Capgemini’s Central Finance (CFIN) blog series part-1 focused on the overview of functionality & use cases of CFIN. In Part-2, we will cover the technical aspects as well as key lessons learnt from our global experience.
Central Finance is an evolving, and fast growing, deployment option for S/4HANA or S/4HANA Finance with major innovations seen in each release. Until S/4HANA release 1511, it was being largely seen as a solution to deliver consolidated finance reporting.
However, 1610 and 1709 delivered new functionalities such as Central Payments and Receivables Management and there is now a genuine use case for CFIN to exist in the end state architecture without moving to single instance ERP.
With growing interest in the market (we cannot quote exactly what we’ve heard but we believe the CFIN licence sales are now in the multiple 100s) this is the right time to give a flavour of what to expect in an implementation.
What is the process for posting the documents?
Below is the technical architecture of CFIN:
Initially, we need to do the following activities:
- activate the business function FINS_CFIN and standard configuration for posting the FI/CO document in CFIN
- define the mapping actions – ‘Keep’, ‘Clear’, ‘Map’. ‘Map’ action could be key mapping (long-lived master data), value mapping (configured data), cost object mapping framework (short lived master data) in CFIN
- implement notes in source SAP systems to enable the standard CFIN structures. In S/4HANA 1709 system the standard CFIN structures are available without implementing notes
If we simply draw the overall process on a board, below are the 4 steps:
Step-1: Initial load – It is a mandatory step, post which only the online replication can be performed. The sequence is first FI postings (via RFC) and then pure CO postings (via SLT & AIF). It needs to be marked as complete in table VCFIN_SOURCE_SET, even if there is no initial load. We can define the period from which line items/balances need to be transferred.
Step-2: Financial documents replication (reposting) – Post initial load, whenever any document is posted in source SAP systems, business functions in table TRWPR triggers the function modules, which in turn populates the data from tables (BSEG, BKPF etc.) to the below CFIN tables:
- AUFK (cost objects)
- CFIN_ACCHD (financial documents)
- COBK (CO secondary postings)
CFIN system has the same set of tables as well. With the help of replication mechanism for the said tables in SLT, the data gets replicated from source to target system. SLT can do replication both real time as well as periodically (based on configuration) but the idea of CFIN is real time replication. SLT checks and replaces the data according to the mapping framework and MDG and then feeds the data to accounting interface for posting the universal journal.
For non-SAP sources, the replication happens via SLT only. There are no table triggers involved and it happens with customization efforts in both sources and target system.
Step-3: Error handling – AIF is included in CFIN license. It handles errors pertaining to FI/CO documents & cost objects replication, initial load of CO documents. Initial load of FI balances & FI documents is directly through RFC and not AIF.
Step-4: Audit trail – It is available not only to SAP source systems but with customization to non-SAP systems as well.
What are the various limitations of Central Finance and how do we get around them?
Central Finance is still an evolving deployment option and there are some limitations to be aware of. Below are some of these along with some workarounds to consider (based on the current release S/4HANA 1709 FPS01):
- Source systems – SAP ECC 600 or later versions are supported. For the below SAP releases (4.6C, 4.7 and ECC 500) or non-SAP-systems, customer specific implementations and workarounds are needed (snote 2323494).
- Central Payments and clearing – This functionality has been “Released with restrictions”. It can be activated at company code level and RX/PX can be paid in CFIN with no replication back to source systems. It needs to be requested from SAP for usage by raising incident at component FI-CF-APR. This works for SAP source systems, however for third parties’ sources, a correction note needs to be applied. The key limitations being that after technical clearing the reports in source systems don’t indicate the actual clearing status and it needs to be reported in CFIN, Central down payments with SD/MM is not supported etc.
- Taxation – tax reporting is evolving in CFIN and there are still restrictions in a majority of the tax reports, a key example being missing Business Partner details for certain SD invoices in Advance Return for Tax on Sales/Purchases report. Our guidance is that special attention and thorough testing is required over and above ECC and S/4HANA when used in CFIN. For details, refer to the link here.
- Projects System- Projects and WBS master data and postings can now be replicated in CFIN via ALE. There are some exclusions such as networks, replication of changes in master data and customer enhancements, planning, budgeting etc. not supported. We recommend carrying out any master data changes in the source system. Also, planning, budgeting, settlement and results analysis needs to be performed in the source system. For details, refer link here
- PCA – Internal PCA postings can now be replicated to CFIN as FI postings. The deleted data in the source system via 0KE1 doesn’t get replicated to CFIN and can lead to reconciliation issues. We recommend in such a case to delete the data in CFIN directly. For details, refer link here.
- COPA – A replication from costing based CO-PA in a ERP source system into an account based CO-PA in the Central Finance is not supported out-of-the Box. Central Finance does not directly replicate CO-PA documents from the source system. With S/4 HANA 1709, a tool set is provided to facilitate this that includes CO-PA mapping tool and enhanced replication features. During CO document replication, CO document line items posted on reconciliation objects will be replaced by PA segments in central system. PA segment is generated from characteristics brought from source system. In SAP standard, only FI document lines items which are posted to a profitability segment in the source system will be posted to a profitability segment in the target system. Other postings that do not have a profitability segment, for example goods issues, related attributes, such as sales order and sales area, will not be available in the replicated posting and will also be missing from the resulting ACDOCA line items. In such scenarios source system BAdIs can be used to populate the source postings without profitability segments in the target system by deriving the Customer & Product relevant characteristics. With S/4 HANA 1709, cost centre assessment to CO-PA, Sales order settlement to CO-PA and Activity allocation to CO-PA is supported for replication into Central Finance. However Top-down distribution, which is an internal CO-PA document is not yet supported. (snotes 2184567, 2320298).
- Document splitting – Document splitting can be activated and handle basic splitting rules in the target system. This is not fully supported by SAP yet and it depends on the customer-specific requirements to how much extent it can work. We recommend to factor in custom developments in the project plan to make it work. (snote 2184567).
- Asset accounting – the asset accounting subledger replication is not supported yet in CFIN; replication in CFIN is only general ledger postings currently (as at 1709). However, a good workaround to consider here moving the operative process from source to target; direct capitalisation could move to CFIN; indirect capitalisation project settlement and capitalisation could be stopped in the source system and moved instead to target.
- Settlement rules – Settlement rules assigned to cost objects are not replicated but they can be directly created in CFIN.
What are our key lessons learnt from our experience?
We have significant global experience in Central Finance. We thought it would be useful to consolidate some of the major lessons learnt which may help you in your consideration of CFIN.
- Interest in CFIN is growing in the market; proof of concept (PoC) initiatives can help to test key scenarios. Some organisations are choosing to undertake a proof of concept (PoC) to help develop the business case and the programme plan and roadmap for the full implementation. This also allows the validation of key data mappings and the target reporting model.
- Agile methodology of project delivery works well for delivering a CFIN PoC. This generally requires a small but highly experienced team. Basis and ABAP resources also have critical roles to play in the PoC project.
- Understand the potential source system impact. There is an impact on source systems too; the application of a, sometimes substantial, number of OSS notes (see OSS note 2323494). We need to start this activity early and factor in the basis effort and the regression testing effort in source systems.
- Understand the target reporting and data model. CFIN allows us to be creative in mapping and transform data to the target data model. User fields can be extended in source system and then replicated in target or they can be directly added in the target system as well. There will be requirement for development of functional modules in order to deliver this (see OSS note 2543899).
- Understand how to connect the non-SAP source systems. There are no standard connectors available for initial load and online replication for postings of non-SAP source systems. This will require customisation in both source and target system. There are external organizations (like Magnitude) that provide connectors to do this, which should be considered.
- Powerful BAdIs inbuilt in CFIN can handle complexity. Dedicated BAdIs are available in CFIN to handle complex mapping actions, CO secondary postings etc. For example, I the case of CO-PA data replication, they can help enrich source postings without profitability segments, to populate characteristics in the target system through derivation of Customer and Product relevant characteristics.
- Understand the initial load process; it is quite different from online replication. We need to ensure that the batch jobs are optimised to avoid any performance issues (especially during CO secondary postings load). There might be requirement of SAP support for it as well. The recommendation is to have factor additional time for this activity in the project plan for it. In some scenarios, Initial load can be considered as an opportunity for data migration with lesser efforts (no LSMWs). The best practice is to do it in 3 steps:
- Copy of production system with test CFIN system
- Live production system with test CFIN system
- Live production system with live CFIN system
- Evaluate the need for full license of MDG. Full license of MDG provide the benefits in case there is huge master data which is dispersed in multiple complex ERPs. Based on rules set in configuration, system can come with proposals for master data mapping to prevent duplicity in target system. System can also push master data from target system to source systems. Implementation timelines for it need to be factored in the project plan.
- Understand the effort in replication of PCA, PA, PS postings. There are still limitations for replication of PCA, PA and PS postings and there might be requirement for custom programs.
- Understand the effort for switching on document splitting. Document splitting can be switched on the target system but depending on the customer-specific side-conditions, there might be requirement of custom developments.
In the final blog of the series we will discuss Central Finance as part of the overall S/4HANA roadmap.
For more information, get in touch with Capgemini CFIN experts today!