Choosing an MES is rarely decided on features. It is decided on five trade-offs nobody likes to put in writing.
Functional comparison grids all look alike: every vendor ticks nearly every box. What really separates the solutions lies elsewhere, in structural trade-offs best settled before the first demo.
1. Who owns the item and routing master data?
The ERP, the PLM or the MES? Only one answer is acceptable per data object, and it must be written down. Architectures where two systems "can" create an item produce duplicates within six months.
2. What does the shop floor do when the MES is unavailable?
A paper fallback, local operation at the workstation, or a production stop? That answer drives the network architecture, the sizing of the workstations and, often, the real project budget.
3. Is traceability an obligation or a convenience?
In pharmaceutical or food production, traceability gates product release: it has to be blocking. Elsewhere, recording after the fact is usually enough. The two cases call for neither the same product nor the same user experience.
4. Who will change the configuration three years from now?
If the answer is "the integrator", total cost of ownership will run two to three times the initial estimate. If it is an internal team, check that the tool really is configurable without development, and budget the training from the start.
5. How much customisation will you accept?
Every bespoke screen is debt payable at the next version upgrade. We recommend listing explicitly the screens where specificity is justified by competitive advantage, and sticking to the standard everywhere else.
Once those five answers are written down and endorsed by operations management, choosing the vendor becomes almost mechanical. When they are not, no comparison grid will save the project.
Back to the blog