Your Food Cost Percentage Is Only the Beginning

Understanding what you should have used—and what you actually used

By Jeff Fritz | The Operator’s Table

“We ran a 32% food cost.”

That number deserves a follow-up question: What should it have been based on what you sold?

If your theoretical food cost was 30%, there is a two percentage point gap to explain. If it was 27%, the same 32% actual cost tells a very different story.

I believe operators have to take the time to understand both numbers. Comparing your food cost to a budget or another restaurant can provide perspective. Comparing actual cost to an accurate theoretical cost tells you much more about the execution inside your own operation.

In full service, that same discipline belongs behind the bar—with liquor, wine, and beer measured separately.

Theoretical cost starts with what you sold

Theoretical cost is the ingredient cost you would expect for your actual sales mix if every item were prepared to its established recipe and portion specification.

Theoretical cost dollars = Sum of each item’s recipe cost × quantity sold

Theoretical food cost % = Theoretical food cost dollars ÷ net food sales × 100

It is not an average of the food cost percentages printed on your recipe cards. An item that sells 500 times must carry more weight than one that sells 20 times.

The foundation has to be accurate: current ingredient prices, tested usable yields, complete batch recipes, correct portions, and POS mappings that capture modifiers, substitutions, and add-ons. Include cooking oil, sauces, garnishes, and accompaniments where applicable. Normal trim and cooking losses belong in tested recipe yields; avoidable spoilage should not be hidden in the standard.

Use comparable ingredient prices and valuation methods on both sides. Otherwise, a purchasing-price difference can look like a portion-control problem. Food ingredient cost also needs to stay separate from prep labor for this comparison; labor can be evaluated alongside it when assessing total profitability.

What actually goes into theoretical food cost?

Let’s use a grilled chicken entrée priced at $22, served with roasted potatoes, vegetables, and sauce. Every ingredient needed to deliver that plate should be included. The following prices and yields are illustrative.

At a $22 selling price:

$3.85 ÷ $22 × 100 = 17.5% theoretical food cost

That entrée should consume approximately $3.85 in food ingredients every time it is sold, assuming the established portions and preparation methods are followed. The table rounds each component to cents for readability; a production costing system should retain precision until the final total.

Why yield matters

Paying $3.60 per pound does not mean a pound of cooked, usable chicken costs $3.60. If only 80% of the purchased weight becomes usable cooked product, the effective cost is $4.50 per usable pound.

Ignoring yield would value the 6-ounce cooked portion at just $1.35 instead of approximately $1.69. That understates the cost by about $0.34 per plate—or approximately $340 over 1,000 plates.

Normal trimming and cooking losses belong in the recipe’s tested yield. A tray of chicken that spoils because too much was prepared is a separate operational loss. Building that avoidable waste into the standard would conceal an opportunity to improve.

Small ingredients still count

The sauce is not simply an estimated forty cents. Its batch recipe should identify and cost every ingredient—stock, cream, butter, aromatics, and seasonings—and measure the finished usable quantity after cooking.

Oil, butter, garnishes, and included bread also consume inventory. Leaving them out makes theoretical cost look better on paper while creating an apparent variance when those ingredients appear in actual usage.

Count each ingredient once. Butter included in the sauce recipe should not also be charged again as sauce butter elsewhere on the plate.

This example measures food ingredients. Prep labor, payroll taxes, utilities, rent, and other operating expenses still matter when evaluating profitability, but they are tracked separately from theoretical food ingredient cost. Packaging should also be tracked consistently in its designated expense category.

Discounts reduce net sales and can raise the theoretical cost percentage without changing the ingredients used. Extra portions, substitutions, and add-ons change ingredient requirements and must flow through the recipe and POS calculations.

From one plate to the restaurant’s theoretical cost

An individual plate’s percentage is only one part of the calculation. The restaurant’s theoretical food cost depends on how many of each item guests actually purchased.

For this simplified menu, assuming full-price sales:

$6,460 ÷ $25,200 × 100 = 25.6% theoretical food cost

This is a sales-weighted calculation. Simply averaging the three items’ food cost percentages would give the wrong answer because they sold in different quantities. These figures establish what sales should have required. Next, inventory tells us what the operation actually consumed.

Actual cost starts with inventory—not just invoices

Purchases alone do not tell you what you used. Some purchases remain on the shelf, and some product used this week was purchased last week.

Actual cost dollars = Beginning inventory + net purchases + transfers in − transfers out − ending inventory

Actual food cost % = Actual food cost dollars ÷ net food sales × 100

Use the same period, category boundaries, count timing, and sales denominator for actual and theoretical. Net purchases should reflect credits and returns.

Employee meals, comps, tastings, remakes, and waste need consistent treatment. A comp can consume product without producing net sales. Record those events and show them as explained usage, or include them in an explicitly defined adjusted theoretical calculation. Do not count the same event twice.

Returning to the menu example above, suppose inventory and purchasing records show actual food usage of $7,200 for the same period:

The gap is calculated before rounding the displayed cost percentages. That $740 needs to be reconciled through documented waste, employee meals, comps, portion differences, inventory errors, or other causes.

How much of a food-cost gap is acceptable?

Restaurant365 publishes a typical food variance target of 1.5–2.5%, depending on the operation. In its comparison against sales, this is a percentage point gap between actual and theoretical cost—not a percentage of theoretical ingredient spending.[1]

That is a useful benchmark, not a universal allowance. My suggested operating goal is to work toward two percentage points or less, investigate sustained results above the published range, and keep reducing preventable losses even when the total falls within target.

The $740 in our example requires explanation. It is not automatically theft, and it is not automatically recoverable profit. Some may be documented meals or waste; some may be counting, pricing, or recipe errors. The purpose of the analysis is to find out.

Bring the same discipline behind the bar

Calculate liquor, wine, and beer cost against their respective net sales. A blended beverage percentage can conceal a problem in one category.

Diageo Bar Academy identifies 1–2% of inventory as an acceptable industry norm for bar variance and recommends keeping variance below 2%.[2] This is an inventory-shrinkage benchmark, not a two percentage point allowance on beverage sales. Its guidance does not establish separate numerical standards for liquor, wine, bottled beer, and draft beer.

For a clearly defined internal usage measure, track:

Unexplained usage % = (Actual usage − sales-based theoretical usage − documented nonsales usage) ÷ actual usage × 100

Calculate by product in consistent units, or aggregate at consistent ingredient costs. Label the denominator: software may calculate variance differently, so percentages are comparable only when their definitions match.

For perspective, 2% unexplained loss on $5,000 of actual beverage usage is $100. Against $25,000 in beverage sales, that is 0.4 percentage points—not two points. That difference matters when setting expectations.

Find the dollars, then coach the behavior

Review the largest dollar variances first. Recount the item, check receipts and transfers, confirm recipe yields and sales mappings, then observe preparation or service. Weigh portions. Check pours. Look at what is being discarded and why.

A lower actual cost than theoretical deserves investigation too. Underportioning, incorrect counts, or overstated recipes can create a favorable-looking report while guests receive an inconsistent experience.

I believe this works best in an environment of open communication, accountability, and recognition. Staff need to understand the standards, have the tools to follow them, and feel comfortable recording a mistake. Punishing honest waste reporting encourages hidden waste and weaker information.

Count consistently, review weekly, assign corrective actions, and recognize measurable improvement. Track both total variance and the unexplained remainder. Recording a loss explains it; it does not recover the money.

Your actual cost tells you what was consumed. Your theoretical cost tells you what your sales should have required. Understanding the gap gives you a place to act.

— Jeff Fritz

Helping Operators Build Stronger, More Profitable Businesses.

Sources

1. Restaurant365, “Actual vs Theoretical Analysis – Troubleshooting Variances.” https://docs.restaurant365.com/docs/actual-vs-theoretical-analysis-troubleshooting-variances

2. Diageo Bar Academy, “A Guide to Effective Inventory Management.” https://www.diageobaracademy.com/en-us/home/bar-management/a-guide-to-effective-inventory-management

Benchmarks are operating references, not mandatory standards. Category-specific control targets and the explicitly defined unexplained-usage formula above are recommendations for this article.