Food cost mistakes vs the right method

The no-count mistake: how you leak money every month

A mid-market restaurant with $120K monthly food sales that doesn't audit inventory is leaving money on the table every month unseen. Here's how: January you report spending $12K on ingredients against $120K in sales—30% food cost. Looks good. But while that happened, your cold storage inventory went from $3K to $5.2K—you bought more than you cooked. Real January COGS was $12K purchases minus $2.2K inventory growth, meaning $9.8K actually got eaten. That's 8.2% food cost, not 30%. Next month when inventory dips, you'll see 36% due to compensation. Without monthly counting you think you're floating 30-36% when you're really steady at 28%. That optical illusion makes you believe margins are healthier than they are. Multiply the mistake by twelve months and you make wrong calls on pricing, suppliers, even viability. The damage compounds because you're benchmarking yourself against phantom numbers.

Untracked shrink: the 3% nobody sees

Shrink is what you throw out: spoiled vegetables, expired meats, rejected plates, kitchen waste—and what silently vanishes. Mid-market restaurants typically shrink 2-3.5% of ingredient consumption. Sounds small. It's not. If your monthly ingredient inventory is $25K and shrink runs 3%, you're tossing or donating $750 monthly. That's $9K yearly, silent, no budget line. Twenty restaurants combined: $180K vanished annually by not tracking shrink. Worst part is it gets normalized: 'that always happens,' chefs say. What always happens is NOT COUNTING. When Masterestaurant audits shrink week-by-week, we find locations at 5%, and when recipes are reaudited and staff rerained, it drops to 2%. The difference is $400 monthly per location—money that existed but was invisible. Weekly shrink control with scales and logs is the cheapest lever in the kitchen: thirty minutes weekly, zero cost, potentially $2K to $5K yearly per location freed.

Lost discounts: money you don't see because it never enters

Your supplier offers 5% off for on-time payment, or $200 bonus for volume. Where does that money land in your calculation? If you don't standardize it, half your team logs it in the invoice and half notes it separately, and your COGS drifts inconsistently month to month. Real example: a five-pizzeria group had total invoices of $18K but unlogged discounts worth $1.2K—often on-time payments that partners knew had to happen, but the amount never hit paperwork. When untracked, those $1.2K don't subtract from COGS; your food cost reads 6.7% high. Over a year that's $14.4K you counted as 'expense' when it was discount. Fix: three ledger lines: gross invoice, discounts (categorized: on-time / volume / return), net purchases. Without it, your 12-month series has noise and you can't tell if supplier prices rose or your negotiation discipline fell.

12-month trend: why one good month means nothing

One month at 28% food cost tells you almost nothing—could be noise, a high-sales month, a supplier running slower. But 12 months of 28%, 29%, 28.5%, 29%, 31%, then 30%, 28.5% tells the truth: your steady state is 29% with seasonal peaks. That's insight. If the series were 28%, 29%, 30%, 31%, 32%, 33%—rising trend—you'd see something is breaking. An accountant sees the average (30.2%) and says 'acceptable'. An operator tracking the series sees the upward slope and acts BEFORE it hits 34%. Your 12-month trend is also your market-change detector: when you see costs started rising four months back, you trace it (when did the supplier change? When did you launch that new dish?) and make surgical calls. Masterestaurant found restaurants where one distributor was charging 12% premium with no reason—a single phone call freed $4K yearly.

12-month trend: why one good month means nothing — in practice

Without the series, they'd kept overpaying silently. When Masterestaurant arrives and audits real food cost (physical inventory, net purchases, weighed shrink) versus what was reported via estimation, the average gap is 4.2 percentage points. That means: reported 29.5%, actual 33.7%. On $140K monthly food sales, that 4.2 points is $5,880 monthly—$70K yearly difference between what they thought they spent and what they actually spent. A Mexico City pizzeria running 2,500 pies monthly calculated food cost at 31%; after audit, 36.2%. The issue was nobody had asked for beginning and ending physical counts, so they used 'estimated average' from old months plus current purchases. When we subtract actual beginning inventory (weighed that day) + purchases − actual ending inventory (weighed at close), we found they consumed MORE ingredients than their sales accounted for. Kitchen wasn't being robbed—they were pricing promotions off-menu, serving heavier portions 'because it looked right', making unlogged donations.

Audited COGS versus estimation: the gap that matters

From audit they learned cost was the villain, not the supplier. Six weeks of recipe and portion work brought it to 30%. That's money that existed inside the business—you just had to count right to see it. **No inventory counting:** $8K in monthly purchases looks like 32% food cost, but if inventory grew from $3K to $5K, actual COGS is $6K and the % is 24%. Two points lost by ignoring the balance sheet. Run that annually and you'll believe your margin is 8 points better than it actually is. **Normalized shrink:** 3% shrink 'always happens' goes untracked. On a menu with $25K monthly in ingredients, 3% is $750. Twenty restaurants × $750 × 12 months = $180K yearly that nobody sees. The right method weighs damaged cans, wilted vegetables, daily waste—it's more than you think, and with exact numbers you can attack it. **Discounts that vanish:** your supplier gives you $300 this month for on-time payment.

Five mistakes costing $1K+ each

Is that subtracted from purchases or left in the invoice? Without a rule, half your team won't record it, and your food cost drifts 0.8 points with no visible reason. **Last month's inventory:** you close the books on the 30th but count inventory on the 15th of next month. That two-week gap, with daily sales and new purchases, contaminates the calculation. Close date must be fixed and synchronized with your accounting close, or it doesn't exist. **Ignored trend:** food cost was 31%, then 29%, then 31%, then 33%. That's not noise—it's that beef prices spiked week 3 of August and nobody adjusted menu pricing. With 12 months of data you see the pattern and react next month. Without it, you wait a quarter to 'review costs'.