PERANCANGAN METODE PERAMALAN PERMINTAAN PASAR UNTUK PERBAIKAN FORECAST ACCURACY DAN KEBIJAKAN INVENTORI DI PT MEPROFARM PHARMACEUTICAL INDUSTRIES

• Main Author: FAUSTANTA NIM 13414072, VANIA
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/31441
• Institution: Institut Teknologi Bandung

The Indonesian government law about National Healthcare Insurance (Jaminan Kesehatan Nasional or JKN) causes a surge of demand for generic medicine, which in turn causes pharmaceutical companies to do business in such a way in order to obtain maximum income and ensure customer satisfaction. PT Meprofarm Pharmaceutical Industries is one of those Indonesian pharmaceutical companies that produce various kinds of generic and patented medicine. <br /> <br /> Recently, PT Meprofarm has been experiencing stockouts or back order events that cause inventory to be exhausted which makes the company unable to fulfill customers’ demand on the due date and causes late delivery. PT Meprofarm currently does not have a theoretical system of demand forecasting and inventory management in place that can make predictions of future demand based on historical sales data. The current forecast accuracy in PT Meprofarm is 61,77% (with the error level in the amount of 38.23%, which is calculated using the Mean Absolute Percentage Error or MAPE method), which is considered high, especially when compared to the company’s Key Performance Indicator (KPI), which is 25%. Therefore, this study is conducted to increase the company’s forecast accuracy and use the results to estimate the right inventory policy. <br /> <br /> The improvement in forecasting methods is done by testing three different time-series methods to determine which method generates the smallest value of MAPE after having classified the studied products or items into clusters based on data similarity first using hClust method. The Q Model of inventory is used to estimate the best inventory policy for each item using Microsoft Excel VBA Macro as a counting tool. The improvement results in 1.73% increase in forecast accuracy which is then used to estimate economic manufacturing quantity, reorder point, safety stock, service level, and expected total annual inventory cost that vary based on items considered.