The supply chain approach to planning and procurement management

Hewlett-Packard Journal, Feb, 1997 by Gregory A. Kruger

A common situation is the introduction of a new product. Suppose the chosen point of reference is measuring demand uncertainty as real-time customer orders versus forecast. How do we manage a new product introduction? A viable option is to use collective business judgment to set the demand uncertainty even though there is technically a sample size of zero before introduction. Prior product introductions or a stated business objective of being able to handle demand falling within [ or -] [Delta] of the plan during the early sales months can be used to establish safety stocks. In fact, the organization can compare the inventory costs associated with different assumptions about the nature of the demand volatility. Estimates of average inventory investment versus assumed demand uncertainty obtained from the statistical models can help the business team select an introduction. strategy.

Effect of Minimum Buy Quantities and Desired Delivery Intervals

In most cases, there are constraints on the order sizes we place to our suppliers, such that replenishment orders are not exactly the difference between the theoretical order-up-to level and the inventory position. These constraints may be driven by the supplier in the form of minimum buy quantities or ourselves in the form of economic order quantities or desired delivery frequencies. The net effect of all such constraints on order sizes is to reduce the periods of exposure to stock-outs.

For example, suppose the factory's plans predict needing 100 units of some part per week. Further suppose that the ordering constraint is that we order 1000 units at a time determined by either the supplier's minimum or our economic order quantity. This order quantity represents ten weeks of anticipated demand. Once the shipment of parts arrives from the supplier, there is virtually no chance of stocking out for several weeks until just before the arrival of the next shipment. Given this observation we see that safety stock requirements actually decrease as purchase quantity constraints increase (see Appendix V).

Although safety stocks decrease, average on-hand inventory and the standard deviation of on-hand inventory both increase. See Appendix III for formula derivations of the average and the standard deviation of on-hand inventory.

Effect of Review Period

Analysis of the equation for the standard deviation of demand uncertainty given above shows that as the review period R increases, [[Sigma].sub.X] increases, thereby driving up safety stock. This makes sense because the safety stock is there to provide the desired confidence of making it through R weeks without a stock-out. However, note that the service level metric itself is changing. For R = 1, the service level gives the probability of making it through each week without a stock-out. For R = 2, the service level gives the probability of making it through two weeks, for R = 3, three weeks, and so on. Increasing review period therefore has an effect similar to that of minimum buy quantities. When operating at longer review periods, purchase quantities to the supplier are larger, since we are procuring to cover R weeks of future demand and not just one week of future demand. To keep the average weekly service level at the desired goal, safety stock would actually have to be throttled back as the review period increases because of less frequent periods of exposure.