Walk into almost any small-to-mid-sized manufacturing facility in Michigan, and you’ll find the same scene: good people working hard, decent equipment running well, and software that’s quietly making everything harder than it needs to be.
It’s not always obvious. The plant doesn’t grind to a halt. Orders still ship. But somewhere between the front office and the shop floor, hours are being burned, decisions are being made on stale data, and one person retiring could take years of critical process knowledge with them. The cost is real; it’s just spread thin enough that nobody writes a check for it.
If you’re running a manufacturing operation and you’ve had that nagging feeling that your systems are working against you instead of for you, this is for you. Here are five signs your manufacturing software is holding you back, and what to do about it.
1. Your “System of Record” Is a Spreadsheet on Someone’s Desktop
You have an ERP. You probably also have a folder somewhere (on a shared drive, or worse, on one person’s local machine) called something like “MASTER SCHEDULE v7_FINAL_USE THIS ONE.xlsx.”
Sound familiar?
This happens because the official system doesn’t quite do what people need it to do, or it’s too slow, or nobody was ever trained on it properly. So someone built a workaround in Excel that actually works, and now that spreadsheet is the system of record, even if nobody wants to admit it. The ERP is what the auditors see. The spreadsheet is what actually runs the floor.
The problem isn’t that the spreadsheet exists. The problem is that it lives outside your controlled system, it doesn’t update automatically, and the second the person who built it goes on vacation (or gives their notice) you’re flying blind. Worse, any decisions made off that data are only as good as the last time someone manually updated a cell.
Good looks like this: one source of truth, accessible to everyone who needs it, updated in real time or close to it. That doesn’t mean ripping out your ERP. Sometimes it means integrating the data that’s stuck in spreadsheets back into a real system, or building a lightweight custom tool that fills the gap your ERP can’t reach.
2. Your Systems Don’t Talk to Each Other
Your ERP knows about the order. Your MES knows what’s running on the floor. Your shipping system knows when it went out the door. But none of them know what the others know.
This is the integration gap, and it’s one of the most expensive problems in manufacturing because the cost doesn’t show up as one big line item. It shows up as:
- The customer service rep who has to call the floor supervisor to get an order status because they can’t see it in the system
- The production planner who re-keys the same job information into three different systems every morning
- The quality manager who can’t link a customer complaint back to a specific machine or shift without pulling records from two places and matching them by hand
Consider a metal fabrication shop that runs a job scheduling system, a separate quality management tool, and an ERP that was implemented a decade ago. None of them are integrated. Every time an order moves through production, it gets touched in all three systems manually. At 40 jobs a week, that’s hours of re-keying every day, plus the mistakes that come with it.
When your systems are connected, data entered once flows where it needs to go. A job created in your ERP populates your scheduler. Quality holds update automatically. Shipping pulls confirmed completions. That’s not magic, it’s integration, and it’s entirely achievable without replacing every piece of software you have.
3. Your “Real-Time” Data Is Actually From Yesterday
You have reports. Maybe you even have a dashboard. But when you look at it in the morning, you’re looking at what happened yesterday, or worse, what a system thinks will happen based on a plan that was set last week.
End-of-day summaries made sense when reporting required someone to physically pull records and compile them overnight. That era is over. But plenty of manufacturers are still operating as if it’s 1998, making decisions today based on data that was already 12 hours old when they sat down with their coffee.
In a real-time world, the delay matters. A machine goes down at 8 AM. By the time the production manager’s daily report runs and someone reviews it, it’s 4 PM and you’ve lost half a shift without anyone making a conscious decision to address it. Capacity problems that could have been caught at 10 AM compound all day.
True real-time visibility doesn’t require a seven-figure MES implementation. For many shops, it starts with making sure your floor data (machine status, job completions, scrap counts) flows into a central view that’s updated continuously or at minimum every 30 minutes. The investment is modest. The operational clarity it buys is significant.
When something goes sideways on the floor today, how long before the right person knows about it? If the answer is “end of shift,” you have a real-time data problem.
4. Only One Person Knows How to Do That
Every plant has at least one. The person who built the scheduling macro. The one who knows how to get the legacy CNC controller to talk to the ERP. The one who remembers why the work instructions are formatted the way they are, because they were the one who set it up 12 years ago.
When that person is in the building, everything works fine. When they’re out sick, on vacation or, and this happens, they give their two weeks’ notice, you realize how much institutional knowledge is walking around in one person’s head instead of living somewhere accessible.
This isn’t a people problem. It’s a systems and documentation problem. Process knowledge that exists only in someone’s memory is a single point of failure in your operation. And in a tight labor market, it’s also a retention risk. If one person’s absence can destabilize production, that person knows it.
The fix is part process improvement, part software. Documented workflows. Standard operating procedures that are actually maintained and accessible from the floor. Systems configured so that the logic is transparent, not buried in a spreadsheet formula that only the original author understands. When the process lives in the system, it survives personnel changes.
5. Someone Spends Hours Every Week Re-Keying Data Into Reports
This one should be extinct by now. It isn’t.
Somewhere in your operation (Finance, Ops, or Manufacturing) there is a person whose Friday afternoon consists of pulling data from one system, cleaning it up, and pasting it into another system or a report template so the right people can review it on Monday. They’ve been doing this for years. Everyone accepts it as normal.
It’s not normal. It’s expensive, it’s error-prone, and it’s a symptom of systems that weren’t designed to work together.
Think about what that time actually costs. Five hours a week of a $60K/year employee manually compiling data is roughly $7,500 per year, and that’s before you count the errors. A transposed number in a cost summary, a missed row in a shipping reconciliation, a production count that’s off because someone pulled from the wrong tab. Manual data transfer introduces mistakes at every step, and those mistakes compound downstream.
Automating that data flow (connecting the source systems so the report builds itself) is typically a weeks-long project, not a multi-month overhaul. In most cases, the ROI is measured in months, not years. The person doing the re-keying can spend that time on something that actually requires their judgment.
What to Do About It
If you read through those five signs and recognized your operation in one or more of them, the good news is that none of these are unsolvable. They’re not even unusual, these are the exact problems that show up in shops all across Michigan.
The place to start is not by buying new software. The place to start is by understanding what’s actually broken and why. Sometimes the right answer is integrating what you already have. Sometimes it’s a targeted custom tool. Sometimes it’s a process change that costs nothing but time and clear thinking. Rarely is it “throw out everything and start over.”
Here’s a practical first step: map your biggest friction points. Where does data have to be touched by human hands before the next person can use it? Where do people work around the official system because the system doesn’t match how the work actually flows? Where does a single person’s absence create an outsized problem?
That map is the starting point for a real conversation about what to fix first, in what order, and what it will actually take.
At Top Shelf Systems and Engineering, that’s exactly the kind of conversation I have with manufacturers. No sales pitch, no pre-packaged solution. Just an honest look at where your systems are working against you and a practical plan for fixing it.
I offer a no-pressure intake process (a simple questionnaire that helps you articulate the problem) followed by a call to talk through what we’re seeing. From there, we figure out together whether and how we can help.
If any of this hit home, reach out. The worst outcome is a 30-minute conversation that helps you think more clearly about your operation.