← Back to all posts Manufacturing & shop floor

MES software: how to choose one (and when you do not need one)

A buyer-side guide to MES software: the four categories of vendor, the eight criteria that actually decide the outcome, what it really costs, why implementations fail, and an honest test of whether your ERP already covers it. Written by an ERP partner, including the cases where we would tell you not to buy.

Also in: Deutsch Nederlands

Short answer: choosing MES software is less about feature lists than about two questions most vendors skip. First: is the shop floor really your bottleneck, or is the pain a layer below in planning or master data? Second: does your ERP already cover this, which for people-driven production it often does. Get those two right and the rest of the selection becomes straightforward. Get them wrong and you will buy a precise instrument for measuring the wrong thing.

This page is written from the buyer’s side. If you are still working out what a MES is, read what is a MES first; if you are weighing which layer to invest in, MES vs MRP vs ERP has a five-question diagnostic. This page assumes you have decided you need one and are choosing.

Four categories of MES vendor

The market looks fragmented until you notice it clusters into four types, each with a predictable strength and a predictable weakness.

CategoryStrong atWeak atFits when
Industrial automation MES (from automation and control vendors)Machine and PLC integration, process control, high-frequency dataCost, business-side integration, agilityThe floor is machine-driven and heavily automated
Specialist independent MESShop-floor depth, analytics, OEE, scheduling detailNeeds integration with your ERP; duplicate master dataManufacturing is your core complexity and you can carry the integration
ERP-native shop floor (a module of your ERP)One database with planning, stock and finance; no integration projectDeep machine connectivity, second-level captureProduction is people-driven and you want one system
Industry-specific systems (print, food, metal)Fits your trade out of the box, familiar vocabularyLocks you into one sector; weaker outside its nicheYour industry’s specifics outweigh breadth

There is no winner in that table, only a fit. What we do see: companies who put “the floor is machine-driven” and “we are people-driven” in the wrong box pay for it for years.

The eight criteria that actually decide it

Ignore feature matrices for a moment. In the projects we have seen succeed or struggle, these decided the outcome.

  1. Operator usability at the workstation. Can someone wearing gloves, in a noisy hall, report a step in two taps without reading anything? This is the single highest-weighted criterion and it is almost always underweighted in selections, because the people doing the selecting are not the people doing the tapping.
  2. Integration with ERP and planning. Where do orders come from and where do results go? Every integration you avoid is a project you do not run and a synchronisation you never debug.
  3. Machine connectivity. Do you genuinely need PLC and sensor data, or do people report? Be honest here: this criterion alone often decides your category.
  4. Quality and traceability depth. Control points during operations, measurements, blocking on failure, lot and serial genealogy, recall reporting.
  5. Configurability without code. Can your own team add a check or change a screen, or is every change a vendor ticket with a rate card?
  6. Deployment and hardware. Tablets, mounts, scanners, network coverage in a hall with metal everywhere. This is a real budget line that gets discovered late.
  7. Total cost over five years, not licence price. See below.
  8. Track record in your industry, with references you can actually call. Not logos on a slide: names and phone numbers.

What MES software really costs

Licence cost is usually the smaller half of the bill. A realistic five-year view includes:

  • Licences or subscription, per user or per workstation.
  • Implementation and configuration, typically the largest line for specialist systems.
  • Hardware: tablets, industrial mounts, scanners, label printers, plus network coverage on the floor.
  • Integration with ERP, planning and machines. This is where specialist MES projects grow, because the integration is a project in itself.
  • Training, including the reality that shop-floor staff turn over and training is never one-off.
  • Annual maintenance and support, often a percentage of licence value.

The pattern worth knowing: a specialist MES commonly reaches six figures once integration and hardware are counted, while an ERP-native shop-floor module can be a fraction of that, mostly because the integration work does not exist. That is not a claim that one is better. It is the reason the categories are priced so differently, and why the “do I need a separate MES” question is worth real thought before you shortlist.

Our opinion: most MES selections weigh the wrong things

We are an ERP partner, so read this knowing what we sell. It also means we have watched a lot of these decisions from close by.

The demo lies, politely. Every MES demos beautifully, because demos run on clean data with a presenter who knows the shortcuts. The floor does not run on clean data with a presenter. Insist on a pilot at one real work centre, with your bills of materials, your operators, and a week of actual production. What survives that is what you are buying. Vendors who resist a pilot are telling you something.

Weight operator usability above everything else. A MES that operators avoid does not produce partial data, it produces wrong data, filled in from memory at the end of the shift, and that is worse than no data because you will trust it. Every other feature is downstream of whether people actually use the thing.

Be suspicious of the OEE promise. OEE is the most-demoed and least-earned metric in this market. It is genuinely valuable, but only once your reporting is complete and honest. Buying a MES for OEE puts the dashboard before the data. Get accurate reporting first; the percentage will still be there next quarter.

And the one that costs us work: if your planning is unreliable or your master data is a mess, do not buy a MES this year. You will get a very precise measurement of chaos. Fix the layer below first. We have told prospects this and lost the project to someone who told them what they wanted to hear. We would still say it.

The honest test: does your ERP already cover this?

Before you shortlist anything, run this check. If your ERP has a shop-floor app, does it do:

  • Digital work instructions at the workstation
  • Start, pause and stop timing per operation
  • Quantity and scrap reporting at the moment it happens
  • Quality checks as a step, with blocking on failure
  • Barcode scanning for components, lots and serials
  • Direct flow-through to stock, planning and cost price

If yes, you have the MES role covered for people-driven production, without an integration project. That is the situation for most mid-sized manufacturers, and it is why we so often end up recommending the boring answer: switch on what you already own, run it for a quarter, and revisit.

You need a separate MES when the answer is genuinely: deep machine and PLC integration, capture at the second across many machines, or validated production under regulatory requirements. Then buy the specialist, connect it to your ERP for planning and finance, and do not pretend either one can be the other.

For reference, in Odoo that shop-floor role is covered by the Shop Floor app on top of Manufacturing, with Quality and IoT alongside. We describe what it does and where it stops in what is a MES, including the parts Odoo does not do well.

Frequently asked questions

How do you choose MES software? Decide which layer you need first, then weigh operator usability highest, followed by ERP integration, machine connectivity, quality depth, configurability, hardware, five-year cost and industry references.

What types of MES software are there? Industrial-automation MES, specialist independent MES, ERP-native shop-floor modules, and industry-specific systems. Each has a predictable strength and weakness.

How much does MES software cost? Licences are the smaller half. Count implementation, hardware, integration, training and annual maintenance. Specialist projects often reach six figures; ERP-native modules cost a fraction because the integration does not exist.

Do I need MES software if I already have an ERP? Often not, for people-driven production. A separate MES earns its place with deep machine integration, second-level capture, or validated production.

Why do MES implementations fail? Undefined processes, an operator interface designed for managers, and buying a MES while planning or master data is still unreliable.


Want a second opinion before you shortlist? Book a free Odoo scan - we go through your floor, your planning and your data, and tell you honestly whether you need MES software at all, and what we would not sell you.


Read more: Manufacturing software: the categories · Production planning and scheduling · What is a MES? · MES vs MRP vs ERP · MRP system explained · Odoo for manufacturers · Can Odoo run your production floor? · The TARGET method

Frequently asked questions

How do you choose MES software?

Start from the problem, not the product. Decide which layer you actually need (ERP base, planning, or shop-floor execution), then judge candidates on eight criteria: operator usability at the workstation, integration with your ERP and planning, machine connectivity, quality and traceability depth, configurability without code, deployment and hardware, total cost across five years, and the vendor's track record in your industry. Weigh operator usability highest, because a MES that operators avoid produces no data at all.

What types of MES software are there?

Roughly four categories. Industrial-automation MES from automation vendors, strong on machine and PLC integration. Specialist independent MES, strong on shop-floor depth and analytics. ERP-native shop-floor modules, weaker on machine integration but sharing one database with planning and finance. And industry-specific systems built for one sector, such as printing or food, which fit closely but limit you elsewhere.

How much does MES software cost?

Licence cost is usually the smaller half. Expect the full picture to include licences or subscription per user or per workstation, implementation and configuration, hardware (tablets, scanners, mounts, network coverage on the floor), integration with ERP and machines, training, and yearly maintenance. Specialist MES projects commonly run into six figures once integration and hardware are included, while an ERP-native shop-floor module can be a fraction of that because the integration work does not exist.

Do I need MES software if I already have an ERP?

Often not. If your ERP has a shop-floor app covering digital work instructions, start and stop timing per operation, scrap and quantity reporting, quality checks and barcode scanning, that covers the MES role for people-driven production. A separate MES becomes worth it when you need deep machine integration (PLC, process control), capture at the second across many machines, or validated production in a regulated industry.

Why do MES implementations fail?

Three recurring reasons. The processes were not defined first, so the software was asked to fix an unclear process. The operator interface was designed for managers rather than for people wearing gloves in a noisy hall, so reporting is done from memory at the end of the shift. And the MES was bought while the underlying planning or master data was still unreliable, which produces a very precise measurement of chaos.

Recognize this from your own setup?

A 30-min scan turns hunches into a concrete view, what stays standard Odoo, what becomes custom, what doesn’t need code at all.

Get in touch ← Back to blog