Run menu engineering six weeks before the new menu prints: recost, classify every dish, decide what changes and project the margin before you commit.
By Rob Bruni, founder of OpsEdge Hospitality, twenty years running restaurants, bars, clubs and resorts. Updated November 2026.
Every menu change is a chance to fix the margin of the whole menu, and most restaurants waste it by deciding what changes on taste and habit, then pricing it at the last minute. Running menu engineering before you print turns a seasonal change into the highest-return hour of the quarter. This guide walks through a spring menu change step by step: pulling the data, recosting every plate, classifying the menu, deciding what changes, projecting the result before you commit, laying out the page and measuring after launch, with a worked example of an eight-entree menu.
Once a menu is printed, every price, dish and position is locked in for three months or more. If the analysis happens afterward, it can only tell you what you should have done. Run it six weeks before the change and it shapes every decision: which dishes stay, which get repriced or re-specced, which new dishes earn their place, and where each one sits on the page. The menu engineering guide covers the method itself; this is how to use it on a real menu change.
Pull at least eight weeks of item sales from the POS, by category, so one busy weekend does not skew the mix. Then recost every plate at today's invoice prices, with the right yields. Costs drift between menu changes, and an analysis on last season's costs gives last season's answer. The recipe costing guide and the recipe yield guide cover how.
Here are eight entrees at a full service restaurant, eight weeks of sales, with every plate recosted:
| Entree | Price | Plate cost (recosted) | Margin | Sold in 8 weeks | Mix | Quadrant |
|---|---|---|---|---|---|---|
| Ribeye | $42 | $17.60 | $24.40 | 610 | 12.7% | Star |
| Roast chicken | $26 | $7.30 | $18.70 | 820 | 17.1% | Star |
| Burger | $17 | $6.40 | $10.60 | 1,140 | 23.8% | Plowhorse |
| Salmon | $31 | $11.40 | $19.60 | 380 | 7.9% | Puzzle |
| Short rib | $34 | $12.90 | $21.10 | 700 | 14.6% | Star |
| Mushroom risotto | $23 | $5.10 | $17.90 | 290 | 6.0% | Puzzle |
| Pork chop | $29 | $12.40 | $16.60 | 250 | 5.2% | Dog |
| Fish and chips | $22 | $8.10 | $13.90 | 610 | 12.7% | Plowhorse |
Average margin, weighted by sales, is $17.15. With eight entrees the average mix is 12.5 percent, and we set the popularity threshold at 70 percent of that, 8.8 percent. An item at or above both is a Star; popular but below the margin line is a Plowhorse; high margin but below the popularity line is a Puzzle; below both is a Dog.
The Menu Engineering Matrix and Sales Mix Analyzer takes price, plate cost and units sold for up to 50 items, calculates contribution margin, mix and the thresholds from your own menu, sorts every item into Star, Plowhorse, Puzzle or Dog, and suggests actions for each. Excel and Google Sheets, with a 20-item worked example and companion tools for food and bar menus.
See the Menu Engineering Analyzer Recipe Cost Card SystemFree calculator
Enter up to ten items with price, recosted plate cost and units sold for the same period. It works out the weighted average margin and the popularity line and sorts every item into Star, Plowhorse, Puzzle or Dog. The eight entrees above are loaded.
| Item | Price $ | Plate cost $ | Sold | Margin | Mix | Quadrant |
|---|---|---|---|---|---|---|
Popularity line is 70 percent of an equal share of the mix. Compare items in the same category only, entrees against entrees.
Every item gets one decision, and the quadrant tells you which kind:
Make as few changes as the analysis supports. Changing everything at once makes it impossible to tell what worked.
Before anything goes to print, estimate the new mix and recalculate. Assume the burger loses a few percent of its sales to the price increase, the featured salmon takes some orders from the burger and the fish and chips, and the new dish takes the pork chop's sales plus a few from the roast chicken:
| Current menu | Spring menu (projected) | |
|---|---|---|
| Entrees sold in 8 weeks | 4,800 | 4,766 |
| Average margin per entree | $17.15 | $17.74 |
| Extra margin a year, same traffic | about $18,400 |
Same guests, a few dozen fewer burgers, and about $0.59 more margin on every entree sold. That is the payoff of doing the analysis before the printer. If the projection is flat or negative, revisit the decisions before you commit, not after.
Before launch, the kitchen gets new recipe and plate spec cards for every changed dish, the POS gets the new prices and buttons, and the floor gets a tasting and a short script for the dishes you are promoting. Then measure: run the same analysis on the first four weeks of the new menu and compare it with the projection. The items that moved the way you expected stay as they are; the ones that did not get one more adjustment. Track food cost weekly through the change, because a new menu is when actual and theoretical food cost drift apart.
Five free calculators turn these guides into results in a few minutes: prime cost, food cost, labor cost, break-even and pour cost. Each one shows the gap to the benchmark for your concept in dollars a year. No signup.
Open the free calculators See the benchmarksAbout six weeks. That leaves time to recost, classify the menu, develop and test new dishes, project the new menu's margin and send it to print, rather than analyzing a menu that is already locked in.
At least four to eight weeks of item sales from your POS, by category, so a single busy weekend or a special does not distort the mix. Use the same period for every item you compare.
Raise the margin without breaking what guests like about it: a modest price increase, a tighter portion spec, a cheaper side or garnish, or pairing it with a high-margin add-on. Avoid big changes to the dish itself.
Remove it or replace it with a new dish built to have at least the category's average margin, costed and tested before it goes on the menu. Make sure the replacement does not simply take orders away from your Stars.
Run the same menu engineering analysis on the first four weeks of the new menu and compare average margin per item sold and the mix against your projection. Track weekly food cost through the change as well.
