Yield variance: where the raw material actually went

How to calculate yield variance in production, in units and in rupees. Standard versus actual yield, worked through one batch, and mistakes that hide losses.

Yield variance is the gap between what a batch should have produced from the raw material you put in and what it actually produced. In units: standard output for the actual input − actual output. In rupees, multiply that shortfall by the standard cost per unit of output. If 1,000 kg of input should give 940 kg of finished goods and you got 910 kg, you are 30 kg short, and those 30 kg have a rupee value you can name.

Why does yield variance matter for a small factory?

Raw material is usually the largest cost in a batch, and yield is where it disappears without anyone noticing. Stock registers show what went into production and what came out. They do not tell you, on their own, whether the gap between the two was normal.

A small daily shortfall seldom gets looked at. Over a month of batches it adds up to raw material you paid for, carried GST on, and then lost to spillage, over-trimming, moisture loss, rejects or plain recording errors. Yield variance turns that silent loss into a number per batch, so you can ask where it went.

What is standard yield and what is actual yield?

Standard yield is the output you expect from a given input when the process runs normally. You set it from your recipe or bill of materials and from your own record of good batches. It is not the theoretical maximum, and it is not a figure copied from someone else's plant.

Actual yield is what you weighed or counted at the end of the batch, as saleable finished goods. Rework and rejects do not count as output until they have actually been recovered.

Standard yield is usually written as a ratio:

Standard yield ratio = standard output ÷ standard input

How do you calculate yield variance in units?

Two equivalent ways, and it helps to know both.

Output view: what should this input have made?

Standard output for actual input = actual input × standard yield ratio
Yield variance (units of output) = standard output for actual input − actual output

Input view: how much input should this output have needed?

Standard input for actual output = actual output ÷ standard yield ratio
Yield variance (units of input) = actual input − standard input for actual output

A positive result means an unfavourable variance, because you lost more than planned. A negative result means a favourable one, because you got more than planned. A favourable variance is still worth checking, because it often points to a weighing or entry error rather than a better process.

How do you convert yield variance into rupees?

Value it at standard cost, not at the price you happened to pay this week and not at selling price.

From the output view: units short × standard material cost per unit of output
From the input view: excess input × standard purchase cost per unit of input

Both should give the same rupee figure. If they do not, check your arithmetic and the cost you used.

Can you work one batch through, start to finish?

The numbers below are illustrative, for a spice-grinding unit making chilli powder.

Item	Standard	Actual
Raw chilli input	1,000 kg	1,000 kg
Chilli powder output	940 kg	910 kg
Standard cost of raw chilli	₹180 per kg	—

Step 1: Standard yield ratio. 940 ÷ 1,000 = 0.94

Step 2: Standard output for the actual input. 1,000 × 0.94 = 940 kg

Step 3: Yield variance in units of output. 940 − 910 = 30 kg short (unfavourable)

Step 4: Standard material cost per kg of output. (1,000 × ₹180) ÷ 940 = ₹180,000 ÷ 940 ≈ ₹191.49 per kg of powder

Step 5: Yield variance in rupees (output view). 30 × ₹191.49 ≈ ₹5,745

Cross-check from the input view:

Standard input for 910 kg of output = 910 ÷ 0.94 ≈ 968.09 kg
Excess input = 1,000 − 968.09 ≈ 31.91 kg
31.91 × ₹180 ≈ ₹5,744

The two results differ by a rupee only because of rounding. This batch used about 32 kg of chilli more than it should have, and that cost roughly ₹5,745. That is the figure to take to the production floor. "Yield was a bit low" is not.

If this unit runs one such batch a day, the loss adds up across a month. The next question is not "is that a lot?" but "which step lost it?"

Where does the raw material actually go?

Once you have the number, trace it. Work through the batch stage by stage and weigh at each point where you can.

Receiving: was the input weight at the godown the same as the supplier's invoice? Short receipts show up later as poor yield.
Moisture and cleaning: stones, stems, dust and water lost in drying are real losses. They belong in the standard if they are normal, and in the variance if they are not.
Process loss: spillage, dust from the grinder, product left in machines and hoppers at changeover.
Rejects and rework: off-spec output kept aside and never recorded, or recorded twice.
Recording: a batch logged as 1,000 kg because that is the usual figure, when the scale said 1,020 kg.

A simple batch sheet with a weight at each stage usually explains most of the variance within a few batches.

When does this calculation not apply as written?

Multiple raw materials in different proportions. If you use several inputs and their mix changes from batch to batch, part of the gap is a mix variance, not a yield variance. Split the two before you blame the process.
By-products with value. If the bran, husk or offcuts are sold, record them as output at their own value. Treat them as loss and the yield will look worse than it really is.
Discrete assembly. If you assemble 100 units from counted parts, a yield calculation in kg is the wrong tool. Track component usage and scrap per unit against the bill of materials instead.
Price changes. If the raw material cost moved this month, put that in price variance. Keep yield variance at standard cost so the two do not blur.

What are the common mistakes?

Setting the standard from your best ever batch. Then every batch shows a loss and people stop reading the report. Base the standard on normal, good batches.
Never updating the standard. A new machine, supplier or recipe changes what normal looks like. Review the standard when the process changes.
Valuing at selling price. This inflates the loss and mixes margin into a cost figure.
Netting batches together. A good batch and a bad batch averaged together hide both. Calculate per batch first, then total.
Ignoring work in progress. Material sitting half-processed at month end is not lost. Count it before you call it variance.
Not adjusting purchasing. If your real yield is consistently below the standard, your raw-material requirement is understated. That feeds straight into your reorder point and you will run short.

How does yield variance connect to the rest of your inventory?

Yield sets how much raw material each unit of finished goods really consumes. If your system still plans with the standard yield while the floor runs below it, raw-material stock will drift from the books. Every stock-take will then show an unexplained shortage. Recording actual input and output per batch keeps the raw-material and finished-goods ledgers honest. It is also one of the stronger reasons to move production records out of Excel and into inventory software. The same discipline of recording usage against a standard sits behind food costing in a restaurant kitchen.

FAQ

What is the difference between yield variance and price variance? Price variance asks whether you paid more or less per kg than planned. Yield variance asks whether each kg you bought turned into as much finished goods as planned. Keep them separate. A cheaper supplier can look like a saving on price while quietly costing you more through poorer yield.

Should I value yield variance at purchase price or at finished-goods cost? Use the standard cost of the raw material, either per kg of input or per unit of output. Both give the same rupee figure if you use them correctly. Do not use the selling price. That mixes lost margin into a cost number and makes the variance look bigger than it really is.

How often should I calculate yield variance? Calculate it for every batch if you can, or at least every week for continuous production. A monthly figure is useful for the accounts, but by then the cause is hard to trace. Batch-level numbers let you tie a loss to a shift, a machine or a lot of raw material while people still remember what happened.