Process Design in the DACH Mittelstand — the TO-BE Discipline, Not the AS-IS Capture

Process design is the methodical discipline of drafting target processes — distinct from analysis (as-is) and implementation (ERP, RPA, workflow engines). The ABPMP BPM CBOK lists it as a dedicated Knowledge Area; ISO 9001:2015 clause 4.4 makes target-process determination mandatory. From over 1,200 Dreher Consulting projects we know the pattern: Bitkom 2025 reports 53 per cent of firms struggle to manage digitalisation, and a large share of designed target processes drift in the line within two years when mandate and integration are missing. We recommend a reduced BPMN-light stack — five notation elements, mandated process owners, volume simulation before customising — instead of corporate Six Sigma overhead.

What is process design? Demarcation from process analysis

Process design is the deliberate shaping of target processes — specifying how a process should run in the future. The ABPMP BPM Common Body of Knowledge places it as a dedicated Knowledge Area in the process lifecycle — Design necessarily follows Analysis and is not a substitute for it.

Three terms used interchangeably must be kept apart:

1. Process design vs. process analysis — Analysis captures the as-is state (bottlenecks, cycle times, interface breaks). Design drafts the to-be state. A "design workshop" producing sticky-note as-is captures is methodically an analysis session.
2. Process design vs. process modelling — Modelling is the notation in BPMN or EPC. Modelling without prior design produces diagrams without architectural decisions.
3. Process design vs. process implementation — Implementation is the build in ERP customising or workflow engines. Design says how; implementation says with what.

The methodological backbone is ISO 9001:2015 clause 4.4, which turns target-process determination into a mandatory QMS clause. Amendment 1:2024 adds the climate-action context but leaves §4.4 untouched.

Why process design matters in the DACH Mittelstand in 2026

First — the ERP landscape is shifting. The DSAG Investment Report 2026 records continued modernisation momentum — around 37 per cent of SAP users plan to move to S/4HANA by the end of 2027. Target-process design is among the decisive preparatory disciplines — yet often started only after the RFP has gone out. The ERP standard then designs the process.

Second — digitalisation pressure is rising. The Bitkom "Digitalisation of the Economy 2025" study (603 firms) reports: 53 per cent struggle to manage digitalisation, 82 per cent describe the current crisis as also a crisis of hesitant digitalisation; data protection and the skills shortage rank among the biggest hurdles.

Third — the drift problem pushes. From over 1,200 projects we know: many Mittelstand firms launch to-be designs, but only a fraction keep them alive without drift in the line after two years. Success factor one is the link to a documented process map plus a mandated process owner — not the choice of suite.

BPMN-Light, EPC and SIPOC — the Mittelstand stack

The OMG standard BPMN 2.0 specifies more than a hundred symbols; the full notation only pays back in much larger organisations. From over 1,200 projects we recommend a reduced five-element stack:

• Task — the executable activity.
• Gateway — decision point (exclusive, parallel, event-based).
• Start and end event — defined trigger and outcome.
• Sequence flow — the temporal ordering.
• Lane — responsibility swim lane per role.

These five cover roughly 80 per cent of Mittelstand processes cleanly. The EPC remains valid in SAP-oriented houses. From over 1,200 projects we see corporate stacks run successfully in 50-to-500-employee firms only rarely. Becker/Kugeler/Rosemann (Springer) is the reference textbook for deeper questions.

Two more tools belong in the stack: SIPOC for one-page scope definition and a value-stream sketch for end-to-end visualisation. Both complement process analysis as as-is groundwork and value stream mapping for the to-be/as-is comparison.

What textbooks leave out — the anti-patterns

1. "The ERP standard will design it for us" is the most expensive reflex

Contrary to the standard vendor sales narrative, we recommend designing target processes before ERP selection — not during customising. From our experience in over 1,200 projects in the DACH Mittelstand we see the sequence inverted repeatedly. The ERP standard then designs the process. In our recent engagements before S/4HANA, Business Central and IFS Cloud go-lives, it was the majority of mandates that ran into budget overruns from this pattern. Companies entering ERP selection with documented design keep the customising budget far better under control.

2. "Process owner pro forma" is mandate-lessness on paper

The ABPMP CBOK names the process owner as the mandate-holder. In the Mittelstand the role is often pro forma — without budget authority, escalation rights or veto power. In our 1,200+ projects: process owners without an extended-management mandate produce target processes that drift in the first quarter after go-live.

3. Simulation before implementation is the exception in the Mittelstand

The VDI 3633 Part 10 guideline describes the methodology for simulating business processes — the German engineering reference for the "digital twin" in process design. In our projects we observe that volume simulation is performed only in the rarest cases. Our take: Running the target process against realistic volumes, seasonal peaks and variant explosion before ERP customising saves noticeably on downstream implementation time. In variant manufacturing and engineer-to-order operations the lever is even larger, because the bottlenecks scale with configuration complexity.

How we set up process design in the Mittelstand (four phases)

In our vendor-neutral ERP and process consulting, process design is — in line with the ABPMP lifecycle logic — the architectural discipline before any system selection:

1. Scope and process owner. Identify three to five end-to-end main processes, name one owner per process from the extended management team, lock scope on a one-page SIPOC.
2. BPMN-light modelling. Train the pilot team on the five-element stack and model one target process per main process across two to four detail levels. Process mining supplies the data-driven as-is comparison.
3. Volume simulation. Stress-test the target process against realistic volumes, seasonal peaks and variant explosion. Bottlenecks become visible before customising starts.
4. Integration with the ERP specification. Include the target process as a mandatory annex in the spec. Downstream improvement runs through the continuous-improvement process, framed by process excellence as the overarching frame.

The difference from corporate offerings: we build no Six Sigma staff function. We deliver the reduced toolkit and the mandate clarity that keeps the design viable beyond go-live — pragmatic instead of corporate overhead.

Practical example: kitchen manufacturer from southern Germany (anonymised)

The kitchen manufacturer called us because management had heard of a substantial budget overrun at a comparable competitor. We agreed the sequence: design first, then specification, then selection.

Three end-to-end main processes were strategic: Order-to-Cash with configurator-driven variant manufacturing, Procure-to-Pay, and Make-to-Order along a highly automated production line. For each, a process owner was named from the extended management team — with budget authority and veto power.

Inside Order-to-Cash, volume simulation surfaced a bottleneck in order clarification that would have doubled lead time at peak. Corrected in design — before customising. In Procure-to-Pay the simulation showed a variant-explosion effect; corrected at the design stage.

Result: customising budget well below what comparable projects cost, go-live on schedule, master-data quality stable. Had we begun the design after selection, S/4HANA would have dictated the architecture.

Common mistakes in process design

• Mistake 1: Conflating design with analysis. "Design workshops" produce as-is captures. Fix: separate time blocks, methodically delimited.
• Mistake 2: To-be design after ERP selection. The ERP standard designs the process; customising explodes. Fix: design before the specification document.
• Mistake 3: Process owner without mandate. The target process drifts after go-live. Fix: owner from extended management with veto rights.
• Mistake 4: Corporate-grade stack in the Mittelstand. These structures rarely pay off below corporate scale. Fix: reduced five-element stack plus SIPOC.

Our take: The errors are rarely methodological. They are organisational and sequence-driven — which is exactly where vendor-neutral consulting has its leverage.

Frequently asked questions on process design

Next steps

Before launching a process-design initiative in the DACH Mittelstand, clarify three things: which end-to-end main processes are in scope, who holds the process-owner mandate, and which sequence relative to ERP selection is planned.

To deepen these topics, see our process management services and our ERP consulting. Current case studies are in our insights; arrange a first conversation via the contact page.

Matthias Müller Senior Consultant for process optimisation, ERP and digitalisation at Dreher Consulting. Across over 1,200 projects in the DACH Mittelstand he guides Mittelstand companies through complex ERP implementations and specialises in integrating process design, target-process architecture and system selection. More posts by Matthias Müller → Book a free 30-minute initial consultation →