Forecast distribution

[Former Member]

Hello experts,

i have a question about forecast distribution.

FC is now calculated in DP as a monthly quantity.

The business requirement is to have forecast released to SNP as weekly quantities, according to a distribution function and to the number of working days in the week.

We have a couple of possible solutions:

1. disaggregation in DP based on another key figure

2. use a distribution function when releasing to SNP

I tried using the second option but it's not working as expected in our system, and while i am waiting for a response from OSS i would like to check the first one.

1. FC calculated as a monthly quantity

2. FC disaggregated at daily level according to another key figure (manually maintained by the user where necessary)

3. FC re-aggregated as a weekly quantity

4. FC released to SNP in a particular day of each week (i.e. first day of the week, usually monday). If the week lies in between two months(3 days in Month1, 2 days in Month2 ), the FC has be released in the first day of the week that lies in the "corresponding" month.

Do you think this solution is feasible using standard functionalities or does it require an enhancement?

[Former Member · Accepted solution]

HI,

Solution is possible..

What issue, are you facing...?

regards

umamahesh

[Former Member · Accepted solution]

Hello Uma Mahesh,

thanks for answering.

The issue we could find in applying this solution would be mass maintenance of the key figure used for daily disaggregating.

The requirement is to have a "flat" distribution for all codes for which this key figure is not maintained, while disaggregating according to the values maintained there if they are not blank.

If we maintain this key figure only for "some" products, what will happen to all the other products? Will they be disaggregated according to a flat distribution function or on the basis of some historical factor calculated by the system?

Then, will DP be able to understand what are the working and non-working days for each location?

Another issue in my understanding would be the releasing of weekly FC in the first day of the week

belonging to to correct month - but this is probably related to an unexpected and incorrect behaviour of the forecast release transaction /SAPAPO/MC90 - we already raised an oss message for that.

Thanks

Mauro

[Former Member · Accepted solution]

HI Mauro,

working and non-working days of each location will be maintained in the calendar in SNP master data...and data disaggregation is done based on the The factory calendar determining which days are workdays is the shipping calendar that you enter in the location master record.

Did i correctly understand ur question..?

[Former Member · Accepted solution]

Hello Uma,

you're right, obviously working and non-working days are maintained in SNP shipping calendar for the location.

Anyway, we have now partially solved our problems with the period split profile/distribution function, and based on the results of the simulation i can say that it is what we need to satisfy the requirement.

I appreciate your help about the disaggregating/reaggregating in DP, but i thing we'll go for the "period split" solution if we can solve some problems we actually have (released quantity differs from total quantity in DP).

Regards

Mauro