---
title: How to cost a manufactured food product correctly, including by-product allocation
canonical: https://naffo.tech/blog/product-costing-food-manufacturing-by-product-allocation
question: How do you calculate the cost of a manufactured food product?
published: 2026-06-17
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: costing, by-product, margin, dairy, food manufacturing, batch cost
reading_time_minutes: 10
license: Free to quote with attribution to naffo.tech (https://naffo.tech/blog/product-costing-food-manufacturing-by-product-allocation)
---

# How to cost a manufactured food product correctly, including by-product allocation

**Question:** How do you calculate the cost of a manufactured food product?

**Answer:** Cost a food product from the actual batch, not the recipe. Take material issued at landed cost, add conversion cost (labour, energy, packaging, direct overhead) absorbed over actual good output, subtract the realisable value of recovered by-product, and divide by the good output quantity. Costing at standard recipe rates hides yield loss, give-away and rework, which is where most of the margin actually leaks.

## Key takeaways

- Recipe cost is a plan. Batch cost is the truth. The gap between them, in rupees, is your margin leak.
- By-product value must reduce the main product's cost, or every product with a saleable by-product looks unprofitable.
- Landed cost, not invoice cost. Freight, loading, transit loss and non-creditable taxes belong in material cost.
- Absorb conversion cost over actual good output, not planned output. Absorbing over plan makes a bad-yield batch look normal.
- Rework carries cost forward. If reworked material is treated as free input to the next batch, the next batch looks better than it is.
- Recompute cost per batch, the same shift. A monthly weighted-average cost tells you what happened, never which batch caused it.

## Key figures

- **4** — Cost layers in a defensible per-kilo product cost (Basis: Landed material cost, conversion cost absorbed over actual good output, by-product credit at net realisable value, and rework cost carried forward.)

## Why your recipe cost and your bank balance disagree

Most food manufacturers price from a recipe cost sheet built once, in Excel, at standard yield and last year's rates. Then margin quietly disappears in five places the sheet does not model: yield shortfall, rework, give-away on the filler, by-product drained instead of sold, and freight treated as an expense rather than as part of material cost.

A defensible product cost has four layers, computed per batch:

**Batch cost of good output**

```
Cost per unit = (Landed material cost + Conversion cost + Rework carried in − By-product credit) ÷ Actual good output
```
Every term must come from the batch record, not from the recipe. The recipe supplies the standard you compare against.

## Layer 1 — landed material cost

Material cost is not the supplier's invoice line. It is what the material actually cost to have available, usable, at your plant.

_What belongs in landed cost, and what does not_
| Include | Why | Exclude | Why |
| --- | --- | --- | --- |
| Supplier invoice value net of creditable GST | Input tax credit is recoverable, so it is not a cost | Creditable GST | Recoverable — including it overstates cost and distorts pricing |
| Inward freight and loading/unloading | Directly attributable to getting material usable | General office overhead | Not attributable to material |
| Non-creditable taxes, cess and duties | Genuinely unrecoverable | Selling and distribution cost | Belongs after cost of goods, not inside it |
| Transit and measurement loss on receipt | You paid for it and did not receive it | Financing cost of credit period | Track separately as working-capital cost, not product cost |
| Quality-rejection rate adjustment | Conditional acceptance changes the real rate paid | Abnormal spoilage from a chiller failure | Charge to the period as an abnormal loss, not to the product |

> **TIP**
>
> In dairy, land the cost on a **fat and SNF basis**, not per litre. Two tankers at the same rupee-per-litre are different costs per kilo of paneer if their fat differs. Rate-per-kg-fat is the only comparison that survives seasonal milk variation.

## Layer 2 — conversion cost, absorbed over actual output

Conversion cost is everything you spend turning material into sellable product: direct labour on the batch, energy (steam, power, refrigeration), consumables, packaging material, direct maintenance, and plant overhead attributable to production time.

**Conversion rate**

```
Conversion rate = Total conversion cost for the period ÷ Total actual good output for the period
```
Refresh monthly. Then apply it to each batch's actual good output. Absorbing over planned output is the classic error — it makes a low-yield batch look normal, because the shortfall silently lands in an unexplained variance nobody reads.

> **WARNING**
>
> Packaging is conversion cost, and it is where costing sheets are most often stale. Film, pouches, caps, labels and cartons move in price frequently and are consumed with wastage on every changeover. A costing sheet using last year's film rate at zero packaging loss can be several percent wrong on a low-margin SKU — enough to make a product you believe is profitable a loss-maker.

## Layer 3 — by-product credit, three defensible methods

This is the layer that decides whether your numbers make sense in a dairy or food plant. Convert 1,000 L of milk into paneer and whey, charge all the milk to the paneer, and paneer looks unprofitable while whey looks free. Both conclusions are wrong.

_Three by-product allocation methods, and when each is appropriate_
| Method | How it works | Use when | Watch out for |
| --- | --- | --- | --- |
| **Net realisable value (NRV) credit** | Credit the by-product's realisable value (sale price less cost to sell) against the batch cost; main product absorbs the balance | One clear main product plus minor by-products — paneer/whey, ghee/residue, bakery trim | Overstating realisable value for a by-product you do not actually sell every week |
| **Sales-value split (joint costing)** | Split batch cost across all outputs in proportion to their sales value | Genuine joint products of comparable value — cream and skimmed milk from separation | Sales-value moves, so cost moves; document the rate source and freeze it monthly |
| **Physical-measure split** | Split cost by weight, volume, or fat/solids content | Outputs with similar value per unit, or when fat is the real cost driver | Nonsense results when values differ sharply — 760 L of whey should not absorb 76% of milk cost |

> **Paneer and whey, NRV method**
>
> Input: **1,000 L** milk at landed **₹52/L** = **₹52,000**. Conversion at **₹18/kg** of paneer output. Output: **181 kg** paneer, **755 L** whey with a realisable value of **₹2.50/L** net of handling.
>
> By-product credit = 755 × ₹2.50 = **₹1,888**. Conversion = 181 × ₹18 = **₹3,258**.
>
> Cost per kg paneer = (52,000 + 3,258 − 1,888) ÷ 181 = **₹294.86/kg**.
>
> Without the whey credit the same batch costs **₹305.29/kg** — a **₹10.43/kg** overstatement, about 3.4%. On 4,500 kg a month that is roughly **₹47,000 of phantom cost** that would push you to price higher than you need to, or to conclude a profitable SKU is marginal.

> **One rule about by-product credit**
>
> Only credit what you actually realise. If whey goes to drain, the credit is zero and the paneer must carry the full cost. Crediting a theoretical value for a by-product you throw away is how plants convince themselves they are profitable while the bank balance disagrees — and it also removes the incentive to start selling the by-product.

## Layer 4 — rework, and the trap of free input

Rework already carries material and conversion cost from the batch that produced it. If it enters the next batch at zero value, that batch's cost is understated and the original batch's cost is overstated — so yield comparisons between the two become meaningless.

- Value rework at the **cost of the batch that produced it**, capped at net realisable value if it will be sold as a lower grade.
- Carry that value into the consuming batch as an input line, visible on the batch cost sheet.
- If rework is scrapped rather than consumed, charge it to **wastage cost at the value it had when it failed** — not at raw-material rate.
- Track the **rework rate** separately. Rising rework with flat rejection means the process is drifting while QC catches it; costing alone will not reveal that.

## Standard versus actual: use both, for different jobs

_Two costs, two purposes — plants that keep only one always keep the wrong one_
|  | Standard (recipe) cost | Actual (batch) cost |
| --- | --- | --- |
| Built from | Recipe, standard yield, current rates | Batch record: issued material, actual output, actual conversion |
| Used for | Pricing, quotations, planning, budgeting | Margin truth, variance analysis, improvement targets |
| Updated | Monthly, or when a rate or recipe changes | Every batch, same shift |
| Fails when | Used as the margin figure | Used to quote a customer, because it fluctuates batch to batch |

**The variance that matters**

```
Cost variance per unit = Actual batch cost per unit − Standard cost per unit
```
Decompose it into material price variance, yield variance, conversion variance and by-product recovery variance. Each has a different owner: purchase, production, maintenance and sales respectively.

> **EXAMPLE**
>
> Standard paneer cost ₹288/kg; actual for the month ₹296/kg. The ₹8 splits into ₹3 material price (milk rate rose), ₹4 yield (18.1 against 18.5 standard) and ₹1 by-product recovery (whey drained on two batches).
>
> Three owners, three actions, one number. Without the decomposition, the entire conversation is "costs went up" — which nobody can act on.

## Margin: the four cuts worth having

- **Gross margin per SKU** = realisation − actual batch cost. Ranked worst-first, monthly.
- **Contribution per unit of constraint.** If your bottleneck is press hours or filler minutes, rank SKUs by contribution per bottleneck hour, not by percentage margin. The highest-margin product can be the worst use of a constrained line.
- **Customer margin after freight, discount and credit period.** Include the financing cost of the credit you extend. Large customers frequently rank lower than assumed.
- **Channel margin.** Distributor, institutional, retail and own-outlet realisations differ enough that a blended SKU margin can hide a loss-making channel entirely.

## Making this operational rather than an annual spreadsheet exercise

1. **Land the material cost at GRN** — freight, loading and rate adjustments captured on the receipt, not journalled separately at month-end.
2. **Stamp the recipe version on the batch** so a cost step-change can be attributed to the recipe change that caused it.
3. **Compute batch cost at batch close**, using actual good output and actual by-product recovered — the plant sees cost per kg the same shift.
4. **Refresh the conversion rate monthly** from the actual period cost and actual output.
5. **Publish a worst-first variance list weekly** with the four-way decomposition and an owner against each line.
6. **Re-price on actual, quote on standard.** Never confuse the two in the same conversation.

This is the costing layer that sits directly on top of the batch chain in [the ideal food ERP workflow](/blog/ideal-erp-workflow-food-manufacturing), using the metrics defined in [yield, wastage and by-product formulas](/blog/yield-wastage-formulas-food-manufacturing). naffo.tech's manufacturing module implements it natively: landed cost at GRN, versioned recipes, batch-wise material issue, by-product allocation that reduces the main product's effective cost, and per-batch costing that posts to double-entry accounting without a separate month-end exercise.

## Frequently asked questions

### How do you calculate the cost of a manufactured food product?

Cost per unit = (landed material cost + conversion cost + rework carried in − by-product credit) ÷ actual good output. Take every term from the batch record rather than the recipe: material at landed cost including inward freight and rate adjustments, conversion absorbed over actual good output, and by-product credited only at the value you genuinely realise.

### How should by-product value be allocated in dairy costing?

For a clear main product with minor by-products — paneer and whey, ghee and residue — credit the by-product's net realisable value against the batch cost so the main product absorbs the balance. For genuine joint products of comparable value, such as cream and skimmed milk from separation, split the batch cost in proportion to sales value. Never split by physical volume when values differ sharply.

### Should conversion cost be absorbed on planned or actual output?

Actual good output. Absorbing over planned output makes a low-yield batch look normal, because the shortfall disappears into an absorption variance nobody reviews. Absorbing over actual output makes a bad batch show a higher cost per kilo immediately, which is the signal you want on the same shift.

### What is landed cost in food manufacturing?

Landed cost is what material cost to have usable at your plant: supplier invoice value net of creditable GST, plus inward freight, loading and unloading, non-creditable taxes, transit and measurement loss on receipt, and any rate adjustment from conditional quality acceptance. It excludes creditable GST, selling costs and the financing cost of supplier credit.

### Why is my actual margin lower than my recipe cost sheet says?

Usually five reasons compounding: yield below standard, rework treated as free input, give-away above the required overfill tolerance, by-product drained instead of sold, and stale packaging or freight rates in the sheet. Decompose the variance into material price, yield, conversion and by-product recovery, and each gap gets a named owner instead of a general statement that costs went up.

### How often should product cost be recalculated?

Actual cost every batch, at batch close, so the plant sees cost per kilo the same shift. Standard cost monthly, or immediately when a material rate or recipe changes. Quote and price from standard; judge margin and run variance analysis from actual. Keeping only a monthly weighted average tells you what happened but never which batch caused it.

## Related articles

- https://naffo.tech/blog/yield-wastage-formulas-food-manufacturing
- https://naffo.tech/blog/reduce-wastage-food-factory
- 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/product-costing-food-manufacturing-by-product-allocation/markdown