Yield Percentage: Why Your Plate Costs Are Probably Too Low

The gap between what you buy and what you serve

Ask a chef what a kilogram of striploin costs and you will get the invoice price. It is the number on the delivery note, it is the number the supplier quotes, and it is the number that ends up in the recipe spreadsheet. It is also, for any ingredient that gets trimmed, wrong.

The invoice covers the whole primal — fat cap, sinew, chain and all. The guest is served only the part that survives butchery. Somewhere between the loading bay and the pass, a portion of what you paid for stops being sellable, and unless that reduction is written into the costing, every plate built from that ingredient is priced against a cost you never actually paid.

Yield percentage is the name for that reduction. It is the least glamorous idea in restaurant costing, it involves one division, and it is missing from a startling proportion of the recipe cards in professional kitchens. It is also the reason a food cost percentage can look immaculate on the spreadsheet and refuse to behave in the inventory.

How the number works

Divide the edible weight by the as-purchased weight and multiply by a hundred. That is the yield percentage. To convert a buy price into a plate price, divide the buy price by that yield expressed as a decimal.

An 8 kg striploin invoiced at $22.00 a kilogram trims down to 5.8 kg of steak-ready meat. The yield is 72.5%. The edible-portion cost is $22.00 divided by 0.725, or $30.34 a kilogram — an increase of $8.34 on every kilogram you thought you understood.

On a 250 g portion, that is $7.59 of meat rather than the $5.50 the invoice price implies. The dish is understated by $2.09 before you have added a single garnish. At sixty covers of that dish a week the gap is about $125 weekly and roughly $6,500 a year, from one ingredient in one recipe.

The error scales with two things: how much trim the ingredient carries, and how expensive it is per unit. Expensive proteins with heavy trim are the worst offenders, which is unfortunate, because those are usually the dishes you priced most deliberately and the ones carrying the most volume.

Yield loss is not the same as waste

These two get conflated constantly, and the confusion leads to the wrong fix being applied with real effort behind it.

Yield loss is structural and unavoidable. A fish has bones. An onion has skin. You cannot train it away, and no amount of process improvement will get a whole fish to 100%. What you can do is measure it and cost it, so that the price on the menu reflects the money that genuinely left your account.

Waste is avoidable. It is stock that spoiled, preparations that failed, over-trimming beyond specification, and food that went out and came back untouched. That one responds to ordering discipline, storage layout and training, and it belongs in a waste log rather than a yield figure.

The practical distinction: if your best cook and your newest cook produce the same result, it is yield. If they differ by fifteen percent on the same primal, the gap between them is waste. Cost the first, chase the second. Rolling them together means either overstating your prices to cover sloppiness or, more commonly, treating a genuine training problem as an unavoidable fact of the ingredient.

Running a yield test you can trust

The test itself is trivial. Making it representative is the part that takes judgement.

• Weigh the item as delivered, on your own scales, before anything is removed
• Have the person who normally does the job break it down, at normal service pace
• Weigh the usable product, and the trim separately if you intend to reuse any of it
• Repeat three times across different deliveries and average the results
• Record the date and supplier alongside the number

Two details are worth insisting on. The first is that the head chef should not do the test. A yield produced under careful conditions by your most skilled person describes a shift that does not exist, and it will flatter your costings in exactly the way you were trying to correct. The second is the three-repetition rule. Animals vary, crates vary, and a single test locks in whichever variation you happened to receive that Tuesday.

Published yield tables are worth having open as a sanity check — if your test comes back at 90% on a whole fish, something has gone wrong with the measurement. They are not worth using as inputs. Your supplier's specification and your kitchen's butchery standard are what determine your number, and neither is described by a general reference table.

What changes once you have the numbers

Recipe costs stop drifting

Edible-portion cost is the figure that belongs on every line of a recipe costing. Once the ten or fifteen ingredients that carry real trim are corrected, the gap between your theoretical and actual food cost percentage usually narrows sharply — and whatever remains is genuinely waste or loss, which means you can now go looking for it in the right place.

Menu prices land where you intended

Pricing built on an understated cost sets the price too low, and the shortfall is invisible because the arithmetic all checks out. Correcting the input tends to move a handful of dishes noticeably. Those dishes were the ones quietly funding themselves out of your margin. The mechanics of pricing off a corrected cost are covered in our guide to pricing a menu item.

Ordering gets more accurate

The same division works forwards. Needing 12 kg of trimmed product at 72.5% yield means ordering 16.6 kg, not 12 kg. Kitchens that run short on a prepped item despite ordering carefully are almost always ordering the edible weight and receiving the as-purchased one.

Make-or-buy becomes a calculation

A $30 pre-portioned steak looks expensive next to a $22 primal until you notice the primal costs $30.34 edible. Add the labour to break it down and the comparison inverts. Yield is what lets you settle that argument with numbers instead of instinct, and the answer is not always the one chefs expect.

Where operators go wrong

The most common failure is not skipping yields entirely — it is applying them inconsistently. Half the recipes carry a yield adjustment and half do not, usually because two people built them a year apart. The resulting food cost is neither right nor reliably wrong, which makes it worse than no adjustment at all, because you cannot correct for an error you cannot characterise.

The second is crediting trim that never gets used. It is easy to write down a stock credit for bones and a soup credit for vegetable offcuts, and much harder to guarantee that any of it happens on a Saturday night. If the trim goes in the bin, the credit has quietly restored the original understatement under a more respectable name.

The third is treating a yield as permanent. Suppliers change specification, seasons change produce, and a new cook on the butchery section changes the number on the day they start. A yield recorded in 2024 and never revisited is a historical document, not a cost input.

None of this is difficult. It is a set of scales, an afternoon, and the discipline to write the result somewhere it will be used. Against a five-figure annual error on a single ingredient, that is the best-value hour available to anyone running a kitchen.

Frequently asked questions

Which ingredients are worth yield testing first?

Rank by annual spend multiplied by expected trim. Expensive proteins bought whole or as primals come first, then high-volume produce that needs peeling or coring. Anything portion-controlled or pre-prepped can be left alone, because its yield is effectively one hundred percent.

How is yield percentage different from food cost percentage?

Yield percentage is a property of one ingredient — how much of what you bought is sellable. Food cost percentage is a property of the whole operation. Yield is an input into the costings that produce it, which is why an uncorrected yield shows up as an unexplained variance rather than as an obvious error.

Should the yield go on the recipe card or the ingredient record?

The ingredient record. A yield belongs to the product, not to the dish, and storing it once means a re-test updates every recipe that uses it. Yields scattered across individual recipe cards are the reason inconsistent costings appear in the first place.

Does yield matter for a small menu?

It matters more. A short menu concentrates spend on fewer ingredients, so an error on any one of them moves your food cost further than it would in an operation buying across sixty lines. The upside is that a small menu can be yield-tested completely in a couple of afternoons.

What if my yield varies a lot between deliveries?

Use the average for costing, but treat the variability as information about your supplier. Consistent yield is part of what a specification buys, and a wide spread between deliveries is a reasonable thing to raise at your next price review with evidence attached.

Can I just add a flat percentage to cover trim?

It is better than nothing and worse than measuring. A blanket uplift overstates cost on low-trim items and understates it on high-trim ones, so it distorts the comparison between dishes even while the overall food cost looks closer. That distortion feeds straight into pricing and menu decisions.

Run the numbers

Use the free Yield Percentage Calculator to apply everything above to your own figures.