Much of the empirical literature on PRP (Performance Related Pay) focuses on a question of whether the firm can increase firm performance in general and enterprise productivity in particular by introducing PRP and if so, how much. However, not all PRP programs are created equal and PRP programs vary significantly in a variety of attributes. This paper provides novel and rigorous evidence on the productivity effect of varying attributes of PRP and shows that the details of PRP indeed matter. In so doing we exploit the panel nature of our Finnish Linked Employer-Employee Data on the details of PRP. We first establish that the omitted variable bias is serious, makes the cross-sectional estimates on the productivity effect of the details of PRP biased upward substantially. Relying on the fixed effect estimates that account for such bias, we find: (i) group incentive PRP is more potent in boosting enterprise productivity than individual incentive PRP; (ii) group incentive PRP with profitability as a performance measure is especially powerful in raising firm productivity; (iii) when a narrow measure (such as cost reduction) is already used, adding another narrow measure (such as quality improvement) yields no additional productivity gain; and (iv) PRP with greater Power of incentives (the share of PRP in total compensation) results in greater productivity gains, and returns to Power of incentives diminishes very slowly.