FMCG manufacturers in Pakistan: the hidden cost of not having a production ERP

Ask a mid-size FMCG manufacturer in Pakistan what it costs to produce one unit and you will get an answer immediately. Ask how that number was arrived at and the answer is almost always the same: raw material cost, plus a percentage for overhead, plus labour, divided by output.

That calculation is not wrong. It is just missing everything that actually varies.

The three invisible costs

Waste. Material that entered the process and did not leave it as sellable product. In most operations this is measured once, during an annual exercise, and then applied as a fixed percentage forever. Meanwhile the actual figure moves week to week with humidity, with operator experience, with which supplier’s batch you are running.

Rework. Product that failed inspection and went back through. It consumes machine time, labour, and often additional material, but it usually appears in the records as though it were a normal run. The unit eventually sells at full price, so nothing looks wrong. The cost was absorbed silently by the line.

Downtime. Not major breakdowns, which get noticed. The twenty minutes waiting for a changeover, the forty minutes because material did not arrive at the line, the hour lost to a fault that got fixed without a record. Individually trivial, collectively often fifteen to twenty percent of available capacity.

None of these appear in the standard calculation. All three vary. Which means the cost per unit that pricing decisions are based on is a stable number describing an unstable reality.

What this costs in practice

The specific failure mode is not that you lose money on everything. It is that you cannot tell which products lose money.

A manufacturer running eleven SKUs across three lines has, in our experience, two or three that are marginal or negative once real production cost is accounted for. They stay in the range because the blended margin looks acceptable and nobody can see the individual picture. Sales continues to push them, sometimes hardest, because they are easy to sell, which is frequently why they are unprofitable.

The second failure mode is capital. A manufacturer convinced they need another line, when utilisation on the existing lines is sixty percent and the gap is changeover and material staging, is about to spend a great deal of money to buy the problem again.

What a production ERP actually changes

The mechanism is unglamorous: capture events on the floor as they happen, attributed to a run.

Material issued to a run. Output recorded against it. Rejects recorded with a reason. Downtime recorded with a cause. Changeover start and end. None of this is analysis. It is data capture, and it has to happen at the line, in seconds, by the operator, or it will not happen at all.

Once it does, the arithmetic that was previously an estimate becomes a calculation. Yield variance per run. True cost per unit including waste and rework. Machine utilisation with causes attached. Cost by SKU, which is the number that changes commercial decisions.

The implementation risk worth naming

The failure mode of manufacturing ERP is that the floor does not use it.

If recording a reject requires walking to a terminal, logging in, and navigating four screens, it will be recorded at the end of the shift from memory, in round numbers, and the data will be worse than useless because it will look precise. Every deployment we have seen fail in this sector failed here.

The design constraint is that capture must take under ten seconds at the point the event happens. That usually means fixed terminals at each line, large touch targets, no login per transaction, and reason codes as buttons rather than a dropdown. It is a hardware and layout problem as much as a software one, and it should be walked physically before anything is configured.

What to expect

For most mid-size manufacturers, the first month of accurate data produces one uncomfortable surprise and one immediate saving.

The surprise is usually a product that has been sold at a loss for a long time. The saving is usually changeover time, because once it is measured it turns out to be twice what everyone believed, and reducing it costs nothing but sequencing.

Neither of those requires new machinery. They require being able to see the line you already have.