Food Cost Variance Analysis: Find Where Money Disappears
The gap between what your food should cost and what it actually costs is where profits leak. Here's how to measure the variance, investigate the causes, and close the gap.
Food cost variance is the difference between theoretical food cost (what your recipes say you should spend) and actual food cost (what your financial records show you actually spent). A positive variance means you spent more than expected — money disappeared somewhere between the delivery dock and the guest's plate. Systematic variance analysis identifies where, why, and how much.
What Is Food Cost Variance?
Every restaurant has two food cost numbers. Theoretical food cost (also called ideal or standard) is calculated from your standardized recipes — if every dish is prepared exactly as specified, with zero waste and no unrecorded usage, this is what your food should cost. Actual food cost is calculated from your financial records: beginning inventory plus purchases minus ending inventory, divided by sales.
The gap between these two numbers is the variance. A small gap (1-3%) is normal — no kitchen operates at 100% efficiency. A large gap (above 5%) signals operational problems that are actively draining profit.
Food Cost Variance %
A positive result means you spent more than your recipes predict.
Example: 34% actual − 30% theoretical = 4% variance
On a restaurant doing €80,000 in monthly food revenue, every 1% of variance equals €800/month or €9,600/year. A 5% variance is €48,000/year walking out the door. That is not a rounding error — it is a management failure.
How to Calculate Variance: Step by Step
Step 1: Calculate theoretical food cost
Multiply the food cost per portion of every dish sold during the period by the number of portions sold. Sum these to get total theoretical cost. This requires standardized, costed recipes for every menu item — there are no shortcuts.
Theoretical Food Cost
Pull dish-level sales data from your POS system.
Use current ingredient costs, not outdated recipe costs.
Step 2: Calculate actual food cost
Use the standard period formula. This requires an accurate inventory count at the start and end of the period, plus a complete record of all purchases.
Actual Food Cost
Count inventory at the same time of day each period (before opening).
Include ALL purchases — cash buys, emergency runs, and transfers between locations.
Step 3: Compare and investigate
Express both numbers as a percentage of revenue. The difference is your variance. If the variance exceeds your tolerance (typically 3%), it demands investigation.
The 7 Most Common Causes of Variance
Variance does not appear randomly. It follows predictable patterns with identifiable root causes.
1. Over-portioning
The most common cause. A cook who "eyeballs" 200g of salmon might plate 230g. That is 15% more protein per plate. Across 60 salmon portions per week, the extra 30g × 60 = 1.8kg of extra salmon at €22/kg = €39.60/week from one dish alone. Multiply across the menu and over-portioning easily accounts for 2-3% of variance.
2. Waste and spoilage
Ingredients that expire before use, trim waste beyond what recipes account for, and preparation errors that result in discarded product. A busy kitchen that doesn't enforce FIFO rotation can lose 3-5% of perishable inventory to spoilage weekly.
3. Theft
Uncomfortable but real. Industry estimates suggest 4-5% of restaurant revenue is lost to employee theft. It ranges from pocketing steaks to pouring unrecorded drinks. Variance analysis does not prove theft, but a persistent unexplained gap — especially in high-value categories — warrants investigation.
4. Unrecorded comps and staff meals
A table gets a free dessert to apologize for a long wait. The server doesn't record it. The kitchen made the dessert, used the ingredients, but the POS shows no sale. Staff meals that aren't tracked have the same effect — real cost with no corresponding revenue.
5. Receiving errors
Deliveries that are short-weighted or short-counted but accepted without verification. If you ordered 20kg of chicken but received 18kg and didn't catch it, you paid for 2kg you never got. Over a year of weekly deliveries, this adds up.
6. Recipe inaccuracy
Your recipes say the bolognese uses 120g of minced beef per portion, but the kitchen has always used 150g. The theoretical cost is based on 120g — reality is 25% higher. The recipe was never accurate to begin with, which produces permanent variance that no operational improvement can fix.
7. Price changes not updated
Your supplier raised the price of olive oil from €7/L to €9/L three months ago, but your recipe database still shows €7/L. Your theoretical cost is calculated on outdated prices, making actual cost appear higher than it "should" be. This is a data problem, not an operational one.
Variance analysis tells you how much money is missing. It does not automatically tell you why. Resist the temptation to assume the cause — investigate systematically using the framework below.
The Variance Investigation Process
When variance exceeds your tolerance, follow this investigation hierarchy. Start with the largest-impact, easiest-to-verify causes and work down:
- 1.Verify recipe accuracy — are your standardized recipes current? Test-cook your top 5 dishes and compare actual ingredient usage to what the recipe specifies. This eliminates recipe inaccuracy as a false signal.
- 2.Verify ingredient prices — are all costs in your system up to date? Compare your recipe database prices against the last 3 supplier invoices. Update any discrepancies.
- 3.Run category-level variance — break the analysis down by ingredient category (proteins, dairy, produce, dry goods). If variance concentrates in one category, the investigation narrows immediately.
- 4.Check portioning compliance — weigh 10 portions of your highest-cost dishes during a live service. Compare to the recipe spec. Even a 10% over-portion on a €5 protein component costs €0.50/plate.
- 5.Audit receiving logs — pull delivery receipts for the period and compare ordered quantities to received quantities. Spot-check weights on the next 3 deliveries.
- 6.Review comp and waste logs — add up all recorded comps, voids, staff meals, and waste for the period. If the total is less than your variance, the rest is unrecorded loss.
Corrective Actions That Actually Work
Identifying the cause is half the battle. Closing the gap requires action:
| Cause | Corrective Action | Expected Impact |
|---|---|---|
| Over-portioning | Portion scales at every station + daily spot-checks | 1-2% cost reduction |
| Waste/spoilage | FIFO enforcement + waste logs + smaller, more frequent orders | 0.5-1.5% cost reduction |
| Theft | Camera coverage + key-access storage for high-value items | Variable |
| Unrecorded comps | Mandatory POS entry for all comps and staff meals | 0.3-0.8% cost reduction |
| Receiving errors | Weigh and count every delivery against the PO | 0.2-0.5% cost reduction |
| Recipe inaccuracy | Re-test and update recipes quarterly | Eliminates false variance signal |
| Price not updated | Update ingredient costs on every invoice receipt | Eliminates false variance signal |
Proteins (meat, seafood, dairy) typically represent 40-50% of total food cost. A 5% variance in proteins alone can account for 2-2.5% of overall food cost variance. Always investigate proteins first — the ROI per hour of investigation is highest here.
How Often to Run Variance Analysis
Weekly is the minimum cadence for any restaurant serious about cost control. Monthly analysis is too late — by the time you discover a variance at month-end, you have already absorbed four weeks of losses.
- Weekly — compare theoretical vs actual at the total level. Flag weeks where variance exceeds 3%.
- Monthly — run category-level analysis. Identify which ingredient groups are driving variance.
- Quarterly — full recipe audit. Re-test your top 20 dishes, update all ingredient prices, and recalibrate your theoretical baseline.
The weekly check takes 30-45 minutes with a recipe costing system that tracks theoretical cost automatically. Without software, you need accurate POS data, current recipe costs, and a spreadsheet — doable but time-consuming.
What Is an Acceptable Variance?
Industry benchmarks for food cost variance by operation type:
| Operation Type | Acceptable Variance | Action Threshold |
|---|---|---|
| Fine dining | 1-2% | Above 3% |
| Casual dining | 2-3% | Above 4% |
| Fast casual / QSR | 1-2% | Above 3% |
| Catering | 2-4% | Above 5% |
| Bar food | 3-5% | Above 6% |
Fine dining and fast casual have tighter tolerances because their cost structures depend on precision — fine dining due to expensive ingredients, fast casual due to thin margins. Bar food tolerates higher variance because beverage margins (70-80%) subsidize food losses.
More in This Topic
Key Takeaways
- Food cost variance = actual food cost % minus theoretical food cost %. Every 1% of variance on €80K revenue equals €800/month lost.
- A 1-3% variance is normal. Above 5% signals a management problem that demands immediate investigation.
- The top three causes are over-portioning, waste/spoilage, and unrecorded comps — not theft (though theft should not be ignored).
- Investigate systematically: verify recipes first, then prices, then portion compliance, then receiving, then waste logs.
- Run variance analysis weekly at minimum. Monthly is too late to prevent losses from compounding.
Track your food cost variance automatically
Cucinovo calculates theoretical cost per dish from your recipes. Compare against actual spend to find where money disappears. 14-day free trial.
Start free trial