35 THE OUTSIDE-IN PLANNING HANDBOOK | 2023
35
Calculating the Bullwhip Effect
What is the bullwhip effect? It is the phenomenon where slight variations in demand at the point of sale can cause large amplifications to upstream supply nodes. Since the definition, the concept has been taught through games like the Beer Game. Yet, no technology has incorporated the calculation into supply chain planning, and there is no clear definition of how to use the bullwhip factor in decision-making. A high bullwhip translates into bullwhip risk and waste. The first step is visibility. The second is calculation. In the Journal of Business Logistics (2015) Masking the Bullwhip Effect in Retail: The Influence of Data Aggregation outlines the most common calculation highlighted in the research was the coefficient of variation (CV) as measured in “T” periods:
graph). Let the formulas in Excel (“STDEV,” “GEOMEAN”) be your friend! Coefficient of Variation (Sales) “ X ”: Measures the variability of POS sales over time. Standard deviation of POS Mean of POS Coefficient of Variation (Ordered Shipments) “ Y ”: Measures the variability of ordered shipments over time. Standard deviation of ordered shipments Mean of ordered shipments Step 3: Calculate the bullwhip amplification factor . A measure of the extent to which changes in POS are amplified in shipments from the manufacturer to a distributor’s DCs. Utilize “ X ” and “ Y ” from step 2 above. (Y-X) X You now have a straightforward way to calculate how variability at the point of demand is amplified in your upstream orders. Like many KPIs of a business, the most crucial point is to start the measurement process and then work on continuous improvement initiatives to improve the metric over time. Also, take note of aggregation and seasonality effects. As Waller et al. (2015) highlighted, “data aggregation can mask the extent of the bullwhip effect, and that seasonality in the data has a dampening effect..” After defining your bullwhip impact, your focus should shift to possible countermeasures to reduce the bullwhip impact and associated waste.
ikt ) × ] ikt ) × ] 1 T 1 T
S ( O ikt ) / [Σ S ( D ikt ) / [Σ
T ( O t=1
CV
=
'
T ( D t=1
A simple but less accurate calculation is to compare the Coefficient of Variation of channel point of sales to customer orders and customer orders to planning orders (an output of the supply plan) and the comparison of planned orders to purchase orders. Step 1: Create a simple table and chart to show actual demand on an upstream supply node. This example could be a customer DC supplying a set of retail stores. Alternatively, a digital fulfillment center could provide individual digital orders directly to consumers. Simply drawing this picture will often shock companies at their lack of inventory synchronization to actual demand. Step 2: Calculate the coefficient of variation for sales and orders at the bottom of your data set within Excel (See step 1 table and
Made with FlippingBook - Share PDF online