Five Quick Questions

Mark Sandler

Mark Sandler, Chief Development Officer, University of Michigan Libraries, Ann Arbor, MI, USA

1. Does your library assess user behavior?

At the University of Michigan, we don’t always have in place a concerted or ongoing strategy for assessing user behavior, but we’re certainly attentive to the expressed or observed needs of our users, and have undertaken a number of smaller studies around particular services. In general, I’d say the number of users being served remotely makes it harder than ever to maintain a confident grasp of user needs, since those in-library users that we see and talk with may not be typical of the larger community of users and non-users.

2. Why should libraries assess user behavior?

There’s no doubt libraries should be attentive to understanding and satisfying user needs. We’re approaching a time when the most proximate library to a community of users may not be their preferred option for service. As an IP authenticated Michigan user for a free or licensed resource, I can link to that resource from any number of Web pages around the world. Likewise, there are many open online reference services that will respond to inquiries from outside their primary service groups. Hence, libraries increasingly find they are not exclusive providers for users because of locale or organizational affiliation. In such a competitive environment, those that fail to understand or cater to user preferences run the risk of seeing their users drift off elsewhere.

3. How does user behavior impact collection development at your library?

This past year our library carried out a fairly thorough review of print circulations, interlibrary loan borrowing and e-journal/e-book usage. For the first two categories, the data were gathered by user discipline and user status (i.e., faculty, graduate and undergraduate student). In conjunction with selector observations about campus programs, we were able to get a pretty good sense of which collecting areas were experiencing greater than average use or pressure. This information was then used as the basis for the allocation of new base funds, as opposed to our more accustomed pattern of granting across-the-board increases to collecting areas. While such efforts are fraught with complexity — they are approximations of reality at best — it still seems important for librarians to keep asking such questions and trying to make sense of murky indicators of user need.

4. When assessing user behavior, do you take into account varying types of users?

We’re very mindful of the differences between the categories of users we serve, and how difficult it is to design systems and services that satisfy user needs across that broad-ranging spectrum. In a campus environment, that used to mean recognizing the different needs of highly sophisticated faculty users as compared with more novice student users. Increasingly, this simple dichotomous distinction is being complicated by the addition of new layers of continua to describe user efficacy, including among others differences in technological skill and comfort, cultural and language familiarity, extent of interdisciplinarity, and textual vs. media-rich data sources. The net effect of these cross-cutting patterns is that all sorts of users are likely to be insecure seekers of information in some aspects of their work, and libraries now have many more ways in which to disappoint and discourage them, despite doing a good job at meeting some of their needs.

5. What role do you see user behavior playing at your library in coming years?

As more and more of library use is remote, we’ll actually have better and more accessible system-harvested data about users than in the past. And importantly, this will be less about what they say they want or do, and more about their actual habits. It will become easier to trace usage patterns across large and active populations of users, feeding such data directly into reports designed to compare and contrast user behavior by discipline, status, time of year, preferred path to resources, turnaways and failed searches, and many other indicators of preference and satisfaction. The use of more secure authentication certificates will aid in this effort, as will the increased integration in our library gateway systems.

In general, asking librarians if they care about user behavior is like asking Elsevier if they care about customers: The answer has to be “Of course!” That being said, library users are as different as librarians and libraries themselves. Hence, when we start looking, we find a high degree of variability among users rather than a neat clustering of typical behaviors. So, while that sounds like the basis of a good counterargument for throwing up our hands and doing absolutely nothing, it’s really intended as a reminder that when assessing user behavior, concepts like the “average user” or “majority of respondents” shouldn’t distract us from the full range of user responses.