The determination of process mean is important in industries especially items that governed by laws and regulations on net content labeling. Thus, the economic selection of process targeting mainly the optimum process mean is critically significant since it will directly affected the quality characteristic of the item. Depending on the value of quality characteristic, an item can be reworked, scrapped or accepted by the system which is successfully transform to the finishing product by using the Markovian model. By assuming the quality characteristic is normally distributed, the optimum process mean is obtained via probability of the item being rework, scrap and accepted. In this paper, we present the analysis of selecting the process mean by referring to drink bottling production process. By varying the rework and scrap cost, the analysis shows the sensitivity of the Markov approach to determine process mean which maximizes the expected profit per item. Then an improvement of determining the optimal mean estimation is proposed to simplify the iteration process.