How to roll every material's landed cost, plus the labor and overhead, up through a recipe — including the components you make yourself — into a cost per batch and per unit you can stand behind.
Do these first
Real help from real food people
Chef Diego runs a real food plant. If this page didn't get you there, tell us — a person reads every message.
After this you can take a full recipe — every ingredient, the packaging, the labor, and the room it runs in — and roll it up into two honest numbers: what a batch costs, and what a single finished unit costs. That per-unit number is the one you price against, so it has to carry everything, including any component you make yourself before you make the product.
The last two lessons built the pieces this one assembles: the three buckets a product's cost is made of, and the true landed cost of a single material. Now we put them together, recipe by recipe. A recipe tells a person how to make the product. To cost the product you need the same recipe written as data: each material, the exact quantity it takes, and the unit you count that quantity in. That structured version is the . You cost against the BOM, not the recipe card, because only the BOM pins down the quantities.
Turning a recipe into a clean BOM — scaling a batch up or down, choosing the right unit of measure, capturing waste — is its own subject, and its own lesson. Here we start from a BOM that already exists and ask a narrower question: given the quantities, what does the batch cost?
Take each line of the BOM, multiply the quantity by that material's landed cost per unit, and add the lines together. That sum is the batch's material cost.
The word that matters is landed. The last lesson's whole point was that a material costs more than its invoice price once freight, duties, and handling are in. So every quantity here is multiplied by the landed cost per unit, not the invoice price — otherwise the roll-up inherits the same optimism, one line at a time.
Take a barbecue sauce maker running a batch of 200 twelve-ounce jars. The figures below are made up for illustration; the shape is what matters. Start with the materials you buy in, each at its landed cost:
One line is missing on purpose. This sauce also uses a house spice blend — and the maker does not buy that blend, they make it. It cannot be costed off an invoice, because there is no invoice. That line is where roll-up gets interesting.
The blend is a . You cannot drop its raw spice cost into the sauce, because the blend took its own labor and its own overhead to make. You have to cost the blend as its own little batch first, land on a cost per pound, and then bring it into the sauce at that number.
Cost the blend batch — say it makes 10 pounds:
The grinder loses a little, so the batch yields 9 usable pounds, not 10. That makes the blend cost $90 ÷ 9 = $10.00 a pound. This sauce batch draws 2 pounds of it, so the blend enters the sauce's BOM at 2 × $10.00 = $20.
That $20 already carries the blend's own labor and its own overhead. Pulling a made-in-house component into its parent recipe at its fully built cost — not its raw material cost — is the heart of a .
Do not count the blend's labor twice
The $20 blend line already includes the hour it took to make the blend. The sauce batch's labor, coming up next, is only the cooking, filling, and packing of the sauce — not the blend. Add the blend's hour again down there and you have double-counted it. Each hour of work belongs to exactly one level of the recipe.
With the blend costed, the materials are complete: $220 bought-in plus the $20 blend is $240 of material in the batch.
Materials are one of the three buckets. The batch still has to carry its direct labor and its share of overhead — the same two buckets from the first lesson, now counted for this specific run.
Add the three buckets and you have the batch cost:
The $450 is the cost per batch — useful on its own for a production run, and it is what sits in finished-goods inventory until the jars actually sell and it turns into COGS. (That inventory-cost principle is the same one the first lesson drew; IRS Publication 538 covers how inventory cost is identified for tax.)
For pricing you need the cost of one jar, which means dividing the batch cost by the number of jars. The trap is which number. The recipe is written for 200 jars, but a real run does not hand you 200 sellable ones — a few are lost to spillage, a filling misstep, and whatever a QA check pulls. What you can actually sell is the batch's , and say this run yields 190 good jars.
Divide by the yield you can sell, not the count you hoped for:
The lost jars still cost money
The ten jars you could not sell did not cost nothing — their materials, labor, and overhead are already in the $450. Costing on 200 pretends those ten were free and spreads the batch too thin. Costing on 190 puts their cost where it belongs: on the jars that survived. Twelve cents a jar, about five percent, is the difference between the number you hoped for and the one you can defend.
So the roll-up lands on two numbers you can stand behind: $450 a batch, and $2.37 a good jar — every material at landed cost, the blend pulled up at its own built cost, labor and overhead counted once each, and the result spread across the jars you can actually sell.
That $2.37 is true for exactly one set of prices: the ones behind it today. Move one material's landed cost — the tomato paste supplier raises prices, a fuel surcharge lands on a freight bill, the spice blend gets more expensive — and the roll-up shifts. Worse, it shifts twice for a nested component: a costlier spice changes the blend's cost per pound, which changes the sauce that uses the blend.
Costing one recipe by hand, once, is a good exercise, and you should do it — it is the fastest way to feel where your money actually goes. Costing every recipe, pulling each made-in-house component up at its current rolled cost, and redoing the whole roll-up every time a material's landed cost moves is the kind of arithmetic a system built for food carries for you. Bettr Manager, for one, computes a product's cost straight from its recipe, rolls nested components up first, and re-costs the moment a material's landed cost changes. However the math gets done, the number is only ever as honest as the landed costs, the labor, and the yield you put behind it.
Keeping that number current as your input prices swing — so a margin call uses today's cost, not last quarter's — is the next lesson.
Do this in your operation