Business Intelligence by and for SAP – what was, what is and what is yet to come

It has become incredibly easy to integrate relevant business information from SAP into a Business Intelligence tool with just a few simple steps. Read on to learn more!

As of Netweaver 7.4 and OData, SAP has been providing simple endpoints for BI evaluations. You will also learn which additional benefits a HANA database offers for your application.

In the introduction to my last article, I recounted a short history of SAP and third-party business intelligence solutions used. The insight that meanwhile more and more SAP users are attracted to Microsoft front end solutions is still valid. Generally, SAP-related business intelligence has really picked up speed and does not slow down – on the contrary. I have isolated the following three trends for me:

Maybe it is ignorance, lack of knowledge or selective perception, but I have always considered SAP BW a terrifying, monstrous application. Do not get me wrong: I believe that for many companies there were simply no alternatives available on the market at a given (past) time. What was available were similar application monsters at best. However, this had little or nothing to do with agile solutions, self-service applications, data discovery and other current buzz words. My personal forecast is that those who currently still do not use SAP BW will not introduce it in a significant scope. Companies with existing architectures are certainly facing the challenge of how to handle the problem of extensions based on this technology.

My subjective view is: Check new (additional) possibilities and scrutinise the further expansion already now.

SAP has implemented two key topics here. On the one hand, tapping of SAP data via OData can be realized from Netweaver 7.4 onwards. This requires some settings by knowledgeable administrators and possibly the implementation of an own, suitable OData endpoint. However, these are peanuts (also financially) in comparison to the driver software (Theobald, Cubeware…) required in classical data integration.

The second option is the migration of the existing database (SQL, Oracle, MaxDB) to HANA. To avoid any misunderstandings: I do not mean the migration to S/4 but the operation of the existing SAP ERP in HANA. I am, of course, aware that this is a more substantial project which no one will take on for reasons of business intelligence alone. The improved access to HANA, however, would be a great windfall gain from the migration of the database.

Admittedly, this sounds strange coming from me as a supporter of data warehouse solutions. Nevertheless, the current trend is towards the use of in-memory information. Both of the above-mentioned variants of data integration (OData and use of HANA) lend themselves to the use of volatile query results. A good application idea is near real-time information which does not concern a too large table sphere in SAP, has a foreseeable complexity in the query and provide a manageable amount of result values. Although these are some limitations, the relatively large number of sensible applications remains.

The speed refers both to the proximity to real time, but also to the speed in the implementation of possible applications. The simplicity lies in the ability of access. In the simplest case, I need an OData URL from my SAP system and Excel or Microsoft Power BI. The agility refers mainly to the design possibilities of the analytical views on my SAP environment. All in all, real progress that is already on the way.

SAP is becoming more open. The option of using OData is already a first relevant step, which, with HANA, is implemented even more consistently in the openness of the solution. SAP is becoming readable and analysable easily.

Business intelligence solutions by SAP and non-SAP providers are, in fact, on an equal footing. With the new access options via OData and HANA there is no difference anymore in interface protocols for SAP and non-SAP BI products.

Much of what was is certainly still justified and thus also endures – especially the classic data warehouse approach. However, those who will only follow the beaten tracks in the future will increasingly lose efficiency of their solution possibilities, and will not be able to use some newer approaches and extensions such as machine learning or real-time analytics.

This article was also published in german in the e3 magazine, May 2017.