Types of Forecasting Methods in M3 DMP
M3 DMP supports two forecast methods:
Calculated Base Forecast
The Calculated Base Forecast method uses a forecast calculated from the sales history using statistical methods as a baseline. Manual adjustments are added to this basis to make up a total forecast. When changes in the sales history leads to changes in the calculated forecast, these changes are reflected in the total forecast. This method is useful when a statistically generated forecast provides a fairly accurate estimate of future demand. Manual adjustments are used to account for deviations caused by exceptional actions or circumstances such as promotions, closure of sales channels or introduction of new products.
A Calculated Sales Forecast could be generated using the Forecasting module in M3 BE The Total Forecast is the sum of the Calculated Sales Forecast imported into M3 DMP and the Manual Adjustments made in M3 DMP. If the Calculated Sales Forecast is altered due to changes in the historical data, the Manual Adjustments are added to the new Calculated Sales Forecast. The Total Forecast is consequently changed as well.
Example
In January the Calculated Sales Forecast for the M3 BE accumulator Invoiced Qty. (UCIVQT) is as shown below. The Manual Adjustments are made in M3 DMP yielding the shown Total Forecast:
UCIVQT | Mar | Apr | May |
---|---|---|---|
Calculated Sales Forecast | 900 | 900 | 1000 |
Manual Adjustment | 0 | +300 | -200 |
Total Forecast | 900 | 1200 | 800 |
In February the Calculated Sales Forecast is different from above due to changes in the sales statistics. The Manual Adjustments made in M3 DMP are added to the new Calculated Sales Forecast yielding a different Total Forecast:
UCIVQT | Mar | Apr | May |
---|---|---|---|
Calculated Sales Forecast | 1000 | 800 | 900 |
Manual Adjustment | 0 | +300 | -200 |
Total Forecast | 1000 | 1100 | 700 |
If the Dataset contains more than one accumulator, the calculation of Total Forecast is performed in the same way for each accumulator.
The Calculated Base forecast has been divided into 2 types of Forecasting:
- Standard Forecasting
- Total Forecasting
Manual Forecast
The Manual Forecast method depends on all forecast values to be entered manually. Sales statistics or calculated sales forecasts can be used as a starting point but any future changes in the sales statistics or calculated sales forecast will not be reflected automatically in the total forecast. This method is useful when the demand pattern cannot be predicted by statistical data.
The forecast method is version specific and is defined in the Version properties dialog under the Settings tab. When a Dataset is imported, this setting determines the forecast method for the Dataset. The chosen forecast method for a Dataset can be viewed in the Dataset Properties dialog.
Total Forecast values are entered manually in M3 DMP, by using features like Advanced Calculation to create a starting point from historical or calculated data. The Total Forecast is not affected by future changes in the Calculated Sales Forecast.
Example
In January the Calculated Sales Forecast for the M3 BE accumulator Invoiced Qty. (UCIVQT) is as shown below and the Total Forecast corresponding to the example above is entered in M3 DMP:
UCIVQT | Mar | Apr | May |
---|---|---|---|
Calculated Sales Forecast | 900 | 900 | 1000 |
Total Forecast | 900 | 1200 | 800 |
In February the Calculated Sales Forecast is different due to changes in the sales statistics. This does not affect the Total forecast:
UCIVQT | Mar | Apr | May |
---|---|---|---|
Calculated Sales Forecast | 1000 | 800 | 900 |
Total Forecast | 900 | 1200 | 900 |
When importing a Dataset, all budget version(s) are imported along with any historical or calculated data. The forecast values are displayed as they are stored in the budget version(s). When exporting the Dataset, each budget version is exported to the budget version it was imported from.