This paper discusses the key collaborative features of scientific visualization systems. An experiment with several groups of users has been conducted to evaluate the usability and effectiveness of collaborative tools in such a system. In particular, we compared audio and textual communication support for information exchange as well as the use of graphical indicators. Our analysis revealed the importance of the choice of the conversation means and, surprisingly, the little influence of indicators. Furthermore, the experiments have demonstrated that the system's usability can be significantly improved through appropriate grounding and awareness support. User satisfaction therefore increases, which in turn can potentially also increase the quality of the collaborative work.