Custom Code and Technical Debt: What Actually Drives S/4HANA Migration Cost

of SAP customers cite extensive testing as a top challenge in custom code migration. That burden scales directly with how much undocumented custom code a system carries. (smartShift)

AI-assisted tools are good at answering one question: is this code still called, and by what. They're far less reliable on a second one: does the business still need what this code does. That's a judgment call about process, not syntax. Take a custom pricing routine that hasn't fired in eighteen months. It might be genuinely dead, or it might exist for a customer segment that only transacts once a year. Getting that distinction wrong in either direction has real cost. Delete active logic and something breaks in production. Keep everything "to be safe" and you've just recreated the technical debt problem inside the new system.

The organizations getting the most reliable outcomes from AI-assisted code assessment pair it with structured business-process owner sign-off on every object flagged for retirement. The AI output gets used to focus a scarce group of experienced reviewers on a manageable shortlist, rather than asking them to review 22,000 objects from scratch, or trusting an automated recommendation with no review at all.

Custom code assessment is where a lot of migration budgets go wrong early, and where an outside technical review earns its keep fast. It's hard for the team that wrote the code to be fully objective about what's safe to retire, and a rushed or overly cautious call here sets the tone for everything that follows. Orpington Technologies leads S/4HANA migrations built around exactly this kind of independent code and technical-debt assessment, whether as a standalone review or as part of the broader project.