Menu Engineering 101: Stars, Plowhorses, Puzzles, and Dogs

A seafood restaurant in Baltimore had 38 menu items. The owner knew some dishes sold better than others. What he didn't know: 6 items were generating 52% of his profit. 14 items were losing money on every plate. He'd been running the restaurant for 7 years and had never run a menu engineering analysis.

Menu engineering is the single highest-ROI exercise a restaurant can do. It takes about 3 hours. It requires a POS report and a recipe costing spreadsheet. The result is a clear picture of which dishes make you money and which ones are just taking up space on the menu.

Industry Data & Sources:

Bureau of Labor Statistics data shows the restaurant industry employs over 12.5 million people with annual turnover rates around 75%.

The National Restaurant Association provides operational benchmarks and best practices for restaurant management.

The 2×2 Matrix

Menu engineering plots every dish on two dimensions: popularity and profitability. This creates four quadrants:

                    HIGH PROFITABILITY
                           |
        PUZZLES            |           STARS
        (Low popularity,   |    (High popularity,
         high profit)      |     high profit)
                           |
LOW POP ------------------+------------------ HIGH POP
                           |
        DOGS               |        PLOWHORSES
        (Low popularity,   |    (High popularity,
         low profit)       |     low profit)
                           |
                    LOW PROFITABILITY

Stars: High popularity, high profit. These are your winners. Protect them. Promote them. Never run out of ingredients for them.

Plowhorses: High popularity, low profit. These sell well but don't make much money. The goal is to increase their profitability without killing their popularity.

Puzzles: Low popularity, high profit. These make good money when they sell, but they don't sell enough. The goal is to increase their popularity.

Dogs: Low popularity, low profit. These are taking up menu space, inventory dollars, and kitchen attention. Most should be cut.

How to Run the Analysis

You need two numbers for every dish:

Popularity score: The number of units sold in a given period (usually a month). Calculate the average units sold per menu item. Any item selling above the average is high popularity. Below average is low popularity.

Profitability score: The contribution margin per dish (menu price minus food cost). Calculate the average contribution margin across all items. Above average is high profit. Below average is low profit.

Example from a casual Italian restaurant:

Average units sold:

$$\frac{340+285+410+62+195+140}{6} = 239$$

Average contribution margin:

$$\frac{\text{USD}11.79+\text{USD}13.89+\text{USD}10.59+\text{USD}16.19+\text{USD}8.09+\text{USD}7.59}{6} = \text{USD}11.36$$

Classification:

• Spaghetti Bolognese: 340 > 239 (high pop), $11.79 > $11.36 (high profit) → Star
• Chicken Parmesan: 285 > 239 (high pop), $13.89 > $11.36 (high profit) → Star
• Margherita Pizza: 410 > 239 (high pop), $10.59 < $11.36 (low profit) → Plowhorse
• Seafood Risotto: 62 < 239 (low pop), $16.19 > $11.36 (high profit) → Puzzle
• Caesar Salad: 195 < 239 (low pop), $8.09 < $11.36 (low profit) → Dog
• Tiramisu: 140 < 239 (low pop), $7.59 < $11.36 (low profit) → Dog

What to Do With Each Quadrant

Stars: Protect and Promote

Don't touch the recipe. Don't change the price unless costs force it. Do make sure these items are in the best positions on your menu (top-right of each panel, where eyes land first). Do train servers to mention them. Do ensure you never 86 a Star because you ran out of an ingredient.

Plowhorses: Increase Margin Without Killing Volume

Plowhorses are tricky. They're popular. Customers love them. But they don't make enough money. You have three options:

Raise the price. A $0.50-1.00 increase on a high-volume Plowhorse adds up fast. The Margherita Pizza above sells 410 units a month. A $1 price increase adds $410 in monthly profit. That's $4,920 a year from one price change.

Reduce the food cost. Smaller portion of the expensive ingredient. Cheaper substitute for a component. Renegotiate with the supplier for the key ingredient. If the Margherita's food cost drops from $2.40 to $2.10, monthly profit increases by $123.

Add a premium version. Keep the original Plowhorse. Add a "Deluxe" version at a higher price point. The original stays for the price-sensitive customers. The upgrade captures the ones willing to spend more.

Puzzles: Figure Out Why They Don't Sell

Puzzles make good money per plate. They just don't sell enough plates. The fix is usually one of three things:

Bad menu placement. Move the Puzzle to a high-visibility position. The top-right of a panel. The first item in a section. Put a box around it.

Bad description. Rewrite the menu description. Make it sound irresistible. "Seafood risotto" is boring. "Creamy arborio rice with jumbo shrimp, sea scallops, and mussels in a saffron-tomato broth" sells.

Server doesn't suggest it. Train servers to recommend the Puzzle. Run a contest: whoever sells the most Puzzles this week gets a $50 gift card. Servers will suddenly remember it exists.

If none of these work after 4-6 weeks, the Puzzle might actually be a Dog in disguise. Cut it.

Dogs: Cut or Rework

Most Dogs should come off the menu. They're losing money or barely breaking even, and nobody orders them anyway. The Seafood Risotto example above is actually a Puzzle, not a Dog. But if it were a Dog, the math would be clear.

Exceptions: a Dog might stay if it's a loss leader that drives traffic, or if it uses ingredients that would otherwise go to waste. But these are rare. In most cases, cutting a Dog frees up inventory dollars, prep time, and menu real estate for something that actually makes money.

The Menu Redesign

Once you've classified everything, redesign your menu layout:

1. Stars go in the prime positions (top-right, first item in each section)
2. Plowhorses get price adjustments or cost reductions
3. Puzzles get better placement and descriptions
4. Dogs get cut (with rare exceptions)

One restaurant did this exercise and cut 9 Dogs from a 42-item menu. Their overall food cost dropped 2.3 points because they stopped carrying inventory for dishes that sold 3-4 times a week. Their average check went up $1.80 because the remaining items were better positioned. Total annual impact: $47,000 on $1.1 million in revenue.

Pair With Price Optimization

Menu engineering tells you which dishes to promote and which to fix. Our menu price optimizer tells you exactly what price to set. It calculates the profit-maximizing price for each dish based on your costs and target margins, then applies psychological pricing tactics to make those prices feel right to customers.

---

Related reading: The Psychology of Menu Pricing · How to Calculate Recipe Cost · Restaurant Menu Engineering: Complete Guide