Fifth-generation (5G) cellular communication systems aim to obtain a higher data rate, decreased latency time, higher performance even at high mobility speeds, decreased system complexity, lower transmission cost, and an increased system capacity and coverage area. Many of these goals can be achieved because of the studies devoted to the physical layer of 5G cellular networks. In this respect, solutions to the problems of beamforming (steering and precoding or precoding and combining) and channel estimation that are encountered in the physical layer of 5G cellular networks are the key points to achieve the aforementioned goals. Thus, a two-stage beamforming method is proposed in this study.The proposed method is a suboptimal method that minimizes the difference between outputs obtained when fully digital and hybrid beamforming methods are used. The analytical results, which are validated through simulations, demonstrate that the proposed method is an effective solution and, hence, the preferred beamforming approach for 5G millimeter wave band-based wireless systems.