Libraries support data-driven decision making

The following post is part of an ongoing series about the OCLC-LIBER “Building for the future” program. A Dutch version of this blog post is also available.

The OCLC Research Library Partnership (RLP) and LIBER (Association of European Research Libraries) hosted a facilitated discussion on the topic of data-driven decision making on 7 February 2024. This event was a component of the ongoing Building for the future series exploring how libraries are working to provide state-of-the-art services, as described in LIBER’s 2023-2027 strategy.

This image shows three women seated at a table working at computers.
Photo by S O C I A L . C U T on Unsplash

The OCLC RLP team worked collaboratively with members of the LIBER Research Data Management  and Data Science in Libraries working groups to develop the discussion questions. Like our earlier discussion on research data management, we tried to keep things practical, asking participants to share about current and future efforts, and to contribute their thoughts on the role and value of the library in supporting data-driven decision making. Small group discussions were facilitated by generous volunteers from LIBER working groups and OCLC.

The virtual event was attended by participants from 35 institutions across 15 countries from Europe, North America, and Asia. Despite many regional and national differences, there were several key themes that surfaced across the seven breakout discussion groups, which is synthesized below.

What does “data-driven decision making” mean for libraries?

We asked participants this question in a virtual poll, and we reached fairly strong consensus that data-driven decision making means “using evidence to inform decisions and evaluate their outcomes.” While we framed this discussion using the phrase “data-driven,” we recognize that others prefer “data-informed” or “data-conscious.”

Indeed, while the conversations recognized the value of using quality data to inform decisions, we also heard cautionary comments that data should be considered as a decision support tool. Data should be used within context, and users should not use data to the exclusion of other qualitative ways of knowing.

Online poll responses to question about the meaning of “data-driven” decision making

How are libraries supporting data-driven decision making?

There are dozens of ways that libraries are supporting data-driven decision making. We heard from participants who described collective collections efforts, where a group of libraries is working together to manage their combined holdings, to support collection retention decisions, and more. Additionally, borrowing statistics can be used to inform both collection development and weeding decisions.

Beyond collections, participants described analyzing library building usage data (such as gate traffic and wifi usage) to measure the busyness of spaces, to inform space management decisions.

Participants also described the growing role of the library in research analytics, in support of institutional goals. In the UK, the library is usually responsible for managing data about the institutional scholarly record, for reporting to the national Research Excellent Framework (REF) assessment exercise. Elsewhere, library workers are supporting institutional efforts to understand research productivity, progress toward open research goals, and identify potential collaborations. And, of course, libraries are creating specific roles to manage a wide variety of data and make it available for reuse, the topic of a recent LIBER interview with Matthias Töwe, Data Curator at ETH Zurich Library.

Supporting data-driven decision making is challenging

Libraries are awash in data. Several participants described the feeling of being overwhelmed by all the data available, with the sheer volume making it challenging to manage, clean, and use effectively. At the same time, it can be difficult to even know what data is available, because it is spread across many silos within the organization. Greater organization and transparency are necessary.

Collaboration is required, regardless of scale. Multi-institutional collective collections analyses demand significant investment and commitment from a wide variety of stakeholders across many institutions and library units. Even when seeking an answer to local operational questions, where , as one participant noted, “we need certain bits of data from other people,” library workers must apply social interoperability to get work done.

Users asking for data and reports are often unable to clearly articulate what they need. This is apparently such a widely felt pain point that it was the #1 response to our online poll about the tensions and challenges of collaboration around data-driven decision making. One small group discussed the need to repurpose “reference interview” skills to interview data consumers in order to clarify the questions they are seeking to answer.

What is the value proposition of the library for data-driven decision making?

We asked the small groups to discuss the overarching value proposition of the library in supporting data-informed decisions, and several themes emerged across the group discussions:

Libraries know metadata. The skills and knowledge that metadata librarians hold about library data is invaluable for managing collections. . . and more. This metadata expertise is clearly a strength, but one that may be easily overlooked, requiring improved messaging to non-library audiences. One participant expressed concern that library expertise is too easily dismissed because it was seen as “just books,” without recognizing the transferability and value of these skills, such as experience with complex enterprise systems, proficiency with data management, and the consistent application of rules, standards, and policies.

Libraries use data to responsibly steward resources. Shared print and collective collections activities rely upon aggregated library holdings data to make decisions about collections development, retention, and long term and cost effective stewardship of the scholarly record. Several participants also described how data about both collections and library building usage has been leveraged to make decisions about future space utilization. Libraries also need “to show that we are making good use of [campus] resources, so that they will continue to fund us.”

Research support services that extend beyond the library are highly visible to other campus stakeholders. Library support in areas like research data management, research intelligence, and managing data for national reporting requirements, in alignment with campus strategic priorities, often offer the greatest visibility to non-library stakeholders. For example, participants from the UK and Hong Kong described the central role of the library in collecting the scholarly record of the institution, to support national reporting requirements and provide analysis of the output and impact of institutional scholarship. A Canadian participant described their creation of a bibliometrics librarian who now leads an informal network of business intelligence officers across the university, providing decision support about compliance, assessment, and funding. Libraries are also exploring how they can define a set of indicators that will provide insights into open research activities, as described in a recent RLP webinar presentation by Scott Taylor at the University of Manchester.

What are some strategies libraries can use to demonstrate this value proposition?

Library leaders must advocate for the library’s role. We heard many examples of libraries providing institutional decision support. However, it can still be a challenge for non-library stakeholders to recognize the library as strong contributor, and participants echoed a concern we heard in the previous facilitated discussion on research data management: “People don’t think of the library.” Library leaders should be relentless in advocating for the knowledge and skills of library workers, guiding campus partners to conceptualize the library in new and modern ways.

Use a “purpose tree” to codify value and communicate internally and externally. A UK participant in one small group discussion shared how she had created a purpose tree for her metadata team, which included a high level vision and strategy statement about their activities and how they contribute to library and university strategy. The document helped demonstrate that catalogers weren’t just “sitting in the corner and going through books,” but that they played a vital role in stewarding quality metadata, supporting an array of business needs. The other small group participants expressed sincere enthusiasm for this idea, and it seems to offer a framework for team building and strategic goal alignment.

Visualizations and data storytelling is required. A strong theme throughout the small group discussions was that the data itself is not enough. Librarians must also develop data storytelling skills and leverage visualizations in order to effectively communicate and create enthusiasm for the data findings.

Library workers must upskill, both individually and in teams. Library workers bring significant skills to managing data, but they often lack training in data analysis, including tools like PowerBI and Tableau. Participants shared many stories of how they are acquiring these skills. For example, a Hong Kong participant described how her institution formed an interest group to explore data analysis and build skills, enabling participants to learn from each other in a supportive environment. Another participant from the Netherlands described a similar effort, where their local working group is learning data visualization skills and building a broader community of practice. In general, participants expressed the need not only for the upskilling of existing staff, but the future onboarding of staff members with mature technical data analysis skills.

Word cloud summary from event polling

We concluded the event by inviting participants to share one word about how they felt, and they reported feeling inspired, informed, and encouraged.

Join us for the upcoming facilitated discussion on AI, Machine Learning, and Data Science

The next discussion in this multi-part series on state-of-the-art services will take place on 17 April, where we will collectively explore the challenges and opportunities of AI, machine learning, and data science. The session will focus on the ways that research libraries are using (or want to use) advancing technologies to improve library workflows, metadata, and more. By facilitating structured small group discussions, we are inviting participants to ideate and share about their future visions for AI and data science, while also purposefully exploring the challenges libraries face in leveraging emerging technologies responsiblyRegister today to save your spot.