While libraries are reshaping their digital services, the work practices of academic researchers across all disciplines are changing. There is evidence of profound changes in the information behaviour of so-called digital natives, but seasoned scholars, too, seem to be adopting new ways of conducting their work. This change poses a challenge to research libraries: we have constantly to update our knowledge of our customers. People’s information needs are interwoven with aspects of their everyday life. It is, therefore, not adequate to study information behaviour in relation to a single digital service, or even within the context of a library. Recent research has introduced methods that take into account users’ real life context and deepen our understanding of users’ needs. This paper presents a qualitative analysis of researchers’ needs in relation to the Finnish Digital Library, Finna. Our motivation to study this group of Finna’s users arose from the fact that all Finnish research libraries were to join Finna before 2015. Thus, we had to take into account a large end-user group with variable needs. We adopted a contextual approach and posed the following research questions: When does a researcher need a digital library? How does s/he use a digital library in the research process? We based the analysis on data from two independent studies: 1) a user survey provided responses from 129 researchers; and 2) a contextual study provided interviews with 12 researchers from different fields of science. We reflected the findings from these two studies, first against each other, and then to the concept of ‘Reconfiguring Access’ by Dutton (2010). This concept depicts the coevolution of technology and social processes in the context of research. The results did not indicate any dramatic findings or differences between fields of science. The digital library was only an information resource among others. The analysis suggests that as researchers use different information tools in tandem, they also compare the relevance of their functionalities, e.g. search results, at every turn. The analysis detected only few indications of a wider reconfiguration of use, i.e. how researchers use digital resources for purposes other than information retrieval. These notions of collaboration, observation, analysis, and distribution, which can be seen as signals of change in researchers’ work practices, may be transferable in the development of smarter information environments.