How many of us in the production and inventory control field moan about the forecast. Many feel that they could do better than the sales or marketing guys, but it is their job. Typically the responsibility falls to marketing to do the forecast by family group and use all the fancy  forecasting techniques to do it. Sales do a forecast by item by talking to their customers. This can then be summed into families and compared to the marketing forecast. This is the recognised way of doing it as then the trends and economical factors are allowed for in the marketing forecast and the sales team identify special requirements and changes in the market place. As long as these two groups talk to each other when they have prepared their forecast, a final joint forecast considering both view points can be created. We do need one set of numbers to present to the sales and operation meeting and to run the business.

One of the ways we can improve our forecast is to follow one of the basic forecasting principles: a forecast for a group of items is always more accurate than a forecast for each of the individual items.
With this in mind, we would forecast a family of items, which for example could be a brand of coffee, packed into several jar and tin sizes. The forecast is prepared for the total kilograms of coffee to be sold. A bill of material is then used to divide the total quantity into the various packs. This bill is created by looking at the past history of the item. It might be found for example that 50% of the last years sales was sold in 2Kg tins, 20% in 1Kg tins, 25% in 500Gm jars and the remaining 5% in 200Gm tins. From this information a planning bill can be created as shown in figure 1, which will tell us the expected sales per pack size, in weight of coffee, for a forecast of the total weight of coffee to be sold.

Figure 1

If now the total forecast for coffee sales is 200 000Kg in March, then the sales can be expected to be as follows:

 2Kg Tin 200 000 X 0.50 =  100 000 Kg
 1Kg Tin 200 000 X 0.20 =  40 000 Kg
 500gm Jar 200 000 X 0.25 =   50 000 Kg
 200gm Tin 200 000 X 0.05 =   10 000 Kg

To use this approach but to obtain the number of packs for master scheduling and materials planning, we just need to reason through the quantities.
To make it easy for ourselves, using the information of past sales, we can say that for every 10Kg of coffee sold the following tins were used.

2Kg Tin  10 X 50% = 5Kg   = 2.5 Tins
1Kg Tin 10 X 20% = 2Kg  = 2.0 Tins
500gm Jar 10 X 25% = 2.5Kg = 5.0 Jars
200gm Tin 10 X  5% = 0.5Kg = 2.5 Tins

Therefore to create the data for a BOM, we need to reduce this information down to 1 Kg of coffee being forecast. Then these quantities become, 0.25, 0.2, 0.5, and 0.25 tins per Kg of coffee forecast.

Our Planning bill then becomes;

Figure 2

Using this Planning BOM, The total sales forecast for coffee can be forecast and the correct balance of packing can be planned to support the coffee being produced.
Their would be an individual BOM for each pack size packing material, calling up the Tin, Lid, Label and carton etc.
Think this logic through and experiment with old data. This approach can be an aid in improving the forecast and materials planning. It is typically used where a single product is made in a number of colours, or in various packs.
 
 
 
 

August 1999