The purpose of this research is to better understand how the privacy and audience component of feedback influence individual and team performance. A modified version of the Multiple Errands Tests (MET), the Team Multiple Errands Test (TMET), was developed for and used to evaluate teams. In the present task, three team members were given individual shopping lists and a team shopping list within a virtual mall. The group had to work together to ensure that all of the items were purchased. An expert system offered feedback to participants during the task, and feedback was given publically or privately (privacy variable) and directed at the group or at the individual (audience variable). Individual performance and team performance scores across four sessions conducted by 10 teams were not significantly impacted by the feedback modality. However, initial analysis of communication patterns over trials and team strategy used suggest trends that would be worth exploring with additional participants. This work does demonstrate the value of the TMET test bed as a new methodology for studying team behavior.