Synchronizing Supply Chain Operations with Consumer Demand Using Customer Data
Kiely, Daniel A., The Journal of Business Forecasting Methods & Systems
Identifies, compares, and contrasts various types of demand data streams that flow through in different stages of supply chains ... demand planning systems driven by forecasts based on point-of sale data are best suited for synchronizing supply chain plans with consumer demand...where POS information is not available, other customersupplied data can be used asa proxy.
For many years, monthly factory shipment data was the only type of non-syndicated demand information available to vendors for forecasting product consumption. Although factory shipment data does not provide the best consumption forecasts, vendors were left with few alternatives. Other demand data, better suited for forecasting consumption, was either difficult or expensive to acquire.
However, since the late 1980s, developments in systems technology and improvements in supply chain management have enabled retail customers to provide their vendors with better data. Two key technological developments that have enabled customers to transfer and vendors to process consumption data are Electronic Data Interchange (EDI) and high-speed, batch-processing forecasting software. Coincident with the advent of these innovations was the evolution of cooperative customer-vendor business programs including Vendor Managed Inventory (VMI), Continuous Replenishment Planning (CRP), and Collaborative Planning, Forecasting and Replenishment (CPFAR) which became the hallmarks of 1990s supply chain management. The combination of new information technology and cooperative supply chain partnerships has made possible the sharing of consumption-based forecasting information in near real time. The results of these developments have been dramatic: improvements in product-forecast accuracy, reductions in supply chain inventories, and efficiencies in product distribution.
This article reviews demand data streams that are transmitted by customers to their vendors. Throughout, a customer is defined as a retailer that sells finished goods to the end-users, consumers. A vendor is defined as a manufacturer that produces, distributes and sells a finished good to a customer. Figure 1 illustrates the vendor's forecasting objective, as the author sees it, to synchronize production, distribution and inventory levels with consumer purchasing activities for the purpose of optimizing supply chain efficiency.
DIFFERENT CUSTOMER DEMAND DATA STREAMS
Figure 2 shows that, instead of monthly factory shipment information, vendors have begun to use four other types of demand data to drive demand planning systems. These alternative data streams are (1) customer forecasts, (2) consumer purchases, (3) customer warehouse withdrawals, and (4) customer orders. Collectively, these customer-supplied data can be used to form the basis for bottom-up product forecasts which, when aggregated and rolled back up the supply chain, more accurately predict independent demand than do factory shipment-based forecasts.
Figure 3 shows that independent demand is the requirement for items that is influenced by factors that are external to the firms that comprise the supply chain. These external factors bring about random variation in demand for such items. Consequently, independent demand forecasts are typically projections of historical demand patterns. As such, it is assumed here that independent demand is derived from point-of-sale (POS) based consumption data, since consumption is outside of the control of suppliers, vendors, and retail customers. The primary reason for using customer-supplied POS information, as opposed to factory shipment data, is to drive the demand planning system with independent demand. Demand planning systems driven by POS-based forecasts are best suited for synchronizing supply chain plans with consumer demand.
Dependent demand, on the other hand, is directly related to, or derived from the bill of material structure of other items (e.g., raw materials, component parts and manufacturing inventories) or distribution requirements (e. …