Theoretical vs Actual Food Cost: Why the Gap Is Costing You Money
The number your recipes say each dish should cost, and the number your stock count actually shows you consumed, are almost never the same. The gap is where the money goes.
What each number actually measures
Both measure your food cost percentage. The difference is what they trust as a source of truth — your recipes, or your stock count.
If you only run one of them, you're flying blind in the direction you didn't measure. Both together let you ask the only question that matters: where exactly is the gap coming from?
| Metric | How it's built | What it tells you |
|---|---|---|
| Theoretical food cost | Recipe cost × quantity sold (from POS), summed across all dishes | What you SHOULD have spent on ingredients, given a perfect kitchen |
| Actual food cost | Opening stock + purchases − closing stock, divided by food sales | What you DID spend on ingredients, in reality |
Theoretical is a model. Actual is what happened. The model is wrong somewhere — either the recipe doesn't match the dish, or the dish doesn't match the recipe. The gap tells you to look.
Where the gap comes from
Five causes, in roughly the order operators usually find them. Most kitchens have a mix of all five at any given time — the goal is to isolate the dominant one each week.
- Over-portioning — The single biggest cause in most kitchens. A 10g over-pour of cheese on every pizza, multiplied by 200 pizzas a week, is a real number. Scales and standard scoops fix this for under $50 of equipment.
- Waste that wasn't logged — Spoilage, breakage, dropped plates, burned dishes. If it left the kitchen but didn't get rung up or logged as waste, it shows up as actual food cost without any matching theoretical.
- Comps and staff meals — Free meals to unhappy customers, owner meals, shift meals for staff. The food really left the kitchen, but the POS shows no sale — so theoretical undercounts and actual overcounts.
- Theft — Unpleasant to consider but real, especially in cash-heavy operations. The signature is a persistent one-sided gap that doesn't move with menu mix changes.
- Recipe drift — The cook learned to make it slightly differently than the recipe card says. Sometimes that's an improvement; sometimes it's an extra 5g of an expensive ingredient. Re-build the recipe from what's actually leaving the pass, not what's written down.
Healthy gap vs real leak
Some gap is unavoidable — recipes aren't perfectly precise and small waste isn't worth chasing. The threshold between noise and leak is where you should act.
| Variance | What it usually means | What to do |
|---|---|---|
| 0 – 1.5% | Normal noise: minor portioning variance, unlogged 50g spoilage | Nothing. Spend your time elsewhere. |
| 1.5% – 4% | One identifiable cause, often portioning or unlogged waste | Audit one cause this week. Fix it. Re-measure next week. |
| 4%+ | Multiple causes stacked, or one large structural issue | Stop and investigate. A 4% gap on $200k annual food sales is $8,000. |
Track the variance as a percentage of food sales over time. A flat 2% week-on-week is fine. A 1% drift each week for a month is a developing problem worth catching at week two, not week six.
A 4-week playbook to close it
If you've never measured the gap, here's the minimum-viable cadence to get a real number in 30 days and start closing it.
- Week 1 — Get an honest stock count — Pick a day. Close the kitchen 30 minutes early. Count and value every ingredient at current purchase price. Anything you skip is noise added to next week's gap.
- Week 2 — Build or refresh your recipes — Pull your top 10 highest-volume dishes. Re-measure each ingredient with a scale, not by feel. Update the recipe costs with this month's purchase prices.
- Week 3 — Run both numbers for one full week — Stock count Monday morning and Sunday night. Pull POS sales by item for the week. Compute theoretical and actual side by side. Write the gap down — this is your baseline.
- Week 4 — Pick the dominant cause and fix it — Audit one of the five causes above. Pick the one whose fingerprint matches your gap pattern. Make one change. Measure again next week. Repeat monthly.
The deeper how-to is in the food cost percentage guide. If you'd rather not run the spreadsheet, BasilBook computes the variance automatically from your versioned recipes, POS sales, and stock counts.