resource after a trial, document the reasons why—this may help in the future if you are asked to review the resource again or find an alternative supplier.