Written by Peter M. Färbinger, E-3 Magazine, Data Migration International

Thomas Failer, Founder and Group CEO of Data Migration International 

The new conversion will be long-lasting as it will affect all ERP data – consequently, this “data conversion” extends far beyond the boundaries of the SAP community. SAP will not be able to reach this milestone alone.

Data lakes are failed attempts of the past. Similar to the metaverse of Facebook founder Mark Zuckerberg, it will create a new metaroom for data. Analysts from Gartner have already mapped out the route with the terms “data fabric” and “composability”. Data Migration International, DMI, from Switzerland is on the right track with JiVS as a unique tool for managing this data conversion.

“The success of digital companies depends to a great extent on the modernization of their IT,” suggests DMI founder and CEO Thomas Failer at the start of his interview with E-3 Magazine. “It depends in particular on the clean separation of the different layers in the technology stack, starting with the hardware level through to the display and usage level, also known as the ‘experience layer’. Only when the layers are cleanly separated can they be managed and operated independently of each other. This results in far greater flexibility. Silos are removed, while the individual parts of the different layers can be better combined together and reused. In my opinion, this separation is an essential requirement, and Gartner refers to it as a ‘composable business architecture’. This means a company is able at any time to flexibly adapt its resources to changing conditions and recompose them.”

SAP logically recognized the need for digital transformation years ago. But due to its R/3 past, the world market leader of ERP systems was too focused on algorithms – business processes programmed in ABAP. SAP recognized that a new ERP also needs a new innovative data platform. This resulted in the development of the in-memory database Hana. SAP attached considerable importance to streamlining the system itself: New algorithms were programmed and numerous ABAP tables were deleted. Too little attention was given to composability, and Gartner’s “data fabric” concept was poorly customized with SAP Data Hub. The sensible separation, orchestration, and harmonization of layers in the technology stack was thwarted by the unfortunate tying of S/4 and Hana.

Little importance was attached to the ERP data itself: the Data Hub project was poorly set up and failed. Feeling frustrated, many experts left the ERP provider. With Hana, SAP has created a very interesting data platform for the operational management of ERP data, but this concept doesn’t go far enough. In the SAP universe, there’s no holistic view of business data in a B2B2C environment. In this regard, SAP partner DMI has a wider horizon, great foresight, and therefore a considerably better concept for the upcoming data conversion.

“The opportunities outweigh the risks,” says Thomas Failer referring to the upcoming transformation for existing SAP customers. “The transformation to SAP S/4 Hana is a unique opportunity for companies to fundamentally modernize their IT, by introducing the layer model we spoke of that cleanly separates the individual layers. The standard here is the cloud, namely from the infrastructure through to the operation and management of cloud-native applications and services that manage rather specific tasks and can be continually recombined with each other and reused. It’s this new way of providing, operating, and leveraging IT that distinguishes the cloud. It should be considered independently of the issue regarding where this modern type of IT will be implemented and operated. As this can very well be in a company’s own computer center. In this respect, SAP with its new software generation is actually a cloud company, regardless of whether or not the investors and shareholders have already internalized this.”

Data sovereignty is key

The added value of each IT system – ERP, SCM, HCM, CRM – comes from the data, which is why the data conversion is also the focus of all further considerations. What is DMI therefore planning over the coming months to give companies better support with this transformation?

Thomas Failer: “Simplification, automation, integration – these are the most important points on our roadmap. We’ll soon be making several announcements on upcoming innovations based on our product JiVS IMP, which we’ll present in detail in one of the upcoming issues of E-3 Magazine. This will enable existing SAP customers to transition more easily and quickly into the new SAP world, and allow them to accelerate and optimize their digital transformation.”

With modernization comes a unique opportunity to stock take and clean up. The current SAP systems have been running for 20 years or more, and were continually updated, migrated, and enhanced. “Imagine how many company codes, plants, business objects, and data will no longer need to be processed and modified in a future S/4 environment,” DMI CEO Failer points out. “Based on our assumptions and our experience with customers, we’re usually talking here about 90 percent of data and 50 percent of business objects that – unlike before – can be managed outside of SAP. Even Hercules couldn’t have done a better job clearing out King Augeas’ stables! But what’s essential here is the distinction between processing and using data. After all, historical information will of course still be needed together with its business context.”

Holistic data view

Some data must no longer and may no longer be changed. However, this data is still of value, and even more so considering the analysis opportunities in a fully digitized company. “First by clearly separating the layers as I mentioned, existing SAP customers can fully exploit their information and findings obtained through their analyses,” explains Thomas Failer and adds: “Everyone is aware of the value of data. But show me a Big Data project or analytics project that has succeeded in extracting all the data – also and especially including the historical data – from silos and archives and harnessed its potential to create value. There’s a lot of talking about it, but few managers who are actually consistently and decisively planning and taking action in this direction.”

A survey conducted by IG SAP in Switzerland is quite insightful: CIOs know that data is important but only 5% are actually paying any attention to it. For Peter Hartmann, head of the study at IG SAP Switzerland, it’s a fact that the approach for RISE with SAP with regard to a holistic view of business processes, data, and technologies is already established in many companies’ IT/SAP departments. Here, an integrated synergy with SAP can appear positive.

However, DMI CEO Failer has observed that many Big Data projects, perhaps even most of them, are silted if – with the exception of some use cases such as predictive maintenance – they largely take place outside of the ERP landscape. “They usually resulted in projects where data lakes were set up,” summarizes Failer and clarifies: “But as a central data repository doesn’t benefit from a business perspective alone or only marginally benefits, this is very rarely shouted from the rooftops.”

It’s crystal clear from DMI’s perspective: data, whether structured or unstructured, requires a context in order to have meaning. Only then is it possible to decide whether this meaning is relevant to the business or not. Thomas Failer adds: “Depending on the scale of the meaning, the success of every type of Big Data project or analytics project depends on whether the data managers are on board from the start. This includes data scientists and expert users who generate and process this data. And naturally, the CIOs need to be sponsors of these projects and take overall responsibility. In other words: At an organizational level, these projects need to be considered, planned, and implemented holistically. This is all the more important as project teams and data managers tend to think in silos.”

Global data transformation

Transformation projects in greenfield, brownfield, or bluefield style have come to a halt. The following situation has emerged: Existing SAP customers are looking for personal and specialist resources that are lacking in the market. They want solutions that will allow them to scale back their legacy systems as much as possible rather than decommissioning them completely. They’re looking for methods to largely transform the historical data and migrate it to the new system world, although they still need unrestricted access to it along with its business context. “And they limit the projects alone to the SAP data, rather than also taking the historical data from non-SAP systems into account from the outset,” warns DMI founder Thomas Failer.

If existing SAP customers assume that the context of the data has grown together with the application level, there’s no risk. “A separate data layer for historical information should be similar to data fabric. By this I mean a type of skillfully woven fabric that clings perfectly to the wearer’s body. The different threads of the fabric can come loose again and bind back together if the figure changes. The strength of a separate data layer is its fitting precision and customizability, and therefore is the opposite of a Gordian knot. In other words, there’s a separate data layer that resembles a fabric, not only consisting of the data itself but also of its context,” Thomas Failer defines the Gartner-coined term “data fabric” for the SAP community.

“In fact, integration is an important aspect, but not the only one,” Failer describes the current situation. “After all, to make data accessible across the company for analyses and broad user groups, the systems that host the data must be interconnected. And the number of connections grows exponentially in relation to the number of systems to be connected. This is why, as far as effort is concerned, the number of systems that need to be integrated with each other certainly plays a role in this. In other words: Yes, integration capability is important, but an equally decisive factor is reducing the number of systems that are to be integrated.”

For legal reasons, ERP users cannot and may not modify the data for defined periods of time. The reason why most existing SAP customers hold onto their legacy systems is because of the business context. If data has no business context, neither financial authorities nor auditors will consider it to be evidential. And whether this concerns SAP or non-SAP data, in order to harness its value in the new environment as well, the historical data needs to be transformed. Thomas Failer emphasizes: “Compliance and the business value of historical data – these are the very reasons behind the time-consuming, cost-intensive transformation projects that delay innovation, if they aren’t essentially approached any differently than before and if the overall IT isn’t modernized.” 

“Neither data lakes nor integrations alone will solve the fundamental problem,” summarizes Failer. “I think this is one of the main reasons that SAP managers and CIOs are only now recognizing the business case for an S/4 transformation.” A separate data layer is indeed the solution if it is designed in such a way that enables the separation of the rigid connection between the historical information and the legacy applications and systems. “Earlier we referred to the Gordian knot. This kind of data layer has to be struck through, just like Alexander the Great did when he knew the knot was impossible to untangle. This is why he chose a completely different approach,” explains Thomas Failer. The S/4 transformation could be the ideal pilot project for introducing a new, application-independent yet context-aware data layer.

Context information

Thomas Failer: “Context information is the basis of a separate data layer. A large part of a company’s treasures are held on the databases of ERP systems. But it’s the context that transforms the raw material into jewels and this is located at application level. At the same time, many transformation project planners consider this treasure to be ballast rather than a propellant and as a cost-driver when transitioning to S/4, which to make matters worse also draws the project out.”

DMI are convinced: The benefits only materialize once we separate the historical data and its business context from the legacy systems. Thomas Failer: “By moving to the context-aware data layer, not only does the historical data remain accessible, and by this I mean all structured and unstructured information. The same applies to its context, which doesn’t remain in the source systems as a passive asset, but instead is activated for use. The possibilities and advantages stemming from this are huge.”

The transformation to SAP S/4 is only one scenario where JiVS IMP provides support as the core of this context-aware data layer. “Just think about the many other different use cases where data needs to be exchanged quickly and accurately between systems and applications,” says Failer, referring to the many possibilities for using JiVS IMP. This includes, in particular, Big Data, analytics, and IoT, as well as the acquisition and sale of companies and parts of companies, the consolidation and harmonization of computing centers, and application and system landscapes.

“With all these topics, historical information plays a decisive role outside of its source systems. That’s my core message for the specialist departments and the CFO,” emphasizes Thomas Failer who goes on to explain: “Specialist departments, the CIO and CFO initially take a critical view of separating operational and historical data. They tell us they want to have the entire data inventory in the new system as well, with all the negative consequences that we’ve already discussed. This is a catch-22 situation. In order to be agile, fast, and be able to make changes and optimize processes without a great deal of effort, the historical information needs to be literally stored and managed outside of both the old and the new SAP landscape. This is why I’m absolutely convinced: Once the users have understood that, despite this separation, they can access all the information in the same or even a higher quality in the back end, then there’s nothing standing in the way of an SAP S/4 transformation based on our approach and with the support of JiVS IMP.”

About the author

Peter M. Färbinger, E-3 Magazine

Peter Färbinger, Publisher & Chief Editor of E-3 Magazine
B4Bmedia.net AG, Freilassing, Germany.
Contact: pmf@b4bmedia.net | Tel.: +49 (0)8654 77130-21