As do many of the research done in the field of computer-mediated communication (CMC), this paper relies on testing with student groups. Goal of the paper was to compare the fit of CMC tools to idea generation and problem-solving tasks. Idea generation tasks require conveyance of information (i.e. low feedback requirements and high parallelism) and problem-solving tasks require convergence of information (i.e. high feedback requirements and low parallelism). Problem-solving requires participants to recall specialized knowledge, discuss and debate alternative points of view. Thus, the goal is to gain shared understanding, whereas in idea generation it is to share information. In idea generation it is not necessary for the people to focus and discuss the same piece of information. The results show that the CMC tools work better on idea generation, since the tools enable high parallelism.