Pricing 5 min read

Menu engineering: how to identify your most profitable dishes

Menu engineering sorts your dishes by profitability and popularity. Here is how to apply it and what to do with the results.

Menu engineering is a framework for analyzing your menu by two dimensions: how profitable each dish is (gross profit per sale) and how popular it is (how often it is ordered). The combination tells you where your menu is working and where it is quietly costing you money.

The four categories

Every dish on your menu falls into one of four quadrants:

Stars: high profit, high popularity. These are your best dishes. Protect them, feature them prominently, and do not change them unless something is broken.

Plowhorses: low profit, high popularity. Customers love these dishes but they are not earning much. The goal is to improve their margin without alienating customers who order them regularly.

Puzzles: high profit, low popularity. Good margins but not selling well. The question is whether the problem is price, placement, description, or the dish itself.

Dogs: low profit, low popularity. These items cost you in ingredients, prep time, and menu real estate. Most should be removed unless there is a specific strategic reason to keep them.

How to run the analysis

For each menu item, you need two numbers: gross profit per sale (selling price minus food cost per portion) and sales count over a defined period (typically 4 weeks).

Calculate the average gross profit and average sales count across all items. Items above average on both metrics are Stars. Items above average on profit but below average on popularity are Puzzles. Items below average on profit but above average on popularity are Plowhorses. Items below average on both are Dogs.

What to do with each category

Stars: feature them at the top of each menu section. Give them prominent placement and professional photography if you have it. Train staff to recommend them.

Plowhorses: reduce portion size slightly and add a lower-cost component, or raise the price by $0.50 to $1.00 and monitor whether orders drop. Many customers have low price sensitivity on items they order habitually.

Puzzles: improve visibility. Move them higher in the menu section, rewrite the description to emphasize value, or highlight them in server recommendations. If orders still do not pick up after 4 to 6 weeks, reconsider the dish.

Dogs: remove them unless they serve a specific purpose (e.g., a vegetarian option required to serve a segment of your customer base). Removing a dog from the menu rarely causes customer complaints — if it were missed, it would have been ordered more often.

Common mistakes in menu engineering

Using food cost percentage instead of gross profit. A dish with a 20% food cost but a $5 selling price generates $4 gross profit. A dish with a 35% food cost but a $24 selling price generates $15.60 gross profit. The higher-margin dish earns nearly four times more per sale.

Not repeating the analysis. Menu engineering is not a one-time exercise. Run it every three to four months, especially after price changes or seasonal menu updates.

Removing dishes based on popularity alone. A low-selling dish that generates high gross profit (Puzzle) is worth keeping and marketing. A high-selling dish with low gross profit (Plowhorse) is worth fixing.

Frequently asked questions

How do I get sales count data? From your POS system. Most modern POS platforms can export a sales mix report by item over any time period.

Do I need software for menu engineering? No. A spreadsheet with gross profit per item and sales count is enough to categorize your menu. The analysis is the insight, not the tool.

Should I apply menu engineering to my delivery menu separately? Yes. Delivery orders often have a different mix than dine-in — certain items travel better, and pricing differs. Run the analysis separately for each channel.

Try it on your own menu.

Cost a dish in minutes. No spreadsheets.

Cost your first dish