Why your actual food cost never matches the number your recipes say it should

A fast-casual operator we work with had calculated food cost at 28 percent from her recipe cards. She priced the menu off that number, built her ordering schedule around it, and used it as the benchmark when adding new items.

When we built her first proper monthly close, her actual food cost was 34 percent. On $88,000 in monthly sales, that six-point gap was $5,280.

Nothing was stolen. The recipes were accurate. The food was being lost in four specific places that recipe cards do not capture.

Theoretical cost versus actual cost

Theoretical food cost is what food should cost if every portion is exact, every ingredient is used, and nothing is lost between the delivery dock and the customer’s plate. It comes from recipe cards and current supplier pricing.

Actual food cost is what the books show after invoices are entered. It includes everything the theoretical number assumes away: prep waste, spoilage, portions running heavier than spec, and comps that went unrecorded.

The gap between them is food cost variance. A well-run kitchen keeps variance under 1 to 2 percent of sales. A kitchen running 6 points over target is losing $880 for every point, every month, on this volume. That figure accumulates month after month before it shows up in a conversation about margins.

Four causes that recipe cards miss

Portioning inconsistency on proteins. Proteins are the most expensive ingredient on most menus and the hardest to portion by eye under pressure. A recipe calling for a six-ounce chicken breast will run anywhere from five and a half to eight ounces on a busy line. When we measured against actual point-of-sale ticket counts for this operator, the kitchen was averaging 7.2 ounces per chicken portion. At 5,800 monthly covers and $3.90 per pound, that 1.2-ounce overage was $1,840 per month.

Prep waste that never gets costed. When a cook butchers a case of strip loin, the trim goes in the trash. That trim came off beef invoiced at full price. Recipe cards are built on the post-butchering yield. The food cost line in the books includes the whole case. The difference is waste, and it belongs in the food cost calculation, not in a general expense bucket where it disappears.

Spoilage and dead stock. Tuesday’s produce order arrives when last week’s lettuce is still in the walk-in. By Thursday, some of it goes in the trash. A dairy order pulled for a weekend special that sold out early leaves four days of unused product. These losses are real food cost. They appear in the books as purchased and never appear as sold.

Comps and voids that do not get entered. A manager comps a table because a dish came out cold. The food was made and sent. If no comp is entered in the point-of-sale system, the revenue is removed but the reason for the food cost disappears. The next report treats those ingredients as unexplained loss, and the cause goes unfixed.

What one month’s variance looks like

Here is how the gap broke out for the operator above.

Each line has a fix. None of them appears in a recipe card.

Why this matters

The direct cost is in the table. Six points of food cost on $88,000 in monthly sales is $63,360 annualized. For a restaurant targeting a 10 to 15 percent net margin, that gap is the margin.

The less visible cost is in pricing decisions. A restaurant that believes food cost is 28 percent will price new menu items to a 28 percent target. If actual food cost is 34 percent, those prices are set 6 points low from the first week. The menu runs, sales look healthy, and the Profit and Loss report runs soft without a clear explanation.

Recipe-based food cost is a planning tool. It tells you what food should cost if every input runs right. Actual food cost tells you what food did cost. Managing only to the recipe number means managing to a figure that has never matched what the kitchen actually produces.

How proper food cost tracking works

For restaurant clients we work with, two numbers update every week: actual purchases and theoretical usage.

Actual purchases come from invoices entered as they arrive. Theoretical usage comes from the point-of-sale system: the count of every item sold that week, multiplied by the recipe cost for that item. Most restaurant point-of-sale systems can produce this report in a few minutes.

The gap between the two is the variance. We break it out by category: proteins, produce, dairy, dry goods. A variance over 2 percent on any category gets a closer look that week, not at month end. When a cause is identified, it gets corrected before it compounds into the next period.

Best practices for restaurant operators

• Enter every purchase invoice the day it arrives, not at the end of the week. Food cost calculated on partial invoice data is an estimate, and decisions made on partial data are off from the start.
• Pull a theoretical usage report from the point-of-sale system weekly and compare it to actual purchases. The report takes about five minutes to run and shows which categories are drifting.
• Record every comp and void in the point-of-sale system the same shift it happens. Comps are real food cost. Tracking them separately lets you watch the total as a percentage of sales and hold managers accountable to a target.
• Weigh proteins at the station at least once a week during service. Not as a formal audit. A brief spot check resets portioning habits without a formal conversation.
• Price new menu items before they go on the menu. Divide the recipe cost by your target food cost percentage to find the minimum price. Anything below that number is a losing item from the first ticket.

Three questions worth asking

1. What is the difference between my theoretical food cost and actual food cost this month, and what is driving the gap?
2. Are comps and voids tracked as a separate line in the books, or are they absorbed into the overall food cost number?
3. Is actual food cost compared to theoretical usage weekly, or only when the monthly books close?

If those answers are uncertain, the kitchen is being managed to a number that may have very little connection to what food actually costs.

Send a screenshot of last month’s Profit and Loss report along with your last three weekly purchase totals. We will review whether the books are separating theoretical from actual, and whether the variance is being tracked or absorbed into the overall food cost percentage.