When businesses upgrade to SAP S/4HANA, their Accounts Payable teams hope for less manual work, more reliable and useful data and better controls against costly errors like duplicate and overpayments. Our research shows that amongst the upgrades, AP users won't find all the controls they need using SAP S/4HANA as a standalone solution for risk mitigation in payables.
We analysed our customers' ERP data — both pre- and post-S/4 HANA migration — and found a consistent truth: even the most modern ERP systems, including S/4 HANA, still need additional control layers to catch payment errors.
In this article, we’ll break down why this is the case and how Xelix plays a vital role in catching mistakes, errors and risk before money leaves the door.
We analysed data directly from our customers’ ERPs, including those who have already transitioned to S/4 HANA, and compared the number of payment errors we found. Our data set included over 1.2 million invoices from 187 different ERP systems, over an average time period of 25 months. Despite a recent migration to S/4 HANA, we found that duplicate payments and invoice errors equally impacted businesses regardless of ERP system(s) in use.
On average, across all systems, duplicate payments amount to 0.2% of total supplier spend per customer.
The chart below compares duplicate payments as a percentage of spend for an average company, with varying levels of AP controls, to the percentage found in the data of our customers using S/4 HANA and other ERPs. It shows that even after invoices have been processed on S/4 HANA, duplicates still account for 0.22% of spend. To put into numbers, if a business using S/4HANA spends £8 billion, the data shows they should expect to pay out £17.6 million in duplicate payments.
The chart below shows the % of invoices containing duplicates for the average company (this includes companies with varying levels of AP controls) compared to the average duplicate rate in our customers' data, looking at S/4 HANA and other ERPs. Similarly to the first dataset (duplicates as a % of spend), we can see that S/4 HANA does not catch a higher volume of duplicate invoices than other ERPs.
Payment errors are a consistent challenge for all organisations. A single duplicate invoice can lead to unnecessary losses, and overpayments can strain cash flow and relationships with vendors.
Organisations increasingly see the need to proactively monitor transactions and prevent these errors from happening in the first place. They’re moving away from costly recovery audits and implementing smarter upstream controls.
Unfortunately, many believe an upgraded ERP will provide enhanced AP controls. Even with newer ERP systems like S/4 HANA, the complexity and volume of transactions cause these errors to continue.
As you can see from our customer data, S/4 HANA alone falls short when it comes to preventing duplicate payments.
Here are a couple reasons why:
Manual processing accounts for on average, 60% or more of duplicate payments*.
The limitations of S/4 HANA’s native controls
SAP S/4HANA's invoice posting workflows use exact-match validations for key information such as Vendor ID, invoice number, gross amount and posting date. To check for duplicate invoices, S/4HANA uses a special marker (the REPRF field) in each vendor's information to help the system flag if it sees the same invoice more than once.
This approach works well for catching identical copies of invoices, but it might miss invoices that are very similar but not exactly the same.
S/4HANA often won't detect errors such as:
For example, S/4HANA might not flag a keying error that replaces a "0" with an "O" or transposes digits (e.g., 1,234 to 1,324) if tolerance limits in OMRM are set too broadly.
There is a workaround to these limitations by building custom logic via ABAP enhancements, which can be done by a consultant provided by SAP or your own internal employee.
This means added resource and time on top of an already complicated implementation.
Additionally, even with customisations, this custom logic is still rule-based, so it can’t take into account real-life errors that don’t follow rules: OCR misreads from scanned invoices, human input errors etc.
Even a smart ABAP developer can’t predict every possible variant or anomaly, so the custom logic quickly becomes either too loose, missing fraud and errors, or too strict, spamming users with false positives.
Not to mention, rule-based ABAP logic also doesn’t adapt so it won’t learn from past duplicates or patterns.
*This data is supported by research from organisations such as APQC, Ardent Partners, Deloitte, and PwC, which highlight manual data entry and lack of automation as primary drivers of payment inaccuracies.
Xelix goes beyond S/4 HANA’s datapoints
Xelix enhances this process by moving beyond simple rule-based checks and applying advanced Machine Learning (ML) techniques to detect duplicates more effectively.
Instead of relying on exact matches, we analyse over 500 datapoints to assess similarities and differences between invoices and to understand long-term patterns within an organisation’s invoicing data. This enables us to identify duplicates that traditional ERP checks might miss, such as those with formatting inconsistencies, OCR errors or data entry mistakes (supplier or AP side).
These datapoints are processed through Machine Learning models trained to quantify duplication risk. We assign a “risk rating” - high, medium or low risk - to each duplicate pair in the platform, which represents how confident we are that we have identified a true duplicate. Low-risk invoices are filtered out, while high-risk cases are surfaced for your review, reducing false positives and ensuring greater accuracy in duplicate detection.
On day one, Energy Transfer found +$1.7m to recover in a historic audit with Xelix, and in the year that followed, they prevented more than half a million dollars in over payments.
Duplicate vendor records can account for as much as 30% of duplicate invoices*.
Vendor consolidation limits in S/4 HANA’s Business Partner model
SAP S/4HANA's Business Partner (BP) concept doesn't enforce strict vendor consolidation. For example:
Data cleansing and BP merging often requires manual intervention, creating a significant workload for master data teams. The task can also be neglected or left to the end of the year, meaning you’ve likely suffered losses due to unclean data in the meantime.
You can also use custom logic to overcome these BP model, but as is the case with enhancing duplicate controls, the rules-based method has limitations.
*This finding is supported by industry studies from APQC, IOFM, and leading consulting firms, which emphasise the importance of maintaining accurate, deduplicated vendor data to prevent invoice duplication and related payment errors.
Xelix monitors your vendor data for you
Xelix enhances S/4HANA's capabilities with AI-driven vendor data monitoring:
With the help of Xelix, retailer EG Group optimised their vendor file from 22K to 15K, removing inactive or duplicate suppliers to focus on building strong relationships with active vendors.
By combining S/4HANA's centralised BP approach with Xelix's intelligent monitoring, organisations can reduce risks associated with duplicate vendor data such as duplicate payments.
Xelix is an ERP agnostic, bolt-on solution, which means set up is easy and works with any system. By layering Xelix onto your ERP, you can enjoy peace of mind, knowing you have best-in-class controls that prevent leakage through Accounts Payable.
Due to the bolt-on nature of the tool, implementation is simple and time-to-value is quick. Most companies see ROI within one month.
Join the likes of Kraft Heinz, EG Group and BMI who adopted Xelix alongside their S/4 HANA migrations, giving them greater control and visibility in a high-risk period.