Friday 22 May 2020

SAP Analytics Cloud for Planning : Integration is key!

When starting a project to enable automation in your budgeting and forecasting processes, one of the key topics is integration with the new planning software. In this article, Wouter Vanhoutte (SAP Planning & Analytics Architect at Micropole) will focus on the topics of data integration and security.

 

Data integration

  The first thing you think about is starting with a sound set of actual data, to kick off your planning process. At the same time a process needs to be in place to update the actual data in order to create a new forecast version. The outcome of your planning process will be your final budget and forecast versions, and your ERP system also requires this plan data for actual/plan comparison’s and to leverage additional operational planning capabilities.  In a more integrated planning scenario, the actual data would have to come from various sources and applications, in order to setup an integrated planning process.

 

In a typical on-premise setup where data integration is handled by a datawarehouse and where SAP Business Planning & Consolidation ( BPC ) serves as your planning tool embedded into SAP Business Warehouse (BW), actual data was loaded using the extraction, transformation and loading (ETL) capabilities and datasource concepts of SAP BW.  The datasource concept on itself was already an extensive form of automatic integration, since it was delivering standard programs which transformed the relational data schemas of SAP ECC into a flat datastructure suitable for loading into a datawarehouse. This was already reducing the complexity of setting up an ETL process towards your datawarehouse and hence integrating the data into your planning solution. For sending your final plan version back to SAP ECC, one needed to setup quite a complex retraction process, which required extensive knowledge of ABAP programming.   

 

With the evolution of in-memory databases and cloud solutions, the integration architecture with a solution like SAP Analytics Cloud for planning ( SAC ), is greatly simplified. Where the datawarehouse served as an intermediary to create a flat datastructure suitable for import into SAP BPC, a modern ERP system like S/4HANA is working with a much more simplified datamodel which is suitable for direct integration into SAC. E.g. when a customer purchases S/4HANA cloud, it comes with business content inside SAC for planning and exporting the plan data back to S/4.

 

Another important topic is integration of your plan data into your reporting and analytics solutions. With SAP BPC embedded the integration was happening inside SAP BW.  BW delivers infoproviders as the basic data structures for storing your planning data, on which Bex queries are build to connect to various reporting tools. BPC standard did not integrate that easily into SAP’s portfolio of reporting solutions, since BW infoproviders were automatically generated with technical names, and a MDX engine was used to report on BPC standard data inside the Excel embedded front-end tool, the EPM add-in. Since other reporting solutions of SAP were not using this MDX engine any longer, data needed to be extracted from the BPC cubes into other BW infoproviders.

 

As SAC is a solution which integrates the three pillars ( analytics, planning & predictive ) , integration with analytics is delivered as a standard feature. Let’s take a look at example architectures of on-premise and cloud solutions :

 

 

 

Data integration archirecture of SAP BPC – Source : diagrams.net

 

 

Data integration archirecture of SAC – Source : diagrams.net

 

In the second scenario where SAC is used as a planning tool on top of your S/4HANA system , data can imported directly from S/4HANA or shown live in SAC.  From a technical perspective there is no need to load your data into BW/4HANA in order to start your planning process. However, in almost every scenario a datawarehouse will be in place to integrate various datasources and perform complex data transformations and calculations.  In that case you can integrate your SAC planning models with BW infoproviders, importing the required actuals but still exporting your planning data to S/4HANA.

In the case your SAP ERP is not migrated yet to S/4, no standard export function exist. You will have to go back to the traditional setup where you export your plan data to BPC or another BW infoprovider and retract from here to SAP ECC.  A good thing is that SAP has provided tight integration with BPC, i.e. live connections and write back into BPC embedded, and import/export functionality for BPC standard.

The scenario on how you integrate SAC planning into your landscape, will largely depend on your current architectural setup, but the good news is that SAP is supporting various options for import, export and live connections.

 

Security

A concern of some companies might be the data privacy when connecting with SAC to live sources. Well, in fact your live data will never travel to the SAC servers. Your browser connects to SAC, SAC is sending back the metadata about the story you are launching, and then the browser makes a cross-origin resource sharing ( CORS – required technique to overcome limitation in a web application to access resources from a different origin ) call to your on-premise system to get the data. So the data only travels from your on-premise system to your browser.  Your on-premise needs to be properly configured in order to make this happen.

 

When your users are working in S/4HANA, and then shifting to your planning screens, you want to provide them with a fluent experience, without having them to log on to another system.  In order to have this working smoothly , SAP has foreseen various possibilities to have single sign-on ( SSO ) working together with your S/4 or netweaver systems.

SAC provides Security Assertion Markup Language ( SAML – an open standard for exchanging authentication and authorization data between parties ) to enable SSO, simplifying not only authentication to SAP Analytics Cloud but also to connected data sources from your landscape. It is also possible to leverage your existing authentication mechanism, like Kerberos or client certificate authentication. This is achieved by choosing the “none” authentication option in the SAC connection.

 

Another point of attention is the seperate set of roles in SAC to control access of users to models.  When using the SAML framework for connectivity, one can map user SAML attributes to roles in SAC. This way, users can be assigned SAC roles centrally from within your on-premise system. There is also the option to dynamically have the users created in SAC.

 

Conclusion

The alert student will have noticed that integration is one of the strong points of SAC. Various options exist to integrate your data from transactional and datawarehouse solutions, even beyond the SAP software stack.  Your goal should be having a seamless integration of your planning process with your transactional system. By the various connection and authentication options SAC is delivering, a high degree of integration can be achieved. 

There will always be a sound basis configuration required in your on-premise to enable the integration with your on-premise systems. SAP is also providing SAC planning content for S/4HANA where the data integration with S/4 is included in the content. The following blog is constantly updated with updates on SAC content : https://blogs.sap.com/2016/11/18/sap-businessobjects-cloud-content/