Aggregation of the Forecast to product level

[Former Member]

Hi,

In the Demand Planning We have a requirement that say there are 2 products A and B. We will execute the forecast at the product level and we get forecast say 200 ifor product A and 300 ifor product B for next month.

Now i will run the forecast at the product group level(taking these 2 products in one selection) and i will get the forecast say 500. Now the requirement is that this forecast is to be disaggregated to 200 for product A and 300 for product B . Client do not like to run the forecast at the product level.If we run at the product group level,this should be disaggregated to the product level in such a way it shhould equal to the forecast when we run at product level. Forecast profile is same for product and product group level. All the other parameters such as historical period,forecast periods etc are same at the product and product group level.

Please advise what are the settings required.

Thanks

Siva

[Former Member · Accepted solution]

Hi Siva,

If I understand correctly what you need is either you forecast at aggregate level or details level, forecast at details level should be same in both case.

This is more of chicken or egg come first situation. If you really like the forecast at the aggregate level to disaggregate same as forecast at detail level than I would suggest to forecast at lower level.

Reason being, Even though you use same forecast profile you will get different forecast numbers at lower level and aggregate level due to different sale pattern (Sales history) at detail level in case you are using complex forecast strategy.

You may take following approach to get some what near by results if not exact.

1. calculate "detail proportion factor" for all periods you want to forecast.

2. Forecast at product group level and dis aggregate based on this detailed proportion factors.

Hope this helps.

Regards,

Harshil Desai

[Former Member]

Siva

As Harshil mentioned, statistical forecast results at product group would be different from statistical forecasts at product level due to differences in time series patterns (different trends, seasonalities and therefore statistical 'fit').

There is no way to do what you are saying - run forecast at product group level and disaggregate based on forecast at product level WITHOUT running a forecast at product level.

What you could do is to first run the forecast at product level saved say Key figure FCSTKF1 and then run the forecast at product group level in the actual statistical key figure FCSTKF2. In the planning area definition, disaggregate FCSTKF2 based on FCSTKF1.

In this scenario, statistical forecast at product level is used only for disaggregation (product mix), with actual forecast (product volume) based on product group.

Rishi Menon

[Former Member]

Thank You all. Closing the thread.

- Siva