Enabling Sales Forecasting Using Dataset

Outcome

The process results in an activated dataset that can be used in sales forecasting.

When dataset forecasting is enabled the following files are affected:

• System calendar file (CSYCAL)
• System period file (CSYPER)
• Division file (CMNDIV)
• Comparing forecast methods file (OSSCFO).

Sales forecasting using datasets can be used when you want to forecast in dimensions other than a fixed combination of item/warehouse. For example, sales forecasts can be used in collaborative demand planning in M3 Demand Planner.

Before you start

• The company must have an idea of how to use dataset forecasting; what forecast methods to use, etc.
• Dataset must be enabled. You can do this parallel to this process.

Follow these steps

1. Define Forecasting Settings for Company and Division

   The forecasting settings for company and division consist of the following:

   • A system calendar per division must be defined, in 'System Calendar. Open' (CRS900). You create it by pressing F13 twice, entering From and To values and then pressing F13 again.
   • A period table must be defined for every year and period type used, in 'System Calendar. Open Period' (CRS910). Note: If you want to use weekly forecasts, you must create a period type 9 table. This table must follow the standard week definition in a normal Gregorian calendar, starting the week on a Monday.
   • A system period must be defined for each period type in 'Company. Connect Division' (MNS100/F). If MPM is run in period type 1 and the number of periods per year is 12, this means that production has period length of one month. COM might have period type 2 and the number of periods is 52, meaning weeks. MPM period is the same as forecast period.

2. Define Distribution Curves and Indices

   The seasonal curves and indices that should be used in the dataset forecasts must be defined, in 'Seasonal Curve. Open' (CRS915) and 'Trend Curve. Open' (CRS916). Seasonal and trend curves control how values are distributed over a period.

3. Define Forecast Method

   The forecast method that should be used in the dataset must be defined, in 'Sales Forecast Method. Open' (OSS430). Forecast methods specify how demand should be calculated, and can for the most part be defined by the user with the help of control parameters, check parameters, limits and optional calculation options.

4. Activate Forecast for Dataset Accumulators

   This activity can be performed parallel to the dataset enabling process. When you connect an accumulator field to the dataset, in 'Dataset. Connect Accumulator Fields' (OSS402/E), you can specify whether a sales forecast should be calculated for this accumulator field. Most accumulator fields and key combinations can be forecasted.

   If you want to use advanced forecasting the dataset must contain transaction type 34 and accumulator field UCDEMA (Demand).

5. Connect Forecast Method to Dataset Header

   This activity can be performed parallel to the dataset enabling process. The forecast method, defined in 'Sales Forecast Method. Open' (OSS430), is connected to the dataset in 'Dataset. Open' (OSS401/E).

6. Activate Dataset

   Before the dataset forecasting can start to work, the dataset must be activated, in 'Dataset. Open' (OSS401). How to activate the dataset is further described in the Enabling Dataset process.

   Note: You cannot change the forecast method and other data once the dataset is active. If you want to make changes, you must first deactivate the dataset.