Question:
Hi
We are planning to use like modeling for new products. I had the following doubts and am not sure what to do .
We are planning at product family/Region...however our like modeling is based at material level. The old product is phased out and a new product substituted.We also use APODPDANT for proportional factors based on history.
If we run the forecast at the product family/region level , how does the product substitution occur ?? If the product level substitutions are considered and a forecast is genearted correctly at the product family/region , how is this disaggregated as obviously the new product has no history and no proportional factors maintained.
Did anyone face similar issues and how did you resolve it ?? I was wondering if we have to calculate the new proportional factor manually for each level....it will be a nightmare or can we just copy the history from the old products history in the proportion of phase in/phase out ?? Even this would be difficult has the phase in /phase out ratios will change periodically.
Any help/suggestion is welcome.
Thanks
Answer:
If you are running APO DP 3.0, then once you create the Like model which is always dependant on the product. Then you need to assign it.
How you set up your like model assignment controls how the product substitution is carried out.
Suppose you have OLD Product 123 and new product 456.
- First you create like model definition for Product 456 based on the history of 123.
- Next you need to assign the like model.
- Suppose you configured the Like Model assignment to be based on two characteristics Product and Region.
- Then you need to assign the Like Model to the relevant regions ie North America and Europe.
- You need to ensure that you check the "Material Forecast" box in the Master Forecast Profile
- Then when you run the STAT batch jobs, you need to ensure that the aggregation level is at Product and Region by doing so, then the like model will be carried at each level. Proportional factor does not apply here. Because the level you generate the forecast is the level that you assigned the like model.
If you run the Stat fcst at just product level, NO like modelling will occur. Because the level of the characterisctic combination does not match the assignment of the Like Model.
I hope this helps.