This article studies the algorithmic project News values at Swedish public service radio, from the perspective of datafied managerialism. Drawing on ethnographic observations and interviews with managers the study shows how the project, that outwardly works to automate news-sorting algorithmically, was employed to generate data about a number of internal journalistic activities, for a variety of purposes. Data was perceived of as a type of capital that could engender, amongst other things, increased knowledge about the internal workings of the organization, thus making it easier to audit its activities, and to standardize the practice of news-valuation throughout SR. Importantly, these goals were not planned in advance, but emerged over the course of the project. The results show how longitudinal approaches to algorithms and data-collection could benefit journalism studies, as they provide a more comprehensive picture of how data are operationalized in journalistic organizations.