Written by: JJ Tan, Founder, Jelly
Key Takeaways
- The menu engineering matrix classifies dishes into Stars, Plowhorses, Puzzles or Dogs using contribution margin in pounds and menu mix against a 70%-of-equal-share popularity threshold.
- Contribution margin (selling price minus food cost) is the key profitability metric, and focusing on food cost percentage alone can reduce total profit.
- Re-running the matrix every 60–90 days, and monthly when prices move quickly, keeps dishes in the right quadrant as costs change.
- Keeping accurate, up-to-date cost and sales-mix data is the main operational challenge for UK restaurants using the matrix.
- Jelly automates invoice scanning, live dish costing and POS sales-mix integration so the matrix stays current without manual spreadsheets.
Keep Your Matrix Current Automatically
What Is The Menu Engineering Matrix?
The menu engineering matrix is a dish classification framework developed by Michael Kasavana and Donald Smith at Michigan State University in 1982. It evaluates every item within a single menu category on two dimensions: profitability, measured by contribution margin in pounds rather than food cost percentage, and popularity, measured by each dish’s share of units sold.
The result is a two-by-two grid that guides concrete menu decisions. It shows what to protect, reprice, reposition or remove. Kasavana summed up the logic clearly when he wrote that “perhaps the oldest misconception in food service management is that food cost is directly related to profitability.”
How The Four Menu Engineering Quadrants Work
Each dish in a menu category falls into one of four quadrants based on whether its contribution margin and menu mix sit above or below the category thresholds. The list below shows what each quadrant means in practice, with a UK example and the action it calls for. A dish can move between quadrants as costs or demand change, which is why the matrix needs regular updates.
- Star has high contribution margin and high popularity. Example: a well-priced weekend special that sells out every service. Action: protect quality, give prime menu position, and test a modest price rise.
- Plowhorse has low contribution margin and high popularity. Example: a popular pub classic such as fish and chips with a thin margin due to ingredient cost. Action: re-cost the plate, trim garnish, or apply a small price increase.
- Puzzle has high contribution margin and low popularity. Example: a well-margined lamb shank that rarely gets ordered. Action: improve menu description and position, rename if needed, and have front-of-house recommend it verbally.
- Dog has low contribution margin and low popularity. Example: a slow-selling, low-margin side dish with its own specialist ingredient. Action: remove or rework unless it serves a strategic purpose, such as being the only vegan main.
How To Calculate Contribution Margin For A Menu Item
Contribution margin (CM) is the money a dish leaves after its food cost is deducted from its selling price. It drives menu profitability decisions because, as the Kasavana method states, you deposit cash, not percentages.
Contribution margin = selling price − food cost
Food cost % = food cost ÷ selling price × 100
Worked example: an £18 main with £5.40 in food cost produces a contribution margin of £12.60 and a food cost percentage of 30%. A different dish priced at £12 with £3.00 in food cost has a lower food cost percentage (25%) but contributes only £9.00, which is £3.60 less per plate. Focusing on food cost percentage alone would favour the cheaper dish, while contribution margin favours the £18 main. A dish with a 40% food cost and £12 margin generates more profit than a dish with a 25% food cost and £6 margin.
For a deeper look at building recipe cost cards in a spreadsheet, see Menu Profitability Spreadsheet: Excel Food Cost Guide.
How To Calculate Menu Mix And Popularity
Menu mix measures each dish’s share of total units sold within its category.
Menu mix % = item units sold ÷ total category units sold × 100
The Kasavana–Smith popularity threshold defines a dish as popular when its menu mix reaches at least 70% of the share it would receive if sales were spread evenly across all items in the category:
Popularity threshold = (100% ÷ number of items) × 70%
For a category with 8 dishes, the equal share is 12.5% and the threshold is 8.75%. The 70% factor exists because real menus never sell every item in perfectly equal proportion; a strict 100%-of-equal-share benchmark would flag almost every item as underperforming.
Worked arithmetic: if a category sells 1,500 mains across 8 dishes, the equal share is 187.5 portions per dish. Multiply by 0.70 and the popularity threshold is 131.25 portions. Any dish selling fewer than 131 portions in the period is classified as low popularity.
The profitability threshold is the sales-weighted average contribution margin: total CM across all dishes divided by total units sold. This distinction matters. In an eight-item example with £14,953 total CM across 1,730 units, the weighted average is approximately £8.64 per dish, which differs from the unweighted mean of the eight individual margins. Competitors and generic guides routinely gloss over this step.
Worked Example: Classifying A Menu Into The Matrix
This example applies the formulas to eight named mains over one calendar month. The popularity threshold is 131 portions, as calculated above. The figures are illustrative and should be replaced with your own POS and invoice data.
Across the eight dishes, the weighted average contribution margin is about £8.64. Roast Chicken and Beef Burger sit well above this line and sell strongly, so they classify as Stars. Fish and Chips, Mushroom Risotto and Veggie Wrap all sell in high volumes but sit close to the CM threshold, so they fall into Plowhorse territory.
Lamb Shank and Pan-Fried Sea Bass earn strong margins but sell in low volumes, which makes them Puzzles. Prawn Cocktail sells slowly and carries a low margin, so it sits in the Dog quadrant.
The Veggie Wrap sits just below the CM threshold but above the popularity line, which makes it a borderline Plowhorse. A borderline classification based on one month of data is rarely enough justification for removing an item. It is a better candidate for a small price or recipe adjustment.
Read across the pattern and three of the eight mains are Plowhorses, with two of them within roughly £1 of the £8.64 CM threshold. That cluster is where a modest re-cost or price adjustment can move the most dishes into Star territory, which is a more productive first move than tackling the two Puzzles. For a downloadable worksheet to run this calculation across your own menu, see Menu Profitability Spreadsheet: Excel Food Cost Guide. For broader strategy on using the matrix to increase gross profit, see Menu Engineering Strategies To Increase Gross Profit.
How Often Should You Re-Run Your Menu Engineering Matrix?
The matrix gives a snapshot of current performance, and it goes stale as soon as a supplier invoice lands with a new price. Prices in the restaurants and hotels division rose 3.8% in the 12 months to September 2026, and individual ingredient movements can be far sharper. Some suppliers have raised prices by up to 40%, driven by energy and transport costs throughout the supply chain. A 5% rise on a key protein takes money off a dish’s contribution margin without any change to the menu itself.
A 2% rise in food costs can eliminate up to 50% of profit margin if not offset by pricing or menu adjustments, and by the time a quarterly review surfaces the problem a formerly profitable dish may have traded at a loss for 90 days.
Recommended practice is to re-run the matrix every 60–90 days at minimum, and monthly when ingredient prices are volatile or a menu change is imminent, with an additional pass after any notable supplier price or menu change. A Star in January may become a Plowhorse by July if protein costs have moved. That cadence is only achievable if the matrix is fed by live invoice-level costs and POS sales mix, rather than manual spreadsheet entry last updated months ago.
Keep Your Matrix Current Automatically
Keeping Menu Engineering Matrix Profitability Live With Jelly
That cadence only works if the cost data is live, which is the problem Jelly is built to solve. It keeps dish costs and margins current from invoice data automatically, so the matrix reflects this week’s supplier prices rather than last quarter’s.
The foundation is automated invoice scanning. Capture invoices via email or photo and Jelly digitises every line item, including quantity, SKU, price and tax, without manual entry. Because ingredient costs update with every new invoice, live dish costing keeps the gross profit margin for every dish current. A red percentage appears when a dish drops its margin, and a green one when it improves. Work that used to take 28 minutes to cost in a spreadsheet takes about 3 minutes in Jelly’s Kitchen section.
The Price Alert feature flags every supplier price increase or decrease, by how much and from which supplier, so chefs have the evidence to negotiate better rates or claim credit notes before a margin problem compounds. The Flash Report then shows the effect on gross profit margin daily, weekly or monthly, calculated from invoice costs and POS sales, so the impact is visible the same week rather than at month-end.
The Sales Mix / Menu Engineering feature pulls item-level sales directly from POS integrations with Square, EPOS Now, Lightspeed and Toast, and shows which dishes are most popular and most profitable in real time. Connecting any supported POS takes about five minutes, and replaces 2–5 hours of weekly manual work. One operator improved gross profit from 65% to 72% within 12 weeks on approximately £500,000 in revenue. Populu lifted GP from 68% to 72% across 16 locations.
Stuart Noble, Head Chef at Cairn Lodge Hotel, put it directly: “Price hikes were crushing our margins — I felt helpless. With Jelly, every dish cost is up-to-date at my fingertips. We slashed food costs by 5% in a month.” Murat Kilic, Chef-Owner at Amber in East London, saves £3,000–£4,000 per month and says simply: “Jelly keeps my business alive.”
NetSuite is a broad, generic business management suite whose menu engineering content is largely theory-led, and its restaurant operations module requires separate configuration of POS menu item linking and recipe records, plus manual data entry, to produce dish-level costs. It is not designed around the invoice-to-dish-margin workflow that hospitality kitchens need. Resy is a reservation-led platform rather than a costing engine. Its own guidance recommends operators use dedicated inventory tools that upload invoices directly and provide real-time recipe management, functionality that Jelly delivers natively.
Jelly closes the loop from invoice line item to live dish margin automatically, at a flat £129/month per location with first-week onboarding. For a full comparison of menu engineering software options in the UK market, see Menu Engineering Software Comparison UK 2026.
Conclusion
The menu engineering matrix rests on three mechanics: contribution margin in pounds, menu mix against the 70%-of-equal-share popularity threshold, and the four quadrants of Stars, Plowhorses, Puzzles and Dogs, each with a clear decision rule. The arithmetic is straightforward. The operational challenge is keeping the cost data behind it accurate as supplier prices move week to week.
With ingredient costs moving as sharply as the inflation figures above show, a matrix built on stale spreadsheet data only tells you what used to be true. As the cadence section set out, that means re-running the matrix every 60–90 days at minimum, and monthly when prices are moving, with an extra pass after any major supplier or menu change.
Jelly closes that gap by keeping costs and sales mix live, as set out above, so the matrix reflects today’s prices rather than last quarter’s. That combination of live data and clear decision rules turns the matrix into a practical management tool for UK restaurants, pubs and boutique hotels.
Frequently Asked Questions
What Is The 70% Popularity Threshold In Menu Engineering?
The 70% popularity threshold is the minimum menu mix percentage a dish must reach to be classified as popular in the Kasavana–Smith matrix. It is calculated as 100% divided by the number of dishes in the category, multiplied by 0.70. For an eight-dish category, the equal share is 12.5% and the threshold is 8.75%.
The 70% factor exists because real menus never distribute sales perfectly evenly, so a strict 100%-of-equal-share rule would classify almost every dish as underperforming. The threshold is a managerial convention rather than a statistically validated universal benchmark, and it shifts whenever dishes are added to or removed from the category being analysed.
How Often Should I Re-Run My Menu Engineering Matrix?
Most operators should re-run the matrix every 60–90 days, and move to a monthly cadence when ingredient prices are volatile or a menu change is imminent. Add an extra pass immediately after any significant supplier price change, menu change or seasonal transition.
Ingredient costs move continuously, and a dish’s contribution margin can fall without any change to the menu itself. Waiting for a quarterly or annual review means a formerly profitable Star can trade as a Plowhorse for months before anyone notices. Operators with high purchasing volatility, or those in the middle of a menu change, should review more frequently. The cadence is only practical when cost data is fed automatically from live invoices rather than manually updated spreadsheets.
What Should I Do When A Supplier Price Rise Turns A Star Into A Plowhorse?
The first step is to quantify the impact. Calculate the new contribution margin using the updated ingredient cost and compare it against the current weighted average CM for the category.
If the dish has dropped below the average, the right response depends on why. When demand is strong and the price sits comfortably within the range for that category, a modest price rise is usually the cheapest fix, and you can test whether volume holds. If the dish is already near the top of its price band, focus on reducing plate cost through recipe, portion size or garnish instead.
Negotiating with the supplier or switching ingredient takes longer, and works best when the affected protein drives several dishes at once. A dish that has been a Star for years deserves a considered response, but it also cannot be left to erode margins indefinitely. Price Alert tools that flag the change the day the invoice lands give operators the reaction time to act before the damage accumulates.
Who Should Own The Menu Engineering Matrix — The Owner Or The Head Chef?
Both roles need access to the matrix, but they use it differently. The owner or finance manager needs the headline GP margin, the quadrant classifications and the trend over time. That information informs pricing strategy, investment decisions and multi-site comparisons.
The head chef needs the dish-level detail. They focus on which ingredients drive cost, which dishes are borderline, and where a recipe adjustment or supplier negotiation would have the most impact. In practice, the matrix works best when it is owned jointly. The chef supplies accurate recipe data and flags supplier changes, and the owner or operations manager sets the action priorities per quadrant.
A shared platform where both roles see the same live figures, rather than the chef working in a spreadsheet and the owner waiting for an accountant’s report, removes the information lag that causes most menu engineering to fail in execution.
How Does Jelly Keep The Matrix Current?
Jelly keeps the menu engineering matrix current by automating the two inputs it depends on: dish costs and sales mix. Every supplier invoice is scanned and digitised to line-item level, including quantity, SKU, price and tax, without manual entry.
When a new invoice arrives with a changed ingredient price, every dish recipe that uses that ingredient is re-costed immediately, and the gross profit margin updates in real time. The Price Alert feature flags every price movement so operators can act on it the same week rather than discovering it at month-end.
On the sales side, Jelly integrates directly with Square, EPOS Now, Lightspeed and Toast via real-time API, and pulls item-level sales data the moment a transaction completes. The Sales Mix / Menu Engineering feature combines live costs and live sales to show which dishes are most popular and most profitable at any point, the same two dimensions the Kasavana–Smith matrix requires, without any manual assembly.
The result is a matrix that reflects today’s supplier prices and today’s sales, rather than last quarter’s spreadsheet.
