There's some­thing mis­sing here – when pur­chase orders are in­com­plete

When analyzing data in checks or audits, you often find that not everything which looks like a problem really is one. You may frequently check whether compulsory entries in the data are missing, and then possibly deduce problems in data consistency – sometimes your conclusions prove to be false.

To avoid this and not report too many "false positives" back to a department, it is a help if you understand the system in which the data are generated, combined with the business processes.

We will take a look at this in this blog post referring to analysis of order documents from SAP® based on the case "Purchase orders without purchase order items".

This focuses on the tables EKKO (order document headers) and EKPO (order document items); data-based screenshots use the desktop version of ACL™ Analytics software.

SAP® images are based on the transaction ME23N ("See order document"). Here the user sees the major content of purchase orders (purchase order number, supplier, document type, item, goods receipt, invoice receipt, etc), or can branch to it.

Fig. 1: Display of a purchase order by SAP® transaction ME23N

Data appearing in this mask are saved in the above tables, i.e. focus on EKKO and EKPO, the purchasing document headers and the purchasing document items (plus, in the EKBE table in the case of the order history).

For a data analyst directly accessing the entire data set (the "raw data"), and not going the round-about way through the SAP® GUI, the picture is usually less of a color mixture:

Fig. 2: The EKKO table opened in ACL™ Analytics software

Sce­nario of pur­chase orders with­out pur­chase order items

To start data analysis the analyst would check data consistency. The theory behind this is that in a consistent data set corresponding purchase order items must exist for each purchase order. By comparison of purchase order headers with order lines ("unmatched joins" in technical terms) you discover that there are six purchase orders for which no purchase order items exist.

Fig. 3: Purchase orders without order lines

Does that need to be a problem?

To get into it further, the analyst looks at one of the purchase order numbers concerned by transaction ME23N. Very clear, above the header, is the text "Held standard purchase order". That is presented as follows:

Fig. 4: Held purchase order


In SAP® it is possible to save purchase orders temporarily (to hold them), and possibly to continue processing them later. These may then also be incomplete, i.e. not all compulsory entries need to be filled in. From a data point of view you can identify such purchase orders – or exclude them from analysis – by the MEMORY attribute in the EKKO table. If an "X" is entered, this is a "Held purchase order". For held and incomplete purchase orders it is not possible to book goods or invoices received or to release them.

Overall there are 14 purchase orders in the data that are now excluded for the following analysis or are to be looked at separately. The EKKO table was filtered for all entries in which the EKKO_MEMORY field is not blank, enabling them to be looked at separately.

Fig. 5: In all there are 14 held purchase orders

Con­crete con­clusion

For your own reports based on order documents – no matter whether you do this with ACL™ or extract data from SAP® systems by the SE16 or SE16N transaction – you should remember that held (and thus possibly incomplete) purchase orders can be contained in the data, and exclude them if necessary. The "Incomplete" attribute does not only refer to missing purchase order items of course. It can mean the absence of a whole variety of content.


General conclusion

Sometimes it is not all that easy to trace back to the base data from the content shown in SAP® transactions. For example, while the supplier number shown in ME23N also appears identically in the EKKO table, it takes a lot of imagination to associate the text "Marked purchase order" in the SAP® mask with the table field EKKO_MEMORY. Occupying yourself with these details is certainly not relevant for all checking and analyzing activities – but it can enhance the quality of analysis and the conclusions that are drawn.



There are further examples in which data are missing or fields are not filled in where you would expect it, like supplier number or material number. I will be looking at this later in a separate blog post. 

Comments (0)
Be the first who comments this blog entry.
Blog login

You are not logged in. Please log in to comment this blog entry.

go to Login