Reduce Restaurant Food Cost: 7 Tactics for US Kitchens

The short version. Your food cost is over 32%? There are seven levers you can pull this week without touching your menu prices. At La Verrerie I went from 36% to 28% in 90 days — that's $28,000+ a year recovered on $800,000 of revenue. The method: recipe cards, purchasing, waste, and COGS tracking in parallel. In that order.

Before you pull any lever, understand the number

Food cost is the indicator that separates a restaurant that runs from one that slowly bleeds out — often without the owner noticing until the year-end P&L. It's the ratio between what you spend on raw ingredients and what you bring in from food sales.

If you're doing $5,000 in food sales on a Tuesday and your ingredients cost you $1,600, your food cost is 32%. Simple math. Hard to control when you're running 60 covers solo with three vendors and a BOH that turns over every six months.

Before reducing, make sure you're calculating it correctly — specifically, that you're using COGS (Cost of Goods Sold) and not just eyeballing invoice totals. The formula:

Actual food cost (%) = (Opening inventory + Purchases − Closing inventory) ÷ Net food sales × 100

And critically: track food cost and beverage cost separately. This is the US industry standard. Beer, wine, and spirits have their own COGS line, their own benchmarks (typically 18-24% for bar programs), and their own levers. Mixing them together is one of the fastest ways to lose visibility on where the problem actually is.

For the full calculation method, see the food cost guide for US restaurants.

What level should you be aiming for?

Direct answer: between 28% and 32% for a full-service or casual dining concept. Below 28% for fine dining where price positioning justifies it. Above 35%, you need to act now — not at the end of the quarter.

These thresholds aren't gospel. They reflect what we see in operations that are financially healthy — ones that pay their staff, service their rent, and have enough left to actually pay the owner. A 38% food cost doesn't kill you in a month. It bleeds you slowly, quietly, until the accountant calls.

Benchmarks by concept type (2026)

To know where you actually stand, start by calculating your theoretical food cost per dish — not just the rough COGS number.

The 7 levers to reduce your restaurant food cost

These aren't abstract ideas. This is the sequence I ran at La Verrerie in 2016-2017. Some showed impact within a week. Others took 30 days to hit the numbers.

Lever 1 — Rebuild every recipe card with current prices

This is the starting point. Without precise, current recipe cards, you're flying blind. Every dish needs a defined portion weight, a calculated ingredient cost at today's prices, and a selling price aligned with your target margin.

The trap most US operators fall into: recipe cards built at opening and never updated. Sysco issues quarterly price updates. US Foods sends price change notices. An avocado or beef commodity spike can move your food cost 2-3 points before you've noticed. If your cards still run Q4 prices and it's Q2, your theoretical food cost is fiction — and you're making menu pricing decisions on bad data.

Rebuilding cards takes time. On 40 dishes, expect 3 to 4 days. Do it once correctly, then set up a system to keep them current.

Lever 2 — Separate your food cost and beverage cost immediately

If you haven't done this yet, do it before anything else.

Food cost and beverage cost have different benchmarks, different vendor relationships, and different levers. A craft cocktail program running at 22% beverage cost can mask a 36% food cost if you blend them. You need to see each line clearly before you can fix either.

Set up separate COGS tracking in your accounting: food purchases vs. beverage purchases, food inventory vs. bar inventory. Run the calculations separately. If your food cost is at 34% and your beverage cost is at 20%, you know exactly where the work is.

Lever 3 — Get ahead of Sysco and US Foods quarterly price changes

This is the lever most US indie operators miss.

Sysco and US Foods (and most regional distributors) push quarterly price updates — but they also make mid-cycle adjustments on volatile commodities: beef, poultry, cooking oils, avocado, citrus. These land in your inbox as a price change notice or quietly appear on your next invoice.

The operators who catch it fast are the ones who photograph every invoice and compare line by line to the previous one. The operators who don't — discover the drift 30 days later in the COGS calculation.

Two practical habits: first, designate one person to do a 5-minute invoice scan on every Sysco or US Foods delivery — flag any item that's moved more than 5% from last order. Second, tie that scan directly to your recipe card updates. If ground beef goes up $0.40/lb, every dish with ground beef needs its food cost recalculated before the next service.

Lever 4 — Train the BOH on portion weights

This lever is underrated. When every line cook plates an extra 0.5 oz of protein on every dish, across 60 covers per service, you're losing several food cost points per week without a single traceable cause.

US restaurant industry turnover in BOH runs above 75% annually (NRA data). Every new hire brings their own muscle memory about "how much is enough." Without standardized portion weights that are actively trained and reinforced, your actual food cost will always drift from your theoretical.

Put a scale on the line. Set portion weights in writing on the recipe cards, visible in the kitchen — not just in the POS or a binder nobody opens. Check compliance weekly for the first month.

Lever 5 — Set up weekly COGS tracking

Monthly COGS is the minimum standard. By the time you see a drift after four weeks, it's already cost you. Weekly COGS with a partial count of your high-movement, high-cost items gives you the visibility to correct within the week.

You don't need a full physical count every week. Focus your weekly count on the 10-15 items that represent 70-80% of your ingredient cost: proteins, seafood, produce, dairy. Those are the items where Sysco price changes and over-portioning show up fastest.

Full physical count monthly. Targeted spot check weekly. That rhythm catches problems before they become expensive.

Lever 6 — Streamline your menu

Fewer menu items means less dormant stock, less spoilage, less complexity in the walk-in cooler. Every dish you 86 permanently simplifies your purchasing, reduces your standing order complexity, and cuts the number of items that can go out of code.

The US food truck and fast-casual segments understand this better than full-service restaurants. A focused menu of 12-18 items is easier to manage, faster to execute, and typically runs a lower food cost than a menu of 40. Keep what sells and margins well. Cut the rest without apology.

To see which dishes are working, cross your food cost percentage by cover count. The dishes that are both high-food-cost and low-volume are the ones to revisit first.

Lever 7 — Track food cost weekly, not at year-end

The final lever is visibility. A food cost number discovered in the year-end P&L is endured. A food cost number tracked weekly is managed.

This doesn't require software to start. A spreadsheet with your weekly COGS calculation — opening stock, deliveries received, closing count, food sales — gives you the number in 20 minutes. What matters is the habit, not the tool.

The goal: know your food cost by Monday morning for the previous week. If it moved more than 1.5 points vs. your target, find out why before Friday.

Case study — La Verrerie, 2016-2017

When I took La Verrerie out of formal insolvency in 2015, I had $390,000 of revenue and margins I didn't fully understand. The business was climbing. But the cash flow didn't match the growth. I eventually ran the numbers properly. Food cost was at 36%. Way too high for a hotel-restaurant doing bistronomy with a $32 average ticket.

Three months of work, in order:

Recipe cards first. I recalculated 48 recipe cards with current supplier prices — not the pricing from two years earlier. On some dishes, the gap between assumed theoretical food cost and actual ingredient cost was 5 to 8 points. That alone explained a lot.

Purchasing second. I compared three suppliers on flour — a high-volume line for a place doing house bread and pastry. Found an 18% saving by switching, equivalent quality. No miracle, just a comparison I hadn't run since we opened.

Parallel: portion training and weekly inventory. Set portion weights in writing. Ran a partial weekly count on proteins and dairy. And removed three menu lines that weren't moving and were tying up cash in dead stock.

Result after 90 days: food cost at 28%. Eight points recovered. On $800,000 of annual revenue the following year, that was +$28,000 of gross margin per year — without raising a single price.

Common mistakes

• Mixing food cost and beverage cost. You can't fix what you can't see. Separate them.
• Running COGS only once a month. Too slow to catch vendor price drift or portioning issues in time.
• Renegotiating every vendor at once. Focus on the 3-5 lines that weigh most in your purchasing. Diluting the effort produces nothing.
• Doing portion training once and never reinforcing it. With US BOH turnover above 75%, portioning habits fade in 3 weeks without active reinforcement.
• Confusing lower food cost with cutting quality. The 7 levers above don't touch your product. They correct waste, over-portioning, and purchasing that isn't working.
• Ignoring quarterly vendor price updates. Sysco's Q1 price update can move your food cost 1-2 points without a single operational change. Catch it at the invoice, not the P&L.

Conclusion

Food cost is something you steer — not something you discover.

Three things to carry out of this:

1. Start with the recipe cards. If they're wrong or stale, everything else is built on sand. 40 cards recalculated takes time. It changes the numbers.
2. Catch vendor price changes at the invoice, not the P&L. Sysco and US Foods move prices quarterly. The operators who catch it fast protect their margin. The ones who don't find out 30 days later.
3. Separate your food cost and beverage cost. You can't fix what you can't see clearly. Run both calculations separately from the start.

36% to 28% in 90 days — it's doable. Not with shortcuts. With order.

Ready to see your food cost updated automatically from your Sysco and US Foods invoices? Book a 20-minute demo — or learn more about how Onrush handles food cost tracking for US indie restaurants.

Written by Cyril Quesnel, founder of Onrush. 20 years in restaurant kitchens in France, two turnarounds. Building the tool I wish I had.