If you have ever pulled a production report from your ERP and had your floor supervisor look at it like it came from a different planet, you already know the problem I am talking about.
Your ERP says one thing and your MES says another. Your floor supervisors are keeping their own spreadsheets because they do not trust either system. And somewhere in the middle of all that, real decisions are getting made on incomplete and/or wrong information.
This is not really a technology problem. It is a business problem that technology created, and the fix requires thinking about it that way.
Why This Happens in the First Place
Let us tell the truth about what is happening. Your ERP and your production floor systems were probably never designed to talk to each other.
ERPs - your SAPs, your NetSuites, your Epicors - are built around transactions. Purchase orders, invoices, inventory counts, financial records. They care about what happened and what it cost. Production floor systems - your MES, your SCADA, your PLCs - are built around operations. Machine states, cycle times, yield rates, downtime events. They care about what is happening right now.
These are fundamentally different ways of seeing your business, built by different vendors, using different data structures, often implemented years apart by different teams with different priorities. Nobody sat down at the beginning and said “let’s make sure these two systems can share data seamlessly.” They were each solving a specific problem in isolation.
And for a long time, that was fine. You could run a manufacturing business with a gap between your financial data and your operational data. People filled the gap with meetings, phone calls, and spreadsheets. But that gap has a cost. And as your business grows, that cost compounds.
What the Gap Actually Costs You
Here is what I see when I walk into a manufacturing operation with a disconnected ERP and production floor:
Decisions made too late
By the time a production issue works its way through the system and shows up in an ERP report, the opportunity to fix it has passed. You are managing the aftermath instead of the problem.
Inventory that does not match reality
Your ERP says you have 500 units of a component. Your floor says you have got maybe 300, and some of those are questionable. Which number do you trust when you are promising a delivery date to a customer?
Double data entry
Somebody - usually someone who has better things to do - is manually re-entering production data into the ERP. That is expensive, slow, and introduces errors every single time.
Finger-pointing between departments
Operations blames finance for not understanding the floor. Finance blames operations for not keeping accurate records. Both are right. Neither is getting what they need.
I have sat across the table from plant managers who are genuinely good at their jobs, running genuinely good operations, and they cannot give me a clean answer to “what was your actual cost per unit last month?” - not because they do not care, but because the data to answer that question is living in three different systems that do not speak the same language. That is the gap. And it is fixable.
What Actually Needs to Happen
I want to be honest here. There is no single button you push to make this problem go away. But there is a logical sequence of steps that gets you from where you are to where you want to be.
Step one: Define your critical data points
Before you touch any technology, get clear on what decisions you actually need to make, and what data those decisions require. Not every data point on the production floor needs to flow into your ERP in real time. Start with the ones that drive your biggest business decisions: yield by line, downtime by machine, actual vs. standard cost per job. Pick five. Get those right first.
Step two: Find where the data actually lives
This sounds obvious, but it is almost always messier than people expect. The data you need might live in your MES, or it might live in a PLC that has never been connected to anything, or it might be sitting on a clipboard in a supervisor’s office. Map it before you try to move it.
Step three: Decide on your integration approach
This is where the technical choices come in. The most common options for mid-size manufacturers are:
- Direct API integration between your ERP and your MES, if both systems support it. This is the cleanest solution when it is available.
- A middleware or integration platform (tools like Boomi, MuleSoft, or lighter-weight options) that sits between your systems and handles the translation.
- A data warehouse or data lake that pulls from both systems into a central location, where you can build reporting and analytics on top of it.
Each approach has tradeoffs around cost, complexity, and how real-time you need the data to be. There is no universal right answer. But there is a right answer for your specific situation.
Step four: Do not skip data governance
This is the step most companies skip, and it is why integration projects fail even when the technology works. If your ERP calls a product “SKU-1042” and your MES calls it “Part A-1042-Rev3,” your integration will move data successfully and it will still be meaningless. Before you connect systems, you need to agree on shared definitions. What is a “unit”? What counts as downtime? These are people problems, not software problems.
Step five: Start small and prove it
Pick one line, one product family, or one key metric. Get that integration working cleanly. Show people a dashboard that pulls from both systems and tells a coherent story. The worst integration projects I have seen tried to connect everything at once and collapsed under their own complexity. The best ones started narrow, proved value fast, and expanded from there.
A Word on “Just Buying New Software”
I want to address the solution a lot of vendors will try to sell you: replace your ERP with something newer that has better floor integration built in.
Sometimes that is the right answer. But more often - especially for manufacturers who have spent years customizing their ERP to fit their specific processes - a full replacement is a multi-year, multi-million dollar project that solves the integration problem by burning everything down and starting over.
In most cases, the better path is to make your existing systems work together. You have already paid for them. Your team already knows them. The data is already in there. What you are missing is the connective tissue, which is a much more tractable problem than a full platform replacement.
Where to Start If This Is Your Problem
If you are reading this and nodding along, here is my honest recommendation: do not start by shopping for software. Start by getting clarity on the problem.
What are the three decisions you make every week that would be better if your ERP and floor data were aligned? What is the cost of the gap you have today? In time, in errors, in missed opportunities?
Those answers are what should drive your integration strategy. Not a vendor’s product roadmap.
At SDS, our Discovery engagement is specifically designed to help manufacturers get that clarity. We map your data, identify the gaps that are actually costing you, and give you a prioritized roadmap - so you are making an informed decision, not just throwing technology at a problem you have not fully defined yet.
If this is something that your company is struggling with, let’s talk. Book a free 20-minute conversation ->
Erik Henry is the founder of Simplified Data Solutions, helping manufacturers in the $50M-$500M range close the data gap and achieve measurable ROI.