Benders decomposition entails a two-stage optimization approach to a mixed integer program: first-stage decision variables are optimized using a polyhedral approximation of the problem's projection; then a separation problem expressed in the second-stage variables is solved to check if the current first-stage solution is feasible; otherwise, it produces a violated inequality. Such cutting-plane algorithm can suffer severe drawbacks regarding its convergence rate. We review the battery of approaches that have been proposed in the literature to address these drawbacks and to speed-up the algorithm. Our contribution consists in proposing a unified framework to explain these techniques, showing that in several cases, different proposals of the literature boil down to the same key ideas. We complete this review with a numerical study of implementation options for Benders algorithmic features and enhancements.