Content area
Full Text
As readers of this journal are quick to recognize, an accurate forecast of sales is vitally important to business planning. Sales forecasts filter through most organizations impacting on production schedules, inventory control, personnel planning and a host of other business decisions. In many business environments, fluctuations in general business conditions make forecasting a difficult task. These general economic conditions, often referred to as "prime movers" because of their overall importance, include such measures as GNP, the unemployment rate, the rate of inflation and, of course, the level of interest rates, among others. For manufacturers of durable goods, these prime movers are especially important. However, incorporating the effect of certain prime movers into the sales forecast is even more difficult because of their lag effects.
The effect of the "prime movers" is frequently taken into account in management's forecasts, though in a subjective way. In fact, judgments are an important component of many firms' forecasting efforts. In this paper we show how such judgment-based-management forecasts improve when combined with quantitative forecasts. The methods that we suggest are relatively easy to understand and to implement.
The company that we use as a vehicle to explain the process is the Delfield Company, a leading manufacturer of foodservice equipment. Delfield uses a production planning system comprised of a master production schedule (MPS) and the corresponding material requirements planning (MRP) system. The master production schedule is a statement of what the company expects to manufacture and is directly dependent on an accurate sales forecast. The MRP system is the computer based planning system that determines material requirements, and then schedules production to meet changing demand. The MRP system is thus driven in large part by the sales forecast.
BACKGROUND
A monthly product sales history of seven and one-half years was used to evaluate various models. One product was chosen as the basis for the development of the models. The product selected was a three-door-reach-in freezer which is considered an "A" item with regard to inventory control, meaning that it represents a large cost when held in inventory. The sales history of the three-door freezer is typical of many durable goods and shows a great deal of volatility. (See Figure 1.) (Figure 1 omitted) Forecasting models for other products in...