Rewarding, essential, potentially taken for granted and relatively under-theorised, research supervision plays a key role in higher education in empowering students to become researchers. Students need to be able to engage with and work with their supervisors in a research dialogue focusing on areas of mutual interest. Supervisors can worry that they have not been in touch with students enough, or that they are doing too much of the work – translating ideas into research activities, suggesting reading and clarifying. Students often laugh at the idea that supervisors could do too much, but it is necessary to guide students into autonomy and away from dependence. The supervision role is much more clearly defined as a professional relationship then that of tutor, friend or colleague and it relies on more than goodwill and spare time. Students at all levels need guidance, modelling and managing so that they can start to develop as independent researchers. Students need more carefully organized direction, when a supervisor has a more casual approach. The paper discusses the difference in understanding of supervising functions in the academic society and features defined by students. The list of features required and functions defined for the supervisor is developed based on collection of students’ opinion during last five years and evaluating of this collection according to Delfi procedure for expert evaluation. Analysis of student’s opinion about supervisor’s work gives very interesting results on definition of different types for advising activities, that students expect from their research supervisor. These results can be interesting not only for supervisors, but also for students themselves and for institutions, which regulate functions and learning outcomes at the study program definition level. Moreover, author has been researched also the correspondence of supervisors to the defined features in the division of Applied Computer Science of Riga Technical University.