Using SAP Machine Learning to Automate Payment Matching and Invoi

Written by Anurag Barua | May 21, 2020

Is your organization’s Accounts Receivable department waging a regular battle to manually reduce an ever-increasing pile (figuratively, not literally) of customer payments but not able to clear your customer invoices in a timely and/or accurate manner?

If your answer is ‘yes’, then you are not in a minority. If your enterprise operates in a high-transaction environment (such as retail, fashion, etc.) that consists of hundreds of thousands and possibly millions of customers that transact with your SAP transaction system(s) on a continuous basis, it is critical that incoming customer payments be applied to the appropriate open invoices.

Keeping Your Backlog Cleared

Why is it important to keep this backlog cleared regularly and speedily?

So that you can recognize revenue in a timely manner
Invoices that are not cleared can ultimately cause a slowdown in sales by generating credit blocks
You want to drive efficiencies in your financial operations. As an example, you want to keep your Days Sales Outstanding (DSO) as small as possible. In a world of manual matching, the DSO will be a high number compared with any industry-standard benchmarks

Unfortunately, the existing matching process in your SAP transaction system (whether it is S/4HANA or ERP or R/3) has been and continues to be in general, manual.

Therefore, a strong business case has existed to automate this important business process. And that is what we set about doing. One of our earliest, most popular, and ‘lighthouse’ Machine Learning (ML) apps as part of our Leonardo Foundation* was the SAP Cash Application.

Its algorithms analyze historical data (including payment advices, electronic bank statements, customer master data, bank master data, accounting documents, etc.) on the path to ‘learning’ matching criteria that are then applied to incoming payment advices and to automatically clear invoices.

*For those of you not familiar with SAP Leonardo, it is the set of tools and services that enable concepts such as Artificial Intelligence (AI), Internet of Things (IoT), ML, Blockchain, Predictive Analytics etc. and helps organizations embed intelligence in their operations & helps transform organizations into ‘intelligent enterprises’. And for those of you that are new to Machine Learning, it is a key component of Artificial Intelligence and is essentially a set of algorithms that generates models to make predictions, without any coding, but based on training data that is fed to these algorithms.

SAP Cash Application

Now let me explain the main ways by which SAP Cash Application automates payment matching with invoices:

Automatic retrieval of relevant data from payment information that comes in both structured and unstructured (including paper invoices, PDFs, etc.) formats
Automatic matching of lockbox payments to open Accounts Receivables

Additionally, this app tries to provide the missing information in partially complete fields to maximize the matching process
Conveniently define matching criteria using the enhanced reprocessing rule engine for predictable transactions and ML for the more unusual and complex ones
A transparent decision-making process that informs the user the basis for matching. A detailed log provides auditable details of the inferences made by the ML algorithms
Payment proposals that satisfy confidence thresholds are automatically cleared

From the above, it is easy to infer how and why this app significantly reduces manual effort thus freeing up staff to work on more value-added activities, increase accuracy and auditability of invoice clearing, and ultimately improving a few KPIs in your financial operations.

What Version of SAP Do You Need?

A question that is frequently asked by customers is what version of SAP do you have to be on in order to leverage this app.

If you are on S/4HANA on-premise version 1709, you are well covered. If your organization has implemented any version of S/4HANA Cloud, Essentials Edition (formerly called the Public Cloud or Multi-tenant Edition) or S/4HANA Cloud, Extended Edition (formerly called the Private Cloud or Single-tenant Edition), you can utilize this app too.

However, if you are on ECC 6.0, the situation is somewhat different. If you are on EHP7 SPS 14 or EHP8 SPS, you will need to use a connector made available by our Innovative Business Solutions Organization (IBSO) as a Repeatable Custom Solution (RCS). If you have any questions regarding product availability, coverage, licensing, etc., please contact your SAP Client Partner.

A Few Resources

Interested in learning more about this exciting app? If so, here are some links that will provide you with more details:

View full post