Menu prices set by gut feel leak margin every shift. Here is the operator’s method: true plate costs, contribution margin, and the four-quadrant matrix that tells you what to promote, reprice, or retire.
July 16, 2026
Most menus are priced by feel — a glance at what the place down the street charges, a rough tripling of ingredient cost, a number that “sounds right.” That worked when food costs were stable. It does not survive the way ingredient prices have moved over the last few years. Menu engineering is the fix: a structured look at what each item actually costs, what it actually earns, and how often guests actually order it, so you can decide what to promote, reprice, rework, or retire. None of it requires a consultant. It requires accurate plate costs, a sales-mix report from your POS, and an hour with a spreadsheet — or a platform that does the spreadsheet part for you. Here is the whole method, the way an operator would run it.
Everything downstream depends on knowing what a dish costs to serve — not roughly, exactly. That means costing the full recipe: every ingredient at current invoice prices, plus the garnish, the bread and butter, the to-go container, the ramekin of sauce nobody writes down. Update those costs when invoices move, not once a year. If your ingredient prices live in the same system that tracks your recipe-level inventory, plate costs update themselves as invoice prices change; if not, block an afternoon each quarter and cost every item by hand. Operators are consistently surprised twice: once by the dish that costs far more than they assumed, and once by the humble item quietly running the best margin in the building.
The classic rule of thumb says target a food cost percentage somewhere in the high twenties to low thirties, and as a portfolio average that is a reasonable sanity check. The mistake is managing individual items to it. You bank dollars, not percentages. A pasta dish at a lean food cost percentage might contribute a few dollars per plate; a steak at a much higher percentage can contribute several times that. Chasing percentages tells you to push the pasta. Contribution margin — menu price minus plate cost, in dollars — tells you to sell the steak. Percentage keeps you honest at the category level; contribution margin makes the item-level decisions.
Classic menu engineering, formalized by hospitality researchers in the early 1980s, plots every item on two axes: popularity (how often it sells relative to its category) and profitability (contribution margin versus the category average). Two axes make four quadrants, and each quadrant has a job:
| Quadrant | Profile | What to do |
|---|---|---|
| Stars | Popular and high margin | Protect them: keep quality consistent, feature them, never let them slip |
| Plowhorses | Popular but low margin | Reprice gently, trim plate cost, or re-portion without hurting value |
| Puzzles | Profitable but slow | Rename, reposition, have servers talk about them, move them on the page |
| Dogs | Unpopular and low margin | Retire them or reinvent them completely — sentiment is not a strategy |
Pull a sales-mix report for the last full quarter — every item, units sold, menu price. Add plate cost for each item and compute contribution margin. Within each category, mark items above or below the category average on both popularity and margin; that sorts everything into the four quadrants. Then make one decision per item — protect, reprice, reposition, or retire — and write it down with a date so the next pass can measure what happened. The whole exercise fits in an afternoon, and the second pass is faster because the plumbing already exists. If the spreadsheet feels heavy, this is precisely the report a modern restaurant platform should hand you out of the box.
A handful of presentation choices reliably help. Dropping the dollar sign softens the “paying” signal — a line that reads 24 lands easier than one shouting $24.00. Skipping the neat price column keeps eyes on descriptions instead of turning the page into a price-shopping ladder. A deliberately premium item at the top of a section anchors everything below it as reasonable. Boxes, whitespace and first-and-last placement steer attention toward the items you want to sell — spend them on stars and puzzles, not on whatever the kitchen likes cooking best. And write descriptions that earn the price: origin, technique, specificity. None of this rescues a badly costed menu, but layered on sound math it is free margin.
Your menu now lives in more places than a laminated page. Orders through your own commission-free online ordering carry no commission, so direct digital prices can simply match your dine-in menu — and still earn more per order than any marketplace sale. Marketplace listings are a different animal: after a large commission comes off the top, list-price parity means every marketplace order sells at a steep effective discount. Many operators run modestly higher prices on third-party apps to claw some of that back — we walk through the arithmetic in commission-free vs third-party delivery. Whatever you decide, decide it deliberately, per channel, and revisit it whenever commissions or volumes change.
Menu engineering combines sales mix and plate cost. On Novaryq, sales, menus and recipe costs share the platform across locations, and multi-location menu management lets operators test a price change at one site before broader rollout. Software plans are priced per location and separate costs remain disclosed; a month-to-month option is available. See pricing.
It is the practice of analyzing every menu item on two axes — how often it sells and how much contribution margin it earns — and then acting on the result: promoting stars, fixing low-margin favorites, repositioning profitable slow movers, and retiring items that do neither. It was formalized in the early 1980s and still works because it is disciplined arithmetic on data your POS already collects.
Most full-service operations land somewhere between the high twenties and mid thirties as an average, but the range varies widely by concept — a steakhouse naturally runs higher than a pizzeria. Treat the percentage as a portfolio-level sanity check and manage individual items by contribution margin in dollars, because dollars, not percentages, pay the rent.
Review your sales mix monthly and run a full menu engineering pass quarterly. Reprice little and often as ingredient costs move rather than saving everything up for one large, guest-visible jump. Digital menus make small adjustments easy to ship immediately, which is one more reason to keep direct online ordering under your own control.
Many operators price marketplace menus modestly higher to offset commissions, which is generally permitted, though rules vary by platform and region — check current marketplace policies before you change anything. Orders through your own commission-free ordering site can stay at dine-in prices, since no commission comes off the top.