This is the end of digitization!

3 reading min.
Prof. Dr. Nick Gehrke

written by

Prof. Dr. Nick Gehrke

Judgement in data analytics.jpg

Part IX of the series: “Digitization of auditing SAP Order-to-Cash Processes”

Data and process analysis have the potential to bring great benefit to auditing. Nonetheless the questions are: how is it possible to pursue and analyze the enormous and extensive results? Does one have to pursue each and every case that emerges from data analysis? Often, there are way too many results and false positives! Frequently auditors are overwhelmed by the amount of results. In the following, I would like to introduce an approach on a possible way forward for Professional Judgement in Order-to-Cash Audits.


1. Digitization of auditing SAP order-to-cash processes
2. How to audit master data in order-to-cash processes
3. Auditing of order-to-cash processes: sales orders and deliveries
4. What's wrong with these sales invoices?
5. Who cares? Auditing incoming payments
6. On how to find exotic processes
7. SAP data structure in order-to-cash: Who cares?
8. Way to go: Auditing real SoD in order-to-cash
9. This is the end of digitization!


Before you proceed reading on the details of the indicators, I would recommend you read the concept of indicators in part 2 of the series first.


Divide et Impera

"Divide and rule" – says the Latin scholar. This simply means: If a task is too difficult, divide the task into subgroups and take care of the subgroups that are easier to process until all the subgroups have been edited. What does this mean for interpreting a data analysis for the auditor?

Overall I have implemented 125 indicators for detecting process weaknesses (through all processes of purchasing, order-to-cash, fixed asset and inventory, as well as cross process). In this series various indicators of order-to-cash were presented. During the implementation, exactly the described problem of the too numerous results came up and therefore a solution had to be designed. I explain it by an example - let's take the following indicator:


Late customer payment

This indicator is associated with the audit objective of saving opportunities.

There is the risk that customers pay late and high receivables lead to a lack of liquidity or loss of interest.

The criterium for this indicator is:

The document will be marked because a customer paid later than 1,5 times of average payment time in days. 

Imagine, that 1854 documents were affected, which were marked by this indicator. The auditor quickly rejoices in this number, he would never have done the data analysis! How should one investigate 1854 indications?

To handle this problem, I have developed the concept of “profiling” which has been implemented consistently in zap Audit. It’s pretty simple: every document which was marked with an indicator is automatically sorted and allocated to a certain group.The creation of groups is dependant on the indicator, as well as the specific analysis objective. The idea behind it is that one only has to look at the result groups and not at every single document.

Let us pursue this example with the indicator "late customer payment". This indicator creates its profiles according to the criterium "days after which the payment was received", this means all documents with payment receipt 0-30 days (profile 1), 31-60 days (profile 2), 61-90 days (profile 3), ... (profile n) after the 1.5-fold payment morale in days are assigned to a profile, a group. For example, it turns out that the 1854 documents are divided into 3 profiles with 3 different day intervals.

The auditor can then concentrate on the profile, which contains particularly many defaulting payers, or even better on the profile in which late payment was made. Through the profiles a special focus is placed on the highest risks or most interesting cases. Random samples lose their methodological value.

It is no longer a question of individual cases, but case groups, which makes the audit much more efficient and also leads to a very good understanding of the audited entity.

In the example given, the question must therefore be considered:

Which customers are in profiles with particularly late payment behavior?

... and thus the examination of the indicator would be completed.


zapliance Newsletter label

Sign up to our newsletter
to receive latest news and updates every two weeks!

zapliance customer brands

Our recommendations to you