---
title: Yield, wastage and by-product formulas every food factory should track
canonical: https://naffo.tech/blog/yield-wastage-formulas-food-manufacturing
question: What are the formulas for yield, wastage and by-product recovery in food manufacturing?
published: 2026-03-04
updated: 2026-08-08
author: naffo.tech manufacturing team (Plant systems and costing)
reviewed_by: naffo.tech implementation desk (Dairy and food plant rollouts)
publisher: naffo.tech — https://naffo.tech
category: Costing & metrics
tags: yield, wastage, formulas, dairy, costing, by-product
reading_time_minutes: 10
license: Free to quote with attribution to naffo.tech (https://naffo.tech/blog/yield-wastage-formulas-food-manufacturing)
---

# Yield, wastage and by-product formulas every food factory should track

**Question:** What are the formulas for yield, wastage and by-product recovery in food manufacturing?

**Answer:** Yield % is good sellable output ÷ total input × 100. Wastage % is actual wastage ÷ total input × 100, excluding recoverable by-product. By-product recovery % is recovered by-product ÷ theoretical by-product × 100. Wastage cost is wasted quantity × material cost plus the value added up to the loss point. Every one of these needs a stated basis (kg per litre, or kg per kg) and a fixed rule on whether rework and by-product count as output.

## Key takeaways

- A yield number without a stated basis is meaningless. "18% paneer yield" only means something as kg of paneer per litre of milk at a stated fat percentage.
- Exclude recoverable by-product from wastage and include it in a separate recovery metric — otherwise a well-run paneer plant looks like it wastes 76% of its milk.
- Cost loss at the value it had when it was lost, not at raw-material rate. Loss after packing carries material, energy, labour and film.
- Theoretical yield comes from the recipe; actual yield comes from the batch. The gap, expressed in rupees, is your improvement budget.
- Fat- and solids-corrected yield is the only fair comparison in dairy, because input quality moves daily and volumetric yield moves with it.
- Give-away (systematic overfill on a filler) is a real loss category with its own formula and usually its own uninvestigated cost.

## Key figures

- **Always** — Fat-corrected basis is required for fair yield comparison in dairy (Basis: Volumetric yield moves with incoming fat and SNF; comparing raw litre-based yields across days compares milk quality, not plant performance.)

## Fix the definitions before you fix the numbers

In most factories, production, QC and accounts each report a different loss figure for the same batch — and each is arithmetically correct, because they are using different definitions. Three questions settle it, and they must be answered once, in writing, for the whole plant:

1. **Is rework output or loss?** Rework is neither: it is a temporary state. Count it separately at batch close and account for it in the batch it is eventually consumed in. Counting rework as good output inflates yield; counting it as loss double-counts when it is later sold.
2. **Is by-product output?** Yes, but not in the yield of the main product. Whey is not paneer. Give it its own item code, its own rate and its own recovery metric.
3. **On what basis is yield expressed?** Output kg per input litre, output kg per input kg, or fat-corrected equivalents. State it in the report header, permanently.

> **WARNING**
>
> The single most common reporting error in dairy is treating whey as wastage. A paneer plant converting 1,000 L of milk into 180 kg of paneer and 760 L of whey does not waste 76% of its input — it recovers most of it as a by-product with its own market. Getting this wrong makes every wastage report unusable.

## The core formulas

**1. Yield %**

```
Yield % = Good sellable output ÷ Total input × 100
```
Good sellable output excludes rework, rejection and by-product. Always publish the basis alongside the number.

**2. Process loss %**

```
Process loss % = (Input − Good output − By-product − Rework − Rejected) ÷ Input × 100
```
This is what genuinely disappeared in the process — evaporation, adherence to vessels, drain, trim. It should be stable batch to batch; instability means process variation, not material variation.

**3. Wastage %**

```
Wastage % = Actual wastage quantity ÷ Total input quantity × 100
```
Actual wastage = spillage + contamination + expiry + unrecoverable rejection. It excludes recoverable by-product and excludes rework that will be consumed.

**4. Wastage cost**

```
Wastage cost = Wasted quantity × (Material cost + Value added up to loss point)
```
Value added includes energy, labour, packaging and overhead already absorbed. A kilo lost at intake and a kilo lost after sealing are not the same rupee.

**5. By-product recovery %**

```
By-product recovery % = Recovered by-product ÷ Theoretical by-product × 100
```
Theoretical by-product comes from the recipe. Recovery below about 95% usually means drain loss at transfer points.

**6. Yield variance vs standard**

```
Yield variance = (Actual yield % − Standard yield %) ÷ Standard yield % × 100
```
Report this signed. A positive variance is not automatically good news — it can indicate under-dosing, moisture retention or a weight-spec breach that QC will catch later.

**7. Rework rate**

```
Rework rate % = Rework quantity ÷ Total output (good + rework + rejected) × 100
```
Rising rework with flat rejection means the process is drifting but QC is catching it. That is a warning, not a success.

**8. Rejection rate**

```
Rejection rate % = Rejected quantity ÷ Total output × 100
```
Break down by reason code. An aggregate rejection rate tells you the size of the problem and nothing about the cause.

**9. Give-away (overfill) %**

```
Give-away % = (Average actual fill weight − Declared weight) ÷ Declared weight × 100
```
Legal metrology requires you to stay above the declared weight, so some overfill is deliberate. Systematic overfill beyond the required tolerance is free product. On a 200 g pack running 2 g heavy, that is 1% of everything you ship.

**10. Packaging loss %**

```
Packaging loss % = Packaging material damaged or scrapped ÷ Packaging material issued × 100
```
Count units for pouches, bottles, caps, labels and cartons; count metres for film. Changeovers are the biggest single driver.

**11. Expiry write-off %**

```
Expiry write-off % = Value of expired stock ÷ Value of finished goods produced in the period × 100
```
Track by SKU, not in aggregate. Expiry is almost always concentrated in a few slow movers that the plan keeps producing on autopilot.

**12. Overall material efficiency**

```
Material efficiency % = (Good output value + By-product value) ÷ Total input value × 100
```
The one number for the owner. It is denominated in rupees, so it cannot be gamed by unit tricks, and it moves only when something real improves.

## Worked example — paneer

> **Paneer batch, 1,000 L milk**
>
> Input: **1,000 L** buffalo milk at **6.0% fat**, cost **₹52/L** → input value **₹52,000**. Standard yield: **18.5 kg per 100 L**.
>
> Actual output: good paneer **181 kg**; whey tanked **755 L** (theoretical 780 L); rework **4 kg**; QC rejection **2 kg**; measured spillage **5 L** milk equivalent.

_Paneer batch — metric by metric_
| Metric | Calculation | Result |
| --- | --- | --- |
| Yield % | 181 kg ÷ 1,000 L | **18.1 kg / 100 L** |
| Yield variance | (18.1 − 18.5) ÷ 18.5 × 100 | **−2.2%** vs standard |
| By-product recovery % | 755 ÷ 780 × 100 | **96.8%** |
| Rework rate | 4 ÷ (181 + 4 + 2) × 100 | **2.1%** |
| Rejection rate | 2 ÷ 187 × 100 | **1.1%** |
| Actual wastage | 5 L spillage + 2 kg rejection (≈ 11 L milk equivalent) | **≈ 16 L equivalent** |
| Wastage % | 16 ÷ 1,000 × 100 | **1.6%** |
| Wastage cost | 16 L × ₹52 (plus conversion on the rejected paneer) | **≈ ₹950 per batch** |
| Yield shortfall cost | 4 kg short × ₹310/kg realisable | **≈ ₹1,240 per batch** |

The instructive part is the last two rows. The visible wastage costs about ₹950; the invisible yield shortfall costs about ₹1,240 — and it appears in no wastage register anywhere. **Across 25 batches a month that is roughly ₹31,000 of product that was never lost, spilled or rejected. It simply never formed.** Yield variance, not wastage, is usually the bigger number.

## Worked example — ghee

> **Ghee from cream**
>
> Input: **500 kg** cream at **60% fat** = 300 kg fat. Ghee is roughly **99.5% fat**, so theoretical ghee ≈ **301 kg** at full fat recovery; a realistic standard allows a **1.5–2%** fat loss to residue and vessel adherence.
>
> Actual: ghee **293 kg**; ghee residue collected **9 kg** (sellable by-product); balance is moisture driven off plus unrecovered fat.

- **Fat-basis yield** = 293 × 0.995 ÷ 300 × 100 = **97.2% fat recovery**. This is the only honest ghee yield metric, because cream fat varies daily.
- **By-product**: 9 kg of residue at a realisable rate is a recovery line, not wastage. Plants that drain residue lose a small but perfectly recoverable margin every batch.
- **Do not** report ghee yield as 293 ÷ 500 = 58.6%. That number falls whenever cream fat falls, and the plant gets blamed for the supplier's milk.

## Worked example — curd and bakery

_Two more common cases, and the metric that actually matters in each_
| Product | Input | Output | The metric that matters |
| --- | --- | --- | --- |
| Set curd, 200 g cups | 1,000 L standardised milk | 4,850 cups filled, 60 cups rejected for seal failure | **Fill efficiency** = 4,850 × 0.2 ÷ 1,000 × 100 = 97% of milk into saleable cups; plus **packaging loss** on the 60 seal failures, which is a machine problem, not a milk problem |
| Bread, 400 g loaves | 100 kg flour + water, fat, yeast per recipe | 168 loaves, 5 kg dough trim, 3 loaves under-weight | **Baker's yield** (loaves per 100 kg flour) plus **dough trim recovery %** — trim reworked into the next batch is recovery, trim binned is wastage |
| Frozen snack, 500 g packs | 250 kg prepared mix | 470 packs, 6 kg broken product, 2.1 kg average overfill | **Give-away %** = 2.1 ÷ 235 × 100 ≈ 0.9% — invisible in every stock report and worth more than the broken product |

## Standard vs actual: where the improvement budget comes from

Theoretical (standard) yield comes from the recipe and the science. Actual yield comes from the batch. The disciplined version of continuous improvement is simply: keep a standard, measure the gap in rupees, attribute the gap, close the biggest one.

**Annualised value of a one-point yield gain**

```
Annual gain = (Yield gain % ÷ 100) × Annual input quantity × Realisable value per output unit
```
Worth calculating before any capex conversation. On 3,00,000 L of milk a year at ₹310/kg paneer, a single percentage point of yield is roughly ₹9.3 lakh — which reframes what a ₹2 lakh press upgrade is worth.

The attribution step is where a system beats a spreadsheet. Yield has to be sliceable by product, recipe version, shift, operator, machine and incoming lot; otherwise you know the gap exists and not where it lives.

## Reporting rules that keep the numbers trustworthy

- **One basis per product, stated in the report header.** Never mix kg-per-litre and kg-per-kg in the same table.
- **Fat- or solids-corrected in dairy.** Compare like with like or you are measuring your milk supplier, not your plant.
- **Rupees next to every percentage.** Percentages set priorities badly; a 0.4% packaging loss can outrank a 3% whey loss in cost.
- **Batch-level granularity retained.** Averages are for trends; batches are for causes.
- **Reason codes mandatory, free text optional.** Anything you cannot group, you cannot fix.
- **Recipe version stamped on every batch.** Otherwise a yield step-change cannot be attributed to the recipe change that caused it.

Next: [the batch-wise control system these formulas plug into](/blog/reduce-wastage-food-factory), and [the dashboard an owner should read in five minutes](/blog/owner-dashboard-manufacturing-business).

## Frequently asked questions

### What is the formula for yield percentage in food manufacturing?

Yield % = good sellable output ÷ total input × 100, where good sellable output excludes rework, rejection and by-product. The basis must be stated — for example kg of paneer per 100 litres of milk. In dairy, correct for incoming fat or total solids, otherwise day-to-day yield movement reflects milk quality rather than plant performance.

### How do you calculate paneer yield from milk?

Divide good paneer output by milk input and express it per 100 litres: 181 kg from 1,000 L is 18.1 kg per 100 L. Compare against a standard yield for the same milk fat percentage, and report the variance. Whey is recorded separately as a by-product with its own recovery percentage, never as wastage.

### What is the difference between process loss and wastage?

Process loss is material that inherently disappears in the process — evaporation, vessel adherence, trim — and should be stable batch to batch. Wastage is avoidable loss: spillage, contamination, expiry and unrecoverable rejection. Mixing them makes the number impossible to act on, because process loss is reduced by engineering and wastage is reduced by discipline.

### What is give-away in food packaging?

Give-away is the systematic overfill above declared weight on a filling line. Legal metrology requires you to stay above the declared weight, so a small margin is intentional, but overfill beyond that tolerance is free product. Give-away % = (average actual fill − declared weight) ÷ declared weight × 100. Two grams on a 200 g pack is 1% of everything you ship, and it appears in no stock report.

### How do you cost wastage correctly?

Cost it at the value the material had at the point it was lost: material cost plus energy, labour, packaging and overhead already absorbed. Costing all loss at raw-material rate understates finished-goods loss substantially, which is why plants often under-prioritise packing-line and warehouse losses that are far more expensive per kilo than intake losses.

### Should rework be counted in yield?

No. Rework is a temporary state, not output. Record it separately at batch close and account for it in the batch that eventually consumes it. Counting rework as good output overstates yield; counting it as loss double-counts once it is sold. A rising rework rate with a flat rejection rate is an early warning that the process is drifting.

## References

- [Legal Metrology (Packaged Commodities) Rules — declared quantity and permitted error, Government of India](https://consumeraffairs.nic.in/organisation-and-units/division/legal-metrology)
- [FSSAI — standards for milk and milk products](https://www.fssai.gov.in/)

## Related articles

- https://naffo.tech/blog/reduce-wastage-food-factory
- https://naffo.tech/blog/ideal-erp-workflow-food-manufacturing
- https://naffo.tech/blog/owner-dashboard-manufacturing-business

---

Published by naffo.tech, an all-in-one business management and manufacturing ERP platform for Indian SMEs — GST compliance, invoicing, inventory, batch manufacturing, yield and wastage control, and double-entry accounting in one system. https://naffo.tech

Markdown source: https://naffo.tech/blog/yield-wastage-formulas-food-manufacturing/markdown