Operations · System Thinking
Your ERP knows everything, except what to do next

Most manufacturing and distribution companies I have worked with run on an ERP system. It records every purchase order, every goods receipt, every production confirmation, every invoice. Ask it how much stock of an item sits in a warehouse in another state, and it answers in seconds.
Yet in the same companies, the daily production meeting still runs on phone calls, WhatsApp messages and a spreadsheet the planner maintains on the side. Expediting is constant. Month-end is chaos.
If the ERP has all the information, why are the decisions still so hard?
Information is not a decision
An ERP is excellent at what it was designed for: recording transactions and integrating data across functions in one database. Before ERP, finance, stores and production each had their own version of the truth. ERP fixed that, and the value of clean, shared data is real.
But knowing what has happened is not the same as knowing what to do next. The hard questions in operations are decisions:
- Which order should this machine run next?
- How much of this item should we keep at that depot?
- When should this order be released to the shop floor?
- What due date can we safely promise this customer?
To answer these, the ERP relies on planning logic built into it, and that logic carries assumptions.
The assumptions inside the planning engine
Most ERP planning modules are built on MRP (material requirements planning) logic. In simple terms, it works backwards from a forecast or a set of orders, using fixed lead times for every item and every step, to tell each department what to make and when.
On paper, it is elegant. In practice, it rests on assumptions that rarely hold:
- The forecast is right. At item level, it usually isn’t.
- Lead times are fixed. In reality, lead times depend on how much work is already in the plant. When the plant is flooded, everything waits longer.
- Capacity is unlimited (or checked only roughly). Plans that look fine on screen can be impossible on the floor.
- Every department should meet its own dates. So gains in one step are rarely passed on, while delays are.
When reality doesn’t match the assumptions, the plan breaks within days. People stop trusting it and fall back on experience and expediting. The ERP becomes a very expensive record-keeping system, while the real decisions are made outside it.
This is not a failure of the software. It is a failure of the management rules the software was asked to automate.
What TOC brings to the table
Theory of Constraints doesn’t replace the ERP. It replaces the planning rules with simpler ones that hold up in an uncertain world:
- Release work by the pace of the constraint, not as soon as an order arrives. This keeps work in progress low, so lead times become short and predictable.
- Protect with buffers, not forecasts. In distribution, keep a stock buffer for each item at each location, sized to cover demand within the replenishment time, and refill it from actual consumption.
- Set priorities by buffer status. Every order or stock item gets a colour (green, yellow, red) based on how much of its protection has been used. Everyone works on red first.
- Quote due dates from the load on the constraint, not from a fixed standard lead time.
Picture the morning production meeting in two versions.
Version one. The ERP report lists 140 open orders, 38 of them flagged “late” against their standard lead time. Sales has already called about six of them. The meeting spends an hour arguing which of the 38 matter most, and the answer depends on who argues best.
Version two. The same 140 orders are coloured by buffer status. Nine are red: they have used most of their protection and will miss their dates unless someone acts. Twenty are yellow. The rest are green, including many of yesterday’s “late” orders, which still have plenty of time. The meeting takes fifteen minutes: what is blocking each red order, and who will clear it today?
Same data, same ERP. The only difference is the rule used to read it.
These rules need surprisingly little data: accurate information at the constraint, current stock and consumption, and order due dates. All of it already sits in the ERP.
A good partnership
Put together, the two do what neither can do alone.
| The ERP provides | TOC provides |
|---|---|
| One clean source of data | The rules for deciding what to do with it |
| Transactions: orders, receipts, stock, invoices | Priorities: which order first, which item to refill |
| Visibility of what has happened | Buffers that protect against what we can’t predict |
| Integration across functions | Measures that align functions to the whole |
In practice, companies often run TOC logic in a light layer next to the ERP: it reads stock, consumption and orders from the ERP daily, calculates buffer status and priorities, and writes back replenishment orders or release dates. The ERP stays the system of record; TOC becomes the system of decision.
The ERP tells you everything about yesterday. TOC tells you what to do today.
Before your next ERP upgrade
If your company is about to spend on a new ERP module, a planning upgrade or an advanced scheduler, ask a few questions first:
- What management rule will this automate? If the rule is “forecast, push and expedite”, automating it will only produce the same chaos faster.
- Do people on the floor trust today’s plan? If not, find out why before buying a more complex one.
- Could simpler rules (release control, buffers, priorities by colour) solve the problem using the data we already have?
Often, the most valuable change is not in the software at all. It is in the rules the software is asked to follow.