Core Platform Rank My Store Family Club Rent & Lease Features Pricing How it works Blog ROI Calculator Restaurant Portal Get Started Free →
Home / Blog / Restaurant Menu Engineering
Restaurant Growth

Restaurant Menu Engineering: How to Find Your Most Profitable Dishes

Your best seller is not always your best earner. Here is how to score every dish on sales and margin, read the four-box matrix, and fix the menu one change at a time.

Stan Liachev, PhD
Stan Liachev, PhD
Founder & CEO, DevourNow | Founder, WorldApp Inc.
· November 16, 2026 · 8 min read
X in f
Restaurant menu engineering: pasta, salads and grilled chicken on a shared dinner table, seen from above

It's Sunday night and the week's sales report is open on the office laptop. The plain cheese pizza sold more than anything else again. It feels like good news until you work out what each one actually leaves after the dough, the sauce and the cheese. The specialty pie that sells half as often might be earning more per order. Restaurant menu engineering is how you find that out with numbers instead of a hunch.

Restaurant menu engineering is a method for scoring every dish on two measures: how often it sells (its share of the menu mix) and how much money it leaves after food cost (its contribution margin). Each item lands in one of four groups, and each group gets a different fix: keep, reprice, promote or cut.

What Is Restaurant Menu Engineering?

The method comes from Michael Kasavana and Donald Smith's 1982 book Menu Engineering: A Practical Guide to Menu Analysis. Kasavana, a professor emeritus at Michigan State University, still writes about the same two tests. An item is popular if its share of sales is at least 70% of an equal share. It is profitable if its contribution margin beats the section's average contribution margin.

No software is required. The whole analysis fits on one spreadsheet tab, with one row per dish. Its value is that it separates "this sells a lot" from "this makes us money", two things a busy owner tends to treat as the same.

Menu engineering formulas
Contribution margin = menu price - plate cost
Menu mix % = units sold of the item / units sold in the section
Popularity line = (1 / number of items in the section) x 70%
Profit line = total contribution margin of the section / total units sold in the section

Why Does Menu Engineering Matter More When Food Costs Keep Rising?

When ingredient costs rise faster than your printed prices, a dish can stay busy while its margin shrinks. The latest public numbers:

  • Restaurant menu prices (food away from home) were 3.5% higher in September 2026 than a year earlier, and are forecast to rise 3.5% for all of 2026, according to the USDA Economic Research Service (updated October 23, 2026).
  • USDA puts the 2006-2025 average at 3.5% a year, and forecasts 2.8% for 2027.
  • The food-away-from-home index rose 0.5% in October 2026 alone, per the Bureau of Labor Statistics.
  • 42% of restaurants surveyed by the National Restaurant Association said they were not profitable last year, as reported by WTOP.
  • The National Restaurant Association projects $1.55 trillion in 2026 industry sales, but only 1.3% real growth. Sales are not the problem for most owners. Margin per ticket is.

Raising every price by the same amount is the blunt answer. Menu engineering tells you which prices can move, which dishes need a smaller portion or a cheaper side, and which ones to stop making.

What Data Do You Need for a Menu Engineering Matrix?

Three numbers per item, for one menu section at a time. Pizzas go against pizzas, not against desserts, because the analysis works best when comparing similar items:

  1. Units sold over a normal period. Skip weeks with holidays, closures or a one-off catering order that would skew the count.
  2. Menu price as the guest pays it. If your delivery menu has different prices, run it as its own matrix.
  3. Plate cost: the cost of every ingredient on the plate, including the side, sauce and garnish. This is the number most owners guess, and the one worth getting right.

Units sold should come from your POS or your online ordering dashboard, not from memory. If you take orders in two systems, add them together first. Owners who use DevourNow's POS integration ($99 setup, then $29 a month) get direct online orders flowing into the same register as walk-in sales, which makes this step a lot faster.

How Do You Calculate Contribution Margin with the Menu Engineering Formula?

Contribution margin is menu price minus plate cost. Then you need two lines to compare every item against:

  • Popularity line = (1 / number of items) x 70%. For six items, that's 16.7% x 0.7 = 11.7% of section sales.
  • Profit line = total contribution margin for the section / total items sold.

Here is a worked example using six 12-inch pizzas at the real prices on Abe's Pizza's online menu in Athol, Massachusetts. The plate costs and units sold are illustrative, not Abe's data, so swap in your own.

12" pizzaPricePlate costMarginSoldMixGroup
Cheese$11.99$3.20$8.7930030%Plowhorse
Pepperoni$14.00$3.60$10.4026026%Plowhorse
Meat Lover$19.50$6.80$12.7015015%Star
BBQ Chicken$16.50$4.60$11.9012012%Star
Hawaiian$15.50$4.10$11.4010010%Puzzle
White Shrimp$19.50$9.40$10.10707%Dog

Total margin across the 1,000 pizzas sold is $10,521, so the profit line is $10.52. The popularity line is 11.7%. The two pizzas that sell most both earn below the line. The cheese pizza leaves $8.79, which is $1.73 short. That's the Sunday-night surprise from the opening.

How Do You Read the Menu Engineering Matrix?

Put each item in one of four boxes, then act on the box, not on your feelings about the dish.

GroupPopularityMarginWhat to do
StarAbove lineAbove lineKeep the recipe and portion steady. Give it the best spot on the menu and the online ordering page.
PlowhorseAbove lineBelow lineRaise the price in small steps, trim plate cost, or pair it with a higher-margin side.
PuzzleBelow lineAbove lineRename it, rewrite the description, add a photo, move it higher, or have staff mention it.
DogBelow lineBelow lineTry one fix for a set period. If nothing changes, take it off.

In the example, a $0.50 increase on the cheese pizza only lifts its margin to $9.29, still below the line. The bigger lever is moving guests up one step. Pepperoni costs the guest $2.01 more and, at these plate costs, leaves $1.61 more. Hawaiian is a Puzzle: it already earns $11.40 per pie and just needs to be seen. White Shrimp has the highest plate cost, so test a smaller shrimp portion before taking it off.

What Does a Menu Engineering Example Look Like on a Real Pizzeria Menu?

Abe's Pizza's live menu shows a problem many independents share: big sections. We counted 25 subs, 28 wraps and 29 club wraps. That changes the math:

  • With 29 items, the popularity line is (1 / 29) x 70% = 2.4% of club wrap sales. Almost any wrap that sells at all clears it, so the test stops separating winners from the rest.
  • Many fillings repeat across formats. Steak bomb, chicken cutlet parm and teriyaki steak cheese each appear as a sub, a wrap and a club wrap. A slow club wrap can still be a strong filling.
  • Prices follow the format, not the filling: most club wraps are $15.99 to $18.99, while the plain wraps start at $12.99.

For a menu built like this, run the matrix twice: once by format (sub, wrap, club wrap) and once by filling across all formats. Cut or merge where a filling is slow in every format, and keep the formats where it earns the most. On an online menu you can also show the top 8 to 10 items first in a long section, so guests don't scroll past your Stars and Puzzles.

See which dishes sell on your own ordering page, with 0% commission on every direct order.

Explore Menu Optimization

Which Menu Optimization Changes Should You Make First?

Start with the changes that cost nothing and are easy to undo:

  • Fix the Plowhorses with plate cost before price. A smaller portion of fries or a cheaper bun no one notices often beats a price rise everyone notices.
  • Move Puzzles up. On a printed menu, put them first in the section or in a box. Online, use them as the section's lead photo.
  • Rewrite Puzzle descriptions. Name the ingredient that makes it worth ordering, not just "house sauce".
  • Test one Dog fix at a time. A new name or a lower plate cost, never both at once, so you know what worked.
  • Change prices on the menu items guests compare least. Raising the item every regular knows the price of gets noticed first.

If you list your menu on Google, the same logic applies there. A Puzzle hidden on page three of a PDF menu won't get found. A free visibility audit shows whether your menu is readable to Google and AI search at all.

How Does Online Ordering Data Make Restaurant Sales Analytics Easier?

Every online order is already itemized and timestamped, so you don't need to count paper tickets. An online menu is also the easiest place to test: you can move a Puzzle to the top or rewrite a description in minutes, then compare the next four weeks with the last four.

The catch is that third-party apps keep most of that data. Orders on your own commission-free ordering page stay with you, and menu and sales analytics are included on DevourNow's Plus and Zero plans. Direct orders also pay out weekly, so a margin change shows up in your bank account, not just on paper.

If you join Family Club, your Stars and Puzzles are the dishes to put in front of local households planning weekly meals.

How Often Should You Re-Run Your Menu Engineering?

Once a quarter is a sensible rhythm for most independents, and again after any menu change or supplier price jump. With USDA forecasting another 3.5% rise in restaurant prices for 2026, a plate cost that was right in spring can be wrong by autumn. Keep the spreadsheet, swap in new numbers, and see which items moved boxes.

For more ways to find where margin leaks out, read how to increase restaurant sales.

Key Takeaways

TL;DR
  • 3 inputs per dish: units sold, menu price, plate cost.
  • 2 tests: menu mix of at least 70% of an equal share (11.7% in a 6-item section), and margin above the section average ($10.52 in the pizza example).
  • 4 groups, 4 actions: keep Stars, reprice Plowhorses, promote Puzzles, test then cut Dogs.
  • $1.73: how far the best-selling cheese pizza sits below the profit line, more than a $0.50 price change can close.
  • 2.4%: the popularity line in a 29-item section, too low to separate winners, so split long sections or score by filling.
  • Every 3 months, or after any supplier price change: re-run the matrix while USDA forecasts 3.5% menu inflation for 2026.

FAQs

Is food cost percentage or contribution margin better for pricing menu items?

Use contribution margin to sort dishes and food cost percentage as a second check. A $28 steak at 38% food cost leaves $17.36, while a $9 side at 25% leaves $6.75. The steak has the "worse" percentage but pays more of your rent.

Should labor cost be part of menu engineering?

The classic method uses food cost only. Add a note on prep time anyway. A dish with a higher plate cost can still be worth keeping if it takes almost no labor, and a cheap dish that ties up a cook for ten minutes may cost more than it looks.

Do delivery app commissions change which box a dish lands in?

Yes. A commission is taken from every order, so the margin on a delivery app sale is lower than the same dish sold in the dining room. A Star at the table can drop to a Plowhorse or Dog on an app. Run delivery as its own matrix with the commission subtracted.

What if a Dog is a dish my regulars love?

Keep it if it brings in guests who order other, more profitable items, such as a kids' meal that brings the family. Just make that a decision, not a habit, and check the full ticket.

Does menu engineering work for small menus?

It works with as few as three or four items per section. With fewer items, the popularity line is higher, so one slow week can move a dish between boxes. Use a longer sales period to smooth that out.

Sources

Stan Liachev, PhD
Written by
Stan Liachev, PhD
Founder & CEO, DevourNow | Founder, WorldApp Inc.

Stan Liachev, PhD, brings over 30 years of software leadership to restaurant technology via DevourNow. He champions local and small-chain restaurants into carving an online space of their own, guiding them to tap into the modern customer base through expert tools, so they don't lose out on commissions.

AI-native restaurant growth platform

See your restaurant's AI & Google visibility score.

Run a free 120-point audit and discover exactly what's costing you orders - then fix it with DevourNow's commission-free growth platform.

The Growth Menu · Weekly

One restaurant-growth playbook in your inbox every Thursday.

Join restaurant operators getting the tactics that lift orders, cut commissions, and win AI search - no fluff.