Migrate from SAP BO to Looker automatically, at a fixed price

 

Migrate from SAP BO  to Looker , Automatically, Package Ellipsys supported the ADEO / Leroy Merlin Group in the massive simplification of its SAP BO platform then in the migration of useful assets to Looker (and Power BI) by automating the process, on a fixed price basis. We share the main aspects with you.     SAP BO  : an uncertain roadmap SAP Business Objet holds more than 5% market share in the BI market with almost 30,000 customers worldwide (6Sense).   However, the platform is navigating in a certain vagueness: Version 4.2 will reach the end of “priority 1” support at the end of 2024. Version 4.3 is expected to be supported until the end of 2027, with standard maintenance ending at the end of 2025.  SAP plans to release the first version of SAP Business Objects BI 2025 in the fourth quarter of 2024. The outlines are not very well known. But what is clarified is that certain components will not be included in the BI 2025 version, including Universe Design Tool (for the construction of UNVs), multi-sourced universes, etc.   Generally speaking, we can consider that SAP's stated desire is to see its customers move to SAP Analytics Cloud (SAC). However, this ambition does not seem to be completely shared.  For what ? :  “Self BI” is popular with business teams, but also with IT, which, willingly or by force, lets the business down. And SAP BO is precisely not a “self BI” oriented solution. This is a very good solution, but perceived as "dated".      Looker, strong growth! Looker, for its part, is doing well, taking advantage of Google's firepower, Looker's very affordable prices and the popularity of BigQuery, Google's Cloud database. Looker is “Cloud native”, easy to use, connected to GitHub for versioning and collaborative work, etc.  Looker's market share is still small, but growth rates are significant.   Migrate from  SAP BO  to  Looker? For our part, we perceive a significant movement from SAP BO towards Looker, in particular given the possibility offered to maintain a semantic layer.  But for those who want to get rid of SAP BO, a monumental project is looming: some companies we work with have more than tens of thousands of BO dashboards in production. If the project must be conducted “paper and pencil”, the bill can be steep. With a more than uncertain result.   Automate  the process! With this in mind, Ellipsys has created an automated migration solution between SAP BO and Looker which is based on its {openAudit} software. This is the {oA.Reverse} feature.   This functionality was implemented on behalf of the ADEO Leroy Merlin Group to reduce the SAP platform which had more than 100,000 dashboards to 8,000, then to migrate it to Looker (and Power BI).   Below, the  6 salient points  of our methodology within the framework of this type of project. We drive them on a fixed price basis.   1.  Migrate  what really needs to be migrated   Why migrate tens of thousands of SAP BO dashboards, when in reality a tiny minority is actually used? Often more than 90% of the SAP BO platform in source has no or no longer use.   {openAudit}, our software, makes it possible to significantly reduce the scope of SAP BO dashboards to be migrated, by allowing massive archives prior to migration. SAP BO dashboards that are obsolete or contain errors are detected. {openAudit} allows them to be archived via the creation of unit BIARs (BO archive format). They may eventually be permanently deleted later. However, we historicize the intelligence of BO (on its different versions). So nothing is lost.    2.  A complete  “reverse engineering”  of the BO platform to be migrated   SAP BO dashboards are often extremely complex given the very long history of the platform.  This complexity must be understood, at the level of the intelligence of the dashboard but also at the level of the sources. The idea will be to collect everything that can be collected with our probes and parsers to capture all the intelligence underlying the SAP BO platform.   3.  A migration of the  “semantic layer”. The semantic layer is an abstraction layer which makes it possible to connect to technical considerations of a database of "business" terms, to allow teams to build dashboards in complete autonomy.    BO is a proven and robust technology which - with its (very) long experience of its semantic layer - has been able to organize information in companies.  The idea is therefore to allow teams to quickly start from an optimized semantic layer in the target technology, to be able to independently build the data visualization layer of tomorrow: to be able to move towards "self BI". Our migration engine {openAudit} will divide the BO Universes (Business Objects) according to each context and fact table actually used in the BO reports used. Our migration engine is based on real use of the source BO platform.  Each BO Universe will be transformed into several Looker Explores based on a star model centered on the fact table. {openAudit} will only include what is used in SAP BO. Each Explore Looker will allow users to create ad hoc queries, explore and visualize data. 4.  Adjust Semantic Layer Migration  to make it more usable by the professions  To make life easier for the profession, it seemed necessary to add a device to our migration engine.  We have enriched our {oA-Reverse} migration engine by dynamically providing a “control table”. It allows you to simply configure the output, to move towards an ideal semantic layer, potentially very close to that of SAP BO.  The first run of our migration engine feeds the control table and produces LookML (Looker modeling language).  The processing is thus “controllable”, which allows the profession to “play” to modify the Explores as it sees fit.  The LookML is versioned.  5.  Lighten  the target platform SAP BO dashboards have accumulated more and more complexity over time.  This proliferation can have real repercussions on the target platform, because there are as many refreshes as there are users. {openAudit} can potentially "move" complex queries into tables and views of the source database or into Looker dashboards.    These complex queries will thus potentially only be played once to reduce costs, facilitate maintenance and speed up consultations.   6.  Validate  non-regression {openAudit} will run BO dashboards and Looker dashboards (after technical migration) with similar prompt values, which will allow mass comparisons with statistical tools.     Conclusion   The decommissioning of SAP is becoming a major challenge for many companies these days: the tool is perceived as being at the end of its life, poorly suited to Cloud use and the growing desire to move towards “self BI” .   Except that these migrations are not easy to implement as these platforms have lasted: they have therefore accumulated complexity. The {openAudit} migration engine makes it possible to automate the transcription from SAP BO to Looker by focusing on analyzing all the subtleties of the source in order to be able to faithfully reproduce it in the target. An interesting option is to start from the semantic layer to create the conditions for a turnkey “self BI”.    {openAudit}'s ability to optimize the BO platform and then factor the intelligence into the target platform allows for a considerable increase in intelligibility and performance. The massification of non-regression tests is valuable to arrive very quickly at an optimum, to make this migration a real success for IT and the business. This ambitious project was carried out on a fixed-price basis on behalf of the ADEO/Leroy Merlin Group with extremely convincing results.