The ERP system is the holy grail of German corporate IT. For decades, the promise has been the same: one system that does everything — finance, procurement, production, HR, controlling. If you had SAP, you had won. Everyone else was working toward having SAP someday. This logic has done considerable damage — not because ERPs are bad, but because the question was framed incorrectly from the start.
Core processes versus secondary processes
The critical distinction that is too rarely made is between core processes and secondary processes. Core processes are those that directly contribute to competitive advantage — the way a company produces, delivers, invoices, or serves its customers. Secondary processes are necessary but generic: expense reporting, leave management, standard bookkeeping. For core processes, you often need specialized systems. For secondary processes, a standard tool will do. The problem: ERPs treat both the same way.
Integration complexity as an underestimated risk
Trying to force everything into one ERP often creates more complexity than it resolves. Customizations cost ten times the license fee. Interfaces become bottlenecks. And when the ERP needs to be replaced after five years, the company is so deeply embedded that a migration is practically impossible. The alternative — a thoughtfully assembled system landscape of best-of-breed tools connected by a clean integration layer — requires more architectural work upfront, but pays off in the long run.
Does this sound familiar?
If you are weighing whether one system really has to do everything, let's look together in a free first conversation at which processes belong together and which are better kept apart.
The Company OS concept
We like to think of system landscapes as a Company OS: an operating system consisting of a central data core, around which specialized applications dock. The data core — today often a modern data warehouse like Snowflake — holds all data together regardless of which system it originates from. The application layer can then change without the core being touched. In this model, the ERP is not the center — it is one of several applications. Thinking about your system landscape this way leads to better decisions: about build vs. buy, about standardization vs. specialization, and about the true costs of IT projects.
The ERP is not the center of the company — the data is. Understanding this leads to better system decisions.