The Menu Engineering Matrix Most Restaurants Ignore (And It's Costing Them Thousands)
Last updated: April 14, 2026
Forty-two percent of restaurant operators reported they were not profitable in 2025 — and most of them had no idea the answer was sitting right there on their menu (source: National Restaurant Association 2026 State of the Industry). Not in a new revenue stream. Not in a staffing cut. In the items they were already selling, just to the wrong people, in the wrong position, at the wrong price.
Menu engineering is one of the most proven, least-used tools in the restaurant industry. A Cornell University study found that restaurants implementing menu engineering practices saw a 10% increase in profitability without adding a single new item or cover (source: eCornell, Menu Design and Engineering). That's not theory — that's operators realizing the money they needed was already on the plate.
This post will walk you through the actual mechanics: how to build your matrix, what to do with each category of item, and how to apply menu psychology so your guests naturally order what makes you more money. No price increases required.
What Is the Menu Engineering Matrix?
The menu engineering matrix was first formalized in the early 1980s by Michael Kasavana and Donald Smith of Michigan State University. It classifies every item on your menu into one of four quadrants based on two variables: contribution margin (how much profit the item generates per sale) and popularity (how often it sells as a percentage of total covers).
The four quadrants are:
How to Actually Build Your Matrix (Step-by-Step)
Most operators understand the concept but never build the actual matrix. Here's how to do it in an afternoon:
What to Do With Each Quadrant
Stars: Protect and Amplify
Stars are your moneymakers. Don't touch their recipe. Don't swap their protein. The most common mistake we see operators make with Stars is assuming they don't need any attention. They do — they need to be promoted. Feature them at the "golden triangle" of your menu (the top right and center-top area, where eye-tracking studies show guests look first), use a graphic box or subtle highlight to draw attention, and train servers to mention them proactively as signature dishes. Stars should also be your first answer on social media and Google menu photos.
Plowhorses: Re-Engineer the Margin
These are tricky. Guests love them — if you remove them or change them dramatically, you'll hear about it. Your options: (a) Quietly increase the price by $1–$2. A popular item with genuine brand equity often absorbs a 10–15% price increase with minimal resistance. (b) Reduce the food cost — use a less expensive cut for the same flavor profile, reduce protein weight by 0.5oz, or slightly adjust portions on high-cost garnishes. (c) Add a high-margin companion — if your $12 pasta is a Plowhorse, offer a $4 add-on (grilled shrimp, extra protein) with a 65–70% margin that significantly boosts the check average when guests say yes.
Puzzles: Fix Visibility, Not the Recipe
If a Puzzle item has a high contribution margin but low sales, the food is usually fine. It's a discovery problem. Move it to a higher-visibility menu position. Rewrite the description with sensory language (not "Salmon Filet" but "Pan-Seared Pacific Salmon with Brown Butter and Capers"). Have servers offer it as a suggestion. Add it to your specials board. Cornell research shows that a simple move from a buried menu position to a prominent one can increase sales of that item by 30–40%.
Dogs: Cut or Convert
Dogs consume prep time, inventory, and server mental load. Cut ruthlessly. If you're holding onto a Dog because "some regulars order it," ask yourself: are those regulars profitable enough to justify the ongoing operational complexity? In most cases, you can redirect those regulars to a Star with good server training. The exception: a Dog that serves an allergy or dietary need that the rest of your menu doesn't address. Keep it, but don't promote it.
Menu Psychology: The Science of Where Guests Look
Eye-tracking research in restaurant settings has consistently found that guests do not read a menu like a document — they scan it. In a multi-page menu, their gaze typically goes to the top right of any given section first, then the top left, then downward. This is called the "primacy and recency" effect in behavioral economics — people remember most what they see first and last. For a single-page menu, the center is the highest-value real estate.
Use this to your advantage:
- Place your highest-CM Stars in the top-right of each section and the first item of each category list.
- Use visual anchoring with "decoy" dishes. A $48 dry-aged ribeye at the top of your entrees makes your $32 NY strip feel like a great value — and your $32 item has a better margin for you. The expensive decoy doesn't need to sell; it just needs to make everything else look reasonable.
- Write descriptions that trigger sensory response. "Crispy" outperforms "fried." "Slow-braised" outperforms "cooked low and slow." "House-made" outperforms "homemade." Research from Toast's industry data consistently shows that descriptive menu language increases both perceived value and willingness to pay by 15–20%.
- Remove dollar signs. Studies show that menus without dollar signs (listing the price as "14" rather than "$14.00") reduce price sensitivity because the currency symbol triggers a "pain of paying" response in the brain.
- Limit choices. Menus with more than 7 items per category create decision fatigue, which causes guests to default to the familiar (often a Plowhorse they've ordered before) rather than exploring Stars or Puzzles. For most casual dining concepts, 5–7 items per section is optimal.
Real Kitchen Example: A Pizza Chain in Denver Adds $140K/Year Without a Price Increase
We worked with a 2-location fast-casual pizza concept in Denver that had been running the same menu for three years. Their COGS was creeping toward 34% — above their 31% target — but they attributed it to ingredient price increases and didn't look further.
When they built their menu matrix for the first time, they found three Plowhorses driving 22% of total covers: a $13 Margherita pizza with a 34% food cost, a $10 Caesar salad with a 39% food cost, and a $5 garlic bread with a 41% food cost. Combined, these three items represented $4,200/week in revenue but were yielding only $2,460/week in contribution margin.
Here's what they changed: The Margherita price went from $13 to $14 (guest resistance: minimal, because it's a simple, premium-positioned item). The Caesar salad protein sourcing was renegotiated with their distributor, bringing food cost from 39% to 31%. The garlic bread was repackaged as a $7 "Loaded Garlic Knots" with an added herb compound butter — same production cost, 40% higher price, and it moved from Plowhorse to Star within 60 days.
The result: COGS dropped from 34% to 30.1% on those items, contribution margin on those three items increased from $2,460 to $3,200/week — a $38,480/year improvement in gross profit without adding a single new menu item or customer.
How Often Should You Run Your Matrix?
Quarterly is the minimum for a stable concept. Monthly is better if you're in a high-volume or high-volatility environment. Ingredients change cost more rapidly than they used to — the NRA reported that food costs are now more than 35% above pre-pandemic levels — which means an item that was a Star in January can slide into Plowhorse territory by April if you're not tracking it. A live menu matrix, updated in your POS or spreadsheet monthly, is one of the highest-leverage activities a restaurant owner can do with a single afternoon.
Connecting Menu Engineering to Your Broader Prime Cost Strategy
Menu engineering addresses the food cost half of your prime cost equation. But prime cost — the sum of COGS and total labor — should be your north star metric. According to industry benchmarks from NOVATAB's 2026 prime cost analysis, the target is 60% or below. Most struggling operators are running at 65–70% or higher, and the leakage is happening in both labor and food cost simultaneously.
The good news: menu engineering can compress food cost percentage by 2–4 points on its own. Combined with tighter scheduling and cross-training on labor, and you can move prime cost from 67% to 61% — a swing that, at $1.5M in annual revenue, is the difference between an $18,000 loss and a $58,000 profit.
Food cost control doesn't start and end with the menu, of course. The most rigorous operators track every cost center with equal discipline — including kitchen consumables that are easy to overlook. If your restaurant uses deep fryers, for example, frying oil is a significant and often unaudited line item. A frying oil cost calculator can show you your annual spend per fryer in minutes — and operators who then apply the same cost-reduction mindset that menu engineering requires (measure, identify waste, change behavior) consistently find $3,000–$8,000/year in oil savings per fryer unit. Every dollar saved in COGS is a dollar that doesn't need to come from the top line. Understanding how to extend frying oil life is one of the fastest-returning operational changes a frying-heavy kitchen can make.
Frequently Asked: How Do I Know If My Menu Is Priced Correctly?
You know your menu is priced correctly when your food cost percentage lands within 2–3 points of your target (typically 28–32% for full-service, 25–30% for QSR), and when your Stars are your most-sold items. If your highest-selling items have the lowest margins, your menu is priced in conflict with your business. Run the matrix, identify which Plowhorses are your top sellers, and calculate what a $1–$2 price increase on just those items would do to annual revenue. The math is usually compelling enough to act immediately.
People Also Ask: How Long Does It Take to See Results From Menu Engineering?
Most operators see measurable improvement in food cost percentage within 30–60 days of implementing menu engineering changes. The fastest results typically come from repositioning Puzzles (high-margin, low-popularity items) to better menu real estate — some operators report 20–35% increases in Puzzle item sales within the first month simply by moving them to a more prominent position and rewriting the description.
Sources
- National Restaurant Association — 2026 State of the Restaurant Industry Report
- Cornell University eCornell — Menu Design and Engineering
- Toast — Restaurant Industry Benchmarks and Labor Data
- NOVATAB — Restaurant Prime Cost in 2026: Definition and Benchmarks
- National Restaurant Association — 2026 Industry Cost Pressures Report
