Forecast Imbalance issue

[Former Member]

Following is a an issue which is hanging for a long time.It's about the Forecast imbalance.

We had,until last year 13 periods and now 12 periods.We plan in months and diaggregare it to weeks.When we do that there is a huge difference is the forecasted values i.e in a week we see 40000 cases and another week zero.For example: in a period there are 4 weeks for 1 week the forcast is 0 for another product product/pack level it shows 300000 and for the follwoing week it shows 2000 and the week thereafter it's 0.That means the monthly forecast is not evenly diaggregated to weeks. Currently we zero out manually and renter the values but it's a herculean task. Can any one tell me why is this happenning so? I have checked the Time based diaggregrationa and the whole planning area it's all fine. I have logged this to SAP also but until now no reply.This is major issue causing major revenue loss.

[Former Member]

Disaggregation follows the setup in the planning area only if there is no existing proportions

I think whats happening is you already have some numbers in the weeks, so when you disaggregate it follows already existing proportions.

What have you got for the time based disaggregation?

if it is P - Data is distributed in time so that each key figure value in the smallest storage bucket represents the same proportion of the value in the aggregate bucket as before.If the key figure values prior to distribution were zero, and if a time stream ID forms part of the storage buckets profile definition, the system checks to see if time-based weighting factors exist for this time stream. If so, the data is distributed according to the time-based weighting factors of the time stream. If no factors exist, the data is distributed equally to each storage bucket.

If you are on SCM5 there is a redisaggregation option on the macro that might help in redisaggregation (dont know if it works on time disaggregation)

You can also do what you are doing manually by automating. Looks like this issue is in production

[Former Member]

Yes,this is a production issue.The time based disaggregation is P and we are on SCM4 upgraded from 3 last year.please provide me your email ID I will send you a snap shot which will give you a clear idea of what's hapenning..

Message was edited by:

Kailash Maisekar

[Former Member]

Harish plz check ur gmail.I've sent the snap shots

[Former Member]

If you have P for the time disaggregation of the KF, then you cant expect the proportions to disaggregate without considering old proportions.

you need to also rethink the business process for entering the values etc

Can you figure out why your weekly existing values are so disproportionate before you enter them? Are you changing them in weekly?

One way you can work around this is to do the following steps

Create a new PA with just the KFs you are in trouble with. Make sure this is in Monthly time buckets

Do the following by background jobs:

Copy the KF from the current PA to the new PA

Zero out the KF in the current PA

Copy back from New PA to current PA

since there is not value in the weekly buckets you will be able to get the proportions

this will be better than manually correcting it

make sure you copy at the correct aggregate level

[Former Member]

I sent this SAP and they asked me to apply a note:410680.Give the situtation above this note really talks about managing our expectations.

Experts:Please throw some light on this note and its validity and how best it can be applied?

Thanks!!

[Former Member]

I sent this to SAP and they asked me to apply a note:410680.Given the situtation above, this note essentially talks about managing our expectations.

Experts:Please throw some light on this note for it's validity and how best it can be applied?

Thanks!!

[Former Member]

Hi group,

Can any one look at the solution part of this note 410680 and suggest some ideas?

Thanks!!

[Former Member]

did you try changing your time based disaggregation to K instead of P

[Former Member]

In which key figure? I am sending you the snap shot.

[Former Member]

in the third post in this thread you have mentioned using P as the time based disaggregation. try changing this to K.

using P for time based disaggregation makes the disaggregation to take preexisting proportions and only if there are no preexisting proportions will the proportions be fresh