Is a 2% Inventory Variance Bad? What's Actually Acceptable

A new manager I was training ran his first variance report, got 2%, and asked me straight up: “is that good or bad?” Fair question. The honest answer is “depends what it’s on.” A 2% variance on paper napkins is nothing. A 2% variance on your ribeye is real money. Same number, two completely different conversations, and you cannot tell which one you are having until you know which item it is sitting on.

So here is what actually counts as acceptable, and how to set a target you can hold the line on.

The short answer #

Acceptable inventory variance is 1 to 3% for food overall, and you want under 2% on high-cost proteins. Liquor tolerates more, 3 to 5%, because of free-pour and spillage. Draft beer runs 5 to 10% thanks to foam and line cleaning. Anything above its band is a process problem, not noise.

That is the cheat sheet. The trap is treating one number as one target across the whole inventory. That is how operators end up chasing zero on napkins while a protein quietly bleeds out the back.

Variance is not food cost #

This is the confusion that wastes the most time, so clear it up first. Food cost percentage is COGS divided by sales. It tells you what you spent. Variance is the gap between what you should have used (theoretical) and what you actually used (counted). It tells you where the money went sideways.

You can run a perfectly fine 30% food cost and still have a 12% variance hiding inside it. The food cost looks healthy on the P&L because the average smooths everything out, and the average is a liar. The variance is the part that says one specific item is leaking. That is the number you act on. Watch only the food cost percentage and you are watching the average, which means you are watching the leak walk right past you.

Acceptable bands by category #

Different categories tolerate different variance for real reasons, not arbitrary ones.

Category	Acceptable variance	Why
Food, overall	1 to 3%	Healthy target for a tracked kitchen
High-cost proteins	Under 2%	Where the dollars concentrate, watch tightest
Produce	3 to 5%	Yield loss and trim are real and variable
Liquor	3 to 5%	Free-pour and spillage build in slack
Draft beer	5 to 10%	Foam and line cleaning eat volume

Produce runs looser because yield is genuinely unpredictable. A case of lettuce is not a fixed number of usable ounces, and it never will be. Liquor runs looser because even a good bartender’s free-pour drifts over a shift. Draft runs loosest of all, because you physically pour beer down the drain every single time you clean a line. None of that is theft. It is the cost of the category, baked in.

How to set your target #

Start looser than you think and tighten where the money is.

1. Run a few weeks first. Get a baseline before you set a target. You cannot judge 2% without knowing what your kitchen normally runs.
2. Tighten on the expensive stuff. Put your real attention on the top 10 cost items. A 1% improvement on a ribeye beats a 5% improvement on fountain syrup every time.
3. Leave the cheap stuff loose. Chasing variance on bread or napkins burns hours for pennies. Let it ride.
4. Do not chase zero. Zero variance usually means somebody is fudging the count, not that the kitchen is perfect. A real kitchen has some variance. The goal is catching the leaks, not a perfect scorecard.

When the number is telling you something #

A single item at double its band is the signal. That is when you stop reading the report and walk into the kitchen.

And I mean walk in. Not send an email, not flag it for the next manager meeting. The cause is almost always something you can see standing on the line: a portion that drifted, a comp button getting leaned on, a vendor shorting a delivery. The report tells you which item. The kitchen tells you why. For how to run the count itself, see how to calculate inventory variance, and for the food-cost side of the same gap, theoretical vs actual food cost.

What this looks like in the calculator #

The inventory variance calculator on this site gives you variance by item in both units and dollars, so you can hold each category against its own band instead of one blanket target. Watch the dollar column. That is the one that tells you which leak is actually worth your Monday morning.

What to do today #

Take last week’s variance report, if you have one, and sort it by dollar impact, not by percentage. The item at the top of that list is your real problem, even when its percentage is not the scariest one on the page. Set a tight band on your five biggest cost items. Leave the cheap stuff alone. And stop measuring success by hitting zero, because zero was never the job. Catching the one leak that matters is.

Sources: BinWise, MarketMan, Sculpture Hospitality, Chefs Resources.