Access to collections is expanded through digitization, but are we saving the "best" volumes, which volumes are the best, and how do we make that decision? Capturing "real" collection data to objectively make and support those decisions is part of Library of Congress (LC) research. Current data suggests that most cultural heritage institutions have digitized less than 10% of their collections, so preservation of the print record is critical for long-term access to this knowledge. This is especially true for 19th and 20th century paper-based materials, where mass production methods resulted in less stable paper. Moving from subjective to objective based data for retention and withdrawal decisions is critical for the robustness of the print corpus and the future of digital collections.