Every S/4HANA programme talks about custom code, but the projects that slip most often slip on data. ECC systems carry years of duplicate customers, inconsistent material masters, half-filled fields and local workarounds. S/4HANA’s simplified data model has no patience for any of it. Data readiness is the real gate between you and go-live — and between go-live and the AI capabilities that make S/4HANA worth the move. This article is part of our complete SAP ECC to S/4HANA migration guide.
Why data is the bottleneck
- Accumulated debt: duplicates, obsolete records and inconsistent conventions built up over many years.
- A different model: S/4HANA simplifies tables and requires the business partner model for customers and vendors.
- Hidden dependencies: reports, interfaces and custom code that silently rely on today’s data quirks.
- Ownership: data quality is a business decision, but it is often left to the technical team.
Start with profiling, not loading
Profile master and transactional data early: completeness, duplicates, invalid values and orphaned records. Profiling turns vague worries into a measurable backlog, and it usually changes the migration plan. It also tells you how much history is worth bringing across at all.
The business partner conversion
Converting customer and vendor masters to business partners is mandatory and frequently underestimated. Resolve duplicates, harmonise address and tax data, and agree number ranges and grouping before conversion runs, not during cutover.
Custom code decides whether S/4HANA runs. Data decides whether it is worth running.
ESS ENN Associates Data Engineering Team
Choosing the migration approach
For new implementations, SAP’s Migration Cockpit and migration objects cover standard loads. Brownfield conversions carry data across in place but still need cleansing beforehand. Selective data transition — moving only chosen companies, years or objects — is often the right compromise, typically delivered with specialist tooling. Third-party platforms are increasingly used for large, complex estates; SAP has, for example, endorsed Palantir AIP for data migration, though results quoted by vendors should be validated on your own data.
Where AI genuinely helps
AI is well suited to the repetitive heart of data work: duplicate detection across fuzzy matches, standardising free-text fields, classifying materials, suggesting mappings and flagging anomalies for review. It shortens the cleansing cycle considerably. What it does not do is decide your business rules — which records are authoritative, which history to keep — so build human approval into every step. See how AI is accelerating S/4HANA migration.
Validate continuously
Run migration rehearsals repeatedly, reconcile counts and values between source and target, and have business owners sign off on samples. The final cutover should be a repeat of something you have already done several times.
Clean data is what makes clean core pay off
A clean, standard S/4HANA core is the foundation for SAP’s AI agents and process intelligence — and those tools are only as good as the data beneath them. Treat data readiness as an investment in what comes after go-live, covered in clean core, SAP BTP and Joule agents. For the code side of the same problem, read migrating custom ABAP code.
ESS ENN Associates supports SAP S/4HANA programmes with AI-assisted assessment, custom-code and data remediation, and testing, through our AI agent development, data engineering and legacy modernisation teams. To assess your SAP data readiness, Contact us.