Since the move to S/4HANA, the question “how do you synchronize data between SAP ERP and EWM” no longer has the same answer, and many projects still get it wrong by trying to activate the Core Interface out of habit.
SAP ERP to EWM master data integration actually hinges on a single architecture choice: is EWM embedded or decentralized? In one case there is nothing to synchronize, because the master data is shared. In the other, the CIF has become unusable, and ALE then DRF take over. This guide untangles the two scenarios and, above all, states who creates what, how, and when.
- The ERP stays the leading system for master data: the flow goes one way only, ERP to EWM.
- In embedded EWM (S/4HANA), nothing is synchronized: EWM reuses the ERP’s shared data. The SCM Product is created in the background, the Supply Chain Units by hand.
- In decentralized, the transfer is real: up to EWM 9.5 through the Core Interface (CIF), then on S/4HANA through an ALE distribution model (IDocs) or the DRF.
- The CIF is not usable for a decentralized S/4HANA EWM: it only creates SCM master data.
Master data in SAP EWM: a single leading system, the ERP
Before talking about transfer, a reminder that clears up half the misunderstandings: master data is maintained centrally in the ERP. The ERP is the leading system, and the flow to EWM always goes one way. “Synchronize” suggests a two-way exchange, which is not the case here.
The ERP Material master becomes the Product master on the EWM side. Concretely, transferring master data from SAP ERP to SAP EWM carries over mostly the descriptions, units of measure, material group, weight and volume. The Plant, shipping point, customer master and vendor master exist on both sides, but the EWM objects are not identical to the ERP ones: only the relevant fields are copied to the corresponding locations and business partners. The Supply Chain Unit (SCU), finally, is a subset of a location’s information, required specifically for EWM.
In embedded, data is shared: a single set of master data, read by both components. In decentralized, it is replicated: a physical transfer copies the data from the ERP to the EWM’s separate database. Confusing the two skews a project’s effort estimate.
Embedded EWM (S/4HANA): there is nothing to synchronize
In an embedded EWM, EWM and ERP live in the same S/4HANA system and share the same database. No master data distribution is needed: EWM directly reuses the data from the ERP side. The question of transfer therefore does not arise.
Two practical nuances that key users often confuse:
| Object | In embedded EWM |
|---|---|
| ERP master data (material, partners) | Shared, read directly, nothing to transfer |
| SCM Product (Warehouse Product) | Created automatically in the background |
| Supply Chain Unit (SCU) | To be created manually |
| Location | Not used in embedded |
The SCM Product is generated in the background, with no intervention. The Supply Chain Unit, however, still has to be created by hand: it is the most common oversight at start-up. And unlike decentralized, Locations are not used in embedded, which simplifies the model.
Decentralized EWM: from CIF to ALE then DRF
In a decentralized EWM, EWM runs on a separate system, with its own database. There, the master data transfer is genuinely real, always from the ERP to EWM. And the method has changed from one generation to the next.
Up to SAP EWM 9.5, based on the SCM platform, master data distribution from the ERP went through the Core Interface (CIF). The CIF historically fed the SCM systems, and it was reused for EWM needs.
With a decentralized EWM on S/4HANA, the picture changes: the CIF is not usable for master data. The reason is specific, and it is what traps projects. The CIF can only create SCM-type master data, whereas the EWM part of a decentralized S/4HANA system reuses ERP-type master data, exactly like an embedded one. The transfer therefore goes through an ALE distribution model (with IDocs) or the Data Replication Framework (DRF, via services). For the first release of decentralized S/4HANA, only ALE was available; it is being progressively replaced by the DRF. On the manual creation side, the Plant and the Supply Chain Units (for the warehouse and the shipping points) have to be entered by hand.
Embedded or decentralized: the decision at a glance
The choice between embedded and decentralized is made for overall architecture reasons, not only for the master data question. But the master data integration effort is a concrete criterion that weighs in the balance.
Embedded EWM
- A single S/4HANA system, shared database
- No master data transfer to configure
- SCM Product automatic, SCU manual
- Tight coupling with the ERP, simple PP integration
- Typical driver: simplicity, integrated scope
Decentralized EWM
- Separate system, own database
- Transfer to configure: ALE or DRF (CIF up to 9.5)
- Plant and SCU to be created manually
- Decoupling from the ERP, warehouse scalability
- Typical driver: automation, high volume, independence
In short: heavy warehouse automation and the need to decouple the EWM pace from the ERP point toward decentralized; simplicity and tight integration with production push toward embedded. Master data synchronization is a consequence of that choice, not its primary criterion. If you are unsure about the deployment mode itself, the guide to choosing between WM, Stock Room Management and EWM sets the frame.
What is created automatically, what by hand
That leaves the most concrete question, and the least often addressed: by object and by deployment, what is shared, created automatically, or to be entered manually?
| Object | Embedded S/4HANA | Decentralized S/4HANA |
|---|---|---|
| ERP master data (material, partners) | Shared | Transferred (ALE or DRF) |
| SCM Product (Warehouse Product) | Auto (background) | Transferred then enriched |
| Supply Chain Unit (SCU) | Manual | Manual |
| Plant | Shared | Manual |
| Location | Not used | Used (route determination) |
Three useful reflexes. The Supply Chain Unit is always manual, in both modes: a warehouse number must be attached to an SCU, and forgetting it blocks the first movements. The distinction between master data and customizing stays clear: this table only covers master data, configuration follows its own rules. And the other half of the contract between the two systems, on the stock side this time, rests on availability groups and the ROD and AFS storage locations.
FAQ: data synchronization between SAP ERP and EWM
Do you need to synchronize master data in embedded EWM?
No. In embedded EWM (S/4HANA), no distribution is needed: EWM directly reuses the shared master data from the ERP part of the same system. The SCM Product is created automatically in the background, but the Supply Chain Units still have to be created manually.
Why is the CIF no longer usable in decentralized EWM on S/4HANA?
Because the Core Interface can only create SCM-type master data, whereas a decentralized EWM on S/4HANA reuses ERP-type master data, like an embedded one. The transfer therefore goes through an ALE distribution model or the DRF. The CIF remained the path up to SAP EWM 9.5.
What is the difference between ALE and DRF for the transfer?
The ALE distribution model transfers data via IDocs; the Data Replication Framework (DRF) transfers it via services. On decentralized S/4HANA, the first release offered only ALE, progressively replaced by the DRF.
Which data must be created manually in EWM?
The Supply Chain Units (for the warehouse and the shipping points) in all cases, and the Plant in decentralized. The SCM Product, on the other hand, is created automatically. Locations are not used in embedded.
Is the ERP Material master copied identically into EWM?
No. The EWM Product master carries over the relevant fields from the ERP Material master: descriptions, units of measure, material group, weight and volume. For the Plant, shipping point, customer and vendor, both sides exist without being identical, only the useful fields are copied.
The lesson fits in one sentence: “synchronize” is the right word only in decentralized, and even there, the means depends on your release. In embedded, there is nothing to synchronize, just Supply Chain Units not to forget. Once the master data is in place, the next step is executing the movements in the warehouse, driven by the RF framework.