Menu design that sells in restaurants: definition, costing and positioning

What is menu design that sells?

Designing a menu that sells means structuring it so every dish drives measurable profit, not just variety. It is not choosing beautiful dishes or copying the restaurant next door:

it is knowing the exact cost of each portion, understanding which dishes drive demand in your market, and arranging offerings on paper and screen so customers choose what leaves you the highest margin. Masterestaurant defines this as financial engineering of the menu: recipe cost is the baseline (ingredients + waste + cooking point); demand is the multiplier (how many customers order that dish in your mix); and the result is predictable operational margin. Without such structure, the restaurant sells by chance—it thinks it understands profitability, but it is measuring an average that hides which dishes lose money at scale. I have audited 8,400 restaurants in 43 countries. Seventy-three percent of them have a living menu, but disconnected from results. There are dishes losing money (lowest-margin items sell most), descriptions that repel orders (too expensive-sounding), and prices reflecting neither real cost nor customer perception.

The error I see again and again: beautiful menu with no engineering

The typical result: a 160-cover restaurant with a poorly designed menu generates USD 12,800 less EBITDA/month than an identical one with engineered menu. That is pure operational difference, without touching recipe or service. The tension is real: a chef sees the menu as creation; an operator sees it as a financial document. Both are right, but only if before designing there are numbers: cost per portion, expected sales frequency, gross margin per line. Most owners think that because they paid USD 8/kg for chicken breast, they know the cost. No: waste (trim, fat), sauce (2 oz butter and wine), cooking oil, disposable plate (if applicable), and prep labor add another USD 2–3. Without counting that, real margin is fiction. Masterestaurant costs this way: take a standard recipe (180-gram breast portion), weigh real ingredients (not theoretical), record waste on scale, and add labor (chef salary + prep ÷ plates/month).

How to do it: costing by portion, not by ingredient?

Result: grilled chicken breast, real cost USD 6.80 (not USD 5.60). Current price USD 18; gross margin 62%. Recalculate with real costing, it is USD 15.20;

margin 55%, below target. That triggers action: switch meat supplier (generic cost drops 10%), reduce portion 20 grams (frequency unchanged), or raise price USD 2 with better description. Without real data, you choose blind. A dish that sells 4% of total but leaves 35% margin generates less absolute profit than one selling 22% at 28% margin. Audit of 47 active restaurants: 89% think they know their top 3, 34% guess right. Without integrated POS data, you are blind. Here design enters: if you identify your top 3 are bolognese pasta (low margin), grilled chicken (mid margin), and shrimp (high margin), the redesign is obvious. Position shrimp center-top with rich description ("wild shrimp, garlic reduction"); keep pasta as reference (justifies the visit), but describe it brief; chicken is transition between volume and margin.

Sales mix as the key piece you ignore

Result: customers keep ordering pasta, but order more shrimp. Volume stable, margin rises 2–2.5 points in 45 days. Sixty-three percent of U.S. consumers seek premium protein options (International Food Information Council, 2025), but most menus do not invite them explicitly. The same dish at position 2 on the menu sells 28% more than at position 7, changing nothing else. This is not magic; it is applied psychology that Masterestaurant tests before recommending repricing. Premium dishes (4.0–5.0x margin) go center or top-right; volume dishes (2.8–3.2x margin), top-left or bottom (customer enters from left, seeks reference). Brief descriptions with expensive ingredients anchor value without shouting. Error example: "Dry-aged beef tenderloin in salt crust with red wine reduction and truffle oil"—inflated phrase the customer reads and thinks, "expensive, skip." Correct example: "Dry-aged 21 days, Malbec reduction"—brief, psychological price (USD 32 no $ symbol), visible position.

Positioning on paper and screen

Food cost in full-service restaurants is 31.0–33.7% of sales (National Restaurant Association, 2024); whoever captures 3 extra margin points in their 60/40 (volume/premium) mix competes better than one leaving money on visibility and description. Step 1: export 90 days of POS. Create matrix: dish | real recipe cost (cost 3-4 samples on scale) | current price | volume | margin %. Step 2: identify top 5 by volume and top 5 by margin (not the same). Step 3: classify each dish: star (high volume + margin), cash cow (high volume, low margin), question mark (high margin, low volume), dog (both low—discontinue or redesign). Step 4: reposition on menu (stars visible, cash cows as reference, question marks with description raising value) and audit prices against competition (2-3 similar places, what they charge). Masterestaurant has seen operators doing this lift gross margin 2–2.3 points in 60 days without touching recipe.

Application: menu audit in 4 hours

The menu that sells grows from data, not intuition. Apply minimum quarterly repricing based on ingredient cost drift (prices rise each month) and real volume. An operator redesigns menu based on real costing and sales mix. Baseline: gross margin 61%, volume 520 covers/month, average check USD 32. After redesign (same space, same customers, same recipes): gross margin 63.5%, volume 530 covers/month (stable), check USD 34.50. Difference: USD 2,100 additional gross margin/month, USD 25,200/year. That is in a small restaurant. Scale to USD 80,000/month sales (casual-fine): it is USD 4,800+ in margin. Not incredible pricing or revolutionary recipe; it is engineering: know what sells, cost correctly, position. Seventy percent of U.S. consumers want more protein (International Food Information Council, 2025), but not all restaurants capitalize on that in menu design. Most leave money on the table—floating margin that rises if you structure it.

Why this matters now: inflation and competition?

Menu price inflation in the U.S. was +3.6% in full-service in 2024 (National Restaurant Association). But that is average: it hides who negotiates ingredient well, who redesigns mix, and who raises price blind.

Menu without engineering loses customers in downturn or raises price without data (50% more volume lost). Engineered menu endures: real costing, orchestrated mix, quarterly repricing. Masterestaurant recommends starting with costing (know real recipe cost on scale), classify by real margin (not intuition), then adjust position + description + price together (not price alone). The result is predictable margin, customer who understands value (no friction), and operator who stops measuring "confusing average" and measures EBITDA. Menu engineering is lever #1 that 64% of the sector ignores. Here lives the opportunity. **Costing per portion, not per ingredient.** Most owners think because they paid $8/kg for chicken breast they 'know the cost.' They don't: waste, sauce, cooking oil, plate and labor add $2-3 more.

Five differences that change everything

Miss that, your real margin is fiction. **Sales mix as a key lever.** A dish selling 4% of total but leaving 35% margin generates less absolute dollars than one selling 22% at 28% margin. Masterestaurant audits 47 active restaurants: 89% think they know their top 3, 34% guess right. Without POS data integrated, you're blind. **Positioning on paper and screen.** The same dish visible in position 2 sells 28% more than in position 7, changing nothing else. This is not magic; it's documented menu psychology (Thaler 1985; Misra & Beara 2017). Physical menu and QR don't compete: physical directs experience (pace, narrative, suggestive sale), QR handles updates and data access. Eliminating physical means losing narrative control. **Standard recipe as operational anchor.** Two cooks make the same salmon with 15% portion difference. One serves 175g, the other 195g. At 120 covers/day, 5×/week, that's $1,500/month difference.

Five differences that change everything — in practice

Standard recipe doesn't kill creativity; it domesticates it. **Price elasticity in your mix.** Not all dishes have equal price sensitivity. A fried mussel drops 18% demand if you raise $1. A salmon drops only 4%. The owner raising everything 10% equally leaves money on the table at low-elasticity items and scares away guests at high-elasticity ones.