Enable Closed Loop Forecasting

This document explains how you prepare for the concept of closed loop forecasting. Some specific settings must be defined for the exchange of forecasting data between M3 Business Engine (M3 BE) and M3 Demand Planner (DMP) to work properly.


This instruction results in an activated dataset that has been prepared for closed loop forecasting. After using the instruction, you are ready to create values in the dataset.

You can also use this instruction to import datasets into M3 Demand Planner.

The following files are affected:

Before you start

There must be a dataset to use in the closed loop forecasting. If there is no suitable dataset, you can create it parallel to following this procedure.

Follow These Steps

  1. Start 'Dataset. Open' (OSS401).

  2. Create a new dataset or select an existing one.

    If you want to change an existing dataset, you must first deactivate it by selecting option 10=Deactivate.

  3. On the E panel, set the Division field as blank if you want to have transactions from more than one division in the dataset.

  4. Check that the Balance check box is cleared.

  5. Add new or change the existing key fields.

  6. In 'Dataset. Connect Accumulator Fields' (OSS402), check that Invoiced quantity (UCIVQT), Sales amount (UCSAAM) and/or Demand (UCDEMA) are selected as accumulators.

  7. In 'Dataset. Connect Dataset Trans Types' (OSS406), check that transaction type 33 (sales budget) is connected to the dataset. You can also use types 31, 32 and 34 (optional). Press F3=End to return to 'Dataset. Open' (OSS401).

  8. When the dataset is created or adjusted, select option 20=Activate.

  9. Create Sales Budget Version

    In 'Sales Budget. Open' (OSS420), create and activate one or two sales budget versions and connect them to the dataset.

    If you want to use a sales forecast, you must have two sales budget versions - one containing the total budget and one containing the manual adjustments.

Parameters to Set in (OSS401/E)

Field The field indicates …
Division … the division that should be used in the dataset forecast. If a division is entered here, only transactions for this division are included. If the Division field is blank, the dataset is considered as central for the company. Having division (UCDIVI) as a key field in the dataset makes it possible to use divisions as selection criteria.
Period type … the period type used in the dataset. The period type controls how the dataset is divided in periods, such as weekly or monthly period types 1–5 (general period types) or 9 (used in forecasting).
F/C/Budget version … the budget version to be included in the dataset. Budget version is a subordinate ID to the dataset and each dataset can have more than one budget version to store different values.
Key 1, 2, 3, etc. … what key fields the dataset should contain. Up to six keys are allowed. The dataset must contain item number (UCITNO), forecast family (FJBUIT) or planning entity (NDCCIT) as key, otherwise you will need a distribution template to be able to implement the forecast in the material plan.
Balance … whether the accumulated values are saved in the database at the specified key level. When using closed loop forecasting, this parameter should not be selected.
Sequence … the order in which the accumulator fields should be entered. The sequence must start with 01 and end with 24 using consecutive sequence numbers in order for the import to DMP to work.

… the fields used to accumulate data into the dataset. A maximum of 24 accumulators can be used. At least one of the following accumulators must be included in the dataset for the closed loop forecasting to work:

Invoiced quantity (UCIVQT)Sales amount (UCSAAM)Demand (UCDEMA).

Field … the filtering fields used to make selections within the dataset, such as indicating that the dataset should only be updated for certain items.
Transaction type … the transaction type that should be used in the dataset. Transaction types 32 (sales statistics) and 33 (sales budget) must be included in order for closed loop forecasting to work. You can also use transaction type 34 (sales forecast).
Related topics