DESIGN OF RETAIL PRODUCT PURCHASE SYSTEM USING MARKOV CHAIN METHOD

• Main Author: HENDRATA NIM 13414052, ANGGIE
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/25594
• Institution: Institut Teknologi Bandung

CV Bukit Mas Abadi is a form of limited partnership in trading. One form of business managed by CV Bukit Mas Abadi is a concept store. Concept store is a type of retail business that sells various types of goods in one place, but the goods sold have a characteristic based on company's theme. Concept store can be classified as a store that sells perishable goods with the value of goods decrease with the age of goods. The main problem that occurs on the Concept Store is the occurrence of understock and overstock. This happens because there is no purchase system to determine the ordering policy. Therefore, in this final project will be designed purchasing system by using Markov Chain method. This method is chosen because it can consider the expected cost in the long run to make a decision on purchasing. <br /> <br /> Stages of work begins with the reduction of the type of goods in accordance with the list of goods according to the Directorate General of Intellectual Property (DJKI), after the reduction is done, conducted ABC analysis to determine the selected product in the design of the purchasing system, then Collection of historical sales data used to determine the demand for goods for each product . The initial stage of designing the Markov Chain model is to determine the number of states by dividing the range adjusted for the average sales of each period. The main part of the design of the markov chain is to calculate inter-state transition probabilities for sales and inventory quantities. The final stage is the calculation of the profit expectations of each period by considering the cost of booking, overstock costs, and understock costs. <br /> <br /> The result of this is the ordering policy for the 35 selected products with the highest profit expectations criteria. Obtained order policy results for 18 products of ordering a number of 3 when inventory amounted to 1, for 10 products that order a quantity of 2 when inventory amounted to 1, and for the remaining 7 products that order a number of 3 when inventory amounted to 2. The author also proposes to maximize profits after implement the ordering policy. With the markov chain method obtained an increase in profit from the existing condition Rp 591.960.709,37 to Rp 730.274.157,79.