In late 2012, it appeared that the University Library, University of Saskatchewan would likely no longer be able to afford to subscribe to the entire American Chemical Society "Big Deal" of 36 journals. Difficult choices would need to be made regarding which titles to retain as individual subscriptions. In an effort to arrive at the most conscientious and evidence-based decisions possible, three discrete sources of data were collected and compared: full-text downloads, citation analysis of faculty publications, and user feedback. This case study will describe the triangulation method developed--including the unconventional approach of applying a citation analysis technique to usage data and survey responses. Such a thorough, labor-intensive, method is likely not practical for analyzing larger, multidisciplinary journal bundles. When it becomes necessary to break up a smaller collection important to researchers in a particular discipline, this technique may provide strong evidence to support librarian decisions as well as involve faculty in the process. [This study was originally presented to the STS Research Forum at the American Library Association (ALA) Annual Conference, Las Vegas 2014.]