Analysis and summary of sales forecasting data

¶ … graph in this exhibit are based on actual observed data from sales, demand, inventory and purchases. A standard purchase level of 300 a[pears to have been established, with purchase levels adjusted for seasonal sales and demand differences to a maximum of 400 and a minimum of 0. Monthly demand scale numbers also match this seasonal trend, with higher numbers in the months of peak demand and lower numbers in the off seasons. Monthly demand scale numbers were determined by observing the relationship between purchases and sales, though in the table shown the sales figure is calculated by using the monthly demand scale figure divided by ten a s multiplier for the purchase level, such that a demand scale figure of 10 results in a sales level equal to purchases. Inventory is determined simply by taking the previous month's inventory, adding purchases, and subtracting sales; an initial inventory of 2000 was given for the first month of observation. The monthly demand scale numbers are further verified by an analysis of the demand graph, which tracks actual demand levels in each of the months of observation.

Exhibit 2

Seasons were established by grouping months of similar demand and sales together resulting in a seven-month average selling season, a tow month high selling season, and a three-month low selling season; the first table in Exhibit 2 simply breaks down the demand levels for each season, the fraction they represent of total annual demand, and an index of demand level taken by dividing the seasonal demand level by the average level the of demand to determine how much above or below average the demand is during each season. Average sales-per-month within each season are also given in this table. These figures were used to make projections for both 2011 and 2012, as shown in the sales and error table to the right; the level of error for 2011 is given (as determined by subtracting the actual sales value by the projected sales value) as a means of establishing the accuracy of projections for 2012. The sales graph shows the projected and actual sales levels in 2011 as a means of visually comparing the accuracy (and divergence) of the projections, while the error graph beneath demonstrates the degree of error in the projections for each month, with distance from the zero line in either direction indicating the degree of error in the projections.

Exhibit 3