Faulty group processes have harmful effects on performance but there is little research about intervention protocols to pre-empt them in higher education. This naturalistic experiment compared a control cohort with an inducted cohort. The inducted cohort attended a workshop, consultations, elected a leader and used tools (a group log and group contract) designed to minimize social loafing, optimize coordination (by boosting good information sampling) and orient group dynamics towards the task. In the absence of inductions, a faulty system of processes was in play and this had a significant impact on group performance. In contrast, the inductions created a buffer. Structural equation modeling showed that the intervention made both group cohesion and conflict beneficial to group performance. The induction protocols enhanced students' individual accountability, a sense of unique responsibility and dissent during group decision-making (which improved its quality). The implication is that inductions help optimize the processes within student teamwork.