There Is No Ideal Food Cost. There Is Only Your Format

There Is No Ideal Food Cost. There Is Only Your Format

In brief

• The 28-35% band you see quoted everywhere has no primary publication behind it. It circulates as professional habit, not as a verifiable measurement.
• The one figure that can actually be attributed to a source — 28-32%, from FIPE's Rapporto Ristorazione — describes Italian table-service restaurants. That is one format in one market, not a global ceiling.
• A pizzeria reading its menu-wide average is looking away from the only decision that matters, which sits in the toppings and never in the dough.
• At a bar, a low food cost is a consequence of the sales mix, not evidence of good management.
• The honest comparison is by menu category, and against your own history — not against an average you found online.

Every restaurant owner has been told this at least once, usually by a consultant or a video: keep food cost under 30%. It is a comforting sentence, because it turns a messy subject into a line you either clear or you don't. The trouble is that the line describes one kind of operation, and the people running a pizzeria or a neighborhood bar use it anyway — reaching bad conclusions about a business that may be doing perfectly well.

Why "keep food cost under 30%" is the wrong question

A food cost of 28-35% is a snapshot of one format, the full-service restaurant, not a ceiling for anyone who sells food. It is worth being precise about where the number comes from, too. That wide band has been passed around trade writing for years with no primary publication holding it up, while the one reference you can genuinely cite — the 28-32% in FIPE's Rapporto Ristorazione — is calculated on Italian table-service restaurants and should be read as Italian market data, not as a worldwide benchmark.

A full-service restaurant has a recognizable cost structure: raw product broken down in house, a kitchen brigade, a dining room with servers holding a table for two hours. Inside that structure, 30% on product leaves room for labor, and labor is the line that decides whether the P&L closes.

Move the same number onto a pizzeria and you are comparing two things that only resemble each other in name. What the guest pays for a pizza is mostly oven, time and service; the ingredients contribute very little of it. Move the number onto a bar and the comparison breaks in the opposite direction, because the margin on a pour has nothing in common with the margin on a center-cut steak.

This is also why other people's percentages deserve some suspicion. A number that comes from an operation unlike yours is not a target; at best it is a hint. We went into that at length in how to read a restaurant savings case study without being fooled by the headline percentage.

Why a pizzeria's food cost lies until you separate dough from toppings

In a pizzeria the aggregate food cost is close to useless, because it averages two economies that point in opposite directions: the dough, which costs a few cents per ball, and the toppings, which on some pies cost more than everything else on the menu line combined.

Say you run a plain margherita and a pie built on imported buffalo mozzarella, dry-cured ham and a finishing oil. To the guest they are the same product — a round pizza — but the first has an ingredient cost your system barely registers, and the second eats a serious share of its own menu price. Average them together and you get a number that describes neither one and tells you to do nothing.

The real decision in a pizzeria is never "lower the food cost." It is a decision per menu line: is this topping worth the price I can actually charge, in this dining room, in this neighborhood, to this guest? Sometimes the answer is yes even at a punishing cost on that single pizza, because it is the pie people come for and it pulls the rest of the menu along with it. Sometimes the answer is no, and then you change the product, the portion or the price — not the aggregate, which is only the consequence.

There is a practical reason to look at toppings line by line, as well. That is where your supplier's price moves week to week, while flour, water, salt and yeast sit still for months. A pizzeria's food cost does not drift upward evenly. It drifts in three or four items.

Why a low food cost at a bar is not an achievement

Behind a bar the margin on beverage is structurally wide, so a low food cost is normal and not a medal. A coffee, an amaro, a cocktail: product cost is a small fraction of the price, and that holds for everyone, from the most carefully run bar in town to the one that has never costed a recipe in its life.

Which leads to a conclusion that sounds like a paradox and isn't: a bar at 18% is not outperforming a restaurant at 32%. It is selling a different mix. And inside that 18% you can have an immaculately run coffee program sitting next to a kitchen losing money on every plate, because beverage volume covers the damage and leaves the average with nothing to say.

So at a bar the aggregate percentage is even less informative than elsewhere. The useful questions are different ones: how much of revenue comes from the kitchen versus the bar, what the kitchen's food cost looks like on its own, and how the mix shifts across dayparts. An operation doing lunch service at noon and aperitivo at seven is running two businesses under one roof, and a single percentage cannot narrate both.

How to build YOUR range instead of the internet's

Food cost by menu category is the number that makes an honest comparison possible, because it compares like with like: appetizers against appetizers, pizzas against pizzas, cocktails against cocktails. It is the first step, and no external average substitutes for it.

Building it is less laborious than it sounds. You need a recipe card for each item, with real portion weights and a yield figure for trim — the as-purchased weight you buy is not the edible portion that reaches the table, and on fish and produce the gap is large. From there you get theoretical cost per line, you group by family, and you end up with a percentage per category you can read without borrowing anyone's benchmark.

The second step is comparing yourself with yourself. Take the same category a month later and look at the movement, not the level. If pizzas went from 22% to 27%, something specific happened — a price list moved, a portion crept up, trim got worse — and something specific can be found. Note that you are above an internet average and you have no direction to search in at all.

The third step is the distance between theoretical and actual. Theoretical cost says what the items you sold should have cost you; actual says what you truly spent on product in the period. The gap between them is the margin that walks out without passing the register: off-card portions, product thrown away, ordering mistakes. We have written at length about how the quantity on an order line decides waste before the delivery ever reaches the kitchen.

Your target range, in the end, does not come from a table. It comes out of those three numbers together: the per-category percentage you can sustain given your labor cost and your rent, the month-over-month movement you consider acceptable, and the theoretical-to-actual gap you are willing to live with.

What changes in practice when you stop chasing a single number

Once you read food cost by category, decisions move from purchasing to the menu — and the menu is where the big levers are. Markup stops being one multiplier applied to everything and becomes a choice per item: high where the guest has nothing to compare against, low where the price is public and checkable, like a margherita or an espresso.

What you do with the expensive dish changes too. Under single-number logic you drop it, or you raise its price until the average comes back in line. Under per-category logic you first ask how much it sells and what it pulls: an expensive item that brings people in and gets two sides and a dessert ordered alongside it is worth more than an efficient one nobody picks. That is menu engineering, and it needs two numbers rather than one — cost and product mix.

Finally it changes how you deal with suppliers, because you know which items are worth a conversation. Negotiating across the entire basket is a way to lose an afternoon; negotiating on the six SKUs that actually weigh on a category in trouble is an hour of work with a measurable effect the following month. It is also, incidentally, one of the better arguments for keeping your supplier list short enough that each relationship is worth having.

None of those three shifts is available while food cost stays one number to be checked against an average. It is the difference between knowing you have a problem and knowing where it is.

The right number is the one that makes you ask a question

If you keep one thing from this piece, keep it in this form: a food cost percentage that raises no question in your mind is doing nothing for you, however good it looks. A restaurant at 29% that cannot say where the 29 came from is worse off than a pizzeria at 34% that knows exactly which three toppings pushed it there, and why.

And there is an uncomfortable corollary. To ask that question you need a current figure, not last quarter's close. A food cost that arrives three months late describes a story that already ended, one you can no longer act on. That is the problem mayo's spend control for restaurants is built around, and it is the same reason a food cost that updates as orders land changes the kind of decision available to you: not another benchmark to chase, but your own numbers, by category, while there is still time to use them.

---

mayo editorial team