R/3 and ECC legacy

But what we and many of my SAP regulars once celebrated as a competitive advantage is turning out to be a leaden legacy in the face of the digital transformation to S/4. The grown SAP ERP monoliths, interspersed with millions of lines of our customer-specific code, now resemble archaeological excavation sites rather than modern IT architectures. Over the years, many of these systems have grown into pure colossi, modified beyond recognition, with often up to 60 percent of the code no longer being used during operation - as I found out - but still causing maintenance costs and harboring security risks.

The challenge of the Abap conversion is far more than a technical diligence task; it is a brutal cut through the history of our Group. While previous release changes were often of a purely technical nature and confirmed the slogan „Never change a running system“ (as editor-in-chief Färbinger never tires of emphasizing), S/4 demands a radical paradigm shift. The switch to Hana and the new ERP data model make many old Abap constructs obsolete. Simple SQL statements that relied on the existence of certain aggregates or indices are coming to nothing, and SAP's so-called simplification items are forcing developers to fundamentally revise or even discard their code.

Please note: Many old functions that continued to run as part of the first S/4 migration, see Compatibility Packs, had to be converted to new S/4 functions by the end of 2025. The SAP Simplification List defines thousands of functional changes and discontinuations for each S/4 release, which significantly determine the system conversion process. It is no longer enough to simply adapt the code syntactically; it must be semantically aligned with the new logic of the „Principle of One“, which rigorously eliminates redundant functions. In this area of tension between preservation and renewal, SAP propagates the „Clean Core“ strategy as the only way out of the modification trap. The philosophy is simple, but the implementation is very painful for us, to say the least! But we see light at the end of the tunnel: a former companion of Professor Hasso Plattner at HPI at the University of Potsdam, Professor Dr. Alexander Zeier, presented an AI system as an exit strategy for Clean Core and Abap modifications.

The digital core of the ERP system must remain clean and free of modifications in order to ensure upgradeability and agility in the cloud. The days of tinkering directly with the SAP standard program (modification) are over. The new model requires a strict separation of standard code and enhancements. For existing SAP customers, this means that they have to transfer their beloved Z programs and screen applications to modern, cloud-native architectures.

But how can this balancing act be achieved without stalling the business processes? The answer lies in a differentiated extension model. Critical extensions that are closely interwoven with the data and transactions of the core can be implemented as „on-stack extensions“ using Abap Cloud. Everything else, especially detached innovations or mobile apps, belongs as a side-by-side extension on the BTP, decoupled from the ERP core. For legacy systems, this often means massive refactoring: code must be analyzed, cleaned up and, if it does not conform to the standard, either wrapped in a wrapper or completely rewritten.

In this storm of transformation, the SAP Customer Center of Expertise (CCoE) has taken on a completely new, existential significance, which led to heated discussions at our SAP regulars' table. In the on-prem world, the CCoE (CCC when I was an active IT manager) was often primarily responsible for first-level support and license management, but now it has to transform itself into the strategic control center of the hybrid IT landscape. The CCoE is no longer just responsible for operations, but for the governance of the entire architecture. It must monitor compliance with the clean core principles and ensure that the specialist departments do not fall back into old patterns and produce uncontrolled growth.