Re-forecasting: How often For Best decisions And Efficiency
RE-FORECASTING: HOW OFTEN FOR BEST DECISIONS AND EFFICIENCY By Donald York Makes a case for why re-forecasts should be event-driven and not calendar-driven…often oversell of a product in one period is followed by an undersell in the next period … recommends a decision rule which states re-forecast only when oversells or undersells exist for three consecutive periods. Many organizations nowadays routinely re-forecast every item or SKU in their line at fixed timed intervals. The interval is most often a month, which could be longer or shorter as the situation warrants. Occasionally, it may be necessary to re-run the forecasts for some items or category of items during the interval, in response to various events (predominantly external events). These reactive mid-interval forecasts, however, are not the issue here, as they are unplanned and clearly justified. In contrast, forecasting at fixed intervals is a pre-scheduled event, initiated at a certain well-defined point in the calendar. Organizations usually re-run this total product or service line forecast at the beginning of each new period, shortly after the most recent period’s actual sales are available. This re-forecast ...
From Issue:
Summer 2005
(Summer 2005)
