Preparing the Forecast Cycle
This chapter explains how the administrator prepares M3 DMP for the forecasting activities in the current forecasting cycle. This includes the tasks of updating Datasets from M3 BE and preparing data and features to make them available to the forecasters.
Outcome
A forecast package for each forecaster using distributed work method is created. The central version is opened for forecasters using central work method. The environment is made available so that the forecasting activities in the current forecasting cycle can happen.
Uses
Each forecaster uses either a forecast package or a central version to perform collaborative forecasting. The administrator (server) version is used to combine the Market Forecast Proposals into a Sales Forecast Proposal and to consolidate the proposal.
Processes
The process consists of two activities. In the first step activity the data structure is adjusted. The server version is opened, which provides access for the forecasters using central work method, while forecast packages are generated for the forecasters using distributed work method (Online and Offline). See About Users.
Activity Description
- Importing
and structuring data
In this process, the administrator imports Datasets from M3 BE and splits them into M3 DMP Datasets, Discarded records and Residual records. If any phantoms are needed to forecast for the period, these are created. This establishes the data foundation for the process of collaborative demand planning using a distributed process flow.
- Adjusting
environment
In this process, the Version parameters are adjusted to suit the forecasting process for the current cycle. The administrator adjusts time marks and time horizons in the active version. The Changeable Time Horizons where forecast values can be entered or changed is defined.
- Generating forecast packages
This process is used to build the tailored forecast packages for the forecasters. The administrator manages user rights to ensure that the right Datasets are included in the right packages. Appropriate filters and favorite views are specified for the Datasets, and to-do lists are created for the forecasters. The administrator can create favorite views for different forecasters. So when these forecasters login, they can view only those favorite views which are assigned to them by the administrator. The administrator finally generates these forecast packages.