$$DECISION PROBLEM $1$: Using Forecasts to Plan for Customer Demand$$

$$Managers use forecasts to plan ahead to ensure the company has sufficient products and resources to meet customer demand for the coming month, quarter, or year. Accurate forecasts enable managers to make better purchasing decisions, as well as to develop more efficient resource, staffing, and scheduling plans.$$

Olde World$$s sales manager must make purchasing plans for the private$-$label products Olde World sells. One of the items the company stocks is SunPro, an add$-$on that customers buy to spray on their windows and doors for UV protection. The sales of this item are seasonal, with peaks in the spring, summer, and fall months. Demand $($in cases$)$ for the past $12$ months at Olde World are listed in the attached table:$$

Month Sales $($$$)$ Month Sales $($$$)$ January $2,000$ July $5,000$ February $2,800$ August $5,400$ March $3,200$ September $6,000$ April $3,900$ October $5,000$ May $4,300$ November $3,800$ June $4,900$ December $3,300$

A$.$ Based on the numbers in the table above, forecast the sales $($by month$)$ from March through December using a $2-$month weighted moving average. Use the weights of $(3/4)$ and $(1/4),$ giving more weight to more recent data.$$

B$.$ Forecast the sales for the months of March through December using exponential smoothing with $=0\mathrm{.}7$ beginning with an initial forecast in January of $$3,000$ and starting error measurement in March.$$

C$.$ Using the mean absolute deviation as your performance criterion $($with error measurement beginning in March$),$ compare the two forecasting methods you used in A and B$.$ Which method would you recommend as a more reliable predictor? Why?$$

D$.$ Using the mean absolute percent error as your performance criterion $($with error measurement beginning in March$),$ compare the two forecasting methods you used in A and B$.$ Which method would you recommend as a more reliable predictor? Why?$$