Profitable menu: criteria to build it, measured before and after the Masterestaurant method

The first number is margin in currency, not percentage

Rank your menu by CONTRIBUTION MARGIN in currency per dish, and only then look at the percentage, because the register does not collect percentages. A USD 6 dish at 68 % margin leaves four dollars; a USD 22 dish at 58 % leaves nearly thirteen, and the second one pays the shift payroll. Input prices in 2026 widen that gap faster: according to the US Bureau of Labor Statistics via CBS News, consumer beef reached USD 5,98 per pound in May 2025, an all-time high, and ground beef hit USD 6,12 in June of that year per the same agency via NPR. Against that, the USDA Economic Research Service puts chicken at USD 2,99 and pork at USD 3,11 per pound in 2024. The concrete decision: reprice your beef dishes first, in currency, not in food cost points. Menu position moves the mix measurably: the first entrée listed in its category carries a 33 % chance of being ordered regardless of price, according to NeatMenu's 2026 menu psychology analysis.

Sales mix is steered from the paper and from the server

A third of that category's tickets is decided before anyone speaks. Now stack the other layer on top. In that 92-seat restaurant in Bogotá, twelve low-margin, high-waste dishes took 41 % of the tickets because those were the ones the server recited whenever a guest hesitated, and the real food cost from the register landed at 36,4 % against a theoretical 29 %. Seven and a half points of difference that had nothing to do with purchasing. Put the highest currency-margin dish at the top of each category and drill it on Tuesday in preshift, with the name and the exact phrase. Fewer than you carry today, and the criterion is not a magic number but how many inventory references your kitchen can hold without waste. That menu ran 74 dishes at a theoretical 29 % food cost that the register contradicted at 36,4 %. Every dish adds purchases, slow rotation and one more line the server must recall.

How many dishes should a profitable menu carry?

And here I was wrong for years: I thought trimming the menu was a kitchen efficiency matter, when the big effect sits in the server's mouth, since a person can only keep six or eight live recommendations going at once.

With protein prices like 2025's — beef at USD 6,51 per pound per the USDA — every slow-rotating beef reference is waste bought at the highest counter price. Cut the dishes below 3 % of the mix and reassign those purchases. If your beverage list is not separated from the food analysis, you are giving away the cleanest margin in the business. Average pour cost on alcoholic beverages sits near 20 % according to BackBar's industry guide, with liquor around 15 %, draft beer around 20 % and wine between 35 % and 45 %. Compare that with a protein dish that starts at 32 % food cost in the best case.

The bar is the margin lever almost nobody reprices

A USD 12 cocktail at 15 % pour cost leaves USD 10,20 in gross margin; your mid-range burger, which Circana via Restaurant Business puts at a USD 14,48 menu price in September 2025 after rising 3,1 % year over year, leaves considerably less. The decision: set a beverage attach rate target per table and pay for it in preshift, not on the printed menu. Benchmarks read differently by size, and whoever applies them the same way across three scenarios gets two of them wrong. In a small single-kitchen venue with a 30 to 40 dish menu, NeatMenu's 33 % on the first item of each category is your cheapest lever: reordering the paper costs one evening of work and needs no system. In a mid-size operation of 80 to 120 seats and two shifts, like the Bogotá case, the real lever is preshift, because that is where those 7,4 points of gap between theoretical and register food cost are born.

How to read these numbers in YOUR operation?

In a group of four venues or more, the governing number is currency margin weighted by units sold per site, because a beef dish at USD 6,51 per pound can be profitable in one location and lose money in another with a different ticket.

Pick your scenario before touching prices. Be honest about where the figures originate before defending a decision with them. Input prices come from the US Bureau of Labor Statistics and the USDA Economic Research Service, which measure the US market; menu prices come from Circana via Restaurant Business and from the National Restaurant Association index, which recorded a 9,0 % year-over-year peak in full-service restaurants in 2022; pour cost is published by BackBar as an industry guide, not a sampled study; and the 33 % position figure comes from NeatMenu. Three limits that matter: these are US data and your local market moves different relative prices; industry averages hide enormous dispersion across formats; and none of them measures your kitchen.

Where these benchmarks come from and what they do NOT prove?

At Masterestaurant we use these benchmarks as a contrast frame, never as a substitute for your own portion count. The printed menu remains the instrument that steers the mix, and the QR code is an operational complement, not its replacement.

The reason is attention physics: paper shows the full category at a glance, and that is exactly where the 33 % probability on the first listed entrée documented by NeatMenu operates; a phone screen shows two or three items and forces scrolling, so position loses force. What would happen if you migrated everything to QR tomorrow? You would lose control of position, the server would lose the paper as a shared script with the guest, and the mix would flatten toward whatever the guest already knew before sitting down, usually the beef dish that costs the most to produce at USD 5,98 per pound. Keep the QR for updating prices between reprints.

Diego F. Parra's criterion on what gets repriced and what gets cut

A dish gets repriced when its currency margin falls below the median of its category, and it gets cut when it also fails to reach 3 % of the mix. That is the order, and the order matters because cutting before repricing kills dishes that were merely mispriced. The underlying tension is real: raising prices pushes traffic down, and the National Restaurant Association index showed that the 9,0 % peak in 2022 came with visible guest resistance. The bridge is charging for differentiation rather than inflation: Datassential reports that 74 % of operators say global flavors let them charge more. And watch the cycles, since Technomic via CSP Daily News measured a 1,9 % year-over-year drop in plant-based menu items in 2024. This week: measure the currency margin of your ten best-selling dishes. First one is a unit-of-measure problem. Nearly all menu profitability talk happens in percentages, and percentage is a trap at low ticket levels: a dish carrying 68 % margin on a USD 6 price leaves four dollars, while one carrying 58 % on USD 22 leaves almost thirteen.

The three differences that actually move margin

The register does not collect percentages. So the first criterion, the one that orders every other, is MARGINAL PROFITABILITY PER DISH expressed in currency, then crossed against the units that dish genuinely sells. Second: sales mix is never a passive reading. A menu is a machine for directing attention, and attention moves through two layers, the paper and the person. Paper —position, boxing, no currency sign, description length— explains part of the behavior; the human layer explains the rest, and wherever we measure both, a server's verbal suggestion outweighs any graphic design decision. That is why a redesign without training returns two or three points of mix and then flattens. Third one touches demand elasticity, and here I was wrong for years. I assumed raising the price of the star dish was the short path to margin, and in two mid-size chains that cost real money: the star is usually the most elastic item precisely because guests know it and compare it.

The three differences that actually move margin — in practice

Reverse the move. Raise price where no external reference exists —signature plates, sides, house drinks— and protect the star as is, since the star is what brings the guest to the door.