Menu Engineering in Practice: How to Find the Dishes Losing You Money (With a Worked Example)

The Problem Menu Engineering Solves

Almost every bar, bakery or restaurant owner knows the menu by heart — what sells most, what guests praise. What almost nobody knows off the top of their head is how much profit each dish actually leaves after paying for ingredients. That is exactly where money leaks: dishes that sell a lot and leave almost nothing, and dishes nobody orders that keep taking up kitchen time, stock and menu space.

Menu engineering is the technique that crosses two pieces of information you already have — how much each dish sells and how much margin each dish leaves — to decide what to highlight, what to adjust and what to drop. Formalized in the 1980s by Kasavana and Smith, it is still the most used method in the industry because it is simple enough to fit in one working afternoon.

If you want the full theory, with the BCG matrix and sales mix, it is in our menu analysis guide. This article is the hands-on version: a worked example, start to finish.

The Two Numbers You Need for Each Dish

1. Contribution margin = selling price − ingredient cost (the dish's food cost). Not a percentage: the amount in currency left from each unit sold to pay rent, staff and your profit. A R$ 79 dish with R$ 48 of ingredients leaves R$ 31; a R$ 24 side with R$ 6 of ingredients leaves R$ 18. The first has a worse percentage margin and a better margin in money — and money pays the bills.

2. Popularity = units sold in the period. Pull it from your ordering system or, failing that, from one typical month of tickets. Avoid atypical months (holidays, vacations, renovations).

For margin you need the cost of each recipe — if you don't have it yet, the article on recipe costing shows how to build it painlessly. For the analysis to hold, use the same period for all dishes in the same category (don't mix starters with desserts: always compare dishes competing for the same guest decision).

Worked Example: Four Bar Snacks

Let's use a fictional bar, Zé's Bar, and only the "sharing plates" category. The numbers are illustrative, but the reasoning is exactly what you will apply to your menu.

• Fries — price R$ 24, ingredients R$ 6, margin R$ 18, 300 sold in the month
• Grilled picanha — price R$ 79, ingredients R$ 48, margin R$ 31, 40 sold
• Chicken strips — price R$ 29, ingredients R$ 14, margin R$ 15, 220 sold
• Cod fritters — price R$ 32, ingredients R$ 24, margin R$ 8, 18 sold

Total: 578 plates in the month. Now two reference calculations:

• Weighted average margin: sum of (margin × units) divided by total units → (18×300 + 31×40 + 15×220 + 8×18) ÷ 578 ≈ R$ 17.40. Above that is "high margin"; below, "low margin".
• Popularity line: the classic rule uses 70% of fair share. With 4 dishes, fair share is 25% (144 units); 70% of that is ~101 units. Sold more than that, it's popular; less, it isn't.

The Matrix: Where Each Dish Lands and What to Do About It

Crossing the two criteria, each dish lands in one of four quadrants:

• Star (high margin, high popularity) — Fries: R$ 18 margin and 300 sales. Protect it: keep quality and portion, give it visual prominence on the menu and never touch the price without testing.
• Plowhorse (low margin, high popularity) — Chicken strips: sells a lot, leaves R$ 15. The dish that most needs attention: raise the price by R$ 2 or R$ 3 (loyal guests rarely notice), cut the recipe cost or pair it with a high-margin item.
• Puzzle (high margin, low popularity) — Grilled picanha: R$ 31 per sale, but only 40 sales. The problem is visibility, not price: move it on the menu, improve name and description, suggest it actively at the table.
• Dog (low margin, low popularity) — Cod fritters: R$ 8 margin and 18 sales. A candidate to go — or to be completely reworked, if it is part of the house identity.

Notice what happened: without changing a single price, you already know the month's effort goes to the chicken strips (218 sales × R$ 2 more = R$ 436 extra profit per month from a timid adjustment) and to the picanha (each extra sale is worth almost twice a plate of fries).

The Most Common Mistakes When Applying It

• Comparing different categories. Dessert will always look "unpopular" next to a main. Analyze within each category.
• Using percentage margin instead of money. A dish with 70% margin on R$ 10 leaves R$ 7; another with 40% on R$ 80 leaves R$ 32. The second pays more bills.
• Forgetting the real cost. Oil, gas, delivery packaging and prep waste belong in the dish cost. A recipe card with a yield factor solves this.
• Doing it once and filing it away. Ingredients go up, menus change, seasons turn. The analysis needs redoing every month or every menu change.
• Dropping the dog without context. Some low-margin dishes bring in the guest who orders the star. Decide with the whole, not the isolated line.

How to Do This Every Month Without Becoming a Spreadsheet Person

The exercise above takes an afternoon the first time and is worth every minute. The hard part is repeating it: pulling sales, updating costs, recalculating averages, reclassifying. That is why most restaurants do it once, improve, and then let it age.

Mise en Place's Menu X-Ray does the structural part of this analysis automatically: you upload a photo or PDF of the menu and, in two minutes, get a 0–100 score across five dimensions — variety, pricing, balance between categories, naming and descriptions — with quick wins that cite your dishes and prices. It is free, requires no signup, and is the ideal starting point before opening the spreadsheet: it shows where to look first.

For those already using Mise en Place as their system, the same score runs on the registered menu and the quick wins become tasks on the Action Plan, with re-analysis whenever the menu changes. Menu engineering stops being an annual project and becomes routine.

Sources

• Kasavana, M.L. & Smith, D.I. Menu Engineering: A Practical Guide to Menu Analysis. Hospitality Publications, 1982 (rev. 2012).
• SEBRAE. Engenharia de Cardápio: Como Aumentar a Lucratividade do seu Restaurante, 2024.
• Abrasel — Brazilian Association of Bars and Restaurants. Food Service Industry Indicators, 2025.