The forecast algorithms in SAP Forecasting & Replenishment calculate forecasts by taking into account regular, week-based DIFs (Demand Influencing Factors). The forecasts can be broken down subsequently into daily values that are then used in the requirements calculation and requirements quantity optimization. Daily values are calculated from the weekly values using weighting profiles for the week (weekly profiles).

The system calculates dynamic weekly profiles from historical consumption data on a daily basis at either the location or location product level (in other words, based on the consumption data for all products or on an individual product in a location). In cases where it is not possible to calculate a dynamic profile (for example, for new products that have no consumption history) you can use a static weekly profile.

Both the dynamic weekly profiles and the static weekly profile show how consumption is distributed across a week. During a specific event, the distribution of consumption during a week can differ greatly from the regular distribution. For example, a retail promotion can result in high consumption of a product at the start of a week, whereas high consumption is usually only registered on the weekend.

You can replace the dynamic and static weekly profiles for a certain period and certain location products or locations with your own customized profiles. If you want to replace an existing weighting profile for a period longer than one week, you can choose between the following weighting profiles:

Because a specific event may last for different lengths of time (for example, a retail promotion may last 10 days the first time and 12 days the next) but with the same consumption pattern each time, you can create a weighting profile that has no fixed duration. You can then scale this to suit different periods and requirements by extending or shortening the time period.

For correction DIF occurrences, you can use a defined weighting profile to distribute a value that is valid for more than one day to the individual days in a validity period. That is, you can take non-work days (days without consumption) into account by distributing the percentage value from these days to the other days.

Since irregular consumption patterns such as the above can occur repeatedly for a certain event (for example, a retail promotion), you can create the weighting profiles as reusable objects by assigning a unique ID to each profile. You create, change and delete weighting profiles as master data. You then assign weighting profiles to location products by entering the weighting profile ID in the appropriate DIF occurrence record.

You can also specify a weighting profile as a default value for a DIF; the system automatically assigns this profile when new occurrences of this DIF are created. The value can be overwritten manually or deleted. If the DIF occurrence is created by the inbound interface, a profile assignment defined in the interface overrides the default value.