What a Meal Really Costs Per Serving

Waste costs more than you'd think

If 30% of a package is bone, skin or peel, the edible part is 70% of the weight — and it carries 100% of the price.

So the cost per edible gram isn't up by 30%. It's 1 ÷ 0.7, which is 43%.

That's the whole idea, and it catches people out because the instinct is to subtract when the arithmetic divides. It also accelerates:

Waste	Multiply the price by
10%	1.11
20%	1.25
25%	1.33
30%	1.43
40%	1.67
50%	2.00
75%	4.00

The rule is 1 ÷ (1 − waste), never 1 + waste.

True cost per serving sits next to If you ignored the waste so you can see the gap on your own numbers. Set the waste to zero and the two agree exactly — which is a handy check that you've entered everything correctly.

Measure your own waste

Published yield figures for the same ingredient disagree, because they depend on the cut, the grade, the season and how carefully somebody trims. There's no honest single answer to "how much of a chicken is meat".

So weigh it:

1. Weigh the package as it comes
2. Prepare it exactly as you normally would
3. Weigh what you actually cooked

The difference over the original weight is your waste. Do it once for the five or six things you buy regularly — it takes an afternoon and the figures stay good for years.

The shelf label is flattering the wrong products

Supermarket unit pricing is calculated on the package weight. It has to be — the shop doesn't know how you'll trim it.

That means it systematically flatters anything with bone, shell, peel or heavy trim. A bone-in cut always looks better on the shelf label than it works out in the pan, and two products can look 30% apart on the label and be identical once you're comparing what you eat.

Cost per 100 (edible) is the figure that actually decides which is cheaper. That's the number to compare between shelves.

Bulk is only cheaper if you eat it

This is why the tool reports two different totals.

What you SPEND is whole packages — what leaves your wallet. What you USE is the value of the servings you actually wanted. At the defaults, four servings from a single pack is $5.71 of food out of $12.00 spent.

The difference isn't a loss and it isn't a saving. It's food in your fridge, and which one it becomes depends entirely on whether you eat it.

That's the honest test for "is the big pack cheaper". A large pack at a lower unit price that half rots is more expensive than a small pack at a higher one — and the unit price on the shelf can't tell you which happened, because that part is about you.

Packages to buy always rounds up, because you can't buy 0.48 of a chicken.

Use one unit and stick to it

Package weight or volume and Serving size have to be in the same unit. Grams for solids, millilitres for liquids, or a plain count — a dozen eggs is a package of 12 with a serving size of 2.

Mixing units is the main way this calculation goes wrong, and it goes wrong quietly: the answer still looks like a plausible price, so nothing prompts you to check it.

Note that Serving size means what reaches the plate, not what you started with. Edible weight in the package and Servings per package both follow from that, so getting it wrong shifts everything downstream.

What this deliberately leaves out

Energy to cook it, and your time.

Both are real. Neither belongs in a grocery comparison — energy costs are small per meal and vary enormously by appliance and tariff, and folding your time in makes comparing two packages harder rather than easier. If you want to account for them, do it once at the level of a whole meal.

The same goes for oil, salt and seasoning. Rarely worth costing individually.

Costing a whole meal

Run each significant ingredient separately and add the per-serving figures.

In practice two or three ingredients account for most of the cost of a meal — the protein, and anything out of season. The rest rounds to noise.

It's worth doing properly once for the meals you cook often. The result tends to be surprising in both directions, and more usefully, it identifies the one substitution that actually changes the number instead of the five that don't.