Slow seller analysis for generic / variant articles

Question

Hi all,
We are having issues getting slow seller analysis to work for generic variant articles.
We have created a generic with variants of a certain colour and size.  We have attached a markdown profile to one of the variants and transferred this to BI (the profile in BI is assigned to the generic at for a specific colour and size).  We have then loaded sales and goods receipt data for the variant article into the relevant cube in BI.
When we execute the slow seller query directly in BI (trans RSRT1) it is correctly retrieving the sales and GR data at the generic / size / colour level, however when we execute the slow seller analysis using transaction /n/MAP/SSM2 we get no markdown proposals generate the the log returns a message that no data could be found.
We have implementation the best practice building block however can't identify why no data is found for our selection criteria.
Does anyone have any documentation (apart from the best practices documentation) on how this needs to be set up?
Many thanks,
Steve

Answer 1

The problem is also described and answered within OSS Note Nr. 0000561477 for Year: 2009. Name: Slow seller analysis not picking up variant articles
Bottom line of this problem is:
I analyzed the data in cube  
0RMP_MC21. The movement data of seasonal merchandise (generic articles)
is expected to be stored in this cube on the level generic/color/size.
This is controlled via MAP foundation. In this case, the slow seller  
report can evaluate the data and propose markdowns. Unfortunately, the
data in your cube 0RMP_MC21 is stored on variant level (material      
number). You can verify this by running the query 0RMP_MC21/          
0RMP_MC21_Q0020. In contrast to the screenshot you provided as an     
attachment, this query doesn't give you values for article 400018 on  
the level color/size, but only for the variants. Therfore, the slow   
seller report won't find the data.                                    
If you ensure the update of the cube according to the MAP foundation  
(0material is filled with generic article), then the slow seller      
analysis will pick up the variants as expected. 
Post is answered!
Regards,
Markus
Edited by: Markus Semar on Jun 18, 2009 6:00 PM
Edited by: Markus Semar on Jun 18, 2009 6:03 PM