Between SAP DMC, Werum, Critical Manufacturing and smaller providers, a sober comparison pays off. Seven criteria we consistently weight in selection processes.
The headline
We regularly accompany MES selection projects and see this: most poor decisions are not made in the choice of vendor, but before it — in an unclear requirements analysis. If the requirements specification is shaky, every vendor is roughly equally good or equally bad. The seven criteria below help to sharpen the selection.
1. Interface to the ERP — especially to SAP
If you use SAP: how deep is the SAP integration? Native S/4HANA connection, IDOC, OData, custom interfaces — all possible, but very different in maintenance and upgrade effort. A superficial decision here costs you dearly in year five.
2. Configurability without programming
Configuration vs. programming is the decisive lever for total cost of ownership. Vendors whose customisation only goes via custom code make you dependent — both on the vendor and on consultants.
3. Validation readiness (for regulated industries)
In medical devices, pharma and food: how does the vendor support validation under GAMP, IQ/OQ/PQ? Are there validation packages? What is the update strategy — does every update invalidate the validation?
4. Machine connectivity
OPC UA is standard, but not every vendor implements it equally. Key question: how many machine types are supported out of the box? How much effort per new machine — in days, not in marketing claims?
5. Scalability: pilot → plant → group
Many mid-market manufacturers start with a pilot line and plan a roll-out to further plants. The vendor’s licensing and implementation logic has to allow that without the economics of scale suddenly looking different from what was promised.
6. Consultant network
Vendor A has a dense consultant network, vendor B only the manufacturer itself. The latter means: every customisation runs via the vendor — which strongly shapes pace and cost. The former gives more choice, but also more variability in quality.
7. Roadmap and financial stability
MES vendors are often mid-sized themselves; some are acquired or change strategic direction. A look at ownership structure, roadmap and customer base helps to reduce the risk of investing in a “soon to be discontinued” system.
Recommended approach
- Requirements analysis — workshop with production, QM and IT, prioritised requirements specification (60–80 requirements, not 400).
- Long list — 6–8 vendors from a market overview, pre-filtered by industry and size.
- Short list — 2–3 vendors based on a standardised RFP.
- Proof of concept — a real demo against your own data model, not on the vendor’s showcase.
- Contract negotiation — using a scoring matrix that weights all seven criteria.
Sticking to this approach not only shortens the selection — above all it ensures that the system afterwards does the job it was bought for.