Most attempts to reduce food costs start in the wrong place. Someone looks at a food cost percentage that has crept up three points, decides the problem is the product, and starts trading down: cheaper cheese, thinner protein, a house olive oil nobody would choose twice. The percentage improves for a month. Then covers fall, regulars stop mentioning the thing they used to order, and the savings disappear into a softer top line.
The cost is almost never in the quality tier. It is in what you pay for the same quality, how much of it reaches a plate instead of a bin, and how many line items you have never re-priced. Those are all fixable without touching the recipe. And you cannot see any of them until every dish is costed out, which is why a recipe costing template that gets you to a real cost per portion is the first tool here, not the last.
This is the full set of levers, ordered roughly by how much money they free up per hour spent. Eight of them are yours to pull. The ninth is not, because import tariffs and duties add to the landed price of what you buy however well you buy, and that needs handling differently.
Key Takeaways
Wholesale food prices overall were 1.9% below their year-ago level in August 2026, while fats and oils were up 21.2% and eggs were down 59.4% (National Restaurant Association, retrieved 2026-09-29). The average tells you nothing. Only line items do.
Re-bid your ten most expensive items rather than your whole order guide. Ten lines usually carry the majority of the spend, and they are the only ones a supplier will sharpen a pencil for.
Restaurants throw away roughly 4% to 10% of everything they buy. That is money already spent, and recovering it costs nothing but attention.
Spec changes beat quality cuts. A different cut, grade or pack format of the same ingredient can drop cost per portion without the guest noticing anything.
Some of your increase is not negotiable. Separate the drivers you control from tariffs, freight and commodity cycles, and stop trying to fix the second group with harder buying.
Boost customer satisfaction with just a few clicks
Most-Loved Features:
- On-demand drivers
- Real-time GPS tracking
- Delivery confirmation photos
- Over 50% of customers report a smoother delivery experience
Why the average wholesale food price tells you nothing in 2026
Headline food inflation numbers are useless for buying decisions right now, because the average is hiding enormous spread underneath it.
In August 2026, average wholesale food prices sat 1.9% below where they were a year earlier (National Restaurant Association, retrieved 2026-09-29). Read that alone and you would conclude your costs should be falling. Now look at what individual categories did over the same twelve months:
| Category | Year-over-year change, Aug 2026 |
|---|---|
| Fats and oils | +21.2% |
| Unprocessed finfish | +16.3% |
| Prepared seafood | +14.8% |
| Fresh fruit | +7.6% |
| Wheat flour | +5.7% |
| Beef and veal | +3.1% |
| Cheese | −4.2% |
| Refined sugar | −10.0% |
| Fresh vegetables | −10.8% |
| Processed poultry | −12.5% |
| Pork | −13.9% |
| Butter | −38.7% |
| Eggs | −59.4% |
Source: National Restaurant Association, wholesale food price data for August 2026, retrieved 2026-09-29.
A fryer-heavy menu and an egg-heavy menu experienced opposite years. The spread between the top and bottom of that table is over 80 percentage points, inside a market whose average barely moved.
Two things follow. First, any cost-cutting plan built on a national inflation figure is guesswork. Second, the spread itself is an opportunity: when butter is down 38.7% and oils are up 21.2%, the kitchen that notices gets a margin the kitchen that doesn’t never sees. Longer term, the US Department of Agriculture forecasts food-away-from-home prices rising 3.5% across 2026 and 2.6% in 2027, with beef and veal up 9.4% and fresh vegetables up 5.7% (USDA Economic Research Service, Food Price Outlook published 2026-09-25). Protein is where the sustained pressure is.
Lever 1: Cost every recipe before you cut a single ingredient
You cannot reduce a number you have not measured at the dish level. A food cost percentage across the whole business tells you that something is wrong; it never tells you which thing.
Costing every dish out gives you three numbers per item: what the ingredients cost, what you charge, and what is left. Almost always, the picture is lopsided. A handful of dishes are quietly unprofitable, one or two are far better than anyone realised, and the item everyone assumed was the problem is fine. Without that, every cut is aimed blind.
The mechanics are more fiddly than they look, because invoice units and recipe units rarely match and trim loss has to be priced in. Our step-by-step walkthrough of building a recipe costing template covers the columns you need, where to get a working spreadsheet, and how to convert a case price into a defensible cost per portion. Do that first. Everything below gets easier and safer once it exists.
Lever 2: Re-bid your ten most expensive line items, not the whole order guide
Trying to re-price four hundred SKUs achieves nothing, because a supplier will not discount an entire catalogue and you will run out of patience by line forty.
Sort your purchase history by total spend, descending. Take the top ten. On most menus those ten lines — usually proteins, dairy, oil and one or two signature items — account for a large majority of food spend. Get written quotes on exactly those specs from two or three other distributors, then take the quotes to your incumbent.
This works because it is specific and because it is credible. “Can you do better on everything” invites a no. “Here is a quote for the same 40-pound case, same grade, delivered the same day, at $14 less” is a question your rep can actually take to their manager. Re-run it quarterly, not annually. Prices on volatile categories move far faster than your contract cycle.
One caution: hold the spec constant while you bid. If the competing quote is for a lower grade or a different pack size, you are not comparing prices, you are comparing products, and you will end up back at trading down without meaning to.
Lever 3: Consolidate suppliers to cut delivery minimums and reconciliation time
Every additional supplier adds a delivery minimum, a possible fee, an invoice to check and another chance of a short or a substitution nobody catches.
Fewer purchasing relationships mean bigger orders per vendor, which is what unlocks volume pricing and gets you under delivery minimums instead of paying to reach them. The administrative saving is real too: reconciling four invoices a week instead of eleven is hours back, and unchecked invoices are where price creep lives.
The tension is perishability. Consolidating dry goods and frozen into one big weekly drop is almost always right. Consolidating produce is often wrong: buying a week of lettuce to hit a minimum, then throwing a third of it out, costs more than the split delivery did. Consolidate where shelf life is long, and keep frequent small deliveries where it is short. If your own vans are already going out, inbound pickups on the return leg can sometimes replace a delivery charge entirely.
Lever 4: Change the spec, not the quality tier
This is the lever that separates cost reduction from cheapening, and it is the most under-used one on this list.
A quality cut swaps a good ingredient for a worse one. A spec change buys the same ingredient in a form that costs less to land on a plate:
-
Whole-muscle instead of portion-cut, where you have the butchery skill to break it down. You are paying for labour you already have instead of the supplier’s.
-
A different cut of the same animal for braises and ragùs, where long cooking makes the premium cut’s advantage irrelevant.
-
Frozen at peak season for anything that gets cooked anyway: berries into a compote, peas, certain fish. The quality argument for fresh applies to raw applications, not all of them.
-
Bulk pack instead of individual portions on high-volume items, provided you will actually get through it.
-
Seasonal substitution inside a dish rather than across the menu: the same dish built on whatever the crop actually is this month.
The guest test is simple. If a blind taste of the finished dish is unchanged, it is a spec change. If it isn’t, you cut quality and should expect to pay for it in repeat business.
Lever 5: Fix yield loss before you blame purchase price
Purchase price is what you pay per pound. Yield is what percentage of that pound reaches a guest. The second number moves cost per portion just as hard as the first, and nobody watches it.
A case that looks 8% cheaper but yields 15% less trim-free product is more expensive, and it will look like a win on every report you have. Trim, peel, bone, shrink during cooking and portion overrun all sit in this gap. So does the produce that arrives two days from turning.
Practical version: weigh in, weigh out, on your five biggest items, once. Record the yield percentage. Put it in the costing sheet permanently, because cost per portion calculated on raw purchase weight is fiction. Then compare suppliers on yielded cost rather than case price. It reorders the rankings more often than not.
Lever 6: Tighten portion control where the guest can’t perceive it
Over-portioning is invisible on any single plate and enormous in aggregate. Two extra ounces of protein on a dish that sells 60 covers a week is a large number by the end of a quarter, and no guest has ever noticed the difference.
The tools are unglamorous and they work: scoops and ladles sized to the recipe rather than whatever is in the drawer, portion scales at the plating station for proteins, pre-weighed portions during prep for the highest-cost items, and photographs of the correct plate where the line can see them.
Where you must not do this is anywhere the portion is the value proposition. If people come for the size of the sandwich, the sandwich is not the place to find 40 cents. Take it from the garnish, the sauce yield and the protein trim instead, and be honest about which is which, because shrinking the thing people came for is a quality cut wearing a cost-control label.
Lever 7: Order to a par level instead of a hunch
Restaurants discard somewhere between 4% and 10% of all the food they buy, mostly through spoilage and overproduction. That is inventory you already paid for, and most of it was over-ordered before it was ever wasted.
Par levels fix the ordering half. For each item, set the quantity you want on hand at the start of a week based on actual usage, count what you have, and order the difference. It sounds obvious and it replaces the real process in most kitchens, which is a walk through the cooler and a guess.
Two habits do the rest:
-
Rotate strictly first in, first out. New deliveries go behind older stock, always. Most spoilage in a well-stocked kitchen is not a demand problem, it is a stacking problem.
-
Review cost of goods against sales weekly rather than monthly. Monthly review tells you what happened; weekly review lets you adjust the next order while it still matters.
Lever 8: Shift volume toward dishes that share expensive ingredients
Cross-utilisation is a purchasing strategy disguised as menu design. Every ingredient used in exactly one dish carries its own spoilage risk, its own minimum order and its own dead inventory when that dish sells badly.
Once the costing exercise from Lever 1 exists, you can see which dishes carry the best margin and which ingredients appear in only one place. Then you move volume: better position on the menu, server focus, and pricing for the strong performers, and a decision about the singletons: either find a second use for that ingredient or take the dish off.
The margin here comes from three places at once. Fewer distinct SKUs means larger orders on each and better pricing. Ingredients used in three dishes turn over before they spoil. And steering demand toward dishes you already make money on lifts the blended margin without repricing anything. Note that pricing is a separate discipline with its own logic. This lever is about which dishes you sell more of, not what you charge for them.
Lever 9: Separate the cost drivers you can’t negotiate
Some of your increase will not respond to any of the eight levers above, and continuing to attack it with harder buying wastes the effort.
Tariffs and import duties are the clearest example. They attach to the goods at the border, they apply to your competitors’ suppliers as well as yours, and no amount of re-bidding removes them, because a distributor cannot discount a tax. The same is true of freight surcharges and commodity cycles like the beef market, which the USDA forecasts up 9.4% across 2026.
The response is different in kind: change origin, change the ingredient, absorb it deliberately, or reprice. Which of those applies depends on what share of your basket is affected and how substitutable it is. We cover that decision in detail in how import tariffs land on your food costs and what actually absorbs them, including who legally pays the duty and the cash-flow gap between paying it and selling the goods.
Worth doing before anything else in this section: work out what percentage of your food spend is actually imported. Operators routinely over-estimate it, and the answer determines whether tariffs are your main problem or a rounding error you have been blaming for a purchasing issue.
How to tell whether any of this worked
Track the gap between theoretical and actual food cost. Theoretical is what your dishes should have cost given what you sold, which falls out of the costing sheet automatically once it exists. Actual is what you really spent. The difference is waste, theft, over-portioning and unrecorded comps.
Watch the gap, not the headline percentage. A food cost percentage can move because ingredient prices moved, because your sales mix shifted, or because you raised prices, none of which tells you whether your operation got tighter. The variance between theoretical and actual strips all of that out. Closing it from six points to two is a real operational win, and it happens without a single change to what the guest receives.
Give the whole programme a quarter. Levers 1, 2 and 5 pay back in weeks. Levers 3, 7 and 8 need a full ordering cycle before the numbers mean anything.
Frequently asked questions
What is a good food cost percentage?
It depends heavily on format. Quick service typically runs lower than full service, and fine dining runs highest because the ingredient is more of the product. Comparing yourself to a single industry-wide number is misleading. Compare yourself to your own trend, and to the gap between your theoretical and actual cost, which is the part you control.
Can you reduce food costs without changing suppliers?
Yes, and usually faster. Yield measurement, portion control, par-level ordering and first-in-first-out rotation all sit entirely inside your four walls and need no negotiation. In most kitchens those recover more than a supplier switch does, because they address food you have already bought.
Does buying in bulk always reduce food costs?
Only when you use it before it degrades. Bulk pricing on shelf-stable goods is close to free money. Bulk on perishables frequently costs more than it saves once spoilage is counted, and it also ties up cash in inventory that isn’t earning anything.
How often should food costs be reviewed?
Review cost of goods against sales weekly, re-cost your dishes whenever a major ingredient price moves, and re-bid your top ten line items quarterly. Annual reviews are far too slow for a market where individual categories move 20% or more in a year.
Where to start
Do these three in order and stop there for the first month. Cost your dishes so you can see where the money actually goes. Weigh in and weigh out on your five biggest items so your costs reflect yield rather than invoice weight. Then take written quotes on your ten largest lines to your incumbent supplier.
That sequence gets you most of the available money without touching a single recipe. The remaining levers are about holding the gains: par levels so you stop buying food you throw away, portion discipline so the savings reach the plate instead of the bin, and clear eyes about which part of your increase was never yours to negotiate in the first place.
