In this article
Your meat is 15% more expensive than it was last quarter. Your menu still reads exactly the same. Somewhere between those two facts sits a decision most kitchens already make without ever running the numbers on it: how many grams did that plate quietly lose this week, and how far can that go before someone notices?
There are only two levers for absorbing a cost rise without losing money: the price on the menu, or the portion on the plate. The first is a decision your guest literally sees printed, every single visit, and can compare against last month. The second is a decision that is printed nowhere — until someone weighs the plate, or more likely, until it just becomes obvious.
Supermarkets have been fighting this for three years under a name that is now a word in its own right: shrinkflation. Same chocolate bar, same box of washing powder, a few grams or millilitres less, same price on the shelf. The instinct to slap that same label on a kitchen is understandable, but a restaurant has no wrapper printed with a weight on it — it has a plate, a guest sitting across from you, and a line that sits in a different place for every dish.
This article doesn't treat that as an ethics debate, because that debate has already been settled and the conclusion doesn't move: treating a guest fairly isn't negotiable. It treats it as arithmetic most kitchens never actually run, even though they're already making the decision — one scoop at a time, often without realising it. How far can a portion shrink before someone notices? How much of a cost rise does that actually cover? And what still has to go on the menu for the part it doesn't?
Further down, you can drop your own five dishes into a calculator — cost, menu price, portions sold per week, and the cost rise you're trying to absorb — and see exactly how far you can safely cut each one, and what's left over for the price list.
Why a portion is the quieter lever than a price
A price increase is the most visible decision a restaurant can make. It's printed, a regular compares it against the menu they remember without even trying, and it sometimes shows up in a review word for word. A portion is printed nowhere. There's no menu line that says "220g" next to the price, and that's exactly why it's the lever a kitchen reaches for first the moment an ingredient gets more expensive — not out of dishonesty, but because it's the one lever that doesn't seem to trigger any resistance.
That instinct is partly right. A price can never quietly change on you — the next invoice, the next receipt, every guest sees it. A portion can quietly change, up to a point. That point isn't a feeling or a guess: it's a measurable line, and mapping that line, dish format by dish format, is exactly what this article does.
The two levers aren't equivalent in what they fix, either. A price increase restores your margin in full, in one move, visibly, for everyone. A portion cut restores it only up to that dish's own safe line, invisibly up to that line, and almost always leaves a remainder that only the price can still cover. This article runs both numbers side by side, so the choice doesn't have to rest on gut feel.
The ultimate guide Menu Engineering & Drinks: 6 Steps to More Margin From menu layout to food cost: everything your menu can earn you, in one guide. Open the guideThe 7 numbers behind a smaller portion
In the order you'll actually need them: first the one formula that applies to every dish, then the line that differs per dish, then what happens once you put the two side by side.
1. A 13% smaller plate covers a 15% cost rise — with zero price change
Here's the one formula worth remembering from this whole article. If your ingredient cost rises by i percent, a portion cut of i / (100 + i) × 100 percent fully offsets it — purely arithmetically, regardless of your selling price or current food-cost percentage. At a 15% rise, that's 15/115 × 100 ≈ 13.04%. It feels backwards that a smaller cut than the cost rise itself would be enough, but the math holds: you're cutting from the new, already-inflated cost, not the old one.
The elegant part of this formula is that it knows nothing about your dish, your price or your margin. It only depends on the size of the shock. An ingredient that gets 8% more expensive needs a portion cut of roughly 7.4%. A 25% shock needs a 20% cut. That's the first number the rest of this article works with, and it's the same sum for a bowl of fries as for a steak.
What the formula doesn't tell you is whether that cut is actually possible — whether a guest would feel it. That's a completely different question with a completely different answer per dish, and that's exactly what the next number is about.
2. The line a guest actually feels
Portion-perception research traces back decades to Weber's Law: people don't notice an absolute change, they notice a change relative to what's already there — and that threshold sits higher than most kitchens assume. A weighed, portioned piece of protein on a plate with few other elements sits at the low end of that threshold: a guest who knows the dish spots a noticeably smaller cut faster than you'd hope.
Underneath that is a second reason kitchens routinely misjudge this: the Delboeuf illusion, named after the Belgian mathematician who first described it. The same quantity looks smaller in a bigger bowl than in a smaller one, and people are structurally worse at judging volume in a bowl or glass than a flat, clearly bordered piece on an open plate. That's not a lack of attention — it's how human perception works, and it's why the exact same portion cut can go completely unnoticed in one format and stick out in another.
Below is what that means for one signature dish in concrete terms: what happens if you do nothing, what the safe ceiling allows, and what it actually takes to fully offset the cost rise — with the part that crosses that ceiling marked separately.
A portioned protein main, 150 covers a week, after a 15% ingredient-cost shock.
The red, hatched section on the third glass is the part of the cut that crosses past the safe ceiling — exactly the part a guest would likely notice on this dish. The amount beside it is what still has to go on the menu, per year, because the portion alone can't cover it any more.
3. Why a bowl hides more than a plate
Not every dish has the same safe ceiling, and the difference isn't about price or the recipe — it's about format. A dish sold by a stated count — six shrimp, eight dumplings — has almost no ceiling at all: a guest can count, and counting is the most exact form of perception there is. A portioned piece of meat or fish sits a little higher, but still low — it's a recognisable, fixed shape that a trained eye catches a difference in quickly.
A composed plate — a main with several elements arranged next to each other — already hides more, because attention spreads across multiple components rather than locking onto one piece. And right at the top sits a bowl or glass: soup, pasta in a deep bowl, a curry. There, the Delboeuf illusion and the general difficulty of judging liquid volume both work in your favour, so the safe ceiling sits well above what any other format allows.
Below are five common formats side by side, each showing how much cost inflation it can absorb through portion size alone before crossing its own ceiling. The vertical line is the cost shock used throughout this article — anything that stays to the left of it can dodge the price list entirely.
Five common formats, each showing how much cost inflation it can absorb through portion size alone. The line is this article's shared shock — whatever falls short of it can't dodge the price list.
The two bars that fall short of the line are exactly the two formats a kitchen reaches for first when it needs to cut — a counted item and a portioned protein — and, as it happens, the two formats that can hide the least. The bowl and the composed plate clear the shock comfortably, with nothing a guest would feel.
4. The trust you're giving up without noticing
Retail's own consumer research is worth a restaurant's attention here. Repeated surveys since 2023 land on the same picture again and again: well over seven in ten shoppers say they've watched at least one product shrink over the past year while the price stayed put, and that share keeps climbing rather than falling. Among younger consumers, roughly four in ten say it explicitly costs the brand their trust once they clock it — and a majority of people who spot it say they've since gone out of their way to avoid the product.
That matters more for a restaurant than it might first appear, because a guest doesn't carry last year's menu around to compare — but they do carry a memory of "this used to be more", and that memory gets spoken out loud in exactly the channel a restaurant is most exposed to: a review. "Small portions for the price" is one of the single most common complaints in restaurant reviews, and it almost never shows up after a clearly announced price rise — it shows up after a change nobody explained.
That isn't because a guest rates a portion change as worse than a price rise. It's because an unannounced change reads as something hidden, while an announced one — however unwelcome — reads as honesty. The gap between the two is rarely the size of the change. It's whether anyone explained it.
5. The law that already exists in France — and doesn't yet in the EU
Since 1 July 2024, French supermarkets larger than 400 square metres have had to flag it: the moment a pre-packaged food item shrinks in quantity without a proportional price cut, that has to be clearly signposted at the shelf for two months, with the real price-per-kilo or per-litre increase stated alongside it. A breach can cost an individual up to €3,000 and a company up to €15,000. It's currently the only national law in the EU that names shrinkflation and attaches a fine to it.
No such requirement exists EU-wide. The European Parliament formally asked the European Commission for EU-wide rules on it in 2024, and the Commission's own position is that shrinkflation is assessed case by case under the general rules against misleading commercial practices — not under a dedicated law with its own name. That's the exact pattern this site keeps describing for rules that are on their way: nothing that legally obliges you today, but a direction that's unmistakable, and a public that's noticing faster than the legislation is moving.
A mandatory menu disclosure like the French supermarket rule doesn't exist for a restaurant anywhere today — a dish isn't a pre-packaged product with a weight printed on a wrapper. But the point behind the French law holds just as well in a kitchen: a change you don't explain eventually becomes whatever explanation a guest invents for it, and that explanation is rarely kinder than the truth.
6. Four levers, side by side, in euros
Shrinking a portion isn't the only quiet lever, and it's worth lining up all four before you pick one. A price increase restores margin in full and in one move, but it's the most visible of the four to a guest — it's printed, and it gets compared. A portion cut restores margin up to this article's ceiling, invisibly up to that ceiling, and almost always leaves a remainder that only the price can cover.
A recipe change — a cheaper origin for the same ingredient, a slightly smaller garnish, a side that moves from house-made to bought-in — sits somewhere in between: invisible to a guest who doesn't know the dish, noticeable to a regular who knows exactly how it used to taste. And trimming the extras — a bread basket that gets a little smaller, an amuse-bouche that quietly disappears — is usually the cheapest move in euro terms, but also the fastest to be noticed, because an extra is by definition the part a guest didn't expect and therefore remembers.
None of the four is automatically the right call. The calculator below combines two of them — portion and price — because those are the two levers that apply to every dish regardless of the recipe. The other two stay a per-dish judgment call no formula can make for you.
7. The blended plan: cut to the line, put the rest on the price
The strongest approach is rarely one lever on its own — it's a blend, and the sum behind it is surprisingly simple. Cut the portion to exactly that dish's safe ceiling, never further. Whatever that ceiling doesn't cover goes on the menu, as a price increase smaller than what you'd have needed without the cut.
For a dish whose safe ceiling is larger than the required cut — the bowl or the composed plate from the earlier number — the answer is simple: the portion alone is enough, the price doesn't need to move at all. For a dish whose ceiling is smaller — a portioned piece of protein, anything sold by count — the remaining price increase is usually far smaller than a kitchen would instinctively ask for, precisely because part of the shock has already been absorbed on the plate.
Drop your own five dishes into the calculator below. Enter the real cost, the menu price and the weekly covers, and it works out the safe cut per dish — and, where that isn't enough, what still has to go on the price.
Drop your own five dishes in
Enter what each dish costs, what it sells for, and how many portions you sell a week. The five rows start out with the numbers of a fifty-seat bistro — overwrite them with yours, and adjust the cost rise at the top to whatever you're actually seeing come through.
Every dish gets its own safe ceiling based on its format. Below, three numbers update live: how much of the total shock you can absorb through portion alone, what still has to go on the price, and which dish carries the biggest share of that.
Shrinkflation calculator
Five formats, one shared cost shock, and per dish what the portion alone covers and what the price still has to do.
One shared shock across all five dishes, so you can see straight away how the same rise plays out completely differently by format.
—
The safe ceiling per format is a guideline drawn from perception research, not an exact rule — a regular who knows a dish well notices faster than a first-time guest. Everything runs in your browser; nothing is sent or stored. This model never cuts past the safe ceiling: whatever that ceiling doesn't cover is always shown as a price increase, never as a deeper, unnoticed cut.
Two things worth knowing as you read this. The safe ceiling is a guideline, not a guarantee — it's based on how people judge portion size on average, not on any one guest. For a dish your regulars know inside out, extra caution is never wasted.
And this model never cuts past that ceiling, even where that would mathematically cover the whole shock. The moment a cut would cross it, the screen always shows that as a remaining price increase instead — never as a recommendation to go deeper anyway. What a guest would notice shouldn't stay quiet; it belongs on the menu.
What to do with this this week, this month and this quarter
Adjusting all five dishes at once is the fastest way to a menu nobody recognises any more. This order works better.
This week — measure the shock, not the fix
- For your three highest-volume dishes, find out how much their core ingredient has actually risen over the last two months — invoices, not gut feel.
- Work out each dish's format: sold by count, portioned, composed, bowl, or sliced — that decides the safe ceiling next.
- Drop the three into the calculator above with their real numbers, and read off how much portion alone can cover.
- Don't act on it yet — know the numbers first, then change the plate or the menu.
This month — cut to the line, never past it
- For dishes where portion alone covers the whole shock, adjust the plate only — leave the price alone.
- For dishes where portion isn't enough: cut exactly to the safe ceiling, and raise the price only for the remaining part.
- Never cut a dish sold by count — its safe ceiling is close to zero, and price is the only honest lever there.
- Change at most one thing per dish at a time — a dish that gets smaller and cheaper-sourced in the same month stacks two invisible changes on top of each other.
This quarter — explain what you did
- Feed your real, updated costs back into your recipe costing tool, so your next menu prices against reality instead of a guess.
- Where a price does go up, just name the reason on the menu or at the table — an explained price rise draws noticeably fewer complaints than a silent one.
- Re-run this check every time an ingredient price moves again — the safe ceiling per format stays the same, but the shock changes every quarter.
- If you notice you've trimmed the same dish's portion more than once already, that's the signal it's time for a price round instead of another cut.
The line isn't an opinion, it's a calculation
Shrinkflation isn't a moral category your restaurant either falls into or doesn't — it's a scale, and every kitchen sits somewhere on it the moment an ingredient gets more expensive and the menu doesn't move. The question isn't whether your portions will ever get a fraction smaller when buying prices turn against you. It's whether that happens inside a line you actually know, or by accident, one scoop at a time, until someone notices.
The good news is that the arithmetic is small. One formula for how much a portion needs to shrink to cover a cost rise, one ceiling per format for how much of that a guest genuinely won't feel, and a simple sum for what the remainder — if there is one — needs to cost on the menu. None of the three asks for anything you don't already have sitting in your invoices and your till.
And once you know that line, it belongs right next to your pricing strategy and your recipe costing tool — two instruments that both assume a portion matching what the menu promises, rather than one quietly drifting a different direction from the price beside it.