This article presents a Heuristic proposal to minimize two situations that occur when machines do not finish processing batches at the same time: 1) bottleneck when a machine takes more time to finish a batch and 2) waiting time that a batch must wait before being processed by the machine. To minimize both situations, batches must have the same processing time. To solve the above, the Heuristic proposal will select the job with the longest processing time and the number of pieces. Each piece will be a batch. The batch processing time is determined by the piece with the longest processing time. To validate the efficiency of the heuristic, information was collected from a case study of a metal-mechanic company. The Heuristic proposal was compared with other heuristics that have been used to solve the batch processing such as: First in-First Out (FIFO), Last in-First out (LIFO), Best Fit (BFF), First Fit (FF), Batch First Fit (BFF), First Fit Longest Processing Time (FFLPT), Decreasing Batch Size (DECR), Processing time to job size ratio increasing order (PIAI) and the Simplex method. The results showed an considerable reduction of bottleneck and waiting time using the Heuristic proposal compared with the others mentioned heuristic and the Simplex method. Keywords: Optimization, Flow shop, Heuristic, Simplex, Metaheuristic.