The primary goal of this session was to help journal publishers of all kinds to optimize their operations. The group of speakers provided a microcosm of the challenges faced by different publishers throughout the STM community. In terms of the size of their operations, publications varied from one weekly journal and 18,500 pages published annually in the case of Neuroscience; to one weekly journal and 22,000 pages in the case of the Proceedings of the National Academy of Sciences of the United States of America (PNAS); 33 technical journals, 22,000 pages, and 220 issues per year in the case of the American Society of Civil Engineers (ASCE); and 10 technical journals, 26,000 pages published, and more than 100 issues released annually (semimonthly to quarterly) in the case of the American Meteorological Society (AMS).

Different fundamental workflows are needed for publishers with such a variety of throughput. Jackie Perry discussed the need for Neuroscience to transition from an “assembly-line” workflow, in which rapid training and production were outweighed by the boredom and burnout associated with that approach, to a “people-designed” workflow, which allows a unified team approach consisting of interested and integrated staff that have much broader knowledge of all aspects of the workflow. ASCE, with its many journals, has opted to forge a single workflow that covers all but one of its journals; this requires rigid schedules and lots of planning but yields maximum efficiency. ASCE has the same month-by-month schedule for each journal, staggered due dates, and standardized deliverables (for example, one due date and one mailing date). The exception or outlier is a quarterly journal that has its own scheduling paradigm and that, according to Angela Cochran, is an enormous time sink, taking up a disproportionate amount of time. Cochran strongly recommends against having titles that are outliers with respect to workflow. The ASCE approach is in distinct contrast with that of AMS, in which production scheduling and resource allocation are customized to the demands associated with each individual journal, which can range from under 1,000 to more than 6,000 pages annually. AMS also maintains separate in-house copyediting and technical-editing staffs; this allows for increased scrutiny of each paper but requires a larger staff than most STM publishers to adjust for the number of journals and pages published. The print schedule for PNAS is fixed; author proof return dictates the online production schedule. Processing and production are handled by two teams of six people each.

Although the contrasting workflows described show clearly that there is no single overarching approach to optimal production for publishers of different kinds, what was most striking in listening to the speakers were the similarities in reporting, metrics, and relationships with vendors that undergird the production workflow of all four publishers. It is those common characteristics that appear to drive the success of the publishers and that other STM publishers would do well to emulate. They include comprehensive reporting and scheduling (knowing where each paper is at any given moment), regular meetings with staff and vendors to make sure that expectations are clear and that production is on track, robust metrics to measure performance throughout the production process, and adherence to predetermined budgets.

The session was engaging, and the takeaway message was that although publishers can use any number of workflow designs effectively, core elements are common to all successful publishers in producing their journals.