Menu Engineering Calculator
Enter one dish's sales and cost alongside your menu averages to see which of the four menu engineering categories it falls into — and what to do next.
What menu engineering actually does
Menu engineering sorts every dish on your menu by two things you already know: how often it sells, and how much money it makes when it does. Plotting those two against each other produces four categories, and each one has a different correct action. It is the most reliable way to raise profitability without raising a single price across the board.
The reason it works is that most menus are carrying two opposite problems simultaneously. Some dishes sell constantly while barely covering their own cost. Others earn excellent margin and almost nobody orders them. Both are invisible if you only look at food cost percentage, because that number says nothing about volume.
Contribution margin is the dollar figure each sale contributes toward rent, wages and profit. It is deliberately measured in dollars rather than percent, because dollars are what actually pay your bills.
The two axes
Popularity
A dish's share of total covers, compared against what it would sell if every item on the menu sold equally. On a sixteen-item menu, equal distribution would give each dish 6.25% of covers. The conventional threshold is 70% of that figure — so anything selling above 4.4% counts as popular.
The 70% adjustment exists because menus are never evenly distributed in practice. A handful of items always dominate, and holding every dish to a perfectly equal share would classify almost the entire menu as unpopular.
Contribution margin
Whether the dish's margin sits above or below the menu-wide average. Note that this is a relative test, not an absolute one: a dish is not judged against some universal target but against the rest of your own menu. That is what makes the method work across concepts as different as a cafe and a steakhouse.
The four categories and what to do
| Category | Profile | Action |
|---|---|---|
| Star | High popularity, high margin | Protect. Hold the spec, keep it visible, do not discount. |
| Plowhorse | High popularity, low margin | Raise price modestly or reduce plate cost. |
| Puzzle | Low popularity, high margin | Reposition, rename, or promote before removing. |
| Dog | Low popularity, low margin | Remove, unless it serves a strategic purpose. |
Stars
Your best dishes, and the ones most often damaged by well-meaning intervention. The temptation is to use a Star as a promotional discount because it is popular. Resist it: you are discounting the dish that generates the most profit per cover. Instead protect the recipe specification rigorously, since portion drift on a Star costs more than on anything else on the menu.
Plowhorses
The category with the most recoverable money. Guests already want these dishes, which means demand is proven and price sensitivity is usually lower than you fear. A modest price rise on a Plowhorse typically produces almost no drop in volume, and the margin improvement flows straight to the bottom line because the volume is already there.
The alternative is re-engineering the plate cost — substituting a component, adjusting a garnish, or improving yield. Rebuild the costing with the recipe costing calculator before deciding which route is less damaging to the dish.
Puzzles
Profitable dishes nobody orders. Before removing one, work through why: is it badly positioned on the menu, poorly described, unfamiliar, or simply priced above what guests expect in that section? Menu position and description change ordering behaviour more than most operators expect, and a Puzzle that becomes a Star is the single highest-value outcome available from this exercise.
Dogs
Unpopular and unprofitable. Removing them is usually correct, and the benefit is larger than the dish's own numbers suggest — fewer items means less inventory, less prep, less waste on slow-moving ingredients, and a shorter menu that is easier for guests to read. The exceptions worth keeping are dishes serving a genuine need, such as a vegetarian or allergen-safe option that lets a whole table book.
A worked example
A sixteen-item menu sold 2,400 dishes last month. The average contribution margin across the menu is $11.20. The popularity threshold is (100 ÷ 16) × 0.7 = 4.4%.
| Dish | Price | Cost | CM | Sold | Share | Category |
|---|---|---|---|---|---|---|
| Ribeye | $32.00 | $12.80 | $19.20 | 180 | 7.5% | Star |
| Burger | $16.00 | $6.20 | $9.80 | 420 | 17.5% | Plowhorse |
| Sea bass | $28.00 | $9.60 | $18.40 | 70 | 2.9% | Puzzle |
| Chicken salad | $14.00 | $5.10 | $8.90 | 62 | 2.6% | Dog |
The burger is the clearest opportunity. At 17.5% of covers it is the menu's engine, but its $9.80 margin sits below the $11.20 average. Raising it from $16.00 to $17.50 lifts contribution margin to $11.30 — above average — and at 420 units generates an extra $630 per month from one price change on a dish guests are already choosing.
The sea bass is the second opportunity. It earns $18.40 per plate, nearly as much as the ribeye, but sells to only 2.9% of covers. Moving it to the top of its menu section and giving it a better description costs nothing and may convert it into a Star.
Getting the inputs right
The classification is only as good as the plate costs behind it, and this is where the exercise most often goes wrong.
- Apply yield percentages. A protein bought at $20 per kilo with an 80% usable yield actually costs $25 per usable kilo. Skipping this consistently overstates margins on exactly the dishes where accuracy matters most.
- Include garnish, oil, seasoning and sauces. They rarely appear on recipe cards and typically add two to four points of food cost.
- Use a full trading period. One quiet week will misclassify seasonal dishes. A month is the practical minimum.
- Compare like with like. Run starters, mains and desserts as separate exercises. A dessert will almost never clear a threshold set by main courses, and classifying it against them tells you nothing.
- Re-cost before you classify. Costings more than a quarter old will have drifted with supplier prices.
Get your food cost percentage and plate costs accurate first. Menu engineering built on stale costings produces confident recommendations that happen to be wrong.
Why total contribution matters more than category
The four-box grid is a starting point, not a verdict. The number that ultimately decides where to spend your attention is total contribution — margin per plate multiplied by units sold.
A Star earning $19.20 across 180 covers contributes $3,456. A Plowhorse earning $9.80 across 420 covers contributes $4,116. The Plowhorse is the more important dish to your business despite the worse category, which is why the burger and not the ribeye is the first thing to fix in the example above.
Rank every dish by total contribution alongside its category. The categories tell you what kind of intervention each dish needs; the total contribution tells you which ones are worth doing first.
Frequently asked questions
What is menu engineering?
Menu engineering classifies each dish by popularity and contribution margin into four categories — Stars, Plowhorses, Puzzles and Dogs — so that each receives the pricing or menu action that suits it.
How is the popularity threshold calculated?
Divide 100 by the number of items on the menu to get an equal share, then multiply by 0.7. On a 16-item menu that is 4.4% of covers. The 70% adjustment reflects that real menus are never evenly distributed.
Should I use contribution margin or food cost percentage?
Contribution margin. Food cost percentage ignores volume and favours cheap dishes, which is how operators end up promoting a low-cost item that contributes far less in dollars than a higher-cost one.
What should I do with a Plowhorse?
Either raise the price modestly or reduce the plate cost. Demand is already proven, so a small price increase usually costs very little volume while improving margin across high unit numbers.
Should I always remove Dogs from the menu?
Usually, but not automatically. A dish that lets a group book — a vegetarian, vegan or allergen-safe option — earns its place through the covers it protects rather than its own margin.
How often should menu engineering be repeated?
Quarterly, and after every menu change or significant supplier price movement. Categories shift as costs and guest preferences move.
Can menu engineering be applied to drinks?
Yes, and it often works better there because beverage margins vary enormously. Run the drinks list as a separate exercise with its own thresholds rather than mixing it with food.