Libraries are for Use

Demonstrating the value of librarianship

Benchmarks of usage: What is good? What is not-so-good?

I’ve been thinking a lot about benchmarks lately. This is because we are in the midst of a complete overhaul of our collection development methods (more about this in another post). This method is going to require more frequent and broader measures than ever before. The problem is that these measures mean little if there isn’t something to compare them with. Take for instance, the cost-per-use (CPU). So a particular resource has a CPU of $5.15. Is this good or bad? Should we merely try to advertise it more or should we cancel?

For the latest budget cuts, we needed a simpler way to rank our resources from most likely to least likely to be cut. So we compared everything against each other. Everything was relative. For this, we used percentiles: resources in the top (highest) 75th percentile (75% or higher) were among those looked at first. But the CPUs ranged from pennies to over $200. What should we consider the benchmark CPU? Ideally, it would be in the pennies range. But that’s not practical. As a librarian, I know there are some valuable resources that will be both expensive and not heavily used.

These thoughts made me wonder how a business person would view this problem. If we were a business, what do we need to change to cover our costs? I know this comes dangerously close to considering the library as a “cost-recovery” service, but it could help us determine some target values of key measures.

So if we were a for-profit, what kind of service would we provide? What would be our business model? I could think of two: flat fee service (Netflix) or per item rented (Blockbuster). Actually, many libraries in early America were the former, charging membership dues to gain full access. So this is not terribly new. But the per-item model is easier to think about, so let’s consider that first.

I wondered what would go into the calculation of a per-item fee, and I thought of the cost of the item to the library, its expected use over its lifetime, and the value to the customer. The first two are easy to estimate, at least at the broadest level. We can examine median & average uses per item for various categories of resources like book circulation, sessions, etc. To calculate expected use of physical materials, we could multiply the total number of first time, say 5 circulations (proxy for number of people) by the average number of total circulations (including renewals), say, 2. Then we can divide the 5-year average cost of books, say $50, by the expected uses (5*2=10) to give us our cost-recovery fee ($5). I know these are very blunt calculations.  After all, there are the costs of processing, storing and re-shelving, etc. But it’s a starting point.

So now we have a benchmark CPU- a cost-recovery – $5 for print books. Thus, if the CPU of books from a particular publisher is $50, then the collection development librarian may want to reconsider this publisher. Of course, we could calculate CPU’s by year of publication and by general subjects, so that the comparisons can be more refined.

A similar method can be used for e-resources. Unlike physical materials, which currently have but one measure of usage, there are often multiple usage measures. The choice should be applicable to all resources being compared. For the purposes of setting a benchmark of all e-resources, the session would be the most common measure. More refined analyses can be made by grouping resources by the kind of final outcome or product (abstract, full-text, song played) and then using that as the measure of usage. Thus, the CPU of an A&I database would be based on the number of abstracts viewed, while in the case of a full-text database it would be the number of full-text views.

Membership fees would be harder to calculate, particularly with e-resources. This is because we currently don’t collect data that could show the number of e-resources that are individual uses. But I believe we could calculate it. Our computers in the library require users to log in (there is a loophole for guests, but it’s not known how many users use it.), which could be tracked via a proxy server. Using data-mining of proxy server logs, we could estimate the number of resources any one user accessed over the course of a set time period. For physical materials, it is a little easier to get per-person usage. Our ILS records the total circulations in each person’s record. Unfortunately, it is a very uninformative number. But this information could be estimated using active circulation data. Daily circulations & renewals could be downloaded (even by user & material type) and over a set time period, averages estimated.

While it would take quite a bit of data-wrangling, I can see the potential to estimate usage of physical & electronic resources by any one person.


# of uses


Cost of usage













A&I Database




A/V stream







The purpose of this exercise was to explore the sources of bench markers. But these only reflect internal usage. Its still just relative. Is a CPU of $5 for a physical book good or bad? Of course, this is a value judgment, which are inherently sticky. We can compare one resource against another, based on internal usage. We might want to compare our CPUs against other libraries. This would let us know if we are getting more or less out of our expenditures than others are. This, of course, would require collaboration, but I don’t think so much so as to be impractical. 

But it still doesn’t answer that vexing question: is it good or bad? That, I’m afraid, is one that must be answered by the stakeholders. Do our administrators think $5 per use is a bargain or a rip-off? What about our faculty and our students? Well, that we may never know. So, what does it mean to us, the librarians?


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s


This entry was posted on August 14, 2014 by in Academic Libraries, Assessment, Collections, Intriguing Ideas.

Join 28 other followers

August 2014
« Jul   Sep »
The Scholarly Kitchen

What’s Hot and Cooking In Scholarly Publishing

Libraries are for Use

Demonstrating the value of librarianship

Scholarly Communication |

Demonstrating the value of librarianship

Library & Information Science Research |

Demonstrating the value of librarianship

Library Collections |

Demonstrating the value of librarianship

Lib(rary) Performance

About library statistics & measurement - by Ray Lyons

Walt at Random

Demonstrating the value of librarianship

The Scholarly Kitchen

Demonstrating the value of librarianship

The Quarterly Journal of Economics Current Issue

Demonstrating the value of librarianship

Texas Library Association blogs

Demonstrating the value of librarianship

Demonstrating the value of librarianship

Stephen's Lighthouse

Demonstrating the value of librarianship


Demonstrating the value of librarianship

Reference Notes

Demonstrating the value of librarianship Truth-O-Meter rulings from National

Demonstrating the value of librarianship

Open and Shut?

Demonstrating the value of librarianship


Demonstrating the value of librarianship

Musings about librarianship

Demonstrating the value of librarianship


Demonstrating the value of librarianship

%d bloggers like this: