SAP ERP as data source for Microsoft BI (Part 1)
A few weeks ago Hilmar Buchta wrote a fantastic blog post about importing data from SAP into SQL Server and explained the principles of SAP DeltaQ and Xtract IS from Theobald Software. http://blog.oraylis.de/2015/04/importing-data-from-sap-using-the-sap-bw-delta-queue-and-theobald-software-xtract-is/ Today I would like to start a series of posts showing real world experience and best practices from one of our BI projects with SAP ERP Retail as major data source.
Scenario Just like thousands of companies worldwide our customer uses SAP ERP for his day to day business like scm, logistics, material management, financials, controlling and is currently implementing SAP POSDM for his retail chain. For one central line of business he currently uses SAP BW reporting. Meanwhile he uses SQL Server 2014 already as data platform for analytical applications and central data hub. So, hey, with this wonderful world of SAP applications, why extract data out of SAP and import it into another database for reporting and analysis purpose?
Single Point of Truth The Data Warehouse as Single Point of Truth (SPoT) needs more data than just from your ERP-System. There are additional data sources like external market data, data from suppliers, data from web shops, machine sensor data and additional internal applications that needs to by analysed altogether. And from our point of view all data sources should be modeled in one common data model in a way, so that the business users are able to easily navigate and find relevant information. This goal is best reached by dimensional modeling. You may want to take a look at resources from Kimball Group, pioneer of dimensional modeling, for this technique: http://www.kimballgroup.com/category/dimensional-modeling-fundamentals/
Benefits But there are even more reasons for a common Data Warehouse Layer implemented on a proven and high performing data platform like Microsofts SQL Server 2014. Just to name a few:
- Performance. State-of-the-art techniques like columnstore index (xVelocity) speed up typical analytical loads.
- Reduced Cost. Popular tools like Integration Services (ETL), Reporting Services, Analysis Services (Cubes) included for a very competitive price.
- End User Access. Integration with user-friendly office tools, e.g. Excel with Pivot Tables/Charts for decision-makers and analysts.
- Usability. Modern self service BI tools like PowerBI that are easy to use and integrated with mobile and cloud environments.
- Choice Of Hardware. No Need for expensive proprietary hardware appliances. Even 10s of Terabyte are no problem (Fasttrack architecture).
To keep it short, our customer is very much impressed how quick and easy he can access relevant business information from the new Data Warehouse on Microsoft BI platform. So stay tuned! I'll share more information on this topic within the next weeks.