General
Your ERP Has the Production Data. So Why Is Everyone Still Scheduling in Spreadsheets?
When every department runs its own version of the production schedule, the problem may not be your ERP. It may be the work required to turn its data into a plan everyone can actually use.
When every department runs its own version of the production schedule, the problem may not be your ERP. It may be the work required to turn its data into a plan everyone can actually use.
A manufacturing manager recently raised a question that will sound familiar to many operations leaders: across five companies and multiple plants, they repeatedly saw the same pattern. Production data lived in the ERP, but teams exported it into spreadsheets to create their actual schedules.
Different departments maintained different versions. Sometimes, even people within the same department had their own spreadsheets.
The ERP was the official system of record. The spreadsheets were where the work actually happened.
It's a revealing example of a problem that extends well beyond manufacturing.
The ERP isn't necessarily the problem
An ERP can hold production orders, inventory balances, bills of materials, routing information, and planned dates. Depending on the system and its configuration, it may also support production planning and scheduling.
But having the data doesn't automatically mean the business has a practical, shared way to coordinate production.
A production schedule has to account for real operating constraints: machine availability, cycle times, labor, material readiness, setup requirements, competing priorities, and changes in customer demand.
Some of these constraints are difficult to model. Others change faster than the official plan can be updated.
So a planner exports the data, adds the missing context, rearranges priorities, and builds a spreadsheet that reflects what they believe can actually happen.
Then another department creates its own version.
Before long, the business has several schedules that each make sense to the people maintaining them, but no single view that everyone can confidently rely on.
The issue isn't simply that people use spreadsheets. Spreadsheets can be excellent planning tools.
The problem is that the spreadsheets become disconnected versions of operational reality.
What happens when every department has its own schedule?
Consider a manufacturer producing custom equipment across several departments.
The ERP says an order is due next Friday. The planning team has put it on Wednesday's schedule. But the machining department is waiting on a material delivery, and assembly is still working from last week's priority list.
A customer then requests an earlier delivery.
The change triggers a chain of questions:
- Can machining complete the parts in time?
- Is the required material available?
- Which other jobs would need to move?
- Will the change affect assembly or testing?
- Who needs to know about the revised commitment?
- Which schedule should everyone update?
None of these questions is particularly unusual. The challenge is getting the right information to the right people, understanding the downstream consequences, and keeping everyone aligned as the situation changes.
Someone checks the ERP. Someone else checks a spreadsheet. A supervisor sends an email. A planner updates a schedule. Another department doesn't see the change until the next meeting.
The result can be expediting costs, idle capacity, overtime, missed delivery dates, and hours spent reconciling conflicting information.
The cost isn't just maintaining spreadsheets. It's the delay between something changing and the entire operation responding to it.
Why simply moving everything into the ERP may not solve it
For some manufacturers, the right answer is to improve ERP configuration, implement finite-capacity scheduling, or adopt dedicated advanced planning and scheduling software.
If the business lacks reliable production routings, machine capacities, cycle times, or material requirements, it may need to address those fundamentals first.
No AI layer can reliably produce an optimized production schedule from incomplete or inaccurate constraints.
But there is another situation worth distinguishing.
The business already has a functioning ERP. Its planning process is reasonably well established. Yet people still spend significant time exporting information, reconciling changes, chasing updates, and communicating what those changes mean for other departments.
In that situation, buying another major system may not be the first or only answer.
The missing capability may be the coordination between the systems, schedules, and people already involved in running production.
A different approach: connect the work around the schedule
At LevelOneAI, we focus on the operational work that happens between business systems, especially when the official record doesn't tell the whole story of what's happening.
For a manufacturer, that could mean creating a shared operational view around existing production orders and schedules.
Rather than replacing the ERP or pretending that every spreadsheet should disappear, the approach starts by identifying where coordination breaks down.
1. Bring the relevant information together
Connect the information needed to understand each job: its official production status, current schedule, material dependencies, latest changes, and relevant communications.
The ERP remains the authoritative source for the records it owns. Existing planning tools can continue to perform specialized scheduling calculations.
The goal is to make the information surrounding each job easier to understand without forcing every team to adopt an entirely new way of working.
2. Surface changes that matter
Suppose a supplier changes a delivery commitment, a machine goes offline, or a customer requests a different ship date.
The important question isn't merely whether a new email or spreadsheet exists. It's which production orders may be affected, what commitments are now at risk, and who needs to respond.
An AI-assisted coordination layer can help connect those updates to the relevant work, identify missing information, and flag potential conflicts for review.
It should distinguish confirmed facts from assumptions. It should also avoid claiming that a job can be rescheduled successfully without the capacity and constraint data needed to establish that.
3. Give departments a shared view of operational status
Instead of asking each department to reconstruct the situation from its own files, teams can work from a shared view of job status, outstanding dependencies, emerging risks, and next actions.
A planner can see which jobs need attention. A production supervisor can see unresolved dependencies. An operations manager can identify where decisions or handoffs are holding up work.
The aim isn't to remove departmental expertise. It's to reduce the manual effort required to keep that expertise coordinated.
4. Keep people in control of production decisions
Not every scheduling conflict can be resolved by an AI recommendation. Moving one job can affect machine utilization, setup costs, downstream capacity, and other customer commitments.
Those decisions should remain with the people and scheduling systems equipped to evaluate the trade-offs.
LevelOneAI's role is to make changes and their implications more visible, organize the evidence, and help the responsible people take action. Updates to authoritative systems should happen through controlled, approved workflows.
What this could look like in practice
Imagine starting with one production area rather than attempting a plant-wide transformation.
The team identifies a recurring problem: schedule changes don't reliably reach all the departments they affect.
A focused implementation could establish a shared view of selected production orders, bring in relevant schedule and status updates, flag jobs with unresolved dependencies, and show who needs to take the next action.
The team could then measure whether it reduces time spent reconciling spreadsheets, shortens the time required to communicate a schedule change, or improves visibility into jobs at risk.
If the results justify it, the approach can expand to additional departments and workflows.
If the underlying problem turns out to be inadequate scheduling logic or missing production data, the business will have a clearer understanding of what a dedicated planning system or ERP improvement must address.
Either way, the goal is to solve the actual bottleneck rather than assume that a bigger software implementation will automatically fix it.
The bigger lesson for manufacturing operations
Spreadsheets aren't necessarily evidence that an ERP has failed. Often, they are evidence that people have found a practical way to handle work the formal system doesn't accommodate well.
The problem starts when those workarounds become the only way to coordinate the operation, and each one creates another place where information can become outdated, decisions can get lost, and departments can fall out of sync.
Manufacturers don't necessarily need to eliminate every spreadsheet or replace every system. They need to understand which workarounds are useful, which create risk, and where better coordination would make the biggest difference.
That's the problem LevelOneAI is designed to help address.
Your ERP can remain the system of record. Your scheduling tools can remain the engines for production planning. The opportunity is to make the work between them more visible, connected, and responsive.
Because a production plan is only as useful as the operation's ability to act on it.