The fastest way to find low-margin dishes is to calculate two numbers for every menu item: its contribution margin (menu price minus food cost) and its popularity (how often it sells). Plot every dish against those two axes, and your problem items reveal themselves instantly. Dishes that sell well but earn little, and dishes that earn well but barely sell, are where your money is leaking. This article shows you how to run that analysis and—more importantly—how to fix what you find.
You can't engineer a menu you haven't costed. For each dish, add up the cost of every ingredient at the exact quantity used in the recipe—including the items people forget: garnish, oil, butter, the slice of lemon, the side of bread.
Here's the math for a single dish:
Example: A burger sells for $14. The beef, bun, cheese, lettuce, tomato, sauce, and fries cost $4.20 total.
That $9.80 is what actually pays your rent, labor, and profit. Food cost percentage tells you efficiency; contribution margin tells you dollars. You need both, because a dish with a great margin percentage that almost no one orders contributes little, and a low-margin dish that sells constantly can quietly drain you.
A practical target for most full-service small restaurants is a food cost between 28% and 35%, but don't treat that as gospel. A $3 plate cost on a $9 appetizer is "high" at 33%, yet the $6 margin might be excellent for your concept. Always look at the dollar figure alongside the percentage.
Export item-level sales counts from your POS for a clean period—at least 30 days, ideally 60 to 90 to smooth out specials and weekends. You want, for each menu item:
Multiply units sold by contribution margin to get total contribution per dish. This single column is the most honest ranking of your menu. The dish at the top is your real workhorse, even if it isn't the one you're proud of.
This is classic menu engineering. Compare each dish to your menu averages for popularity and contribution margin, then sort into four buckets:
Most "low-margin dish" problems live in the Plowhorse and Dog quadrants. Here's how to handle each.
These are the trickiest because customers already love them—you can't just yank them. Options, from least to most disruptive:
The goal isn't to make plowhorses less appealing—it's to keep the appeal while clawing back margin.
Puzzles are profitable dishes people don't notice. Before cutting, try:
If a puzzle still won't sell after better placement, it becomes a dog.
Dogs usually deserve to go. Each one carries hidden costs: extra inventory, ingredients that spoil, longer training, slower kitchen. Before cutting, ask whether the dish shares ingredients with stars (so removing it won't reduce waste) or serves a strategic role (a vegan option, a kids' item). If not, cut it and watch your prep and waste shrink.
Food cost isn't the whole story. A dish can have a great plate cost but destroy your margin through:
When you flag a low-margin dish, check whether the problem is the recipe cost, the yield, or the labor before deciding the fix.
Ingredient prices move constantly. A dish that was a star at $4.20 plate cost becomes a plowhorse when beef jumps 20%. Re-cost your menu quarterly at minimum, and immediately whenever a key ingredient's price spikes. Keep your recipes documented with exact quantities so a price change updates every affected dish automatically.
---
The math behind menu engineering is simple, but doing it by hand across 40 dishes—and keeping yield, portion, and price changes current—is where most small restaurants give up.
That's exactly why we built PrepSheet. It turns any recipe into a costed, scalable prep sheet, automatically tracking ingredient costs, yields, and portion math so you can see your real margins the moment prices change. If you'd rather stop guessing which dishes pay the bills, [join the PrepSheet waitlist](#)—we'd love to have you.