This topic endeavors to present you with guidelines to help you decide how to implement Interchangeability in Demand Planning. Which method you choose is based on your planning requirements.
Whenever you want to plan a new product in Demand Planning, you must have previously created the necessary characteristic value combinations before you start planning. Realignment is one method to do this. However realignment also copies the data from the source characteristic value combinations to the target combinations. It also copies the notes and generated selections (see for instance Selections in Forecasting).
Since copying the data takes much longer than creating the new characteristic value combinations, you should not use realignment to just create new combinations. The better way to do this is to edit the combinations in a spreadsheet file, upload the data into an InfoCube (see Upload Process and Transferring Flat Files), and then generate characteristic value combinations from the InfoCube (see Generating Master Data for Demand Planning).
Data from all the key figures is copied in a realignment run. This means that it is probably necessary to edit the copied data before actually working with the new product.
Since realignment generally run as a background job periodically after all entries have been made in the realignment table, it is not a method that should be used for quick changes. It is not suitable for periodic product changes.
Realignment is suitable for scenarios where a product is replaced for technical reasons and you can expect the demand to be very similar to the predecessor product.
The classic DP lifecycle planning functions give you the largest degree of flexibility when making a demand forecast.
In comparison to realignment like modeling reduces the amount of data in the system – the historical data is only needed for one characteristic value combination. It also permits more complex mapping processes, for example summing the historical data from several existing products to generate the historical data for the new product or using factors from a time series. Using such weighting factors it is possible to introduce seasonal effects into such a model.
Phase-In/Out models are, apart from macros, the only method of way of depicting a gradual controlled change from one product to another.
Although a phase-out profile can prevent you creating demand for a discontinued product, it does not transfer the demand to the successor product as in other applications. To do this use a product split.
This is a powerful method of accounting for short term changes. This function releases the data stored for one location product in Demand Planning to one or more other products at the same location in SNP. In the product split table you can specify that a product split is valid only for a limited period of time. You can thus plan short offensives without having to create characteristic value combinations in Demand Planning.
Since the product split is only executed during release to SNP, you cannot see the changed data in Demand Planning unless you release the data back from SNP to DP. This is of course assumes that the requisite characteristic value combinations already exist in DP, thus negating the advantages of this method.
It is not possible to model a different curve for the successor product in comparison to the predecessor product. For example:
Such a scenario cannot be modeled using product split without a lot of effort. To model this type of behavior consider using promotion planning with cannibalization (see Promotion Planning and Cannibalization Group). These functions cannot be accessed from Interchangeability.
However a situation with constant demand for the second product, as in the periodic demand case in Business Scenarios Involving Interchangeability, can easily be modeled using product split.
This is the classic discontinuation case. In this scenario no demand should be created for product 1 after a particular date. At the same date the demand for product jumps from 0 to the forecast value.
1. In the basic settings set the Aggregated Lifecycle Planning with Like Profiles and the Aggregated Lifecycle Planning with Phase-In/Out Profiles indicators.
2. Create a like profile with product 1 and product 2 as the reference products
3. Assign the like profile to product 2 (and other characteristic values as required)
4. Create a phase-in profile that is constant at 100 % over the whole period. You can use a relatively short period. Set the necessary indicator so that a factor of 100% is used after the phase-in period.
5. Create a phase-out profile that is constant at 0 % over the whole period. You can use a relatively short period. Set the necessary indicator so that a factor of 0 % is used after the phase-in period.
6. Assign the phase-in profile to product 2; the phase-out profile to product 1.
7. Carry out a forecast for both products together in interactive planning.
1. On the Product Interchangeability – Demand Planning screen select the relevant product combination.
2. Select Product Split and choose Create Entry.
The system creates two entries
i. For the period up to the discontinuation date
ii. From the discontinuation date onwards
8. Check the entries, particularly the dates and periods.
9. Release your demand plan to SNP.
Planned independent requirements are created for the predecessor product up to the discontinuation date, and for the successor product after this date.