What is hidden behind the new paradigm in the system architecture

Note: We have used commission links in this article and marked them with “*”. If an order is placed via these links, t3n.de receives a commission.

The ERP is central for companies, but due to its nature and the cumbersome integration it becomes a brake on innovation. A data layer should provide a remedy. But what is behind it?

From software for material requirements planning to the core system in the company – enterprise resource planning applications have developed significantly over the past 30 years. A lot of data runs through the ERP and there are more and more interfaces, for example for e-commerce, customer and supplier management and business intelligence. The ERP has moved from the edge to the center.

Usually this is considered an advantage, but is that also true in this case? The numerous failed ERP projects speak against it. Many companies have had bad experiences with ERP systems. The costs get out of hand, the dependency on one provider increases and customization leads to uncontrollable complexity.

In addition, missing or inadequate interfaces are a headache for IT teams. In order for existing applications to continue to run, companies often have to set up improvised data flows with workarounds. And as soon as a system is renewed, the wild ride begins again.

The elephant in the room: ERP applications are a powerful brake on innovation and dynamism in many companies. The integration effort is high, many processes have to be adapted to the system. This often leads to a fatal self-limitation: The ERP system determines which processes are possible at all. Ultimately, a company’s ability to innovate suffers, precisely because the ERP is draining resources and suppressing initiatives – because “our system cannot do that”.

In addition, there are numerous technical restrictions with a braking effect in everyday operations. For example, the product data for the online shop is often taken from the ERP. From a technical perspective, this is a bad solution because ERP systems do not have performance-optimized databases. So it makes sense to save the product data in a fast database system.

However, it is not enough to simply replace the ERP as the core system with another application – for example with the systems for product information management (PIM) that are widespread in e-commerce. A possible solution must be generalizable and transferable for all industries and company sizes. So it is not helpful to replace the ERP with another application, but to change the data architecture.

If you look at the data flows in a company with a central ERP, you will immediately notice one thing: You are making a detour through the ERP, it (also) works as a data hub. However, it fulfills this function only very inadequately. It therefore makes sense to move the ERP back to the edge and limit it to the core task of resource planning.

The previous ERP data hub is being replaced by software optimized for this in a data layer or data hub. In an ideal architecture, there is a hub in the center that accepts or offers all data available in the company via standardized interfaces. Such a hub has control over all data structures in the company, including account data and access rights.

All other systems such as CRM, CMS, online shops, ERP, MES, financial accounting or merchandise management must connect to the hub and then exchange data with each other via the hub API. Internally, the hub uses an optimized digital structure that is independent of the other systems. This means that he can save any data and connect to other IT systems.

This independence allows a smooth transition in the introduction of such a data layer, which is much easier than with numerous other solutions. In the first step, the data hub is set up separately from the rest of the IT. However, it is not a monolithic application, but consists of different components. It basically consists of a powerful database system, an event handler and a system for API management.

Its great flexibility is also shown in the fact that it does not necessarily have to contain all company data. It is particularly advisable to save data in the hub that is frequently and regularly queried and therefore requires performance. For all other data, the hub only needs information about where it is located. It makes sense to build the entire data layer in the cloud. In this way, it can offer maximum availability and reliability.

Once set up, the hub can be optimized with game data independently of other applications in test mode. As soon as it works satisfactorily, the development team gradually connects all applications to it. During this migration phase, the old and new architecture work in parallel. This makes it possible to pull in the data layer during ongoing operations in the company. The old infrastructure will only be switched off when the entire migration project has been completed.

Switching to a data layer gives companies an important advantage for the further development of IT: Individual applications can be exchanged more easily, as they are only integrated in the hub. This makes migration projects easier.

A data hub also avoids a large part of the usual integration costs. Typically, a new application has to be integrated into four or five other solutions. For example, a new shop in most companies has to be integrated into the ERP, CRM, financial accounting and payment provider. This takes a certain amount of time, has high costs and is dependent on the availability of suitable developers. With the data hub, the integration effort and the associated risk are drastically reduced.

In conclusion, it can be said that an independent data layer brings companies cost advantages and greater IT efficiency. New digital technologies are available more quickly and integrated into the existing IT landscape. This means that companies – not only in e-commerce – are technically positioned for the future and can react more flexibly to changes, both in the business model and in new technologies.

Christopher Möhle

Christopher Möhle is CEO at Turbine Kreuzberg. He supports companies in the development of digital product and service platforms as well as individual web applications. His focus is on technology and implementation.