Build recipe costing for every item
When it may fit: No reliable theoretical cost exists.
Evidence needed: Recipes, yields and current supplier prices.
Trade-off: Needs upkeep as prices change.
Restaurant Intelligence Brief
How to Compare Actual Food Costs With Expected Food Costs
Theoretical food cost is what the menu should cost if every plate follows the recipe. The gap between it and actual food cost is where waste, portioning, theft and recording errors live.
1. The business problem
How far is what we spend on food from what our menu says we should spend?
Every restaurant has an expected food cost, even if no one has written it down: the cost of ingredients per plate multiplied by the plates sold. What is actually spent is usually higher.
The difference is one of the most useful numbers in the business, because it isolates what is lost between the supplier and the guest, separate from changes in supplier prices or the menu.
2. What the evidence can reveal
The comparison needs three figures:
- Calculated Theoretical cost: recipe cost per portion times portions sold, from the point-of-sale system.
- Calculated Actual cost: beginning inventory plus purchases minus ending inventory.
- Calculated Variance: actual less theoretical, in dollars and as a percentage of sales, by category and over time.
A variance has many possible sources, from over-portioning to waste to unrecorded comps. A persistent or growing gap deserves investigation, and a stable one may simply reflect normal trim and prep loss.
3. Investigate the problem
Illustrative A small illustration shows the idea. If recipes imply $4,000 of food for the plates sold in a week and the inventory count shows $4,400 of food was used, the variance is $400, or 10% of theoretical cost. A restaurant would want to know whether that is steady, rising, or concentrated in a few items.
In the synthetic restaurant, Calculated food and beverage cost rose 13.1% against sales growth of 6.8%. Entrées moved from a food-cost percentage of 33.6% to 37.1%, while appetizers and shares changed little, from 28.8% to 29.1%.
Information gap The demo has only actual cost, so it cannot compute a variance. This is a useful reminder that a food-cost percentage alone cannot tell supplier price from usage. The comparison needs recipe costing and inventory counts. Each 1% change in ingredient cost is worth about $9K a year in this example, which indicates how much effort the measurement can justify.
4. Explore the interactive experience
Investigate it yourself.
Restaurant Profitability Lab: a menu-category dashboard, a scenario simulator for covers, ticket, ingredient cost, labor and menu mix, evidence-linked answers from APIS, and a sample brief, all on synthetic data.
The experience below is built on a synthetic restaurant, “Demo Bistro.” Nothing in it is real, so you can explore freely.
- In the menu-category table, find the category whose food-cost percentage rose the most.
- In the Scenario tab, raise ingredient cost by 5%. This is the effect of price alone, with no change in usage.
- Ask APIS what additional data would explain the food-cost increase.
5. Evaluate possible solutions
Closing the gap starts with measurement, then targets the largest sources.
6. What leadership should do next
Questions to answer
- What is the theoretical food cost for the period?
- How far is actual from theoretical?
- Is the gap stable, growing or concentrated?
- Which items or categories drive it?
Data to gather
- Recipe costs and yields
- Point-of-sale item sales
- Inventory counts and purchases
- Waste, comp and void logs
Steps to take
- Cost recipes for the top 20 items.
- Run a weekly variance report.
- Investigate the largest three variances.
- Repeat monthly.
7. Engage Two Pillars
Find Opportunities to Improve Location Performance
Two Pillars can help restaurant owners and operators investigate margins, cost structure, location comparisons, menu economics, utilization and expansion decisions using your own data. A good first step is a food-cost variance review that sets up recipe costing and a repeatable weekly check.
Sources and data notes
- The organization, its figures, and every number in this article and its interactive experience are synthetic and illustrative. They are not real data, a benchmark, or a forecast.
- Reviewed October 2026. This article describes an approach to investigation. It does not replace review by qualified professionals.