Why Is My Actual Food Cost Higher Than My Theoretical?
Last updated: May 8, 2026
Your actual food cost is always going to be higher than your theoretical. That's not a sign of failure — it's physics. Theoretical food cost assumes a perfect world where every chicken breast is exactly the spec weight, every portion is plated precisely to the recipe card, nothing spoils, nothing gets dropped, and no one helps themselves to a shift meal that wasn't logged. None of that happens in a real kitchen. Actual food cost adds all of that reality back in: the trim loss, the variance, the waste, the theft, the receiving errors, and every other imperfect thing that happens between the walk-in and the plate.
The question isn't whether a gap exists. It's how big the gap is. According to MarginEdge's operator research, most restaurants run a gap of 2 to 5 percent between theoretical and actual food cost. A 2 percent gap in a well-run kitchen is essentially unavoidable — it reflects normal trim, prep yield loss, and reasonable waste. A 5 percent gap is worth investigating. A gap above 6 or 7 percent means something is systematically wrong, and it's costing you real money. On a $1 million in annual food revenue, a 5 percent gap is $50,000 walking out the back door every year without explaining itself.
The problem this post is solving is the gap that doesn't make sense. You have good recipe cards. Your theoretical food cost looks reasonable at 30 percent. Your actual food cost comes in at 35 or 36 percent every single week. You can't find the explanation. That 5 or 6 percent difference isn't one problem — it's usually four or five overlapping causes, each small on their own, adding up to something that materially damages your margin. This post diagnoses each of those causes specifically: over-portioning, spoilage and waste, receiving errors, menu mix shift, and theft. Then it gives you a concrete process to close the gap.
One important note before diving in: the actual vs. theoretical analysis only works if your theoretical cost is built correctly in the first place. If your recipe cards haven't been updated to reflect current ingredient costs — and if you haven't bought the same protein from a distributor recently, you know these costs have moved — then your theoretical baseline is wrong, which makes the variance analysis meaningless. Update your recipe costs to current pricing before running this analysis.
Why is my actual food cost higher than my theoretical food cost?
Actual food cost is always higher than theoretical because it includes real-world factors that recipe cards don't account for: over-portioning, spoilage and waste, receiving errors, menu mix drift, and theft. A 1–2% gap is normal. A 4–6% gap costs a $1M restaurant $40,000–$60,000 annually and points to specific, fixable problems in portioning discipline, purchasing, and inventory controls.
First: What Theoretical Food Cost Assumes (and Why That's the Problem)
Theoretical food cost is calculated by taking your sales mix — every item sold across every shift — and multiplying it by the standard recipe cost for each item using current ingredient prices. It assumes every item was made exactly to spec, with no deviation, no waste, no errors at receiving, and no shrinkage. It's a clean mathematical model of a perfect kitchen. Useful as a baseline. Completely fictional as a description of what actually happens.
The Five Root Causes of Food Cost Variance
Over-portioning — the most common and most invisible cause
A cook who consistently plates a 5 oz salmon instead of 4.5 oz is running 11% over cost on every salmon dish they touch, without knowing it and without it showing up anywhere except the variance number. In a high-volume service, that drift compounds fast. Over-portioning doesn't require malice — it's usually habit, rushing, or a cultural norm of "being generous." The fix is plate weight checks, regular audit drops, and portioning tools (scales, portion spoons, ladles) that make accuracy frictionless during a rush.
Spoilage and waste — especially on proteins and produce
Theoretical cost doesn't account for the 86'd halibut that turned before service, the quart of hollandaise that broke and got tossed, or the dozen lemons that went soft before anyone used them. These losses are inevitable at some level, but they're not fixed — they're manageable. The operators who run tight waste numbers do daily walk-throughs of the walk-in, track what gets 86'd and why, and adjust par levels and ordering frequency to match actual usage velocity rather than estimated need.
Receiving errors — you're paying for what you ordered, not what arrived
A produce delivery where the case weight runs 2 lbs light on every box, a protein order where the portion counts are off by 10, an invoice with a price that doesn't match your agreed-upon contract rate — these receiving errors are common and routinely go unchecked. Someone who received and signed off without weighing or counting just absorbed that loss into your food cost without documentation. Designated receiving procedures with actual weight checks and invoice verification catch this.
Menu mix shift — what you sold isn't what you costed for
Theoretical food cost is calculated against your actual sales mix, but most operators run the theoretical calculation against an assumed or historical mix. If your theoretical assumes 40% protein dishes and you actually ran 55% this week because of a special or a server push, your theoretical cost is wrong before variance even enters the picture. Always run theoretical against actual POS sales mix data, not estimated or historical mix.
Theft and untracked consumption — more common than most operators want to admit
Employee theft in restaurants ranges from unlogged shift meals to outright walking out the door with product. Unlogged staff meals alone can account for 1 to 2 percent of your food cost in a busy kitchen where this isn't tracked. Voided checks where food was already prepared, comps not recorded, back-door product removal — these are the higher-end theft scenarios that surface when the variance gets large and everything else has been ruled out. A camera at the back door and a POS system that requires manager approval on all voids aren't punitive — they're cost controls.
How to Actually Run the Analysis
The mechanics of actual vs. theoretical food cost analysis aren't complicated, but they do require your POS data and your inventory system talking to each other — or a manual process that bridges the two.
Theoretical cost: Pull your POS sales mix report for the period (week, typically). For every item sold, multiply the quantity by the recipe card cost for that item at current ingredient prices. Sum that across all items. Divide by total food revenue. That's your theoretical food cost percentage.
Actual cost: Beginning inventory value + purchases during period − ending inventory value = cost of goods used. Divide by food revenue. That's your actual food cost percentage.
The gap between those two numbers is your variance. Run it weekly. Track it over time. A variance that's consistent tells you about a structural problem (portioning, shrinkage). A variance that spikes in a specific week tells you about a specific incident (a large receiving error, an event with product ordered that didn't sell, a shift with unusually high waste).
What to Do When You Find It: Closing the Gap in 60 Days
Before you can trust the variance number, the theoretical baseline has to be accurate. Pull your five highest-volume proteins, your top produce items, and your highest-cost components. Get current pricing from your invoices — not the price sheet from six months ago. Rebuild those recipe costs. Then run the analysis again.
During a regular service, pull five finished plates for each of your top 10 items and weigh them against recipe spec. You're not looking for perfection — you're looking for systematic drift. If your line consistently plates over spec on the same items, that's fixable with portioning tools and a brief team conversation.
Create a one-page receiving log for each of your primary distributors: protein counts, produce case weights, invoice price verification, condition check on temperature-sensitive items. Any order that doesn't match the invoice or the agreed-upon spec gets a call to the rep that day, before you sign. Most distribution errors will be corrected promptly once vendors know you're checking.
Get a waste log on the expo or prep station. Every item that gets tossed goes on the log with a reason. Every shift meal gets logged against the employee. At the end of the week, add up waste by category. Most kitchens that start this process are surprised by the volume — and the pattern tells you what to order less of and what prep quantity to adjust.
Every void on a check after food is fired, every manager comp, every employee meal should require a manager authorization and a reason code in your POS. This creates an audit trail that surfaces both operational errors and potential theft patterns. If you're seeing $800/week in voids on one terminal, that needs a conversation before it becomes a $40,000 annual problem.
The Costs Theoretical Never Captures — Supply COGS
One category that actual vs. theoretical analysis consistently misses is your non-food supply COGS — items that don't have recipe cards because they're not plated items. Cooking oils, cleaning supplies, paper goods, smallwares, and similar consumables all hit your actual cost but aren't represented in any theoretical calculation. For kitchens running fryer stations, frying oil is a significant weekly supply cost that rarely gets tracked with the same precision as food.
If you have active fryer stations and you're not tracking your oil cost to the dollar, running your frying oil cost specifically often reveals it's a $15,000 to $30,000 annual line item for a busy operation — more than most operators estimate. Understanding your fryer maintenance costs in the same detail you track recipe costs gives you a cleaner picture of your true COGS, and often reveals additional savings opportunities in that supply category.
Real Kitchen Example: Atlanta, 90-Seat Southern Casual, $2.4M Revenue
A 90-seat Southern casual concept in Atlanta was running a consistent 6.8 percent gap between theoretical and actual food cost over three consecutive months. Theoretical came in at 29 percent; actual was landing at 35.8 percent. Owner was puzzled — the menu was well-engineered, the recipes were solid, the volume was strong.
A structured analysis over two weeks revealed three overlapping causes. First, their primary protein supplier had shifted portion weights on their 8 oz chicken thighs — invoices still said 8 oz, but actual thigh weights were running 7.1 to 7.3 oz. Every thigh was being logged at recipe spec cost, but the actual cost-per-plate was higher because they were burning through inventory faster than the recipe predicted. This alone accounted for 1.4 percent of the variance. Second, over-portioning on biscuits — their highest-volume side — was running consistently at 1.35 portions per order instead of 1.0. Third, their waste log was essentially non-functional; trash was going out without documentation.
Fixing the receiving process, renegotiating protein weights with the supplier, and implementing portion scoops on the biscuit line closed the gap from 6.8 percent to 2.1 percent in six weeks. On $2.4 million in revenue at 28 percent food cost, that 4.7 percent improvement represented $32,000 in recovered annual margin — without changing the menu, raising prices, or increasing volume.
People Also Ask
What is a normal gap between actual and theoretical food cost?
A normal gap is 1–3% between theoretical and actual food cost. A 1–2% variance reflects unavoidable yield loss, minor waste, and the inherent imprecision of a working kitchen. A 3–5% gap warrants investigation — it usually points to portioning drift, spoilage issues, or receiving errors. A gap above 5% represents a structural problem: consistent over-portioning, significant waste, theft, or a theoretical cost model that isn't reflecting current ingredient prices.
How often should I run an actual vs. theoretical food cost analysis?
Weekly is the right cadence for most restaurants. Monthly analysis is too infrequent — a problem that starts in week 1 has cost you four weeks of margin before you even see it. Weekly analysis catches drift early, when the fix is usually simple (a coaching conversation with a cook, a call to a vendor, a par level adjustment). Most modern POS and inventory platforms can generate the theoretical report automatically if your recipe costs are maintained in the system.
Sources
- MarginEdge — A Restaurant Operator's Guide to Actual vs. Theoretical Food Costs
- Toast — Actual vs. Theoretical Food Cost
- Restaurant365 — Closing the Gap Between Actual and Theoretical Food Costs
- Meez — Actual vs Theoretical Food Cost: What the Variance Reveals About Your Kitchen
- GloriaFood — Actual vs. Theoretical Food Cost: How to Maintain Consistent Profitability
