Inventory shrinkage in restaurants: why 2–10% of revenue is leaking, and how to catch it

Your recipes say food cost should be 28%. Your P&L says food cost is 32%. The gap is four points of revenue. On a $1.5M shop, that's $60,000 a year you'll never see.

That gap has a name. It's called shrinkage, and it's the hidden tax most independent restaurants pay every month without realizing it. Industry estimates put it at 2–10% of revenue across QSR, fast-casual, and casual. The wide band is the tell — operators who measure it sit at the low end, operators who don't sit at the high end, and the difference is mostly whether anyone is looking.

This piece is about what shrinkage actually is, where it hides in a restaurant, and the weekly habit that brings it down from "we don't know" to a real number you can manage.

What shrinkage is, in plain language

Shrinkage is the difference between what your recipes say you should have used and what you actually used. Two ways to express the same thing:

• Theoretical food cost — what your sales mix says it cost you to make and sell what you sold, at recipe-spec portion sizes.
• Actual food cost — what your inventory says you actually used (beginning inventory + purchases − ending inventory).

The gap between those two numbers is shrinkage. It can come from a dozen places, but it always shows up in the same way: actual is higher than theoretical, and the difference erodes margin you assumed you had.

The brutal part is that the P&L doesn't separate shrinkage from food cost. It rolls them together into one COGS line. So when you see food cost at 32%, you don't see "28% recipe + 4% leak." You just see 32% and assume that's what it costs to run the kitchen. It isn't. Four of those points are recoverable.

Where shrinkage actually hides

Six places, in rough order of how much damage they do:

1. Over-portioning. The line cook who's been there three years and is now putting 7 oz of chicken on a sandwich the recipe spec'd at 6 oz is costing you 14% on every chicken sandwich. Multiply by sandwich volume and this is usually the biggest single source of shrinkage in a QSR or fast-casual. Nobody is doing it on purpose. The portion just drifts.

2. Waste — prep and line. Trimmings, items prepped and not sold, items dropped on the floor, items burned. A small amount is normal. A lot is a process problem — usually wrong prep quantities, not staff carelessness.

3. Voids, comps, and "free" food. Voided orders that the food still got made for. Comps without manager approval. Staff meals that weren't logged. Items remade because of order errors. Each of these is a real food-cost expense that didn't generate revenue.

4. Theft. It exists. It's almost never as dramatic as operators imagine — the dramatic version (cash skimming, walking out with cases of meat) is rare. The common version is two ounces of soda extra in a "small," a free drink for a friend, a sandwich rung up at half price. Steady, small, and adds up to real money over a month.

5. Spoilage and expiration. Anything you ordered that aged out before you used it. Usually a buying problem (over-ordering perishables) or a rotation problem (FIFO not being followed).

6. Receiving errors. You paid for a case of 24 and the truck dropped 22. You paid for prime and got choice. Vendors aren't always wrong on purpose, but if no one checks the dock, the errors compound.

The first three usually account for 70%+ of the shrinkage. Fix those before you go hunting in the others.

What "good" looks like

Tight benchmarks for shrinkage as a percentage of revenue:

• Under 2% — best in class. Usually means weekly inventory + portioning audits + a tight comp/void process.
• 2–4% — what most well-run independents land at. Acceptable if it's stable; concerning if it's drifting up.
• 4–7% — a real leak. Worth a focused 30-day project to bring it down.
• Over 7% — you almost certainly have a specific cause, not a thousand small ones. Find it.

Most operators who haven't measured assume they're at 2%. When they actually count, they're usually at 5–6%. The first audit is the painful one. The second is a relief.

A worked example — finding the gap

Take a single-unit QSR doing $125,000 in monthly net sales. The P&L shows:

• COGS: $40,000 → 32% food cost
• Recipe-theoretical COGS (from sales mix × recipe costs): $35,000 → 28% theoretical

Shrinkage = $40,000 − $35,000 = $5,000 / month, or 4% of revenue.

That's the headline number. To make it actionable, break it down by category:

Category	Theoretical $	Actual $	Variance
Proteins	$14,000	$16,200	$2,200 (15.7%)
Dairy	$5,500	$5,800	$300 (5.5%)
Produce	$4,000	$4,400	$400 (10%)
Dry goods	$7,000	$7,200	$200 (2.9%)
Beverage	$3,500	$5,400	$1,900 (54%)
Paper / packaging	$1,000	$1,000	$0

The two categories that jump are proteins and beverage. Beverage at 54% variance is a smoking gun — almost certainly a free-pour or comping problem on the soda fountain or coffee program. Proteins at 15.7% is classic over-portioning. Together, those two categories explain $4,100 of the $5,000 monthly leak.

Now you have a project. Not "fix shrinkage" — that's vague. Specifically: re-train the line on protein portioning, and audit beverage comps for the next two weeks. Two concrete things, both fixable in a month.

That diagnostic — going from "food cost is high" to "two specific categories with two specific causes" — is the entire point of measuring shrinkage by category. The category breakdown turns a vague margin problem into a list of decisions.

The weekly habit that brings shrinkage down

Three pieces, all weekly, all about an hour total:

1. A weekly inventory count, by category

Not a full count of every SKU — that takes four hours and burns out whoever does it. Count the high-velocity, high-cost items only: proteins, alcohol if you serve it, specialty ingredients, and the top 5 dry-goods items by spend. That's usually 25–40 items. Forty-five minutes once a week.

The math: opening inventory + purchases − ending inventory = actual usage. Compare against theoretical usage from the POS-derived sales mix. The variance is shrinkage.

Same day every week. Sunday night or Monday morning works for most operators because the prior week is closed and you can roll the count into a single sit-down with the books.

2. A waste log on the line

A clipboard, a pen, and a rule: anything that gets thrown away or remade gets logged. Item, quantity, reason. No blame, no consequences for staff — just data. Five seconds per entry.

After two weeks the waste log tells you exactly which items are getting dropped, burned, or remade. Usually one or two items account for half the waste, and the fix is a process change: a different prep schedule, a different par level, a different storage container.

3. A weekly comp/void audit

Pull the comp and void report from the POS once a week. Look at every comp over $5 and every void after the food was made. Most will be legitimate. The illegitimate ones — comps without manager approval, repeat patterns from the same employee, voids that look suspicious — are where the theft and "free food" leaks live.

This audit is uncomfortable the first time. After the first conversation with whoever's behind the pattern, it gets a lot less uncomfortable for everyone, including the staff who weren't involved.

What to do this week

If you've never measured shrinkage:

1. Pick this Sunday or Monday. Do one inventory count of proteins + alcohol + specialty items. 45 minutes.
2. Pull last month's COGS from the P&L and last month's theoretical food cost from your POS recipe report (every modern POS has one). Subtract.
3. The difference is your monthly shrinkage in dollars. Divide by revenue to get the percentage.

That single number — your shrinkage % — is the baseline. Repeat the count next Sunday. Compare. The drift will tell you whether the leak is getting worse or you're starting to close it.

Want the worksheet?

We built a free Restaurant Daily P&L Cheat Sheet that includes a weekly inventory variance section — opening, purchases, ending, theoretical, variance — pre-built for the categories above. Plus an 8-page guide with the daily ritual and shrinkage benchmarks by concept.

Get the cheat sheet (free, no credit card) →

If you'd rather not pull theoretical vs actual usage by hand every week, EasyShiftHQ connects POS, payroll, and inventory to flag the variance for you weekly, by category. 14-day free trial, no credit card.

---

Jose Delgado runs a Cold Stone Creamery / Wetzel's Pretzels co-brand at Alamo Ranch in San Antonio. He built EasyShiftHQ because he was tired of finding out two weeks late where the food cost actually went.