Menu design mistakes vs the right method: the numbers that decide

The percentage lies, the till doesn't: the two numbers that run your menu

Ranking your menu by food cost percentage burns profit because a percentage doesn't cover payroll: contribution margin in cash does. Go back to that 180-cover Bogotá restaurant: the seafood risotto ran 29% cost, took 34% of the mix, and returned 11,200 pesos per plate. The grilled sirloin cost 34% and returned 19,400. Across 61 daily risotto units, the gap works out to 8,200 pesos per diverted cover, which is exactly why median pre-tax profit in full service was 2.8% of sales in 2024, according to the National Restaurant Association's Restaurant Operations Report 2024/25. At two point eight percent, shifting the mix a single point toward the right dish beats any purchasing discount you negotiate this quarter. Writing a real description under a dish lifts its sales 27% against the same dish with no description, according to Cornell University's Food and Brand Lab in Brian Wansink's work.

Descriptions aren't decoration: they move 27% in sales

That 27% is not a marketing statistic: it is the cheapest lever there is for rearranging your mix. If your highest-contribution plate returns 19,400 pesos and sells 22 units a day today, adding a concrete description —cut, cooking method, origin, no hollow adjectives— pushes it near 28 units, roughly 116,000 pesos of extra daily margin, 3.4 million a month. One condition holds: the description must live ONLY on the dishes you want to push. Describe all fifteen plates equally and the effect cancels out, returning the menu to the noise it came from. Three scenarios, three different decisions from the same figures. Small restaurant, one location of 40 to 80 covers: you have no analyst, so take the six dishes that make up 70% of your units and work out cash margin by hand, one afternoon; at 2.8% median profit you cannot afford more. Mid-size, 100 to 250 covers or two locations: the quarterly review against real mix is already mandatory, because protein prices move inside the quarter and your printed menu does not.

How to read these numbers in YOUR operation?

Group of five locations or more:

a regulatory threshold appears if you operate in the United States, where the FDA's calorie labeling rule applies to chains with 20 or more locations, and you should design your menu template with that jump in mind before you reach it. Every surplus item on your menu costs you twice: inventory frozen in the walk-in, and customer attention split across dishes you have no interest in selling. I got this wrong for years, treating the menu as a kitchen and graphics problem when it is a financial instrument with typography. The mix gives the hard test: if a dish doesn't reach 3% of units over 90 days, it isn't paying for the space it occupies or the waste it creates, and it goes. In the Bogotá case, the menu —laid out by a design studio, with the risotto in the upper right corner, where the eye lands first— had spent two years pushing guests toward the lowest-contribution plate.

Menu size and the dish that steals the eye

The aesthetics were flawless; the till was not. Your printed menu lost authority because traffic moved off-premises: full service went from 19% in 2019 to 30% in 2024, and limited service from 76% to 83%, according to the National Restaurant Association's Off-Premises Report 2024. Translate that 30%: three of every ten orders at a white-tablecloth restaurant get decided on a screen, with no server suggesting, no upper right corner, and whatever photos and ordering the app imposes. If your financial menu lives only on paper, you optimized 70% of the business and handed the other 30% to an aggregator's algorithm. The same cash-margin hierarchy has to be replicated in the digital menu, and there the order of the first six items matters more than the entire layout of the printed one. Non-alcoholic beverage sales in restaurants grew 30% during 2024, according to Restaurant Dive, and almost no menu treats them as what they are: the highest-contribution category per second of kitchen time, since they never touch the hot line.

Drinks and rising categories: the margin your menu hides

A house lemonade costing 2,100 pesos and priced at 12,000 returns 9,900 without occupying a griddle station. Set that beside the risotto's 11,200, which eats eight minutes of a cook. Over the same period, chicken chains like Raising Cane's and Wingstop grew sales 30% —Wingstop closed 2024 at roughly USD 4.8 billion in system-wide sales, per its own results— on short menus with one dominant category. Concentrating doesn't impoverish your offer: it speeds the kitchen and clears the guest's decision. The figures you just read come from public industry sources —National Restaurant Association, Cornell, Circana, Restaurant Dive, the FDA, Wingstop's published results— and from the reading Masterestaurant and Diego F. Parra apply to them over real operations. There is no primary study here and no proprietary sample, and the limit deserves stating: the 2.8% median profit and the off-premises traffic figures are United States data, so in Bogotá, Mexico City or Lima they serve as a reference for DIRECTION, not for level.

Where these benchmarks come from and how far they reach?

Cornell's 27% comes from field experiments on limited menus. Use them to decide what to measure at home;

the number that governs is always the cash margin of your six best-selling dishes, calculated on your own cost sheet from this month. The underlying split is a unit of measure. Looks-first menu design optimizes percentages; the right method optimizes dollars per cover. A dish at 26% food cost priced at 32,000 returns 23,680; one at 34% priced at 58,000 returns 38,280. The menu that pushes the first looks flawless on the report and loses 14,600 pesos every time a server recommends it. The second split is cadence. A looks-first menu gets revisited when it starts to annoy someone; a financial menu gets reviewed every 90 days against real sales mix, because protein prices, seasonality and guest behavior all shift inside a quarter.

Where the two paths really split?

Four twenty-minute reviews a year beat a six-month graphic redesign. Third comes size. Every surplus item drags a purchase list, a waste line, fridge space and an inventory row behind it.

Going from 45 dishes to 26 is not a matter of taste: it removes nineteen SKUs from the costing system, and that shows up in next month's food cost variance. And there is a fourth, less obvious one: printed menu versus QR. QR is excellent for delivery, for changing prices without reprinting and for reading analytics on what guests look at; the printed menu is what controls service pace, menu narrative and the server's suggestive sell. Whoever kills the printed version to save on printing saves 400,000 a year and loses several million in average check.