Recipe costing: one dish, from box price to plate

Recipe costing is the work of finding out what a single plate of food really costs to make. Not what the main ingredient costs on the invoice, but what everything on the plate costs once it has been trimmed, cooked, reduced, seasoned and garnished. It is the most useful number a kitchen can know, because every pricing decision, every menu change and every conversation about margin depends on it.

It is also the number most kitchens get wrong, usually in the same few ways. The yield of an ingredient is assumed rather than measured. A sauce is listed as a line rather than costed as a recipe of its own. The oil, butter and seasoning are left out because they feel too small to matter. The garnish is forgotten entirely. Each omission is small. Together they can make a dish look far more profitable than it is.

This guide costs one dish completely, from the box price of each ingredient to the cost per portion on the plate. Along the way it covers how to measure recipe yield, how to cost a sub-recipe, what to do about the small things, and how to turn the plate cost into a selling price. The dish and its numbers are illustrative, chosen to show the method, not to recommend a costing.

A recipe costing answers one question: what does one portion of this dish cost in ingredients? The result is the plate cost, sometimes called the cost per portion. Divide it by the selling price before tax and you get that dish's food cost percentage. Take it away from the selling price and you get the dish's cash margin.

A recipe costing is not the same as your overall food cost percentage, which is measured from stock across the whole kitchen and includes waste, over-portioning and everything else that happens in a real service. The recipe costing is the plan; the stock-based figure is the result. Our guide to calculating food cost percentage covers the stock side, and our guide to actual vs theoretical food cost explains how comparing the two shows where money goes.

It is also not a labour costing. Recipe costing measures food. If you want to know whether a dish is worth the time it takes to prepare, add labour as a separate calculation so the food figure stays comparable across the menu.

A menu without recipe costings is a menu priced by instinct. Instinct is often right about what guests will pay and often wrong about what dishes cost, because costs change faster than menus do and because the expensive parts of a dish are not always the obvious ones.

With a costing for every dish, three things become possible. You can price new dishes deliberately, starting from what they cost rather than what the dish next to them costs. You can see which dishes carry your margin and which merely carry your sales. And you can calculate a theoretical food cost for the whole menu, which turns your stock-based figure from a score into a diagnosis.

The example is a starter of roasted beetroot with whipped goat's cheese, toasted hazelnuts, a blood orange dressing and pea shoots. It is a useful example because the main ingredient, beetroot, is cheap, and almost everything that makes the dish worth ordering is not. It is the kind of plate where a quick guess at the cost goes badly wrong.

The recipe for one portion, as the chef wrote it:

Here is the method, step by step. Each step builds on the one before, and skipping any of them is where costings go wrong.

Start from what actually goes on the plate during service, not from what the menu description says. Watch the dish being made, or make it yourself, and write down every component, including the finishing touches. If the line adds a few drops of oil or a sprinkle of salt, they belong in the recipe.

For each ingredient, take the price from your most recent invoice, the pack size and the unit. Use the price you actually pay, after any discount, and before any tax you can reclaim. If you buy the same ingredient from two suppliers, use the one you buy most from, and note it.

Recipe yield is the share of what you buy that ends up usable. Beetroot loses its skin and tops and shrinks a little in the oven. Hazelnuts lose a little in toasting and skinning. Herbs and salad lose their stalks and the leaves that are past their best. Measure yield rather than guessing it: weigh a quantity as bought, prepare it the way the kitchen normally does, and weigh what is left. Do it two or three times and take the average.

Divide the purchase price per unit by the yield. If raw beetroot costs 1.40 per kilo and only 70 percent ends up usable after peeling and roasting, the cost per usable kilo is 1.40 divided by 0.70, which is 2.00. This is the price to use in the recipe, because the recipe measures usable beetroot, not raw.

The whipped goat's cheese and the dressing are recipes in their own right. Cost each one as a batch, divide by the quantity it makes, and use the resulting cost per gram or per millilitre in the main recipe. Never list a house-made component without costing it, because it is often where most of the money sits.

Oil, salt, butter, garnish and finishing touches each cost little, but together they add up, and they are almost always left out. Either cost each one properly, or, if they really are tiny, add a small fixed allowance per portion based on a measured average. The second approach is acceptable for salt and pepper. It is not acceptable for a garnish that costs as much as the main ingredient.

Multiply each ingredient's quantity by its cost per usable unit, and add the results together. That total is the plate cost: the cost per portion of the dish.

Divide the plate cost by the selling price before tax to get the dish's food cost percentage, and take the plate cost away from the price to get its cash margin. Those two numbers, side by side, are what you use to price and promote the dish.

Here are the illustrative numbers for the beetroot starter.

Add them together and the plate cost is 1.81.

Now look at where the money is. The beetroot, the ingredient that names the dish, costs 0.36. The goat's cheese costs nearly twice as much. The pea shoots, a garnish most people would round to zero, cost two thirds as much as the beetroot. This is the pattern that makes guessing so dangerous: the cost of a dish rarely sits where its name suggests.

Suppose the starter is priced at 9.50 before tax. Its food cost percentage is 1.81 divided by 9.50, times 100, which is 19.1. Its cash margin is 9.50 minus 1.81, which is 7.69.

Those two figures answer different questions. The percentage tells you how efficiently the dish turns ingredients into revenue. The margin tells you how much cash it contributes to the business each time it sells. A dish with a low percentage and a small margin may be less valuable than a dish with a higher percentage and a larger one, because the business pays its bills in cash, not in percentages. When you compare dishes, look at both.

If you are pricing a new dish, the costing gives you two starting points. One is a target percentage: divide the plate cost by your chosen percentage to find a price, then adjust it to where similar dishes sit on your menu. The other is a target margin: add the cash you need the dish to contribute to its plate cost. Margin-led pricing often works better for very cheap and very expensive dishes, where a single percentage target would set an odd price. The idea of contribution margin is a useful way to think about it: what each sale leaves over after its own direct cost.

Yield deserves its own section because it moves every number in the costing. An ingredient bought at one price and used at sixty percent yield effectively costs two thirds more than its invoice suggests.

To run a yield test, take a typical quantity as delivered and weigh it. Prepare it exactly as the kitchen does in service: trim, peel, bone, cook or portion. Weigh what is usable, and weigh what was discarded, so the two add up to the starting weight. Divide usable by starting weight to get the yield. Repeat the test on two or three different deliveries, because produce varies, and use the average.

Re-run yield tests when you change supplier, when you change the grade or cut of an ingredient, and when the kitchen changes how it prepares something. A different supplier's beetroot, or a new way of trimming a fish, can change the yield as much as a price rise changes the cost.

Sauces, dressings, stocks, doughs and whipped or cured components are the parts of a menu most likely to be left uncosted, and they are often the parts that carry the most cost. Cost each one as its own recipe, with its own batch size and yield, and give it a cost per unit. Then every dish that uses it pulls that cost, so a change in the price of cream updates the whip, and the whip updates every dish that uses it.

This also stops the most common double error: a component used in several dishes, left uncosted in all of them, understating the whole menu at once.

A recipe costing is a snapshot of prices on the day it was done. Prices move, and a costing that is a season old will understate your costs. Build a simple routine: update the prices of your twenty highest-spend ingredients whenever they move, re-cost your best-selling dishes at least every quarter, and re-cost the whole menu whenever it changes. Our guide to building a food cost calculator in a spreadsheet shows how to lay out ingredients, recipes and menu so a single price change flows through every dish on its own.

Each of these makes a dish look cheaper to produce than it really is, which means the menu is priced on numbers that are too optimistic. The correction is rarely dramatic for one dish. Across a whole menu, it can change which dishes deserve their place.

Once you can cost one dish properly, the rest of the menu is repetition. Start with the dishes that sell most, because they carry most of your revenue and most of your food cost. In most kitchens a small number of dishes account for a large share of sales, so costing them first gives you most of the value quickly. Then work through the rest, and finish with the sub-recipes that several dishes share.

When the whole menu is costed, sort it by cash margin and by number sold. The dishes that sell a lot and earn a lot are the ones to protect. The ones that sell a lot and earn little are the ones to look at first. Our guide on reducing food costs without a new menu covers what to change before touching a dish at all.

Kitchens that change their menu weekly, or cook from what the market has that morning, often decide recipe costing is not for them. It is, but it has to be done differently. Costing every dish in full detail before it goes on a weekly menu is rarely realistic. Costing it well enough to price it sensibly is.

The approach that works is to build a library of costed components rather than costed dishes. If the kitchen regularly makes a handful of sauces, a few starches, a set of dressings and a range of proteins prepared in standard ways, cost each of those once. A new dish then becomes a quick assembly: a protein, a starch, a sauce and a garnish, each pulled from the library at a known cost. The costing that would have taken an hour takes ten minutes.

Keep a few standing dishes on the menu that do not change, and cost those in full. They anchor the menu's margin while the changing dishes rotate around them, and they give you something stable to compare against when the weekly figures move.

Recipe costings decay when nobody owns them. The head chef should own the recipes themselves: whether the costing describes what really goes on the plate, and whether the yields still hold. Whoever does the ordering should own the prices: whether the ingredient list reflects what the kitchen is actually paying. Whoever sets menu prices should own the final step, reading the plate cost and the margin together and deciding what each dish should sell for.

When those three roles meet every month or so, briefly, the costings stay alive. When the whole job sits with one busy person, it is the first thing to be dropped in a hard week, and the menu quietly goes back to being priced by instinct.

Recipe costing itself is kitchen work, and many kitchens keep the build-up in a spreadsheet. Where servearound helps is in what happens to the result.

Each dish lives once in a library and appears on every menu it is part of, so its photos, description, allergens and ingredients are kept in one place rather than retyped for lunch, dinner and the bar. Once a dish's cost is set, the menu report in Analytics shows its units sold, revenue and margin from what actually went out during service, which tells you whether a carefully costed dish is earning what the costing promised. Changing a price happens once and reaches every channel the menu is on. If you would like to see it with your own dishes, talk to us.

Precise enough to change a decision. Weigh the expensive components carefully and measure their yields. For salt and pepper, a measured average per portion is fine.

Cost any special that will run for more than a few days, and any that uses expensive ingredients. A quick costing before a special goes on the board stops a popular special quietly losing money.

Cost the whole batch, including cooking loss, and divide by the number of portions it actually yields in service. Measure the portions the line really serves, not the number the recipe promised.

No. Keep recipe costing to food. Packaging for takeaway and delivery is a real cost, but it belongs in its own line so food cost stays comparable.

Recipe costing turns a dish into a number you can price with. Write the recipe as it is really made, find each ingredient's purchase price, measure its recipe yield, work out the cost per usable unit, cost every sub-recipe separately, include the small things, add the lines for the plate cost, and set it against the price.

Do that for the dishes that sell most, keep the costings current, and compare them with what the kitchen really spends. The answer is rarely where the dish's name suggests, and knowing where it really sits is worth more than any single price rise you will ever push through.