How do news organizations design and implement algorithmically personalized news services? We conducted 16 in-depth interviews with professionals working in European public service broadcasting and commercial quality news media to answer this question. The news business is undergoing rapid transformations regarding how news production is financed, how news is produced and delivered to audiences and how citizens consume news. In all of these changes algorithmic recommender systems play a role. We focus on news organizations’ own personalized news services, and analyze how they define the role of personalization in contributing to the financial success of the organization, in reaching and retaining audiences, and in fulfilling their editorial mission. We interviewed editors, journalists, technologists and business intelligence and publishing professionals to gain a structural understanding of the often conflicting goals of personalization. We found that rather than focusing on increasing short-term user engagement, European quality news media try to use news personalization to increase long-term audience loyalty. In distinction to the “platform logic of personalization”, which uses personalization to produce engagement and sell audiences to advertisers, they have developed a “news logic of personalization”, which uses personalization to sell news to audiences.