Service Mode Run better

How to Reduce Restaurant Food Costs: Find the Leaks First

Use matched records to separate price, yield, portion, waste, sales mix and unexplained use before you change anything.

Two restaurant team members counting dry-store containers with a clipboard and weighing scale.
Conceptual illustration for the Service Mode series.

When food cost climbs, the usual reflex is to call suppliers for a better rate, trim portions or switch to a cheaper brand. Any of those might help, but each is aimed at one cause, and a rising percentage can come from several: higher purchase prices, poorer yields, generous portions, waste, a shift in what guests order, or stock that leaves the kitchen without being recorded. Until you know which, a cut is a guess, and kitchen guesses tend to surface later as complaints, rework or a quieter dining room.

This guide sets out a diagnostic routine for a restaurant that already counts stock and costs its recipes. It does not repeat how to run a stock count or build a recipe card. Instead it shows how to use those records together for one matched period, separate what you can explain from what you cannot, and choose one low-risk change you can measure. Every figure below is an illustrative example for working through the method, not a market benchmark or a target.

Start with one matched period and one boundary

Every number in this review must describe the same days, the same items and the same way of valuing stock. If the closing count was taken on the morning of the 31st but sales run to midnight, or purchases are entered by invoice date while deliveries are booked by receipt date, the gap you find may be nothing more than timing.

Decide the boundary before you collect figures. This example reviews food ingredients and excludes beverages, packaging and cleaning supplies. Apply that boundary to opening stock, purchases, transfers, closing stock and sales. Agree the valuation policy with your accounts team and value expected use on a comparable basis. If the basis changes, explain the effect separately before calling it a usage loss.

  • Period start and end dates, and the time each count was taken
  • Which categories are inside the boundary and which are excluded
  • The valuation basis used for opening stock, closing stock and expected usage
  • Who supplied each figure, so questions go to the right person

Build the inventory value bridge

The bridge converts stock movements into ingredient consumption: what you started with, plus what came in, minus what left to other locations, minus what remained. In the illustrative example below, the restaurant consumed INR 100,000 of ingredients in the period.

Food sales for the same period, excluding taxes collected for government, were INR 300,000. Ingredient consumption divided by food sales is therefore 33.33 per cent. This diagnostic ratio includes waste, staff meals and complimentary food; it is not a claim about their accounting classification. The next step separates those uses so managers can explain the total. A different numerator requires a different label, and a zero-sales period has no meaningful percentage.

Line (illustrative)Value (INR)
Opening ingredient inventory80,000
Add net purchases, after returns and credit notes90,000
Add transfers in5,000
Less transfers out10,000
Less closing ingredient inventory65,000
Ingredient consumption100,000

Compare actual use with what should have been used

Next, work out what consumption should have been. Multiply the quantity of each dish sold by its recipe ingredient value on the same valuation basis. In the example that comes to INR 80,000. Add waste recorded separately during the period, INR 6,000, and staff meals and complimentary items recorded separately, INR 4,000. Explained use totals INR 90,000, leaving INR 10,000 unexplained.

Be strict about where each allowance lives. If a recipe card already includes a trim allowance and the team also logs that trim as waste, the same loss is counted twice and the unexplained figure shrinks artificially. Agree that each type of loss appears in one place only: inside the recipe quantity, or in the waste log, never both.

Expected usage bridge (illustrative)Value (INR)
Expected ingredients for items sold, from recipe cards80,000
Separately recorded waste6,000
Separately recorded staff meals and comps4,000
Explained use90,000
Actual ingredient consumption from the value bridge100,000
Unexplained use10,000

Fix the records before you suspect people

An unexplained figure is a list of questions, not evidence of theft. Rule out recording problems first because they can distort the entire review. Preserve the original record and the reason for each correction so another reviewer can follow the change.

Only once the records are clean does it make sense to look at receiving, portioning, storage and access. Even then, treat the finding as a process gap to close, and discuss it with the team rather than announcing a suspicion.

Include prepared and in-process food on a consistent ingredient-value basis, and remove the ingredients issued into it from raw stock. Otherwise a production batch can look like consumption before the food has been sold. If the records cannot support that reconciliation, narrow the review to a traceable category and record the limit.

  • Unit errors: a case entered as a single piece, grams keyed as kilograms
  • Date errors: a delivery received on the last day but booked to the next period
  • Valuation differences: a changed costing basis, price changes or an old recipe rate compared with current inventory values
  • Production omissions: prepared sauces, stocks, marinades or work in progress left out of the boundary, or counted again as raw ingredients
  • Missing transfers to a sister outlet, the bar or an event
  • Recipe cards that no longer match how the dish is actually made

Check yield against the price you think you pay

Purchase price is not the cost of what reaches the plate. Suppose a kitchen buys 10 kg of a vegetable for INR 600, or INR 60 per kg. After peeling and trimming, 8 kg is usable, so the usable cost is INR 75 per kg. A dish that uses 150 g of the prepared vegetable carries INR 11.25 of that ingredient, not the INR 9 a recipe card would show if it were built on purchased cost. This is a single-ingredient example, not the full cost of the dish.

Do not conclude from one test. Weigh several deliveries before and after preparation and note who prepared them. A poor yield can come from the grade supplied, preparation technique or storage losses, and each needs a different fix: a specification conversation with the supplier, a demonstration at the prep bench, or a change to rotation and storage.

Separate the drivers before choosing a fix

Sales mix deserves particular care. Food cost percentage can rise because more guests are choosing higher-cost dishes, even when every recipe is controlled. That is not necessarily a problem if each cover still contributes enough. Whether to change the menu is a separate decision, best handled through a menu engineering review rather than through cost cutting.

DriverWhat to compareFirst question to ask
PriceInvoice price against the price used in recipe cardsDid prices move, or did the card fall behind?
YieldWeight before and after preparationIs the usable proportion lower than assumed?
PortionPlated weight against the portion specificationIs the specification clear, and is the right tool at the station?
WasteWaste log by item, reason and shiftIs waste concentrated in a few items or a particular shift?
Sales mixShare of sales by dish across two periodsHas the mix moved towards higher-cost dishes?
Unexplained useValue bridge against expected usage bridgeHave unit, date, valuation and production errors been ruled out?

Make one low-risk change and measure it

Choose one cause, one change and one owner. Record the baseline from this review, describe the change precisely, and set a date for the next matched-period check using the same boundary and valuation basis. Examples include re-weighing plated portions of a high-volume dish against the existing specification, adding a weight check at receiving for one heavily used vegetable, or moving the waste log to the pass so it is filled in during service.

Two limits apply to every change. Never reduce a portion silently: if a portion specification changes, update the menu description, brief the team and treat it as a menu decision. Never relax food safety to save money, whether by stretching holding practices or using stock that should be discarded. Receiving checks and storage records that your food safety system already requires can double as evidence for this review.

Put it into practice

Your action checklist

Your progress saves in this browser. No sign-up needed.

0 of 8 complete

From reading to doing

Food cost diagnosis

Open the Food Cost Review tab. Enter the outlet, period, cutoff and valuation basis; complete the inventory value and expected-use bridges; then record the evidence, owner, change, baseline, check date and later result in the action log. Keep waste and staff meals separate only when they are not already included in expected use.

Open the Food Cost Review tab in the launch workbook. Keep a copy for your own business.

Download the free workbook See what’s inside the toolkit →

Take the next step with Keystonne

Explore inventory and purchasing

Explore the inventory and purchasing information, then assess whether the available records fit your own count, recipe and review process.

Explore inventory and purchasing

Sources & further reading

  • BCcampus: Monthly Food Costs ↗

    Primary culinary training text explaining inventory-based food consumption and food-cost ratios, read 7 November 2026. The article uses its own INR example and explicitly includes waste and non-sale meals in the diagnostic consumption ratio.

  • FSSAI Guidance Document: Food Safety Management System for the Catering Sector ↗

    Voluntary FSSAI guidance for catering businesses covering practices such as receiving checks, storage records, approved suppliers and staff training. Applicable regulations prevail over guidance. It is cited here only because the receiving and storage records it describes can also support a cost review; it does not set or endorse food cost targets. The cost arithmetic and diagnostic framework in this article are an original illustrative method.

  • BCcampus: Controlling Food Costs ↗

    Primary culinary training reference for standard yields and portions, read 7 November 2026. The vegetable quantities and prices here are original illustrations, not benchmarks.

Published by Keystonne Grow. Product links connect to the Keystonne ecosystem; check the linked product or service for its current scope.