This paper proposed a two-step approach for reducing power losses using a Modified Mountain Gazelle Optimizer (MMGO). First, the loss sensitivity factor is used to identify the most promising locations for shunt reactive compensators. This approach reduces the search space and improves the efficiency of the optimization process. Secondly, the proposed MMGO algorithm is applied to determine the optimal locations and sizes of the shunt compensators. The performance of the proposed method is evaluated on standard IEEE 33-bus and IEEE 69-bus systems and compared with other optimization techniques reported in the literature. The proposed MMGO produced 34.05% and 35.45% reductions in active power losses in the IEEE 33 and 69 bus systems respectively. Simulation results demonstrated the effectiveness of the two-step approach in achieving significant loss reduction and improving voltage profiles. The findings of this research provide valuable insights for distribution system operators and pave the way for more efficient and reliable power distribution networks.