Theoretical vs Actual Food Cost: How to Find the Gap

Every restaurant has two food cost numbers, and they rarely match. Theoretical food cost is what your sales should have cost if every plate followed the recipe. Actual food cost is what you really used. The distance between them is the most useful number in your kitchen, because it measures everything that happens between the delivery door and the guest's table that your recipes do not account for.

All dollar figures below are illustrative examples, not industry benchmarks.

The two numbers

Theoretical food cost = for each item sold, units sold multiplied by the recipe cost per portion, added up, then divided by food sales.

Actual food cost = beginning inventory + purchases - ending inventory, divided by food sales for the same period. If you have not set up the basics yet, start with how to calculate food cost percentage.

A worked example

One week, food sales of $24,500.

That $1,440 is not one problem. It is a pile of small ones, and your job is to sort the pile.

Where the gap comes from

Work through these in order, because the boring causes are far more common than the dramatic ones.

1. Waste and spoilage. Food that was bought and never sold. A waste log turns this from a guess into a number.
2. Portioning. Plates running heavier than the recipe. A 30 g overage on a protein, repeated all week, adds up quickly.
3. Unrecorded usage. Staff meals, tastings, comps, remakes and spills that never reach the sales report.
4. Receiving and invoice errors. Short deliveries, substitutions and price changes you did not catch at the door.
5. Stale recipe costs. If a recipe still carries last year's price for beef, theoretical cost is understated and the gap looks worse than it is.
6. Counting errors. Inconsistent units, missed storage areas, or counts taken on a different day than the sales period.

Theft is possible, but rule out the other six first. They explain most gaps.

Narrow it down by category

A total gap tells you there is a problem. A category gap tells you where. Split the same calculation into proteins, produce, dairy, dry goods and so on.

Illustrative result: proteins show theoretical $3,100 against actual $3,950, a gap of $850, while produce, dairy and dry goods are each within a few dollars. Now you know to weigh portions, check trim yields and review the protein deliveries, not to chase the whole kitchen.

Then go one level deeper: take your ten most expensive items and compare theoretical usage to actual usage item by item.

A monthly routine that works

1. Count inventory on the same day and in the same order every period.
2. Pull sales by item for exactly the same period.
3. Refresh recipe costs with current invoice prices.
4. Calculate theoretical and actual cost, overall and by category.
5. Pick the one category or item with the biggest gap and investigate it before touching anything else.
6. Write down what you found and what you changed, so next month you can see whether it worked.

Common mistakes

• Comparing different periods. Sales from Monday to Sunday and counts from Tuesday to Tuesday will never reconcile.
• Costing recipes once and forgetting them. Prices move, so recipe costs must too.
• Counting everything in different units each time. Pick one counting unit per item and keep it.
• Ignoring comps and staff meals. If they are not recorded, they show up as mystery gap.
• Fixing everything at once. Change one thing, then measure.

Where software can help

The arithmetic is easy; keeping recipe costs current is the hard part. ChefMind AI's food cost tools let you set a target food cost percentage and see which dishes drift over it, and its recipe costing draws on the ingredient prices in your inventory. It will not find the cause of a gap for you, but it keeps the inputs current.

The short version

• Theoretical is what sales should have cost; actual is what they did cost.
• The gap is the sum of waste, portioning, unrecorded usage, receiving errors and stale costs.
• Split by category, then by item, and fix one thing at a time.
• Track your own baseline rather than chasing a universal target.