Food cost creep: catching supplier price increases before they eat your margin

Supplier prices almost never jump — they creep, a few cents per case, invoice after invoice, until your food cost is two points higher and nobody can say when it happened. Here is how to catch price creep at the invoice line, what to do about substitutions, and how to have the supplier conversation with real numbers in hand.

September 3, 2026

Food cost creep is the quietest way a restaurant loses money. Nobody announces it. The chicken that was fine last quarter is up forty cents a case. The fryer oil came in as a different brand at the same price — a grade lower. A “fuel surcharge” line appeared on the produce invoice in March and never left. Each line is too small to fight over on a busy morning, so the invoice gets signed, the truck leaves, and three months later your food cost percentage is two points higher and nobody can point to the day it happened. That is the defining feature of creep: it is designed — or at least allowed — to stay under the threshold where anyone pushes back. This guide is about moving that threshold: catching price changes at the invoice line where they actually happen, separating honest market movement from drift you can push back on, and deciding, with numbers, when the answer is a supplier conversation and when it is a menu reprice.

Why food cost creeps between invoices

Monthly food cost percentage is a fine scoreboard and a terrible alarm system. By the time creep shows up there, it is blended into everything else that moves the number — sales mix, waste, portioning, theft, a big catering week — so the month-end review turns into an argument about causes instead of a fix. The creep itself happens at a much finer grain, and it wears a few standard disguises:

Line-item drift. A case price rises a few percent with no notice — not once, but in small steps across many deliveries. Any single step looks like rounding; the trend only exists if someone is keeping a dated price history per item.
Pack-size changes. The case that was 6×2 kg quietly becomes 6×1.8 kg at the same price. The invoice total looks identical, and the unit price just rose ten percent. This is the hardest form to catch by eyeballing invoices, because the number that changed is not printed on the page.
Substitutions. The contracted brand is out of stock, so a different one ships at the same price — sometimes a fair swap, sometimes a lower grade. A substitution is not a price change on your item; it is a different item wearing its price tag, and it needs to be recorded that way or your price history turns to mush.
New fee lines. Fuel surcharges, delivery minimums, “small order” fees, case-split charges. None of them touch a unit price, so none of them trip a price alarm — they just land on the invoice total, which is exactly why they are popular.
Contract expiry. The negotiated price you agreed on eighteen months ago lapsed, pricing reverted to list, and nobody on either side mentioned it. Not malice — just what happens when the contract date lives in a drawer.

What a few points of drift actually costs

The arithmetic is worth doing once, because it turns “a few cents a case” into a number that justifies the tracking habit. Take a restaurant doing $20,000 a week in food sales at a 30% food cost. If drift moves that to 32% — well within a drift that goes unnoticed — the difference is $400 a week, roughly $20,000 a year, straight out of profit margin that is thin to begin with. Scale it to your own volume:

Weekly food sales	One point of food cost	Two points, annualized
$10,000	$100 / week	about $10,000 / year
$20,000	$200 / week	about $21,000 / year
$40,000	$400 / week	about $42,000 / year
$80,000 (multi-unit)	$800 / week	about $83,000 / year

These are not projections — they are multiplication. The only assumption is that the drift goes uncaught, which is precisely what happens without a price history. Note also what this table does not say: that every increase is illegitimate. Commodity markets genuinely move, and a supplier passing through a real market increase is business, not misconduct. The point of tracking is that you stop taking anyone’s word for which kind of increase you are looking at.

Track prices at the invoice line, not the month

The fix is a habit, not a project: every invoice that comes off the truck gets its lines recorded into one dated table before it goes to the bookkeeper. Supplier, item, pack size, case price, computed unit price. That is the whole discipline. Done at receiving — the same five minutes as the count and temperature checks — it produces the one thing creep cannot survive: a per-item price history with dates on it. The workflow that works in practice:

Record at receiving, not at bookkeeping. The person checking the truck is already holding the invoice. The bookkeeper sees totals; the receiver sees items — and items are where creep lives.
Always compute unit price — price per kg, per litre, per dozen — never rely on case price. Unit price is the only number that survives a pack-size change.
Compare each line to the last price paid for that exact item from that supplier. A change beyond a threshold you choose — many operators use two or three percent — gets flagged for a look, not a fight. Most flags will have boring explanations. The ones that do not are the money.
Log substitutions as their own item, marked as a sub, rather than as a price change on the original. When the original comes back, your history is intact.
Once a week, scan the flags and the fee lines together. Ten minutes. This is also when you notice the surcharge that appeared in March and never left.
Tie the winners back into your recipes. A tracked price is only half the loop — if the new chicken price never reaches your recipe costs, your menu margins are fiction. Our guide to restaurant inventory management covers that connection in detail.

The supplier conversation: numbers, not vibes

A price history changes the character of the supplier call. Without one, the conversation is “prices feel high lately,” which every sales rep on earth knows how to absorb. With one, it is specific: this item, these five dates, this trend — and on the same call, the three items where you know a competitor’s price. A few things make the conversation productive rather than adversarial. Bring your volume: your annual spend with that supplier is your leverage, and reps genuinely do fight internally for accounts that track. Ask for the contract dates in writing, and put the renewal date somewhere that has an alarm on it. Ask which increases are market-wide pass-throughs and which are list-price adjustments — the honest answer sorts itself into what you accept and what you negotiate. And keep a second supplier warm for your top categories, not as a threat to wave around, but because a real alternative is the only leverage that does not expire. None of this requires being difficult; it requires being the account that notices. Suppliers price to the attention they expect, and the operator with a dated price history gets a different class of pricing than the one who signs whatever lands on the dock.

When the answer is a menu reprice, not a fight

Some increases are real and are not going back down. When your history says a core ingredient has genuinely repriced — market-wide, confirmed across suppliers — the move shifts from procurement to the menu. That does not automatically mean raising prices across the board. Recost the recipes that use the ingredient and look at where the damage concentrates: often two or three dishes absorb most of the hit. Those are candidates for a targeted price adjustment, a portion or spec change, or a quiet rebuild around a different cut — the standard toolkit from menu engineering. What makes any of this possible is, again, the same price history: repricing a menu off month-old costs is guessing, and guessing on thin margins is how a busy restaurant loses money on its best seller.

What your systems should be doing for you

Everything in this guide runs fine on a spreadsheet and a disciplined receiver — that is exactly what the free tracker is for, and plenty of strong operators never need more. But it is fair to name what the tooling should contribute once you outgrow the sheet, because hand-keying invoice lines is only necessary when your purchasing, recipes, and menu live in systems that do not talk. On Novaryq, vendor orders, receiving, ingredient-level recipe costs and the menu sit on one platform with one plan — so a received price flows into recipe costs without re-keying, and an ingredient that repriced shows up in the margin of every dish that uses it. Run more than one location, and the same view works across all of them from one console — including the useful embarrassment of seeing two of your own stores paying different prices for the same case from the same supplier, which is a multi-location conversation worth having with your rep. And because Novaryq offers a month-to-month option alongside the discounted annual term, the reporting has to keep earning its place every month — the same standard this article asks you to hold your suppliers to.

Frequently asked questions

What is food cost creep?

The gradual rise of ingredient costs through small, unannounced supplier price increases, pack-size reductions, substitutions and added fee lines — each too small to trigger a reaction, but compounding into a measurably higher food cost percentage over months. It is distinct from a market-wide price spike, which is visible and discussed; creep is defined by going unnoticed.

How do I track supplier price increases?

Record every invoice line into one dated table at receiving: supplier, item, pack size, case price, and computed unit price. Compare each line to the last price paid for that exact item and flag changes beyond a threshold you choose. Unit price matters more than case price because it is the only number that survives a pack-size change. A spreadsheet handles this fine at one location; at several, it belongs in your purchasing system.

How should I handle a supplier substitution on an invoice?

Record it as its own item marked as a substitution, not as a price change on the product you ordered. A different brand or grade at the same price is a different item wearing the original’s price tag — logging it against the original corrupts your price history in both directions. Substitutions that keep recurring are worth their own conversation with the rep, especially if the replacement is a lower grade.

When should I raise menu prices instead of pushing back on the supplier?

When your own price history — ideally confirmed across two suppliers — shows a genuine, market-wide repricing of a core ingredient rather than one vendor drifting. At that point recost the affected recipes, find the dishes absorbing most of the hit, and adjust those deliberately: a targeted price change, a portion or spec adjustment, or a rebuild. Across-the-board increases in response to one ingredient are usually overcorrection.