Governance of a software project involves the distribution and management of decision rights for significant decisions. A decision right grants authority to make decisions and be held accountable for decision outcomes. Though prior research indicates that the exercise and degree of ownership of decision rights has an impact on software project performance, there has been relatively little direct consideration of what the significant decisions should be or what might constitute an effective underlying specification and management of decision rights during the software project lifecycle. In this paper, a research agenda to reveal such knowledge is presented. This report represents the first output of our work in this area.