Food cost per serving—also called plate cost or portion cost—is the ingredient cost of a single plate, calculated by dividing a recipe’s batch cost by the number of servings the batch realistically yields. It is the bridge between recipe cost (what a full batch costs to produce) and menu price (what a guest pays for one plate), and it is the starting point for theoretical food cost percentage, contribution margin, and item-level pricing decisions.
In this guide, plate cost refers to the ingredient cost of one serving. It does not include labor, rent, utilities, delivery commissions, packaging, or other operating expenses unless those are explicitly added to the recipe specification. The portion math breaks down into three distinct yield stages—ingredient yield (trim), cooked yield (shrinkage), and portion yield (servings produced)—and the number that ends up in the denominator is the one that quietly decides whether a dish is profitable.
For the broader workflow—turning per-serving cost into menu prices and engineering a profitable menu—see our complete guide to restaurant food costing and menu pricing. For the batch-costing method that produces the recipe cost this guide divides, read our step-by-step guide on how to calculate recipe cost, and use our free Recipe Cost Calculator to run the arithmetic in the browser.
Why per-serving cost beats batch cost
A batch recipe cost alone cannot determine the price of one serving. Knowing a batch of bolognese costs $26.40 to produce tells you nothing about what to charge for one plate until you answer one more question: how many servings does that batch actually yield? Divide $26.40 by four servings and the plate costs $6.60; divide it by five and the plate costs $5.28. The same batch cost produces wildly different menu prices depending on the yield assumption, and the wrong assumption silently undercosts every plate the kitchen sends out.
This is why per-serving cost is the operational number, not batch cost. Batch cost is relatively straightforward once ingredient quantities and current supplier prices are known. Per-serving cost is where additional operational gaps appear, because the result depends on trim, cooking loss, and portion consistency. Operators who cost only to the batch level price on a theoretical plate that does not match the plate the guest receives.
Consider the chicken rice bowl used throughout our costing guides. A full recipe cost—chicken, rice, vegetables, sauce, marinade, oil, fried shallots, and seasoning—lands at roughly $2.94 per plate when every ingredient is listed individually. That number is only useful for pricing once you confirm the recipe actually yields the servings it claims. If the spec says four servings but the kitchen plates three and a half after trim and shrinkage, the true cost per serving is not $2.94 but $2.94 × 4 ÷ 3.5 = $3.36—a 14% undercost that compounds across every bowl sold.
How to calculate food cost per serving
The per-serving formula is simple; the discipline is in getting the denominator right.
Cost per serving = Batch recipe cost ÷ Realistic servings produced
Formula cheat sheet — bookmark this for quick reference during costing sessions:
- Cost per serving = Batch recipe cost ÷ Realistic servings produced
- Yield percentage = Usable weight ÷ As-purchased weight × 100
- EP cost per kg = AP cost per kg ÷ Yield percentage (as a decimal)
- Number of portions = Total usable batch weight ÷ Standard portion weight
Step by step:
- Cost the full batch first. Sum every ingredient line—proteins, produce, dry goods, oils, sauces, marinades, garnishes, and seasoning—using the method in our recipe cost guide. Do not bundle the “small” ingredients; oil, marinade, and garnish are exactly where margin leaks, as the callout below shows.
- Measure the realistic yield, not the theoretical one. Cook and plate the recipe the way the line actually does in service, then count the servings. A recipe card that says “serves 6” is a target, not a measurement.
- Divide batch cost by realistic servings. That is your plate cost—the ingredient cost of one plate as the guest receives it.
- Re-cost when supplier prices or yields move. Chicken yield drifts with the cut and supplier; portion creep sets in slowly. Re-measure yield whenever the supplier, cut, preparation method, or portion specification changes, and periodically test high-cost or high-volume dishes for portion drift.
The problem: A food truck costs its pulled-pork sandwich at $2.10 per serving, based on a recipe that claims 5kg of pork shoulder yields 25 sandwiches. In service, the cook portions generously to keep tickets moving, and after trimming fat and bone the 5kg of shoulder actually yields closer to 3.8kg of usable pork—enough for 19 sandwiches, not 25. The batch cost is correct, but every sandwich is silently undercosted by $0.66, and the truck sells a best-seller at a margin far thinner than the spreadsheet shows.
The real-world fix: Stop trusting the recipe card’s serving count. Buy a test batch, trim and cook it the way the line does, then portion it onto the actual service plate and count the plates. If 5kg of shoulder yields 19 sandwiches, divide the batch cost by 19—not by 25. Update the recipe costing with the realistic yield, and the per-serving cost jumps from $2.10 to $2.76. Re-run the pricing from that number, and either lift the menu price or tighten the portion spec until the dish hits target margin.
The three types of yield loss
Yield loss happens in three distinct stages, and keeping them separate is what prevents the most common costing error—counting the same loss twice.
Ingredient yield (trim and fabrication). This is the usable amount of an ingredient that remains after bone, skin, peel, or outer leaves are removed. Whole fish, bone-in meat, and unpeeled produce all yield less usable weight than the raw input. The first step is calculating the yield percentage:
Yield percentage = Usable weight ÷ As-purchased weight × 100
Then convert the as-purchased (AP) cost to the edible-portion (EP) cost—the true cost of the ingredient that actually reaches the plate:
EP cost per kg = AP cost per kg ÷ Yield percentage
The percentage must be converted to decimal form. If raw chicken thigh costs $8.00/kg with an 80% usable yield, the EP cost is $8.00 ÷ 0.80 = $10.00 per kilogram of usable chicken, not $8.00 ÷ 80.
Cooked yield (cooking shrinkage). This is the weight remaining after cooking. Meats shrink depending on cut, composition, and method; stocks and sauces reduce by half or more as water evaporates. There is no universal number—measure the cooked yield of your own spec rather than relying on a textbook average.
Portion yield (servings produced). This is the number of saleable servings a batch actually produces after prep and cooking. It is the denominator in the per-serving formula. For recipes portioned by weight, calculate it directly from the usable batch weight:
Number of portions = Total usable batch weight ÷ Standard portion weight
Plating variance sits alongside portion yield—it is the gap between the portion the recipe specifies and the portion the line actually serves under pressure. A recipe costed at 120g of protein but plated at 150g silently runs 25% over food cost, and this drift compounds across every cover.
Illustrative starting ranges
These broad ranges are included only to show the scale of potential loss. Do not use them as a substitute for measuring your own product.
| Ingredient (as purchased) | Prep step | Typical yield |
|---|---|---|
| Chicken thigh, bone-in skin-on | Bone and skin | 70–80% |
| Beef sirloin, whole | Trim and portion | 75–85% |
| Whole fish | Fillet | 40–50% |
| Onions | Peel | 88–92% |
| Potatoes | Peel | 78–85% |
| Leafy greens | Trim and wash | 70–80% |
The exact percentages vary significantly by cut, fabrication standard, supplier, and operator. The first time you cost a dish, trim and weigh the actual yield yourself; from then on, update the number only when the supplier, cut, or spec changes.
Two ways to account for yield loss
There are two valid costing systems for handling yield. Both work; the danger comes from partially combining them without identifying which losses have already been captured.
System 1—Cost by ingredient yield. Convert each ingredient’s AP cost to EP cost using its yield percentage, calculate the full recipe using usable quantities, then divide by the recipe’s standardized number of portions. If raw chicken thigh costs $8.00/kg with an 80% usable yield, the EP cost is $8.00 ÷ 0.80 = $10.00 per kilogram of usable chicken. A recipe using 0.18kg of trimmed chicken then costs 0.18kg × $10.00 = $1.80, not 0.18kg × $8.00 = $1.44. The batch cost now already reflects what reaches the plate.
System 2—Cost the actual test batch. Use the actual cost of everything put into the batch (at AP prices), then divide by the number of saleable portions physically produced. This captures all losses—trim, cooking shrinkage, and portion variance—in one measured result, rather than calculating each separately.
Avoid double-counting yield loss
The single most important rule: never capture the same loss in both systems. If the ingredient-level yield adjustment in System 1 already accounts for trim or cooking shrinkage, do not reduce the serving count again for that same shrinkage in System 2. Use System 2 only for an additional gap that System 1 does not cover—such as over-portioning, generous saucing, spills, or inconsistent plating.
Most dishes need elements of both. Trim the protein and reduce the sauce at the ingredient level (System 1), then count the actual plates the batch produces to catch plating variance (System 2). Skipping either one leaves a gap between the theoretical plate cost and the plate the guest receives—and that gap is exactly where profitable dishes quietly turn into break-even ones.
One complete worked example: pulled-pork sandwiches
This end-to-end calculation shows how the portion-weight formula and the per-serving formula connect. The numbers come from the food-truck scenario in the callout above.
| Measurement | Amount |
|---|---|
| Raw pork purchased | 5 kg |
| Raw pork cost (AP) | $52.50 |
| Cooked usable pork | 3.8 kg |
| Standard cooked portion | 200 g |
| Realistic portions | 19 |
| Cost per serving | $2.76 |
The cooked yield percentage here is 3.8kg ÷ 5kg = 76%—meaning nearly a quarter of the purchased weight is lost to trim and cooking shrinkage before any sandwich is plated.
First, calculate how many portions the cooked batch produces:
3,800g ÷ 200g per sandwich = 19 portions
Then divide the batch cost by those portions:
$52.50 ÷ 19 portions = $2.76 per sandwich
The same batch cost divided by the recipe card’s claim of 25 portions would have given $2.10 per sandwich—a $0.66 undercost on every single sale.
From per-serving cost to menu price
Once the per-serving cost is accurate, pricing follows directly. At a 30% target food cost percentage, a plate costing $2.94 needs a menu price of $2.94 ÷ 0.30 = $9.80, rounded to $10.00. The pricing step—choosing a target band, sanity-checking against the market, and handling items that cannot hit target margin—is covered in full in our food costing and menu pricing guide, and the way that final price is displayed on the menu (currency signs, anchoring, charm pricing) shapes what guests actually order—see our guide on pricing psychology and currency signs.
To cost a full recipe library and keep per-serving costs current as supplier prices move, start from our free recipe costing template—it carries the yield adjustments and serving-count logic from this guide into a spreadsheet that recalculates every plate automatically.
Accurate plate costs make price decisions easier. Start free with Nommy and publish price changes to your digital menu in minutes—without redesigning or reprinting.
Run the yield math with Nommy’s free Recipe Cost Calculator, then publish the resulting prices instantly through your Nommy digital menu. The costing work cannot be skipped—the bottleneck is never the reprint; it is the arithmetic.
Frequently asked questions
What is the formula for food cost per serving?
Batch recipe cost divided by the number of realistic servings the batch produces. The discipline is in measuring that denominator accurately—using ingredient yield, cooked yield, and portion yield—rather than trusting the serving count printed on the recipe card.
Is food cost per serving the same as food cost percentage?
No. Food cost per serving is a dollar amount—the ingredient cost of one plate. Food cost percentage compares that dollar amount with the menu price:
Food cost percentage = Cost per serving ÷ Menu price × 100
A plate costing $2.76 sold at $9.20 runs at 30% food cost. The two numbers are related but measure different things.
Does food cost per serving include labor?
Usually not. Recipe or plate cost refers to ingredient cost only. Labor and overhead should be considered separately—through prime cost or operating-margin calculations—when assessing whether the final selling price is genuinely profitable.
How is the number of servings calculated from batch weight?
Divide the usable cooked batch weight by the standardized portion weight. For example, 3,800g of cooked pulled pork divided by a 200g portion spec gives 19 sandwiches. That count becomes the denominator in the per-serving formula.