Menu Engineering Template (2026) — The 4-Quadrant Framework Operators Actually Use
A menu engineering template that classifies every item into stars, plowhorses, puzzles, and dogs using contribution margin and item popularity. Includes the formulas, target food cost percentages, and the design moves to make once you know which quadrant each dish lives in.
Last reviewed: 2026-05-07
What you get
- •The 4-quadrant menu matrix with the exact thresholds that separate each category
- •Food cost % and contribution margin formulas with worked examples on real menu items
- •A target food cost benchmark by category (proteins, sides, desserts, beverages)
- •A redesign cadence — when to re-engineer, what to change, what to leave alone
- •5 psychological pricing tactics restaurants overuse and 3 that actually move spend
- •A layout cheat sheet built from eye-tracking studies on real menus
PDF version coming soon — bookmark this page for the full template content.
How to use this template
- 1Pull 90 days of POS sales data — quantity sold and price for every menu item.
- 2Calculate contribution margin per item: menu price minus theoretical food cost.
- 3Plot each item on the 4-quadrant matrix using the popularity and margin thresholds below.
- 4Make one design change per quadrant per cycle. Do not redesign the whole menu at once — you lose the ability to attribute the lift.
- 5Re-run the analysis 60 days after each change and update the matrix.
- 6Repeat quarterly. Menus drift; engineering is not a one-time exercise.
The template
Section by section. Read it once, then write your version under each heading.
The 4-quadrant matrix
Two axes. Item popularity (above or below mix-share average) on one axis, contribution margin (above or below average) on the other. Every item lands in one of four cells.
Mix-share average = 70% × (1 / number of items in category). The 70% factor is the standard menu-engineering benchmark — it accounts for the natural skew where some items always over-perform.
- •Stars (high popularity, high margin) — protect, feature, never discount
- •Plowhorses (high popularity, low margin) — re-cost, re-portion, or raise price 8–12% with caution
- •Puzzles (low popularity, high margin) — reposition, rename, move on the menu, photograph well
- •Dogs (low popularity, low margin) — remove unless they have strategic purpose (allergen coverage, kids menu)
Food cost percentage — the formula and the benchmarks
Food cost % = (cost of ingredients per portion / menu price) × 100. Calculate per item, then weight by sales mix to get the menu-wide blended food cost.
Worked example: a chicken biryani with $3.40 of ingredients sold at $14.95 has a food cost of 22.7%. A grilled salmon with $7.80 of ingredients at $26.95 is 28.9%.
- •Proteins (entrees): 28–34% target
- •Sides and starters: 20–26% target (highest margin lever on the menu)
- •Desserts: 18–24% target — most operators under-price desserts by 15–25%
- •Beverages (non-alcoholic): 12–18% target
- •Beverages (alcoholic): 18–22% beer, 22–28% wine by glass, 16–20% liquor
- •Blended menu food cost target: 28–32% for full-service, 30–34% for fast-casual
Contribution margin — why dollars beat percentages
Contribution margin (CM) is menu price minus food cost in dollars. CM is what funds rent, labor, and profit. A 22% food-cost item at $9 ($7.02 CM) loses to a 32% food-cost item at $24 ($16.32 CM) every single time.
When you re-engineer the menu, optimize for total contribution margin per cover, not blended food cost percentage. The two pull in different directions and the dollar number is the one that hits the bank.
Redesign cadence
Quarterly mini-engineering: re-cost the top 20 items, adjust prices on items where ingredient cost has moved more than 5%, and reposition any item that has shifted quadrant.
Annual full re-engineer: full design refresh, font and layout review, photography update for hero dishes, removal of bottom 10–15% of items by mix-share. New menus also need a 60-day post-change analysis to check the lift was real.
Pricing psychology — what works and what is decoration
Working tactics: omit the dollar sign, end prices in 5 or 9 (not .99), use bracketed pricing (one premium item to anchor the rest of the category higher), and place the highest-margin item as the second item in each section, never first.
Decoy items work: introduce a third option priced 20–30% above the item you want to sell. The middle option's perceived value rises and its mix-share goes up 8–15% within 30 days, repeatedly, in published industry studies.
Tactics that don't move much: price descriptions ('only $14'), color highlights, font weight changes alone. They look like work; they don't move the matrix.
Eye-tracking layout principles
On a single-page menu, the diner's eye lands first on the upper-right quadrant, then upper-left, then sweeps down. Place stars and high-margin puzzles in the upper-right.
On a folded two-page menu, the upper-right of the right-hand page (the 'sweet spot') gets 30–40% more attention than any other location. Reserve it for the highest-CM dish you want to sell.
Avoid leader dots ($14......). They train the eye to scan the price column, which collapses ordering decisions to price comparison and flattens your CM mix.
Pro tips
- 1Never raise prices on plowhorses by more than 12% in a single cycle. Above that, mix-share collapses and the higher price loses you more revenue than it earns.
- 2Photograph stars, never dogs. Photos lift item mix-share by 25–35% — point them at items that already have margin.
- 3Re-engineer one category at a time. Touching the whole menu at once destroys your ability to attribute the change.
- 4If a dog has been on the menu for more than 12 months and has fewer than 5 orders per week, remove it. Long tail menu items eat prep labor and waste.
- 5Calculate portion cost from your invoice prices, not theoretical recipe costs. Recipe books over-state yields by 8–12% versus what actually plates.
- 6Add a single item priced 35–50% above your category average. It rarely sells but anchors the rest of the section higher in the diner's mind.
- 7Track ingredient prices monthly, not quarterly. A 2026 protein market that moves 4 points in a month is normal — quarterly tracking means you re-cost too late.