The e-invoicing requirement has been implemented. However, it is data quality, processes, and systems that ultimately determine whether structured data can be turned into a smart invoicing process.
In recent years, many companies have viewed the implementation of e-invoicing primarily as a regulatory requirement. The focus has been on questions such as: What deadlines apply? What formats are permitted? How can e-invoices be received and processed?
Key Points at a Glance:
- The e-invoicing requirement has been implemented, but the real work is just beginning.
- Structured data is not automatically clean data: errors in format, business rules, and content are still possible.
- Automated processing depends on data quality, processes, and system integration—not on the format alone.
- Those who lay this groundwork now will reap the greatest efficiency gains from e-billing.
With the successful rollout, the goal seemed to have been achieved: PDFs were replaced by structured invoice data, legal requirements were met, and the first e-invoices were received.
But anyone who thought that was the end of the story is now in for a rude awakening. The biggest challenge with e-invoicing isn't the format, but rather the processing of the data it contains.
While e-invoicing does eliminate traditional invoice-processing problems—such as errors in OCR recognition or the manual entry of invoice data—it also brings new challenges to the forefront: Data must be validated, interpreted in a business context, and reliably integrated into existing processes. In the past, the biggest challenge was reliably recognizing invoice data in the first place. Today, this information is available in a structured format. The key factor now is whether it is complete, technically accurate, and stored in the right place. An XML file can therefore be technically perfect in its structure and still not be processed automatically. And even standardized formats, while they create a common basis, leave room for interpretation in practical implementation.
In short: Structured data does not automatically mean clean data. The e-invoicing requirement has digitized invoicing processes. These processes only become smarter when backed by the right data, processes, operating models, and systems.
E-billing has been implemented. Why are so many companies still just getting started?
With the gradual introduction of mandatory e-invoicing, many companies initially focused on complying with legal requirements. Triggered by the European standard EN 16931 and its transposition into national law, they had to select the correct formats, adapt their systems, and ensure they could receive electronic invoices. For many, this seemed to mark the end of the project.
While the technical transition is manageable in many cases, it is only in the day-to-day handling of e-invoices that it becomes clear how well systems, master data, and processes actually work together. Only then does it become apparent where data is missing, where standards are interpreted differently, or where automated workflows fail due to seemingly minor issues.
Discussions with companies and experts on the topic of e-invoicing paint a similar picture: The actual effort involved is often underestimated. Some projects are postponed or even halted because it only becomes clear as the project progresses that the transition involves far more than simply changing a file format. At the same time, the adoption of e-invoicing is proceeding more slowly than expected in many places. In small and medium-sized businesses in particular, existing transitional provisions are often still being used, and necessary adjustments are being postponed. Those who underestimate this effort risk delays that directly impact compliance and liquidity.
““Many companies saw 2025 as the finish line. In reality, though, the real work only begins after that.”
Jens Brettschneider | Business Unit Manager Application Management
ISR Information Products AG
Receiving an XML file is just the beginning of the actual task. Companies must determine which formats are supported, how suppliers submit invoices, what information is needed for a seamless invoice processing workflow, and how this information can be integrated into existing ERP and financial processes. Added to this are considerations regarding consistent corporate standards, complete master data, clearly defined responsibilities, and stable operations.
The more companies seek to automate their invoice processing, the more important these fundamentals become. After all, the success of e-invoicing depends not on the invoice format, but on the quality of the data and the interaction of the underlying processes.
E-INVOICING WHITEPAPER
Not yet familiar with e-invoicing? Our white paper on mandatory e-invoicing provides a concise summary of all the basics.
Why Structured Data Doesn't Yet Mean Seamless Invoice Processing
E-invoicing lays the groundwork for a digital invoicing process.
In the past, the question was: Does OCR reliably recognize the data at all? Today, the question is: Is this data complete, factually correct, and in the right place? That may sound like a minor difference. In practice, however, it means a fundamentally different verification process.
At this point, the actual task changes. Whereas the focus used to be primarily on errors in optical character recognition or manual data entry, today the quality of the transmitted data determines how reliably invoice processes can be automated. This is because even a technically correct e-invoice may contain information that is not usable—or only usable to a limited extent—from a business perspective.
Three Types of Errors Every Company Should Be Aware Of
One of the most important lessons learned from the first e-invoicing projects is that technically correct does not automatically mean business-wise correct.
According to the FAQs of the Federal Chamber of Tax Consultants (as of March 2026) and the BMF letter dated October 2025, there are three clearly distinct types of errors in e-invoice verification:
- Format errors occur when the technical structure of the XML file does not comply with the requirements of the EN 16931 standard. Such a file is legally considered an “other invoice,” which has consequences for input tax deduction.
- Business rule errors occur when required fields are missing or when information on the invoice is contradictory—for example, when the tax amount shown does not mathematically match the net amount.
- Content errors are the most insidious category: The file is technically flawless, but the values are factually incorrect, incomplete, or implausible. In most cases, validation software does not detect these errors.
An e-invoice may fully comply with the required format and contain all mandatory fields, yet still not be eligible for automated processing. For example, internal order numbers may be entered in the field for the customer’s order number, or information may be entered in a location other than the one intended. The invoice is technically valid, but automated processing still fails because the ERP system expects the information to be in a different location.
Validating an e-invoice is therefore an important first step. However, it does not replace a substantive review of the data it contains.
ZUGFeRD in Practice: What Happens When XML and PDF Files Differ?
This challenge is particularly evident in hybrid formats such as ZUGFeRD.
The rule here is clear: For hybrid formats, the XML data set is the legally binding part, not the PDF. If the two differ, only the information contained in the structured data set is relevant for input tax deduction. This is explicitly stipulated in the BMF letter dated October 2025 and applies regardless of what appears on the visually readable page.
For accounting and auditing, this means that anyone who continues to open only the PDF viewer is working on a basis that is not legally binding. Displaying the XML content alongside the PDF view thus becomes an important component of the auditing process—not as a technical gimmick, but as the foundation for a robust audit.
Why Automated Invoice Processing Often Falls Apart Over Small Details
Many problems arise not from complex technical issues, but from small deviations in everyday life.
An incorrectly filled-in data field, inconsistently maintained references, or free-form text instead of structured information are often enough to interrupt an automated invoicing workflow.
A typical example is payment terms. Instead of providing only structured information about due dates or discount periods, invoices often include additional phrases such as “payable immediately without deduction.” People can easily understand this information. Systems, however, must first interpret it and translate it into the appropriate payment terms used by the ERP or financial accounting system.
Standards provide guidance, but not uniformity
There is another aspect to consider: E-invoicing is not a globally uniform standard. Although EU Directive 2014/55/EU establishes a common framework, its specific implementation varies from country to country.
In addition to formats such as XRechnung and ZUGFeRD, there are numerous international specifications and country-specific variations. Companies that work with international suppliers must therefore take different requirements and formats into account.
A standardized format serves as the foundation. However, it does not solve the challenge of validating, interpreting, and integrating data into existing processes. Those who commit to a single format are building a solution for the past. Anyone who wants to process e-invoices automatically needs systems that can adapt to the growing variety of formats.
E-invoicing eliminates many of the problems associated with paper-based processes. At the same time, it reveals weaknesses in data quality, master data, and processes that were often hidden before. This makes it clear that the digitization of invoice processing does not end with the receipt of structured data—it only begins there.
What determines whether e-invoices can be processed automatically?
It is not e-billing that automates invoice processing, but rather the processes that build upon it.
Structured invoice data is the foundation for automated invoice processing. However, whether this results in a fully automated invoice receipt process depends not on the format itself, but on how well the data, systems, and processes work together.
This is because every e-invoice must be placed within the business context of a company. Master data must match, order information must be unambiguously assigned, and payment terms must be correctly interpreted. Only when approval workflows, the ERP system, and financial accounting are fully integrated can the invoicing process be largely automated.
In other words: The XML file provides the information. The process determines what becomes of it.
That is precisely why, following the introduction of e-billing, many companies are now, for the first time, taking a close look at issues that had previously received little attention:
- Which formats should be supported in the long term?
- Through which channels should suppliers submit their invoices?
- What information is actually required for automated invoice processing?
- What corporate standards will apply in the future?
- Is master data maintained in a complete and consistent manner?
- Which processes need to be adjusted or redefined?
The higher the desired level of automation, the more important these fundamentals become. After all, even the best-structured dataset cannot compensate for missing process rules, inconsistent master data, or unclear responsibilities.
E-invoicing doesn’t just reveal these weaknesses—it makes them transparent and forces companies to address them. Yet this is also where the opportunity lies. Companies that use the implementation of e-invoicing as an opportunity to jointly improve their processes, data quality, and system architecture lay the foundation for a more efficient invoice processing system in the long term.
The greatest profit comes only after the launch
So much for the challenges. Now let's look at the other side of the coin.
Once the groundwork is laid, the day-to-day work in the accounts receivable department noticeably improves.
Many tasks that were previously taken for granted are no longer necessary. Invoice data no longer needs to be entered manually or corrected due to typical OCR errors. A zero is no longer confused with an “O,” and line breaks no longer result in incorrectly read order numbers. Instead of having to first recognize information, systems can work directly with the structured data.
This also shifts the focus of the employees. Routine corrections take a back seat. Instead of correcting incorrectly processed invoices, they can concentrate on cases that actually require professional judgment, such as when information is incomplete or there are discrepancies between the invoice and the purchase order.
Companies with a high volume of invoices benefit from this the most. The more invoices that can be processed without manual intervention, the more processing times are reduced and the less effort is required for recurring routine tasks.
Practical experience therefore shows that the true added value of e-invoicing does not come from the introduction of the XML format, but only when companies consistently take advantage of the new opportunities it offers.
Conclusion: E-invoicing is not an end in itself, but rather the foundation for smart invoicing processes
The introduction of mandatory e-invoicing marks an important milestone. However, the experiences of the past few months clearly show that true success depends not on the invoice format, but on how companies handle the structured data.
The key question today is no longer: Can we receive e-invoices? But rather: (What do we do with them?) How do we turn them into a reliable, integrated, and automated invoice processing workflow?
From E-Invoicing to Intelligent Invoice Processing—with ISR Accounts Payable Flow
The very challenges described in this article are the starting point for our work.
As a specialist in enterprise information management, ISR integrates data, documents, processes, and technology in such a way that invoice processing is not only digitized but also becomes a practical tool in day-to-day business operations.
ISR Accounts Payable Flow processes e-invoices in ZUGFeRD and XRechnung formats—and goes a step beyond mere format validation: XML data is displayed alongside the PDF view so that accountants and auditors can spot discrepancies at a glance. Free-text fields, such as payment terms, are processed by the integrated AI algorithm SmartCodingSmartCoding is an intelligent algorithm that assists the user by… More automatically interpreted and translated into processable values. And the validation process includes not only technical format checks but also a comparison with master data and order information.
The result: More invoices processed without manual intervention. Fewer follow-up inquiries, fewer rounds of corrections, and shorter processing times.
Would you like to know where there is still room for improvement in your invoice processing workflow? Schedule a consultation. Together, we’ll examine where data quality and process logic
Please contact us, and we'll look into it with you.
Jens Brettschneider
Head of Business Unit
Application Management
jens.brettschneider@isr.de
+49(0)151 422 05 425
