Back to Tab Blog

How to Reduce Restaurant Food Costs: 12 Ways

Find which part of your food cost moved, then fix it. 12 moves covering vendor prices, receiving, waste, portions, menu mix, and purchasing.

August 8, 2026
How to Reduce Restaurant Food Costs: 12 Ways dark thumbnail
Tab Commerce food cost guide cover showing a fanned stack of vendor invoices with actual food cost at 28.8 percent against a 27.5 percent target
Food cost control

How to reduce food costs in a restaurant without cutting blindly

Start here

Measure the gap, then find which of five things moved: what you paid, what arrived, what got used, what got sold, or what got recorded.

Find the failed branch, then run the matching fix. Tab helps with two of them: Andy AI checks invoice prices and missed rebates, while Tab Card keeps off-invoice purchases tied to the right restaurant.

Get the real number, find the failed branch, then run the fix that matches it.

Contents
  1. Get your real number first
  2. Find where the gap starts
  3. 12 ways to cut restaurant food costs
    1. Recount the expensive stuff
    2. Check invoice prices against your agreement
    3. Chase the credits and rebates you already earned
    4. Tighten receiving and substitutions
    5. Track yield on your big-ticket proteins
    6. Standardize portions at the station
    7. Log prep, spoilage, and plate waste separately
    8. Cut the emergency grocery runs
    9. Simplify low-volume ingredients
    10. Rebuild the menu around margin and mix
    11. Audit discounts, comps, and voids
    12. Give food cost one weekly owner
  4. Your 30-day food cost plan
  5. Where Tab fits around this
  6. Questions operators ask about food cost

Get your real number first

Food cost % = (beginning inventory + purchases - ending inventory) / food sales x 100.

That's period food cost, and it only works if the pieces line up. Count on the same weekday at the same hour, close the period on the same day your sales report closes, and include every purchase, including the card run somebody made on Saturday.

Food cost percentage, worked

($18,400 + $52,000 - $16,900) / $186,000 x 100= 28.8%

  • Beginning inventory$18,400
  • Purchases$52,000
  • Ending inventory$16,900
  • Cost of food used$53,500
  • Food sales$186,000

Actual food cost28.8%

  • Your target 27.5%
  • Gap 1.3 points
  • Gap in dollars for the period $2,418
1.3 points on $186,000 in food sales is $2,418. Assumptions: one 4-week period, food only, no beverage.

Run one clean period before changing anything. Then compare it with your target, built from your P&L: 100% minus labor, occupancy, other operating costs, and required profit.

42% of operators said they were not profitable in 2025. A target borrowed from another concept can make you cut the wrong line.

If you want the math done for you, the free restaurant food cost calculator runs period cost and recipe cost in the browser, which is faster than rebuilding a spreadsheet every month.

Find where the gap starts

Your food cost percentage is one number sitting on top of five separate jobs. When it moves, one branch moved and the other four are innocent.

Where the money leaks

  1. What you paid

    • Vendor price creep
    • Off-contract buying
    • Emergency grocery runs
  2. What arrived

    • Short deliveries
    • Unapproved substitutions
    • Wrong pack size
    • Credits never claimed
  3. What was used

    • Prep waste
    • Spoilage
    • Over-portioning
    • Yield loss
    • Theft
  4. What was sold

    • Comps
    • Voids
    • Discounts
    • Menu mix shift
  5. What was recorded

    • Missing invoices
    • Card purchases never coded
    • Wrong period

Every one of these shows up as the same symptom: food cost percentage up.

Five stages run left to right, and each one leaks in its own way. Paid leaks through vendor price creep, off-contract buying, and emergency grocery runs. Arrived leaks through short deliveries, unapproved substitutions, wrong pack sizes, and credits never claimed. Used leaks through prep waste, spoilage, over-portioning, yield loss, and theft. Sold leaks through comps, voids, discounts, and menu mix shift. Recorded leaks through missing invoices, card purchases never coded, and the wrong period.

Each branch has one check that settles it. Run them in order and stop at the first one that fails.

One settling check for each of the five branches of food cost, with the data source, the owner, and the first action.
Branch The check Where the data lives Who owns it First action
What you paid Invoice price vs your agreed price on your top 20 items by spend Vendor portal, invoices, contract or GPO sheet Owner or controller Pull four weeks of invoices and price-check the top 20
What arrived Case count and weight vs the invoice, plus any substitution Receiving log, scale, delivery tickets Receiving manager Weigh one protein case per delivery for two weeks
What was used Theoretical usage from recipes vs actual usage from counts POS mix report, recipe specs, inventory counts Chef or kitchen manager Yield-test your two most expensive proteins
What was sold Comps, voids, and discounts by employee and reason code POS reports GM Read last week's comp report line by line
What was recorded Every purchase in the period, invoices and card charges both Accounting system, card statements, receipts Bookkeeper or controller Match card charges to receipts for the last 30 days

Branch three is actual versus theoretical: what recipes and sales say the kitchen should have used versus what counts and purchases say it used. That is the gap portion and waste work can move.

The last branch is the one that fools people. If a $2,100 produce invoice landed in the wrong period, or four club-store runs never got coded to the restaurant, your food cost is wrong and your kitchen is fine. Check the paperwork before you call a meeting.

12 ways to cut restaurant food costs

Run these against the branch that failed. Doing all twelve at once is how you get twelve half-finished projects and the same number next month.

1) Recount the expensive stuff

Recount proteins, seafood, cheese, and oil first. A $600 ending-inventory error moves food cost 0.32 points on $186,000 in sales. Use the same people, order, and time each week; if counts still live on a clipboard, compare restaurant management software.

2) Check invoice prices against your agreement

Put four weeks of invoices for your top 20 items beside the agreed prices. Some variance is real market movement; the rest only appears when the agreement and invoice sit together.

If you use Tab Pro, Tab's Andy AI checks invoice lines against vendor agreements and flags price creep or missed rebates for review.

Sort by dollars, not percentage. A 4% protein overcharge matters more than a 30% overcharge on to-go lids.

One line item, priced two ways

One case of one item Before After
Who checks the priceNobody, the invoice gets approvedThe person receiving, at the door
When it gets caughtMonth-end, if everSame week
Contracted price per case$32.40$32.40
Invoiced price per case$35.90$32.40 after the credit
Variance per case$3.50$0
Cases per week2222
Cost of the drift, per week$77$0
Cost of the drift, per year$4,004$0
Illustrative. Assumes one item, 22 cases a week, 52 weeks, price drift caught and credited.

Vendor agreements, scorecards, and how often to re-bid are their own job, and the restaurant vendor management guide covers the cadence so this doesn't become a once-a-year fire drill.

3) Chase the credits and rebates you already earned

Shorts, damage, off-contract prices, rebates, and later statement adjustments are money you already earned. Nobody sends it unprompted, and many claims expire.

Give one person the vendor portal every week. Reading contract terms and invoice lines at volume is also where invoice analysis tools earn their keep, which the AI for restaurants guide walks through with real use cases.

4) Tighten receiving and substitutions

Receiving is where a price problem becomes a cost problem. No signature without a count.

Weigh one protein case per delivery, check pack sizes against the order, and refuse substitutions nobody approved. A swapped brand at the same case price can carry a different yield, which changes your cost per plate without changing a single invoice number.

Where you can staff it, the person who receives shouldn't be the person who orders.

5) Track yield on your big-ticket proteins

Cost per case is not cost per plate. Run a yield test: weigh raw in, weigh usable out, divide.

A whole muscle at $6.80 a pound with 72% usable yield actually costs $9.44 per usable pound. Price your specs off that number, not off the invoice.

Start with proteins because they carry the most dollars.

6) Standardize portions at the station

Scales, scoops, and ladles on the line, with spec cards where the cook can actually see them.

Half an ounce over on a 6 oz protein, 300 plates a week, is about $4,600 a year at $9.44 per usable pound. Nobody in the building can feel that happening.

From case price to plate cost

  1. What the invoice says

    $6.80per pound

    Whole muscle, as delivered and as billed.

  2. What you can actually serve

    $9.44per usable pound

    72% usable yield after trim and cook loss, so the real cost per pound is $9.44.

  3. What half an ounce costs

    $4,600a year, about

    A 6 oz spec, half an ounce over, 300 plates a week, at $9.44 per usable pound.

Illustrative. One protein, one spec, 300 plates a week, 52 weeks. Run your own yield test before you reprice a spec.

7) Log prep, spoilage, and plate waste separately

One waste bucket tells you nothing. Three buckets tell you whether the problem is a recipe, a par level, or a plate the guest didn't want.

Prevention is the only one of those that gets your money back. The EPA calls preventing wasted food the most environmentally beneficial option on its Wasted Food Scale, ahead of composting, because a composted case is still a case you paid for.

8) Cut the emergency grocery runs

The 4pm run to the club store is the most expensive food you buy, and the least visible. It arrives as a card charge, not an invoice, so a food cost built from invoices misses it completely.

Three fixes, in order: set par levels on the items that trigger the runs, name one approved buyer, and make every card purchase get coded to the restaurant the day it happens. Tab Card texts the buyer for the receipt and restaurant after the swipe, so the emergency run lands in the same food-cost period instead of becoming a month-end mystery.

If the runs are really an ordering problem, the best restaurant procurement software roundup covers the tools that handle supplier ordering and par levels properly.

9) Simplify low-volume ingredients

Every one-dish ingredient is a spoilage risk, count line, and storage slot. A $60 case that is half discarded bought $30 of usable food for $60. Cross-use it or reconsider the dish, then confirm the decision with the margin-and-mix test in move 10.

10) Rebuild the menu around margin and mix

Food cost percentage and contribution margin are different arguments. A 38% item making $14 a plate beats a 24% item making $5 a plate, and mix decides which one your guests actually order.

Sell more of the high-margin items by placement, description, and server recommendation before you touch a single price. The menu engineering guide covers how to sort items by popularity and margin without guessing.

11) Audit discounts, comps, and voids

Pull comps by employee and reason code weekly. Read voids after the item fired closely: the food left inventory without revenue. Look for patterns, such as one server with triple the comps of peers, rather than one large Saturday comp.

12) Give food cost one weekly owner

A number nobody owns drifts. Name one person, one day, and one 30-minute review with the same agenda every week.

That 30 minutes has to come out of something. Take it out of chasing receipts, not out of floor time. The point of tightening the back office is putting people back in the dining room, not adding a meeting.

Your 30-day food cost plan

Four weeks, in order, with one named owner. If week 1's count and sales period do not align, every later fix targets a number that is not real.

  • Week 1: count clean and set the baseline.
  • Week 2: check prices, credits, receiving, and substitutions.
  • Week 3: test yields, portions, waste, and menu mix.
  • Week 4: set pars, approved buyers, card coding, and the weekly review.

30 days, in order

  1. Week 1Baseline

    Owner / Controller

    • Run a clean count
    • Close the period against the sales report
    • Calculate actual food cost
    • Set your target from your own P&L
    • Name the weekly owner
  2. Week 2Prices and receiving

    Controller + Receiving manager

    • Price-check the top 20 items against agreements
    • Open the vendor portal and claim outstanding credits
    • Start weighing one protein case per delivery
    • Ban unapproved substitutions
  3. Week 3Usage and menu

    Chef + GM

    • Yield-test the two most expensive proteins
    • Reprice specs off cost per usable pound
    • Put scales and spec cards on the line
    • Split waste logging into prep, spoilage, and plate
    • Sort the menu by margin and mix
  4. Week 4Controls

    Owner + Bookkeeper

    • Set par levels on the items causing emergency runs
    • Name one approved buyer
    • Code every card purchase to a restaurant
    • Audit comps and voids by employee
    • Lock the 30-minute weekly review
Four weeks in sequence. Week 1 is baseline work owned by the owner or controller. Week 2 is prices and receiving, owned by the controller and the receiving manager. Week 3 is usage and menu, owned by the chef and the GM. Week 4 is controls, owned by the owner and the bookkeeper.

When a move works, bake it into next quarter's target. Otherwise you hit the old number and think you're done. If the budget itself needs rebuilding, the restaurant budgeting guide has the line structure and the cadence to hang this on.

Where Tab fits around this

Use inventory, recipe, and procurement systems for counts, specs, yields, and supplier orders. Use Tab to keep off-invoice purchases and invoice errors inside the food-cost picture.

  • After a Tab Card swipe, Tab texts for the receipt, restaurant, and GL code; average submission time is 90 seconds.
  • Coded purchases sync to QuickBooks Online or export to another ledger. The QuickBooks and Xero guides cover the food-cost account setup.
  • Andy checks invoices and vendor agreements for overpricing and missed rebates on the $150-per-location Pro plan.

Tab Card receipt capture

The Tab Commerce card product page, showing the hero and the Receipt Capture and Accounting Automation feature cards, with callouts marking the text prompt after every swipe, receipts back in 90 seconds on average, and purchases coded to a restaurant and a GL code.
Captured from tabcommerce.com/card. Callouts added: text prompt after every swipe, receipt back in 90 seconds on average, coded to a restaurant and a GL code.

Tab keeps missing purchases and invoice errors from distorting the food-cost number you act on.

Questions & Answers

Questions operators ask about food cost

Build it from your P&L: 100% minus labor, occupancy, other operating costs, and required profit. Your trend against your target matters more than another concept's benchmark.

Monthly for accounting, weekly for action. Weekly counts on high-value items catch price and portion problems before the period closes.

Theoretical is what recipes and sales say you should have used; actual is what counts and purchases say you used. The gap points to waste, over-portioning, yield loss, or theft.

Yes. Mix can shift, yields or portions can change, comps can rise, or an invoice can land in the wrong period. Same prices, different number.

Price checks, credits, receiving, yields, waste logs, and comp audits recover money without changing the plate. Shrinking portions should be a later lever, not the first.

start for free

Learn more about Tab

Book a call with a member of our team to learn more, get a demo, and if you'd like, get started with Tab right there.

Platform Demo
Implementation Plan
Onboarding Kickoff
Learn More

Frequently Asked Questions

Who is Tab Commerce?

Tab Commerce is a financial technology company providing the only finance platform built for restaurants. Learn more about us here.

How long does it take to get started?

We're from the restaurant industry and know time is a constrained resource. We've built our onboarding process to get you fully up and running in 10 minutes or less.

How do Tab business credit cards work?

Tab cards work just like any other corporate credit card but with the benefits of powerful spend control software built just for restaurants.

Can Tab Cards be added to Apple Wallet?

Yes, simply add the Tab Card into your Apple Wallet like you would any other card.

Where can I use Tab Cards?

Anywhere that Visa is accepted (most places). Most core vendors accept cards, but some might only be available upon request - we can help you with that.

Can I add my team to Tab?

You can create as many users as you would like with a simple onboarding process and customized roles.

Can I add multiple entities to Tab?

You can create as many entities as you need, and manage them all from a single login.

What are flexible payment terms?

The Tab Card gives restaurants more control and flexibility over their cash flow. Extended payment terms allows restaurants to extend eligible re-payments on their Tab Card by 30 days, providing up to 60 day terms. This is for eligible, pre-approved customers only. Inquire to see eligibility.

Do you offer support?

We provide near-instant support from real humans based in the US and Canada. Support can be accessed via live chat in the platform or by emailing support@tabcommerce.com.

How can I get started?

Click the 'Get Started' button in the top right of this page. You'll receive an email to create your account and book a time to finish onboarding with your dedicated account rep.