Insights · Menu Engineering

Stars, plowhorses, puzzles, and dogs.

October 23, 2026 · Brigade Consulting Group

Every dish on your menu answers two questions: does it sell, and does it earn? Most owners can answer the first from memory. The second is where menus quietly bleed.

Menu engineering is the discipline of sorting every dish by those two measures — popularity and contribution margin (the menu price minus what the plate actually costs to make) — and then acting on what the sort tells you. The method comes from Kasavana and Smith, and it sorts your menu into four crews:

The four crews

  • Stars — sell well and earn well. Protect them. Feature them. Never let a supplier change quietly cheapen a star.
  • Plowhorses — sell well, earn thin. Your crowd loves them, your margins don’t. The fix is cost engineering: re-spec an ingredient, tighten the portion build, or nudge the price a dollar and watch.
  • Puzzles — earn well, sell poorly. Good dishes hiding in bad real estate. Rename them, move them to where eyes land first, put them in your servers’ mouths at pre-shift.
  • Dogs — neither sell nor earn. Cut them, or rebuild them into something else. Every dog on the menu is holding a slot a star could be working.

The honest way to draw the lines

Popularity: a dish is a strong seller if it beats its fair share. Take 1 divided by the number of dishes on the menu, then multiply by 0.7 — that’s the expected share. Beat it, and the dish is popular. Fall short, and it isn’t, whatever your regulars tell you.

Margin: compare each dish’s contribution margin to the sales-weighted average across the whole menu. Above the line earns; below it doesn’t. No dish gets graded on a curve of your affection.

A worked example

Sample numbers, from a twenty-item dinner menu — illustrative, not a client’s:

  • Seared salmon — $28 price, $8.40 plate cost, $19.60 margin, 210 sold last month. Margin above average, popularity well above its fair share. Star. It stays, and it gets the top-right corner of the menu.
  • The house burger — $16 price, $6.90 plate cost, $9.10 margin, 480 sold. It outsells everything and earns below the line. Plowhorse. Don’t touch the recipe your guests love — re-spec the cheese, renegotiate the bun, and test $17.
  • Duck confit — $32 price, $11.00 plate cost, $21.00 margin, 40 sold. Best margin on the menu, nobody orders it. Puzzle. It isn’t a kitchen problem; it’s a placement and description problem.
  • Meatloaf plate — $17 price, $9.20 plate cost, $7.80 margin, 55 sold. Below the line on both counts. Dog. It comes off, and the salmon’s line position gets stronger without it.

One afternoon with ninety days of POS sales and honest recipe costs gets you this sort. The discipline is in doing it quarterly — menus drift just like portions do, and a star in January can be a dog by summer.

Want your menu sorted?

Our Menu Engineering engagement costs out every dish, sorts your stars from your dogs, and hands you a trimmed menu draft ready for print — with the price moves spelled out.

Book an Inspection

← All Insights