Moneyball, the Extra 2%, and What Baseball Management Can Teach Us About Fostering

Innovation in Managing Collections

Greg Raschke

North Carolina State University

Ithaka

March 8, 2012

Moneyball and More...

Baseball to Collections Context

Looking Deeper and Questioning Assumptions

Identifying market inefficiencies.Apply and acculturate significant

innovation.Question long-established wisdom.Test what is “known” with in-depth

analysis, statistical modeling, and new approaches.

Emphasize interpersonal skills in leveraging new knowledge and approaches.

Supply-Side Collections Print-based, unpredictable demand,

and legitimate need for just in case collections

Lead to judging quality by size (as in the ARL rankings) and libraries were then held captive to this standard

Contributed to inelastic demand for journals and a combination of speculative and package buying

Use is secondary to size, dollars expended, and other input measures

Credit to David Lewis (http://ulib.iupui.edu/users/dlewis)

Supply-Side to Demand-Driven

Demand-Driven Collections

Make information easily, widely, and cheaply available

Collections as drivers of research, teaching, and learning

To make special or unique collections held/managed by the library available to the user community and the world

Demand-Driven – Changing Practice

Tension between time-honored role as custodians of scholarship versus enabling digital environment for scholars

Not just PDA – portfolio of approaches, but certainly more responsiveUtilize new tools and techniques to become advanced analystsTruly embrace evidence based decision making

Look at how collections are actually used, not at expressed need

Demand-Driven – More Assumptions

Less tolerance for and less investment in lower use general collections

Resource management based increasingly on use

Modify collecting based on changes in the actual use

Risks of doing nothing – newspapers

Demand-Driven – Assertions

Rewards of adapting – more used and vital than ever

Use based and user driven collecting models will take growing share of budget

Bet on numbers Bet on good and quick Put resources into enabling

digital environment for scholars and custodian role will come out of that strategy

Why So Much Data?

Data analysis is a key component in solving/managing: Increasing pressure for accountability Increasing capability to gather and analyze data Increasing precision in the way we build collections and expend

resources Advocacy

Changing practice and data analysis at NCSU

Serials Review 2009 – Open, Data-Driven, and Real-Time Analysis

Standardized usage data (where available)

Bibliometrics - publication data and citation patterns (e.g LJUR)

Impact factor and eigenfactor User community feedback via

interactive, database-driven applications

Weigh/calculate/quantify user feedback

Weigh price against multiple data points

Usage ((07 usage+08 usage/2)+(publications*10)+ (citations*5)+(Impact Factor)

Community Feedback ((Weighted Ranking x % Match) x Total # Rankings) + 0.1 x # of "1s“

Price/feedback valuePrice/useMerge results to filter out top 20%

and bottom 20%

Looking closer – Finding balanceAn example - a closer look at print item usage

Traditional ILS reporting tools can make this difficult

Advanced analytical tools can help

What types of questions can we ask?

Should Patron-Driven records not purchased be purged after 2 years?How does print item usage break down?Do print items even get used?

If it’s not used after 2 years…

Should PDA records be purged?

Maybe…

We haven’t even hit 50% usage

But what if we take a longer view…

If it’s not used after 2 years…

Things begin to look different

Looking even closer… How does print

item use break down?

Single circ usage is consistently ~14%

Would this change in a PDA only world?

Expenditures to University Data

Expenditures to University Data

Expenditures to University Data

Expenditures to University Data

Measurable Uses of the Collection 2009/2010

Full-text journal downloads* 3,672,600

Database use 1,989,972

Print book circulations/renewals 525,430

Digital collections requests 471,403

E-books 149,815

Reserves** 327,267

Total Uses 7,136,487

* Includes use of NC LIVE full-text content** Includes textbook, print, and e-reserves usage

Measurable Uses of the Collection 2009/2010

From Assumptions to Assertions to Practice Grow/develop/hire analysts. Adapt statistical tools such as SAS software. Partner with digital library/technologists. Develop positive arbitrage. Put resources into enabling digital environment for scholars. Experiment – budget for it, reward it. Work hard to get the faculty to buy into new approaches. Combine analytical approaches with the people skills .

“…there was a bias toward what people saw with their own eyes, or thought they had seen. The human mind played tricks on itself when it relied exclusively on what it saw, and every trick it played was a financial opportunity for someone who saw through the illusion to the reality”.