Enterprise Architecture Management: Why your ERP selection starts with architecture — not software

Mythbuster, eight red flags, a five-question self-test — and why documented end-to-end processes and train-the-trainer programs are essential for SMEs before choosing any system.
Based on over 1,200 projects in the DACH region (Germany, Austria, Switzerland).

From our project experience
A kitchen manufacturer in the DACH region, with around 2,000 employees, came to our workshops with a strategically critical order: 1,500 kitchens for a major international contract project. High order value, tight timeline, complex intercontinental supply chain. Essentially, the kind of business the organization should be mastering. Our first question in the workshop seemed trivial: "How exactly does this contract move through your company?"

The answer the team thought they knew didn't hold up to reality. Sales worked flawlessly in the ERP system. Engineering maintained bills of materials in Excel. Purchasing sometimes created new material entries. Production worked with incomplete and delayed information. Logistics planned containers without consistent transparency. The same data was entered multiple times—never consistently.

The crucial question in the workshop was: "Who actually manages this project from start to finish?" The answer was sobering: no one. Every department optimized its own area; the overall project had no owner. With a project of this magnitude, a structural problem was ruthlessly exposed: four different process variants depending on the region and team, massive breaks in communication between ERP, Excel, and email, a lack of integration of critical areas such as equipment procurement, and high dependencies in the international supply chain.

We deliberately chose to avoid knee-jerk reactions: "We understand the process first—then the system." The entire project order was mapped end-to-end together, from order entry to delivery in the Far East. Only then was the ERP question re-examined. In the end, the key wasn't the system itself. The key was the consistent, standardized process—and a clearly mandated process owner.

Three statements you often hear in the middle class — that are simply not true

In over a thousand projects involving small and medium-sized enterprises (SMEs), we repeatedly encounter the same three statements. Each one sounds reasonable. Each one is wrong.

Myth 1: "EAM is corporate overhead. Too heavy for medium-sized businesses."

Corporate enterprise architecture management (EAM) with its hundreds of pages of architectural models is simply not the right approach for a company with 200 employees. However, EAM for medium-sized businesses can be condensed to just a few pages: a Business Model Canvas, a concise capabilities list, and a clearly documented target state of the processes. What is formalistic in a corporation becomes pragmatic in a medium-sized company—if you know what you can leave out.

Myth 2: "ERP selection is an IT decision."

ERP touches sales, finance, production, service, HR, and reporting. Leaving the project solely within IT management risks turning a strategic business model decision into a mere requirements specification. Dr. Harald Dreher puts it this way: “Trust is the heart of every transformation. We don't just advise; we listen, ask questions, and walk the path together.”

Myth 3: "AI tools can now handle EAM themselves."

We ourselves rely heavily on AI—our SCOReX® model analyzes patterns from thousands of projects and prioritizes risks before any selection. But: AI recognizes patterns; it doesn't assume responsibility. An ERP investment with double-digit leverage belongs in the hands of those who are accountable, who ask questions, and who can challenge decisions. That's precisely why SCOReX complements our experienced consultants—it doesn't replace them.

How we actually work — five steps in which architecture is translated into selection

Based on the EAM preliminary analysis, we derive an ERP selection that starts with processes, not functions. Validated in over 1,200 projects.

1. Business model anchoring. Business Model Canvas as a strategic anchor — nine fields, one page, adopted jointly with management.

2. Process mapping and target architecture. Document current processes in a structured manner, derive target processes from the business model — not from software demos.

3. Capabilities mapping. It separates differentiating factors from commodities. This clarifies what the system absolutely must support and what can be standard.

4. Vendor shortlist based on architecture fit. Not "Which system is the best?" , but "Which system most robustly serves our target architecture?" That's exactly what SCOReX® is designed for.

5. Requirements specification as an architectural document. Not a list of functions, but an architectural specification. Protects you in contract negotiations.

In our projects, this lead time reduces the project duration by up to 30% — because there is less need for revisions, less replanning, and less escalation.

The invisible layer: End-to-end processes before system selection

The most well-known weakness of ERP projects in medium-sized businesses is not the system itself. It is the belief that the process is already documented — when in fact it only exists in fragments.

From over 1,200 projects, we know that if you ask three departments in a medium-sized company how a central order is processed, you'll get three different answers. Four are not uncommon. In the kitchen project described at the beginning, after several weeks of analysis, there were a total of six variations—four main processes, two of which could be quickly eliminated. Only the end-to-end analysis revealed what had previously only escalated: media breaks between ERP, Excel, and email; unclear handovers between sales, design, purchasing, production, and logistics; a lack of integration in appliance procurement; and unclear approval processes for letters of credit and customs.

Why classic BPM alone is insufficient for medium-sized businesses

The traditional business process management approach—capturing, simplifying, and standardizing end-to-end processes—is methodologically sound. However, it encounters a specific limitation in medium-sized businesses: Organizations often lack the expertise to document processes effectively enough to inform ERP selection. Simply using a standard, off-the-shelf toolkit generates unnecessary modeling effort without providing any benefit to the decision-making process.

Our assessment: Selecting an ERP system based on undocumented processes is a gamble — and the house pays.

Train-the-Trainer — how we empower your organization

We don't come to document your processes and then leave. Process ownership must remain within the company. Based on over 100 projects for medium-sized businesses and more than 20 years of experience, we have developed a methodological toolkit that is tailored to your specific context—not a standard method.

It works in three stages:

1. Leading by example. Our consultants model the first critical end-to-end processes together with your process experts — visibly, comprehensibly, with a clear methodology.
2. Take over with coaching. Your process owners model the process themselves; we provide support and correction. The methodology is continuously adapted to your specific situation.
3. Manage independently. Your team records and maintains the ERP-relevant end-to-end processes themselves. We check the quality and ensure that the processes crucial for selection are complete and consistent.

In our experience, this is the only way to permanently establish process ownership—and simultaneously base selection decisions on reliable data. However, the real success factor in such projects is not the toolkit. It is the trust of the process managers . This trust only develops when every practical project question is met with a practical project answer—quickly, without theoretical detours, and with a clear focus on what makes the difference in the specific case.

The one lever that almost everyone overlooks: a Process Owner

In the kitchen project, the most honest answer from management was: Nobody is managing this order end-to-end. That's the typical truth of medium-sized businesses. Only the introduction of a clear process owner—with a mandate spanning sales, design, purchasing, production, customs, and logistics—stabilized the project. The four process variants in different regions were consolidated into a standard process that considers all stakeholders. Bill of materials and variant logic, approval processes including letters of credit and customs clearance, as well as international container logistics, were also standardized.

The effect was measurable in the very next project: significantly less coordination effort, a consistent data basis across all areas, more stable production and shipping planning, and clearer communication with international partners. Only on this basis did we have the prerequisite for a well-founded ERP decision. This is precisely what SCOReX is designed for: It systematically correlates your documented end-to-end processes with vendor architectures—not the other way around.

Eight red flags that will derail your ERP project

From our projects, we have distilled a diagnostic checklist. If three or more points apply to your current project, it's high time to address the architecture and process issues.

Our assessment: Each of these warning signs can be remedied — but only as long as the investment has not yet been approved. Afterwards, corrections will be expensive.

EAM in your industry — five key areas from our experience

Architecture is not the same in every industry. The following key areas come from sectors where we support a particularly large number of projects for medium-sized businesses.

Medical technology

Audit paths, UDI traceability, and MDR compliance are not add-ons—they are architectural requirements from day one. Those who only discover this in the requirements specification will pay twice as much later: in additional modules and in lost time during certification.

Pharmaceuticals & Life Sciences

Validation strategy and GMP compliance should precede system selection, not follow it. A target architecture that considers validation as a cross-cutting issue saves months in implementation—and protects against audit surprises.

Professional services

Here, the architectural question is rarely "which PSA tool ," but rather "how do we define profitable service ?" Utilization, time tracking, and project margins must converge in a single data logic—not three.

Plant and mechanical engineering

ETO, CTO, and MTO mix, variant management, multi-level bills of materials. Standard ERP modules are rarely sufficient here. The architectural question determines whether the configurator, ERP, and PLM even speak a common language.

Delivery & Production

S&OP, supply chain resilience, geopolitical disruption: The architecture must integrate supplier diversification, forecasting, and inventory management into a coherent logic. Those who fail to do so are modeling yesterday's supply chain.

That's exactly what SCOReX® is designed for: It systematically correlates your processes with industry patterns, market trends and risks — before you commit to a system.

Why Dreher Consulting — and not the next AI tool or the next big consulting firm?

Three things specifically distinguish us from both alternatives.

First: 100% vendor-independent. No commission, no partner level, no implementation margin hidden behind the scenes. If we recommend a system to you, it's only because it fits your architecture—not because it increases our margin.

Secondly: 33+ years of consulting experience and over 1,200 projects. A database that no AI tool can match without human input. It is precisely from this database that we developed SCOReX® —as a tool, not a replacement. AI recognizes patterns. Consultants bear responsibility. We consciously combine both.

Thirdly: Ethical AI use as an engineering decision. This is not just marketing hype. AI recommendations are transparently documented, assessed by experienced consultants, and never passed on to management as a black box. A mid-sized company's management must be able to trace the origin of an architectural recommendation—otherwise, they cannot be held accountable.

In addition, there is the simple advantage of being a medium-sized company: you speak directly with Dr. Dreher, not with a junior consultant in the third sub-team.

Next step: 30 minutes directly with Dr. Dreher

In a 30-minute initial consultation, we'll clarify with you what decision you're actually making—and what risks exist—before any time or capital is invested. You'll leave the conversation with a clear next step, not a sales pitch.

Dr. Harald Dreher Managing Director, Dreher Consulting · 33+ years of consulting experience in DACH SMEs · 1,200+ successfully completed ERP and digitalization projects · 100% vendor-independent · Directly available for management and supervisory boards in an initial consultation.  Book an appointment directly with Dr. Dreher →