How to build a successful collection management strategy

Why does a library need a collection management strategy?

Libraries spend a good proportion of their annual resources budget on serials. We know that due to hard economic times, there has been little shift in funding for libraries for the past few years. For 2012 we knew the same would be true, although marginally more positive. Global library budgets were predicted to increase 0.2% on average, and the serials slice of that by 0.5% (global figures from the PCG library budget survey). PCG notes that from their survey, budget increases reported by librarians were largely implemented to meet inflation in prices charged by publishers, and to meet new content demands from their end users.

"Global library budgets were predicted to increase 0.2% on average"

Distributing the designated funds across the library, and specifically across journal holdings, is of critical importance to ensure learning objectives and a return on investment (ROI) is achieved.

Show me the money

Available money is clearly a crucial factor in stimulating systematic/periodic reviews of journal holdings in a library. Another PCG report on journal subscription renewal trends (2005-2011) states budget cuts as the second most popular (or unpopular) reason to cancel journal titles; a trend which has increased in relative importance by 300% from 2006-7.

This is echoed by Duke University’s Scholarly Communications Officer, Kevin Smith, in a recent blog post, “it is increasing prices, coupled with flat library budgets, that account for any rising rate of cancellations” (a relevant example, although in this context he is talking about the relationship between open access and subscription cancellations… perhaps another story to be told there).

In 2010, The Research Information Network (RIN) in the UK, as part of their “Challenges for academic libraries in difficult economic times”, also highlighted serials as a key area  where savings can be made in the face of shrinking budgets (4 key areas were highlighted – serials, the balance of library information resources against staffing, services, and book budgets). While cancelling multiple expensive titles, or an entire big deal may not be an effective solution, several strategies are outlined in RIN’s excellent document.

Internal pressures

There are other pressures on the library to undertake a collection audit, either annually as part of the budgeting/acquisition process, or as a less frequent strategic analysis of holdings. Institutional pressures also exist, as a part of the big picture for a university or college. This is where the strategic plan meets the library.

The library and its resources (including the librarians) are integral to the successful academic output of the institution, and increasingly this is being measured by internal audits, ROI calculations and deep trend analysis. In fact, research carried out and presented in 2011 in the Library Connect newsletter confirms a strong correlation between the expenditure on e-journals, usage of e-journals and research success (read the report, “The use of e-journals and research outcomes: Are they related?”).

An ACRL report published in the C&RL news in June 2012 outlines a key trend in academic libraries in communicating value, and refers to research carried out that directly links the grade point averages of students and overall faculty success with library resources and library instruction.

"The library and its resources are integral to the successful academic output of the institution"

Demanding, savvy users

Another pressure on libraries is from the end-users themselves. Courses change, student intake numbers fluctuate, and departments are never the same size or have the same content requirements.

In addition to this, as more and more of the world moves online, end-users demand instant, seamless access to the content they need, the way they need it, and often won’t compromise (even if it means going outside the library’s content provision). Catering for this range of demands and priorities is no easy task, especially when the goalposts keep shifting year on year.

What’s really at stake?

The ultimate objective of a selection management strategy, looking critically at the library’s holdings on a regular basis, is to make evidence-based decisions on renewals, acquisitions and cancellations. By implementing a clever selection management strategy, the library can balance pressures of both the end-user’s needs and  the institution’s strategic objectives. As well as stabilizing the position of the library, it will also grant the librarian a strong position to demonstrate ROI on their content decisions, and therefore their tenure.

"a strong position to demonstrate ROI on their content decisions"

The Metrics

As more information goes digital and online, more data is produced that we can collect, analyze and interpret. The same goes for journal content, a vast proportion of which is now available and purchased digitally. So what are the metrics that can be put to work in a collection assessment? Click on the image below to expand or right-click to download:

Quantitative measurements

Cost

This old chestnut again. Yes, cost can also be used as a basic determinant in content decisions. It is a quick means to save lots of money but it has serious pitfalls. Basic journal subscription costs don’t take into account the size, frequency of issues, number of pages, number of papers and so on, that a journal may publish during the subscription term.

To add to the complication, subscription costs for titles of similar volume aren’t the same, and vary widely across different publishers. Publisher prices aren’t even static, and big deals (or eDeals) further complicate costs as multiple journals are rolled into one. Format is also an issue when dealing with cost. Often electronic resources are priced differently from their print or hybrid counterparts.

Usage

This has been used for some time, but has returned to the fore again with such developments as the Journal Usage Factor () and the ongoing popularity of live, public metrics as displayed on PLoS journal article pages (far from the closed stats provided by more traditional publishers), in the UK the JISC-led Journal Usage Stats Portal. Other services dealing with electronic journal subscriptions have incorporated extra functionality, such as those recently added to SwetsWise Selection Support.

COUNTER has led the way to some degree, by providing a standard for journal usage statistics (counting unique full-text accesses), but what are they really good for? Well, if you purchase a journal title and the usage count for the year is 0, there is a good chance you will not make a repeat purchase. However, if that title is part of a ‘big deal’ it may not be possible to cancel as there are typically very low swap-in/swap-out rates within these collection deals.

Cost per use

Once you have your costs and usage stats in front of you, it is possible to then make some more informed decisions on value; cost per use (full-text use that is). This seems to work well. Well, to some degree. Aside from being inextricably linked to total cost of that title, it needs a bit of context placed next to it. Some tools populate this for you and allow for additional cost entries.

What about really popular journals like Nature ? They have high subscription costs, but are likely to be read by many users. A niche journal which serves a small, but important department may have a much higher cost per use, but have equal justification for that money.

Perhaps cost per capita is an equal measurement here. A certain department will have solid figures on the members of their department, from student up to senior level, so a valuation can be made on journals which serve a specific department. However, this model does not work when dealing with multidisciplinary titles, neither guarantees any usage from the department it is supposed to serve.

InterLibrary Loans

While these don’t necessarily show value of certain content in your collection, they certainly can be useful to demonstrate any gaps that need filling (and therefore balancing with other decisions), where your students or staff are going to other libraries to access content resources. They can also be useful in showing the buying potential  and for cost savings for pooled resources across institutions.

Other useful quantitative stats

Within an institution itself, there are other metrics that can be referred to. Stats are usually collected on department page accesses, other platform web stats (including repositories), portal access to catalogues and so on. All of these are good indicators to support other key metrics to further inform your decisions.

Qualitative measurements

Impact

We are all familiar with the Impact Factor. ISI (now Thomson Reuters) developed the first one some time ago, but the community still hold onto it as one of the main indices for ranking journal impact. Caution is always advised when dealing with this particular metric, as outlined in the European Association of Science Editors statement in 2008. The fact is that impact may take many different guises.

A great alternative, if we stick with citations as an indicator, is the Scimago Journal Rank, which uses a Google PageRank-style algorithm that adds weighting to the more authoritative sources making the citations, and therefore could be described as more representative. Used properly, these citation-based metrics can give some real depth to a collection analysis and selection validation.

User feedback

This is an invaluable part of the library’s toolkit for making content decisions. Without direct input from faculty heads, subject-specific library staff and even the end-users, the process is likely to fail. As well as surveying or requesting regular feedback on required content, there is also the possibility of matching course reading lists with planned journal holdings to see where the balance may lie.

"all institutions or acquisition librarians will have their own unique pressures"

Is there anything else that might be of use?

The list of influencing factors for making decisions in a collection management exercise can go on to minute detail specific to a single institution or purchasing department.

• Publication outlets - Look where the faculty are publishing most; this could add another layer to making hard decisions about cancellations
• Bounce rates and page depth for journal content – was it just one paper the user was looking for and then left, or did they download several papers form the same journal in one visit?
• Access denied reports – these are usually collected and guarded by publishers in order to upsell content. If you can get hold of them, you can successfully fill gaps in your collection that your users are already trying to access

What about other emerging technologies and metrics?

• Semantic linking – will the fast-growing world of deep semantic lining show relationships between content to give clearer pictures of what journals might be relevant? Or does this rely on the disintegration of the journal as an entity, so that the content becomes king?
• Altmetrics – this term has only been around since 2010, but since then vast amounts of work, investment and online discussions have been poured into these ‘alternative’ ways of looking at the value of research/researchers. It really begs the question, can altmetrics be used by a librarian to judge the relative value of an entire journal title in terms of its impact across the web?
• Workflow tools – other tools that are integral to a researcher’s workflow such as reference management software can also be leveraged for their deep insights into the content publishing, sharing and usage behavior of a group of researchers, such as the Mendeley Institutional Edition, new this year.

Wow, that was a long article

Well, yes it was. But collection and selection management is a long and complex process. There is no right or wrong way to do it, because all institutions or acquisition librarians will have their own unique pressures guiding their decision making. However, there are ways to streamline and simplify the process, particularly on the side of data, which can be pulled into the same place to facilitate a quicker, less time-intensive overview of value.

How do you do it?

What is your strategy, and what stages do you take in the process of selecting content? Do you know other metrics or measurements that have not been mentioned and you would like to share? Please share your tips in the comments section below.

See how other libraries manage their selection management process:
-    A recent analysis of the University of Edinburgh library collection management is highlighted here
-    A statement from Stanford on the key criteria they employ for their selection management
-    A final example from Southwestern Medical Center library

Read about Selection Management services from Swets, including Swetswise Selection Support