Recipe costing: where the GP quietly leaks

· 5 min read

Short answer

A recipe can be costed correctly and still leak margin, because the leak usually isn't in the maths — it's in five places the maths never sees: portioning drift, forgotten garnish, sauce and stock treated as free, sub-recipes that don't update when their ingredients do, and waste that's never logged. Fix the leak points, not the formula.

Most food cost percentage guides — including our own — walk through the calculation: ingredient cost, yield, portion size, divide by selling price. That part is usually fine. Where restaurants actually lose margin is in five places a correct calculation doesn't touch:

1. Portioning drift.

The recipe says 150g. The section serves 180g, consistently, because that's what "a portion" looks like to whoever's plating it tonight. Nobody changed the recipe — the plate just doesn't match it, every single time, and the gap between the two is money that never shows up as a "cost" anywhere.

2. The garnish that isn't costed.

Herbs, a lemon wedge, a drizzle of oil — small enough individually that they get waved off as negligible, and negligible thirty times a service, every service, isn't negligible by month end.

3. Sauce, stock and prep treated as free.

If the mother sauce or the house stock was made from real ingredients and real time, it has a real cost, even though nobody buys "sauce" from a supplier. Recipes that reference a sub-recipe by name without a cost attached are quietly assuming that sub-recipe is free.

4. Sub-recipes that don't cascade.

The stock price of an ingredient changes. Every dish using it should re-cost — including every dish that uses a sauce or a prep item that uses that ingredient, two or three layers down. In a lot of kitchens, only the top-level dish gets touched, if anyone remembers to touch anything at all. This is the same gap covered in

Catching that the price moved is only half the job if the recipes built on it never get told.

Spotting a supplier price rise before it eats your margin

5. Waste that isn't logged as waste.

Trim, spoilage, a dropped plate, a returned dish — if it isn't recorded, the ingredient it used still shows as "in stock" on paper, and the gap between paper stock and real stock reads as a mystery rather than as the specific, findable thing it actually was.

None of these show up as a costing error.

The formula was applied correctly to the recipe as written. The recipe as written just isn't what's actually happening on the pass.

How to find out if it's happening to you, this week:

  • Pick one dish you sell a lot of. Watch it get plated, for real, three times, without telling the section you're watching. Weigh what goes on the plate against what the recipe specifies.
  • Check whether the sauce or stock in that dish has a cost attached at all, or whether it's a name with nothing behind it.
  • Pull up that dish's recipe and check when it was last re-costed against when its ingredients last changed price. If those two dates don't line up, you have the cascade problem.

If the portioning drift is the bigger suspect, the actual yielded weight after prep and cooking loss is what to check first — covered in

The recipe yield calculator

Where FoxEra fits, honestly:

Recipes and sub-recipes are linked, so a price change on a base ingredient recalculates every dish and every sub-recipe that uses it, not just the top layer. It doesn't watch the pass and it can't tell you a section is over-portioning — that's still something a person has to notice by looking. What it removes is the maths getting stale between the times someone looks.

Rather have the software do it?

Core software, saved data, reports, users and David AI text chat are free forever; the metered AI, import, integration and workflow actions listed on Pricing use credits; every company starts with 24,000 credits (the same balance as the £160 top-up), and you can add more from £10 in the app or earn them by watching eligible videos.

Keep reading