Shop floor and MES integration is the single largest cost driver in a mid-market manufacturing ERP program, ahead of quality and traceability and ahead of multi-plant rollout, by ERP Research, an independent research firm. Yet most selection teams decide the vendor on finance and planning, then treat the floor as an integration task for later. That ordering is backwards. Your MES position changes which vendors belong on the shortlist, so it belongs in the vendor choice, not after it. Decide it first and a four-way tier 2 race can narrow to one or two names before the first demo.
The floor is the largest cost driver, and finance is the part teams already know how to buy
General ledger, accounts payable, and consolidation are the parts of an ERP that look alike across manufacturers. They are well-trodden, the configuration is close to standard, and an SI has done them a hundred times. That is why a finance-led selection feels safe. It compares the vendors on the ground every CFO understands.
The floor is where the money and the risk sit. ERP Research ranks shop floor and MES integration as the largest single cost driver in a mid-market program, above quality and traceability and above multi-plant rollout. The reason is that every plant collects data differently, and connecting machines, labor, and scheduling to the ERP is custom work by definition.
That custom work is also the most expensive kind. Holding custom code under 10 to 15% of requirements saves 20 to 40% of implementation cost, by ERP Research. A program that underestimates the floor blows through that cap on the shop floor interface alone. The finding the pillar on manufacturing ERP selection makes once is worth repeating here: the floor decides the budget, so it should decide the shortlist.
Three starting positions decide which vendors belong on the list
There are three places a manufacturer can start, and each one points at a different shortlist. Write down which one you are in before you talk to a vendor.
The first position is no MES at all, running paper travelers and spreadsheets on the floor. Here a system with MES built in saves a second program and a second SI. Favor a vendor whose MES is part of the platform or sold alongside the ERP, so the floor and the ledger go live together.
The second position is a working MES you intend to keep, such as a Rockwell, Siemens, or AVEVA product already on the floor. Here the ERP question narrows to finance, planning, and the interface to the existing MES. The floor is already solved, which removes tier 1's main manufacturing weakness and puts SAP S/4HANA and Oracle Fusion Cloud ERP back in the race on equal terms.
The third position is the hardest: several plants, each with its own MES or none, acquired over time. There is no single answer, and the choice is a program-design question before it is a vendor question.
| Starting position | What it implies for the vendor choice | What it rules in or out |
|---|---|---|
| No MES (paper and spreadsheets on the floor) | Favor a system with MES in the platform or sold alongside, so you avoid a second program and a second SI | Rules in Plex, Epicor Kinetic with Advanced MES, Infor CloudSuite Industrial with Infor MES; a standalone finance-first ERP means a separate MES selection later |
| A working MES to keep (Rockwell, Siemens, AVEVA) | The ERP question narrows to finance, planning, and the interface; tier 1 loses its main disadvantage | Rules tier 1 back in; a built-in MES you would not use becomes cost you do not need |
| Several plants, mixed MES or none | Standardize on one MES-included ERP across the plants, or run a corporate ERP plus a separate MES standard, which is two programs | Above about 20 plants, two programs and two SIs; below that, one MES-included ERP is usually all the company can carry |
The third position needs a rule, because both routes are defensible. Standardizing on one MES-included ERP (Plex or Epicor in discrete, Infor or QAD in regulated process) is one program and one data model, but it forces every plant onto the same floor system. Running a corporate ERP with a separate MES standard below it keeps the floor independent, at the cost of two programs and two SIs. The right default, in our judgment, is the second route above about 20 plants and the first route below it. Sysgenpro's observation that buyers "often overestimate native ERP capabilities" at the plant level applies to this position more than any other (sysgenpro).
Where MES sits differs by vendor, and that location is the decision
MES is not a feature every vendor has in the same place. For some it is the platform itself. For others it is a module sold alongside, an add-on bought separately, or a partner product connected by interface. That location, not a feature checkbox, is what the no-MES buyer is really choosing.
| Vendor | Where MES sits | What that means for a no-MES buyer |
|---|---|---|
| Plex | In the platform; MES is one module of the same system, built from the floor up | One program covers floor and ledger; Rockwell positions it to run standalone beside a corporate ERP or with Plex's own ERP. 700 or more customers across 2,000 or more facilities in 37 countries |
| Epicor Kinetic | Alongside; Epicor Advanced MES is sold with Kinetic for schedule changes and plant floor data collection | Floor and ERP from one vendor, bought as two products that are built to connect, so the interface risk is lower than a third-party MES |
| Infor CloudSuite Industrial | As an add-on; Infor MES is a separate product added to the ERP | One vendor, but the MES is a distinct line item to scope and price; confirm it covers your floor before you count it as built in |
| SAP S/4HANA and Oracle Fusion Cloud ERP | With a partner MES; the ERP connects to a Rockwell, Siemens, or AVEVA product rather than carrying deep plant floor data capture itself | Strong where an MES already exists; a no-MES buyer is signing up for two selections and two programs, which is where tier 1 looks expensive for a small plant |
The pattern is clear. A manufacturer with no MES and fewer than 20 plants gets the shortest program from a vendor that carries the floor in the platform or alongside it. That is Plex or Epicor in discrete work, and Infor or QAD in regulated process. A manufacturer that already runs a Rockwell, Siemens, or AVEVA MES it will keep loses nothing by shortlisting tier 1, because the connectivity is the same work either way. Sysgenpro rates S/4HANA's MES connectivity strong where SAP is already the core system and describes Oracle Fusion's hybrid integration as common for MES.
A finance-led selection tests the ledger and the disruption comes from the floor
The committee that picks on finance tests the close, the chart of accounts, and the consolidation in the demo. Those go live quietly. The disruption at go-live comes from the floor, because that is where the work actually happens and where the data was never clean.
The cost structure shows why. Implementation services run 1x to 3x the first-year subscription, the SI takes 40 to 60% of that, and an external PMO runs $5,000 to $20,000 a month for the length of the program, all by ERP Research. A mid-market tier 2 program runs $300,000 to $3M over 6 to 18 months, against $500,000 to $5M or more and, in programs we have advised, 36 months for a single-region tier 1 program (sysgenpro). The months that overrun are rarely the finance months. They are the shop floor months, when the data capture does not match how the plant runs.
Consider a $500M discrete manufacturer with five plants, 400 named users, and no MES standard. At a tier 2 subscription of $150 per user per month, the software line is $720,000 a year and implementation at 1x to 3x is $720,000 to $2.16M. If the team scoped finance in the demo and discovered the floor in the build, the shop floor interface is the line that pushes custom code past 15% of requirements and adds 20 to 40% to that implementation bill. That is $144,000 to $864,000, as much as a plant's whole share of the program, spent on work the selection never scored.
The timing makes it worse. A survey reported by ERP Today found 31% of implementations took longer than expected and 12% had significant delays. A floor that was treated as a later integration task is the common reason a manufacturing program lands in that 31%. Every extra month carries the PMO at $5,000 to $20,000 and holds the old system in dual-run, a period that runs 12 to 24 months when it is not controlled (Costbench, a cost-benchmarking site). Deciding the MES question in the same committee that decides the ERP is what keeps the floor off the critical path.