The new inventory policy proposed in this research is based on a dynamic order. The newly proposed policy can govern the inventory system since it selects how many items to reorder depending on the system levels. In this research the proposed policy is called a ”dynamic order” it allows the inventory level to reach one of two levels, and the number of reorders is dependent on the inventory level. To solve the inventory optimization problem and find the optimum solution with the lowest cost, the proposed policy has been compared to a well-known inventory policy using genetic algorithm. Computational experiments were conducted to assess 27 experiments with regards to differing lead times, demand, and no. of customers. The results of all the experiments showed that the proposed policy could effectively lead to a 6.7% reduction in cost