How to measure and reduce scrap in production

You close the month and production cost came in above budget. Nobody is quite sure why: flour got a bit more expensive, a batch burned in the oven and several boxes were damaged in packing. All of it blended into an inventory difference that accounting adjusted without asking. The scrap was real, but it was never recorded anywhere, and what isn’t recorded can’t be reduced.

This guide shows you how to measure scrap with simple formulas, split the overrun into what is price and what is scrap, and decide where to attack it. You can do all of it in a spreadsheet.

The three basic numbers

1. Yield: how much good product you get for every unit of input that goes in. Yield = good output / input × 100
2. Scrap rate: what goes in and doesn’t come out as good product. Scrap % = (input − good output) / input × 100
3. Standard quantity allowed: what you should have consumed for what you actually produced. Standard quantity = good units produced × standard consumption per unit

Two rules make these numbers useful. First, measure input and output in the same unit (kg with kg, meters with meters). Second, pick one base for the percentage (input, as above) and stick to it: 10% of input is not the same as 10% of output.

Standard cost, actual cost and the two variances

Standard cost is what a unit should cost if everything follows the recipe: standard quantity times standard price. Actual cost is what you actually consumed times what you actually paid. The difference is the total variance, and it splits in two:

• Price variance = (actual price − standard price) × actual quantity
• Quantity variance = (actual quantity − standard quantity) × standard price

The price variance is explained by purchasing or the market. The quantity variance is the one that matters on the shop floor: it is scrap above normal, expressed in money.

Example: one order worked out in numbers

Suppose an order for 1,000 packs of cookies. The recipe says each pack holds 0.45 kg of product and consumes 0.50 kg of dough; that 0.05 kg gap is the accepted normal scrap (10%). The standard price of dough is Bs 8 per kg.

When the order closes, the plant reports 1,000 good packs, but 540 kg of dough were consumed and purchasing paid Bs 8.40 per kg.

Item	Standard	Actual
Dough consumed	500 kg	540 kg
Good product	450 kg	450 kg
Yield	90.0%	83.3%
Scrap	10.0% (50 kg)	16.7% (90 kg)
Dough cost	Bs 4,000	Bs 4,536
Cost per pack	Bs 4.00	Bs 4.54

Now the variances:

• Total: 4,536 − 4,000 = Bs 536 unfavorable.
• Price: (8.40 − 8.00) × 540 = Bs 216 unfavorable.
• Quantity: (540 − 500) × 8.00 = Bs 320 unfavorable.

Together they add up to the Bs 536. The reading is clear: 40% of the overrun came from price and 60% came from 40 kg of dough lost above normal. If you only looked at unit cost, you would blame the supplier and miss the bigger part.

Where scrap gets recorded

Scrap must be recorded on the order, at the moment and at the stage where it happens, not inferred at month end from an inventory difference. A good record includes:

• the order and the product;
• the stage (weighing, mixing, baking, cutting, packing);
• the quantity, in the same unit as the recipe;
• the cause, picked from a short list (burned, out of spec, broken, expired, spill, machine setup);
• whether it is recoverable (rework) or lost.

Also record what is left over and goes back to the warehouse. If you issued 560 kg and returned 20 kg, consumption is 540 kg; if you don’t record the return, you inflate scrap and inventory stops reconciling.

Analysis by stage and by product

With stage-level records, the example order looks like this:

Stage	Scrap (kg)	% of total
Weighing and mixing	10	11%
Baking	25	28%
Cutting and forming	35	39%
Packing	20	22%
Total	90	100%

Part of the baking loss is normal (water evaporates) and is already in the standard. Where you need to look is cutting and forming: almost 4 of every 10 kg lost come from there.

Do the same monthly, by product: standard scrap, actual scrap and the excess valued in Bs. Sort from largest to smallest excess and focus on the top two or three; they almost always explain most of the problem.

Actions to reduce scrap

1. Start with the stage and product with the largest excess in Bs, not the highest percentage: 30% scrap on a product you barely sell weighs less than 5% on your main one.
2. Calibrate scales and dosing equipment. A small, constant weighing error repeats in every batch.
3. Standardize parameters (temperatures, times, speeds) and write them on the order’s work sheet.
4. Check the first piece after every product change or machine adjustment; much of the start-up scrap is avoided this way.
5. Use inputs in expiry order. Expired stock is scrap too; we explain the difference between consuming by FIFO and by FEFO in FEFO vs FIFO.
6. Separate what is recoverable: if an out-of-spec product can be reworked, record it as rework, not as a loss.
7. Review the standard twice a year, using recorded data, not gut feeling.

Common mistakes

• Measuring scrap only from inventory differences. It mixes theft, counting errors and real scrap, and it reaches you too late.
• A standard that includes all historical scrap. If the standard already allows 17%, the quantity variance is always zero and nothing improves.
• Not separating price and quantity. You argue with the supplier about what is really a plant problem, or the other way around.
• Different units across recipe, warehouse and plant (bags, kg, units) without a clear conversion.
• Not recording leftover returns to the warehouse.
• Forgetting the accounting and tax treatment. How scrap is documented and booked (and whether it is deductible) depends on current regulations; check it with your accountant before you define the procedure.

Monthly checklist

• Every closed order has actual consumption, good output and scrap by stage with its cause.
• Leftovers returned to the warehouse are recorded.
• You calculated yield and scrap rate by product.
• You separated the price variance from the quantity variance.
• You identified the three products or stages with the largest excess in Bs.
• Each one has an action with an owner.

If you want to automate it

All of this works in spreadsheets while volume is manageable. As it grows, an ERP like MKA can record scrap and costs directly on each production order while it consumes inputs by FEFO, so the real cost per lot doesn’t have to wait for month-end close (see production orders).