FastGood Supply Chain Management
PART 1: REVISITED SUPPLY CHAIN
Considering the current physical supply network, it can be seen that FastGood owns 3 factories operating in Indonesia, Thailand and Malaysia while its retailers are spread widely across the South-East Asia region. Due to this, the company is facing issue in managing in current supply chain and is unable to efficiently support its upstream and downstream stakeholders. Due to these issues, the company is facing negative financial downturns and caused operational issues across all of its business segments. Now, due to fluctuating demand of each region, the management is considering to establish new DCs across the region.
DCs are the spatial layout of the freight transport and storage system that can be used for moving goods between production and final customer locations (Onstein & Tavasszy, 2016). The layout of DCs must be done with care as it allows the companies in balancing the customer service levels as well as its logistics costs. The company needs to analyze the logistics costs, service level, number of customers, land availability and labor when deciding on the optimal location and number of DCs to be maintained (Onstein & Tavasszy, 2016). In the given case, the company has 13 DCs spread across the region with 3 DCs in Thailand, 1 DC in Malaysia and 9 DCs spread across Indonesia.
The issue with current supply chain seems to be inappropriate locations of DCs and lacking number of DCs across the region. Specially, the DCs in Indonesia (specifically in Java, Jakarta and Bali) seems to be outnumbered considering the number of customers and products’ demand. The customers are much concentrated in Java, Jakarta and Bali and there seems to be only 2 DCs at this location. Shortage of DCs at this location could mean lower amount of sales and high number of lost customers due to shortage of supplies (Onstein & Tavasszy, 2016). Moreover, in Malaysia there is only 1 DC due to which the customers’ orders are being lost as reported in the case study already. With only one DC, it is not possible for FastGood to cater the demand from huge number of customers in Malaysia, hence adding more DCs here will be beneficial for the company in both Indonesia and Malaysia.
As a recommendation, the following new Map has been developed using Excel Power Map tool.
Figure 1: Power Map For New Supply Chain
The following addition/changes are done in the current supply chain;
- One new DC has been added in Thailand (see DCN1 in Map Above). Total DCs in Thailand are now 4.
- One new DC has been added in Malaysia (see DCN2 in Map Above). Total DCs in Malaysia are now 2.
- In Indonesia two locations of DC has been changed while two new DCs locations were added. Now Total DCs in Indonesia are 10 (see DCN3 and DCN4 in Map Above for new locations).
The main idea of adding DCs in particular areas was done on the notion that the DCs should be concentrated in the areas where the demand for products are more. While, the changes of location were initiated after analyzing the fluctuation of demand at each region. The locations of new DCs are now more concentrated where the demand for products were identified to be more. With addition of new DC’s now the total DCs are 15. This new supply chain will help FastGood to be responsive enough towards fluctuating demand and meeting customer needs at right time.
As the company was facing challenges in terms of lacking number of DCs at key locations and the inability of existing DCs to cover demand of certain regions, it can be assumed that the new proposed supply chain will cater this issue. Since customers’ need and demands are uncertain, the ambiguity in the supply chain became certain. In order to satiate the needs of customers efficiently and effectively, the company needs to identify the key locations where the demand is high and the locations where the demand is uncertain (Onstein & Tavasszy, 2016). It is important for FastGood company to keep working on its supply chain by adding DCs on the locations where the demand is high and by removing it from the locations where there is no demand for products. By doing so, the company will be able to reduce inventory holding costs as well as will see a certain rise in its sales. This is likely because when the DCs will be located near the high demand markets, the company will be able to fulfil customer demand in short time.
PART 2: MOVING AVERAGES
Prediction using a forecasting method is one of the imperative things for an organization (Yaffee & McGee, 2000). Demand forecasting is the process of projecting customer demand for a good or service. In the current scenario, the historical data is used for predicting future data for the detergent. Although there are many ways that can be employed for forecasting the demand, but in the given case we have used moving averages. Moving average is considered to be the best technique in technical analysis as it is used for constructing the trend line and predicting future movement of the values (Klemelä, 2018). This technique can make data clean and aid in minimizing the noise of short-term movements in the data. This technique is called ‘moving average’ because at each stage, a new demanded quantity can be calculated for an upcoming time period (Flemming & Nelis, 2000).
According to the reports of FastGood Company, average inventory of facial cream has increased. Also, it was noted that the company is considering shifting to MRP inventory policy instead of regulatory policy. In order to implement this policy, the management needs to get a grip of forecasted next periods’ demand value by using moving average technique. In the given case, we utilized moving average for predicting demand value for 53rd week. For calculating the predicted demand for facial cream for 53rd week, the moving average of five weeks upto 52 weeks was carried out. Once the moving average was calculated for each week, the chart was plotted for analyzing the movements and visualize the trend line.
Figure 2: Trend Line of Detergent Moving Avg. Demand
From the trend line above, it can be seen that the demand of facial cream from July to December has been fluctuating due to factors like weather unpredictability, changing fashion trends etc. The demand of facial cream has shown somewhat stationary movement. Moving averages for the first few weeks were 47,311; 46,639; 49,896; 50,927; 52,476; 51,831; 50,649 and so on. The line of moving average shows an increasing trend and the value that was predicted for 53rd week is 56,234.
In order to check the accuracy of the predicted value, the degree of closeness of the predicted quantity to the real value must be checked. In doing so, the absolute errors were calculated followed by mean absolute percentage error (MAPE) (U.S. Department of Commerce, 2003). It is an average absolute percent error calculated for each time period minus actual values divided by actual values. The error value is basically the difference between actual value and predicted value however absolute error refers to the modulus of the error used to compute mean absolute error (MAE) and mean absolute percentage error (MAPE). These errors indicate the accuracy of the forecasted results and aid the management in checking the accuracy and efficiency of any forecasting model. In case of demand forecasting in supply chain, MAPE technique is widely used to compute the forecasting model accuracy (Gilliard, 2010). In this case, the mean absolute percentage error (MAPE) is 9% which makes the forecasted results of week 53rd to be 91 % accurate. The forecasted demand of the 53rd week shows that the average monthly demand is rising with time.
PART 3: EOQ
In order to maximize the overall company’s profit, it is mandatory for minimizing the cost of ordering and storing the inventory. The financial metric used for calculating the optimal inventory quantity for ordering and storing is known as EOQ (Muckstadt & Sapra, 2010). It is imperative to understand that the company can incur losses if there is surplus inventory in warehouse or if there is shortage of inventory to meet customers’ demands in times of need. In the given case, the marketing team in Malaysia has already understood that the monthly detergent demand is being shifted to competitors. Due to having less demand for detergent in Malaysia, the company has to keep high levels of inventory in its warehouses in Malaysian DCs causing high levels of storage cost.
The two components of EOQ are holding cost and ordering cost;
- The holding costs are the costs to store the inventory and includes costs associated with storage, space, deterioration, property tax, insurance, deterioration etc. The more the inventory in warehouse, the higher will be its carrying cost (Muckstadt & Sapra, 2010).
- The ordering cost is the cost of creating a purchase order, processing an order and inspecting it. It doesn’t include the purchase price. No matter what size the order is, the ordering cost will have to be incurred with every order. Hence, the more orders are placed, the higher the ordering cost will be (Muckstadt & Sapra, 2010).
For the given case study, EOQ is found as below;
Annual usage in units = 70,000 × 12 = 840,000 units
Ordering Cost = $ 100 per vehicle per shipment
Carrying Cost = 5% of $8 product selling price per case per month ($0.4). For annual cost, it will be 0.4 × 12 = $4.8
EOQ= 2 840,000 100 4.8 = 5916 units
EOQ is the total number of units that FastGood must add to inventory for minimizing its total cost i.e. holding cost, order cost and shortage cost. By calculating, we see that the EOQ was 5916 units. By holding this order size, the company can keep its ordering and holding cost at its minimum.
Moving away from this level of inventory can cause inventory ordering costs and holding costs higher. By looking at the graph below, we can understand that any order point before or after EOQ can increase the total inventory cost for FastGood.
Figure 3: EOQ and Total Cost Curves
Source: (Muckstadt & Sapra, 2010)
At EOQ point, the total cost curve is at its minimum. Any order quantity before the EOQ level will have higher ordering cost but lower holding cos (Muckstadt & Sapra, 2010) t. The holding cost is lower for order below EOQ because there will be less storage required for lower ordered quantity and it will cost less for FastGood to store goods with minimum quantity (Muckstadt & Sapra, 2010). However, since the order cost has to be incurred and it is fixed, so the company will still have to pay for logistics for getting its orders shipped from factories to warehouses. As the order cost is fixed, it will be equally spread over each unit that is shipped causing higher ordering cost (Muckstadt & Sapra, 2010).
Similarly, when the orders exceed EOQ point, the company will have to face higher total cost due to increased holding costs but decreased ordering cost (Muckstadt & Sapra, 2010). This is because when the order will contain high number of units, the ordering cost will be equally spread over large quantity causing lower total ordering cost (Muckstadt & Sapra, 2010). However, the company will have to pay for extra space, deterioration and wear & tear for higher quantity ordered. This will drive the total cost curve upwards as can be seen above. Nonetheless, it can be said that ordering cost and the carrying cost are anti-proportional to each other. If cost of one of them decreases, the cost of other will increase. EOQ is the metric that can balance both OC and CC so that the overall inventory cost for ordered quantity can be kept low (Muckstadt & Sapra, 2010).
In the given case, if FastGood order 10% more than EOQ of 5916 i.e. 6507, the company will have to face more inventory carrying cost. More is the quantity; more will be the associated carrying cost. Due to this, the overall cost of inventory will be higher for Malaysian DCs. Similarly, if the company orders 10% less than 5916 i.e. 5324, it will incur more fixed ordering cost because whether 1 unit is being ordered or more, the ordering cost will remain same. More is the quantity, lesser will be its ordering cost per unit.
In the case above, it means that the company must place orders 14 times per month. While this might be the best number to minimize the cost, the company should see whether its staff can still manage to handle more frequent orders and deliveries or not. Other factors that should be considered include the ties with suppliers and feasibility concern with EOQ (Muckstadt & Sapra, 2010). Ofcourse for placing orders that much frequently, the company would be required to make contracts with its suppliers so that any conflict with their calendars wouldn’t arise. Moreover, as the company is already looking for developing new warehouses in the region, it is important to analyze the variation of demand from location to location in case of new DCs. Holding EOQ might be a waste of resources if the company is holding items in different stocking locations already. In such a case, rather than reordering the goods, the company can consider redistribution of items between different warehouses so that the resources might not get wasted.
Nonetheless, in the given case, it can be said that for maintaining the total inventory cost at minimum, FastGood must re-order 5324 quantity. The calculated EOQ was 5324 with annual demand of 840,000 per annum also means that the company would be required to place order 158 times per year approximately.
REFERENCES
Flemming, M. C. & Nelis, J., 2000. Principles of Applied Statistics: An Integrated Approach Using MINITAB and Excel. Mumbai: Cengage Learning EMEA.
Gilliard, M., 2010. The Business Forecasting Deal: Exposing Myths, Eliminating Bad Practices, Providing Practical Solutions. s.l.:John Wiley & Sons.
Klemelä, J., 2018. Nonparametric Finance. London: John Wiley & Sons.
Muckstadt, J. & Sapra, A., 2010. Principles of Inventory Management: When You Are Down to Four, Order More. Bangalore: Springer Science & Business Media.
Onstein, A. & Tavasszy, L., 2016. Factors determining distribution structure decisions in logistics: a literature review and research agenda. Transport Reviews, 39(2), pp. 243-260.
U.S. Department of Commerce, 2003. Survey of Current Busines. s.l.:U.S. Department of Commerce.
Yaffee, R. A. & McGee, M., 2000. An Introduction to Time Series Analysis and Forecasting: With Applications of SAS® and SPSS®. s.l.:Academic Press.
