Time Series Based Forecast Consumption vs Customer Sourcing

[Former Member]

Dear IBP Experts

In our current IBP setup, we work with a demand planning part that is focused on Product - Customer Levels. The final Consensus forecast is then 'sourced' via the standard Customer Sourcing setup in order to achieve Dependent Customer Demand on Prod-Loc-Cust Level. Then we use the S&OP Unconstrained Heuristic. We are now looking into the Forecast Consumption logic, but we have difficulties to combine this with our Customer Sourcing Logic.

Theoretically, we would work in following way:

1. Source the demand on product-customer level via Customer Sourcing in order to have forecast also on location level.

2. Perform Forecast Consumption on product-location level

3. Run the Heuristic.

However, this currently does not seem possible as the Heuristic (step 3. always performs Custom Sourcing again).

Any recommendation on how to use Forecast Consumption in our case?

Kristof

[lingaiahvanam · Accepted solution]

Hi Kristof,

You can now assign key figures with the following as storage root attributes for input forecast and input sales orders:

• Product-location mandatory (PRDID + LOCID)
• Any additional root attributes optional, for example, Customer (CUSTID). For e.g, PERPRODLOCCUST

The forecast consumption level is flexible and attributes available for Forecast Consumption level are the Intersection of the Attributes of Input Forecast and Sales Order

Example:

Input Forecast: storage level roots PRDID/LOCID/CUSTID + PRDFAMILY

Sales Order: storage level roots PRDID/LOCID/CUSTID + PRDFAMILY

Open/Unconsumed Forecast: storage level roots PRDID/LOCID/CUSTID + PRDFAMILY

Total Demand: PRDID/LOCID/CUSTID + PRDFAMILY

Consumption Level: PRDFAMILY / LOCID

Consumption Mode:

• Master Data with key PRDFAMILY | LOCID + FCSTCONSMODE (Attribute)
• PRDFAMILY + FCSTCONSMODE
• LOCID + FCSTCONSMODE

The attributes available for Forecast Consumption level are the Intersection of the Attributes of Input Forecast and Sales Order.

SAP 1808 can support this level of forecast consumption process.

Best Regards,

Lingaiah