Why your restaurant food cost percentage swings month to month

A restaurant owner we work with had the same question two months in a row. July sales came in at $47,000. Food purchases that month totaled $18,500. Dividing the two numbers gave a food cost of 39.4 percent, which his bookkeeper flagged as a problem. In June, on $46,000 in sales, food cost had been 29.1 percent.

Nothing had changed. Same menu, same recipes, same suppliers. The kitchen had not gone sideways. The calculation had.

What food cost percentage actually measures

Food cost percentage is not purchases divided by sales. The correct formula is opening inventory plus purchases minus closing inventory, divided by sales. That result tells you how much product the kitchen actually consumed during the period. The invoice-only formula tells you how much product arrived at the back door.

When large deliveries land late in a month, or when the walk-in is heavier at month-end than at the start, the two results diverge. That divergence is not a kitchen performance problem. It is a measurement problem, and reacting to it with operational changes produces real disruption in response to phantom signals.

Four reasons the two numbers come apart

Delivery timing varies with the calendar. Most restaurants order on a two-week cycle, and that cycle rarely lines up with month-end. When two large orders land in the last ten days of a month, purchases are elevated for that period. Most of the product gets used the following month. Food cost for the delivery month rises and next month it drops, without anything changing in the kitchen.

Opening inventory is never the same twice. A month that starts with a full walk-in has more product available than one that starts lean. Without counting what was on hand at the start of each accounting period, there is no way to know how much of what was used this month was purchased last month. A restaurant that ran a heavy promotion in late July ends July with less product than usual. August food cost appears low because that product was consumed in August but purchased in July.

Spoilage does not appear without a count. A case of produce that arrives, sits too long, and goes in the trash still appears as a purchase. It never generates a sale. Without a closing count, the loss becomes part of the food cost percentage with no way to identify it separately. A walk-in compressor that ran warm over a weekend can cost $400 to $600 in spoilage, and invoice-only accounting has no way to surface that figure.

Vendor invoices do not always match what was received. Shorted cases, missing items, and pending credit memos mean the invoice is recorded in the books before the discrepancy is resolved. Physical counts catch these gaps. Invoice-only tracking passes them through as valid food cost.

The same August in two formats

Here is what one month looks like under each method.

The $2,200 difference is product that moved from the walk-in at August 1 to the walk-in at August 31. It was purchased in July but consumed across both months. The invoice method treated it all as August’s cost.

Why the gap matters

A food cost percentage that moves five or six points month to month without a clear cause is nearly always a measurement artifact. The temptation is to respond: ask the kitchen manager what went wrong, tighten portion controls, or cut low-margin dishes. Those responses are appropriate when the number reflects actual consumption. They are counterproductive when the number reflects delivery timing.

The downstream effect is on menu engineering. Owners make pricing and portion decisions off monthly food cost. A high-cost month caused by a late vendor delivery should not drive a price change. An accurate cost calculation prevents that mistake.

What a proper food cost process looks like

For restaurant clients, we run a physical count on the last business day of every month. The count covers the walk-in, dry storage, and service stations. Opening count plus the month’s invoices minus the closing count equals food consumed. Consumed divided by revenue is the food cost percentage.

The count does not require specialized software. A clipboard, a consistent path through storage, and a spreadsheet are sufficient. What matters is that it runs at the same point every month, with the same format, so the numbers compare cleanly across periods.

For clients using a point-of-sale system with recipe tracking, we reconcile the physical count against what the system predicts based on sales. That comparison shows how much waste occurred beyond what recipes account for.

Best practices for restaurant operators

A few practices that keep food cost accurate over time:

• Count physical inventory at the close of every accounting period, before the first delivery of the new period arrives and before any prep begins on the new period’s first day.
• Record food invoices on the invoice date, not the payment date. An invoice dated the 29th belongs in that month regardless of when the check clears.
• Keep food purchases separate from cleaning supplies, paper goods, and smallwares. Those items arrive on similar vendor invoices but belong in different expense categories. Including them in food cost inflates the percentage.
• When a delivery is short, note the variance on the receiving copy and request a credit memo before the end of the month. Do not wait until period close to reconcile missing product.
• Log waste as it happens. A daily waste sheet, even a simple one, identifies whether a spike in food cost came from kitchen errors, spoilage, or a delivery problem. That context is essential before any operational response is made.

Three questions worth asking

1. What was your ending inventory value at the close of last month, and is that number recorded in your books as a specific dollar amount?
2. Is your food cost percentage calculated from invoices received, or from product actually consumed after a physical count?
3. When food cost moves more than three percentage points month over month, does your team check delivery timing and inventory levels before looking for a kitchen problem?

If any of those questions do not have a clear answer, the food cost on your Profit and Loss report (the P&L) is tracking your ordering habits, not your kitchen’s performance. Send us your last two months of food invoices and your current P&L. We will tell you whether the variance is a real cost problem or a measurement artifact, and what a clean count process would look like for your operation.