Written by: JJ Tan, Founder, Jelly
Key Takeaways For UK Operators
- Menu engineering software connects your EPOS to capture real item-level sales data and builds accurate plate costs from actual supplier invoices rather than estimates.
- The four-quadrant matrix classifies every dish by popularity and contribution margin, turning subjective menu decisions into data-driven actions on pricing, placement and removal.
- A repeatable three-layer cadence of weekly price alerts, monthly matrix reviews and quarterly menu changes protects margins as supplier prices fluctuate.
- Common pitfalls include missing invoices, unmapped EPOS items and unclear ownership. Addressing these early keeps the matrix accurate and actionable.
- Jelly automates invoice capture, EPOS integration and margin tracking so UK restaurants, pubs and hotels can cut food costs by 3% and lift gross margins by 2 percentage points within three months.
See Jelly In Action For Your Menu
Before You Begin: Roles, Access And Setup
Gather a few essentials before logging in so the first session runs smoothly.
- Admin access to your EPOS account
- Supplier invoices from the last 4–12 weeks
- Recipe and spec sheets for every dish on the menu
- Your current menu with selling prices
- Access to your accounting software (for example Xero)
Ownership usually splits across two roles. The owner, operations manager or finance manager sets up the software, connects the EPOS and owns the monthly matrix review. The head or executive chef owns recipe building and dish mapping and confirms garnish quantities, wastage percentages and batch recipe yields. Cross-functional input matters at key points, such as a chef confirming that a till item maps correctly to a kitchen dish or a finance lead confirming supplier payment terms before a credit note is raised.
Set realistic expectations before starting. Connecting an EPOS takes approximately five minutes, and initial value from invoice scanning, such as price alerts and spending insights, arrives within the first week once suppliers send invoices to a dedicated email address or within 24 hours of photographing invoices into the software. However, the full matrix is only as good as the data feeding it, so budget time for data hygiene in the first two weeks. Treat this as a recurring operating rhythm with weekly price alerts, a monthly matrix review and quarterly menu changes.
Why This Process Matters For UK Operators
Ingredient prices in the UK move constantly. A dish that was profitable last month can be losing money today, and manual plate-cost calculations become outdated quickly because they cannot update in real time as supplier prices shift. Spreadsheet costing is slow, drifts out of date, and cannot connect to POS systems to pull actual sales data automatically.
The outcomes of doing this properly are concrete. You gain live gross profit visibility by dish, faster reactions to supplier price changes, hard data for supplier negotiations, fewer invoice discrepancies and a menu that is deliberately engineered rather than inherited. Most operators underestimate item food cost by 10 to 15 percent because they cost recipes from supplier price lists instead of from received inventory. The software closes that gap by building costs from actual invoices.
The real obstacle to achieving these outcomes is manual spreadsheets and disconnected invoice folders, not another software platform. Software removes manual data entry and spreadsheet drift while still relying on good operational discipline. For further context on the features that drive this process, see Menu Engineering Software Features That Drive Profit.
Step 1 — Connect Your EPOS And Pull In Sales Data
The goal of this step is to feed real sales data into the software so popularity and sales mix reflect actual transactions.
Open Jelly, go to Integrations, sign in to your EPOS, grant permissions and select which categories to sync, for example food and beverages. Jelly integrates natively with Square, Lightspeed, EPOS Now and Toast via real-time API, delivering item-level sales data the moment a transaction completes. Setup takes approximately five minutes and follows the same flow across all four systems.
The most common friction point is admin access. If you lack admin rights to your EPOS account, the integration cannot be authorised, so sort permissions before starting. Jelly flags this requirement upfront.
Success at this step looks like item-level sales appearing in the software, with new items surfacing only from the point the integration was connected. This keeps the mapping clean and free of legacy menu clutter. Connecting a POS automates 2–5 hours of weekly work to get real-time margins and sales mix data.
Step 2 — Map EPOS Items To Real Dishes
The goal here is to link every EPOS item that represents a sellable dish to a recipe so margin calculations stay accurate.
Work through the POS-to-dish mapping screen and link each EPOS item to its corresponding dish in the software. Because mapping only surfaces items sold since the integration was connected, the list stays manageable. You avoid wading through years of legacy till items.
Have your current menu, your EPOS item names and a chef or manager who knows how each till item corresponds to a kitchen dish. Some till items combine multiple dishes or use shorthand names that do not match the menu, and the chef resolves those ambiguities.
Success looks like every food and drink item sold having a mapped dish. You avoid orphan EPOS items and dishes missing from the sales mix. A simple table showing EPOS item name alongside its mapped dish works well as a working document at this stage.
Step 3 — Build Recipes And Plate Costs From Invoices
The goal is to build accurate dish costs from real supplier prices rather than remembered or estimated ones.
Capture supplier invoices by email or photo into Jelly. The software digitises every line item, including quantity, SKU, price and tax, so ingredient costs update automatically as new invoices arrive. Then build each dish in the cookbook by clicking on ingredients already populated from those scanned invoices. Jelly handles unit conversions and maths instantly. What used to take 28 minutes now takes approximately 3 minutes to cost a menu item.
Small details often drive the biggest gaps. Garnishes, oils, seasonings, wastage percentages and batch recipes all add up. Ingredients that operators most often forget include cooking oil, salt, pepper, stock, sauces used in small amounts and garnishes. A pinch of salt is negligible, but 30ml of stock per portion across 200 covers a week matters. Oils, seasoning and garnish collectively add 5–15% to actual plate cost if not tracked.
A worked example in GBP helps make this concrete. A chicken main sells at £18.50. The ingredient costs are £4.20 for the protein, £0.85 for potatoes, £0.60 for vegetables, £0.35 for sauce and £0.20 for garnish. This gives a plate cost of £6.20 and a contribution margin of £12.30, approximately 66%. If the protein price rises by £0.40 on the next invoice, the plate cost becomes £6.60 and the contribution margin falls to £11.90. Jelly recalculates this automatically when the next invoice is scanned, so the change appears immediately rather than weeks later.
Yield also affects true cost. An ingredient that loses 20% weight during preparation costs 25% more per usable gram than the purchase price suggests. Yield percentage is calculated as usable weight divided by raw weight, and the edible-portion cost per kg is derived by dividing the as-purchased cost per kg by the yield percentage.
Success looks like every dish having a live cost and gross profit margin that updates as invoices arrive, with a red or green indicator showing whether margin has fallen or grown. The following table illustrates the output:
| Dish | Selling Price | Plate Cost | Contribution Margin |
|---|---|---|---|
| Chicken Main | £18.50 | £6.20 | £12.30 (66%) |
| Chicken Main (after £0.40 protein rise) | £18.50 | £6.60 | £11.90 (64%) |
For a deeper look at how invoice automation feeds into this process, see Menu Engineering Best Practices & Tools for UK Restaurants.
Explore Jelly’s Recipe And Costing Workflow
Step 4 — Use The Four-Quadrant Menu Engineering Matrix
The goal is to classify every dish by popularity and profitability so decisions become clear rather than subjective.
The two core metrics are popularity, meaning how many of a dish you sell as a share of total category sales, and contribution margin, meaning what each dish contributes after plate cost is deducted from the selling price. The four-quadrant menu engineering matrix traces back to 1982, when Michael Kasavana and Donald Smith at Michigan State University classified items as Stars, Plowhorses, Puzzles and Dogs. The software plots every dish onto this matrix automatically.
Four Menu Engineering Categories And Actions
The four categories, and the primary action for each, are:
| Quadrant | Popularity | Contribution Margin | Action |
|---|---|---|---|
| Stars | High | High | Protect, keep prominent and avoid unnecessary price changes |
| Plow Horses | High | Low | Re-engineer by adjusting recipe, portion or price carefully |
| Puzzles | Low | High | Promote through better description, placement or staff recommendation |
| Dogs | Low | Low | Consider removing or replacing |
Menu engineering guidance recommends running the analysis by category rather than across the whole menu, comparing starters against starters and mains against mains, because different sections have different sales patterns and price points. Success at this step looks like every dish having a quadrant and a named owner for the action.
Step 5 — Turn Matrix Insights Into Menu Changes
The goal is to turn the matrix into specific menu changes that protect or improve margin.
Four main levers sit within your control, and each one carries trade-offs.
- Price: Plowhorse price corrections work best when small and frequent. Increasing a £12.50 plowhorse to £12.95 often recovers margin with almost no guest noticing. Jumping from £12.95 to £14 in one step usually creates more pushback.
- Description: Item descriptions of 12 to 25 words outperform shorter ones when the language is concrete. “Slow-braised lamb shoulder with smoked aubergine and burnt onion” outsells “lamb with vegetables” by a wide margin.
- Placement: Moving a star into the top-right menu position can lift its mix share by 8 to 15 percent in a quarter.
- Removal: Removing a Dog frees kitchen capacity and menu space. Before deleting, check whether the dish serves a dietary need, such as the only vegan main, because removing it could lose whole tables.
Menu Change Example For A UK Chicken Main
Return to the chicken main from Step 3. After the £0.40 protein price rise, the dish’s contribution margin fell below the category average, which would typically move it from a Star to a Plow Horse. A UK pub operator facing this pattern could apply a modest price rise, re-engineer the recipe through portion or ingredient adjustments or accept the reduced margin temporarily while negotiating with the supplier using Jelly’s Price Alert data. The key is to act on the specific dish that has moved quadrant rather than making broad menu-wide changes.
Success looks like a short list of specific menu changes with an owner and a date. You move from a vague intention to “review the menu” to a clear plan.
For pub-specific implementation detail, see Pub Menu Engineering Best Practices: 7-Step Workflow.
Step 6 — Set A Simple Review Cadence
The goal is to make menu engineering a repeatable operating rhythm rather than a one-off project.
The cadence has three layers:
- Weekly: Check price alerts for ingredient increases. Challenge suppliers or claim credit notes using the flagged data. This acts as the fastest lever for margin protection.
- Monthly: Re-run the matrix and check whether dishes have moved quadrant. Any item whose margin has dropped more than 3 percentage points since the last review needs immediate attention.
- Quarterly: Make menu changes and test whether they worked. A 90-day cadence is optimal because it provides enough sales volume for defensible decisions and enough time to see the impact of the previous quarter’s changes.
How To Test Menu Changes Effectively
Compare sales mix and margin before and after each change. For most physical retail changes, allow a minimum of four weeks, and often six to eight weeks for changes involving significant visible alterations to the store environment, for the change to settle before drawing conclusions. Change only a few things at a time to measure the impact of each one. Avoid re-engineering during a holiday week or a new menu launch because the data is noisy.
Success looks like a recurring calendar block for each cadence and a simple record of what changed and what happened.
Set Up Your Menu Review Rhythm With Jelly
Common Mistakes And Troubleshooting
Dirty data is the single biggest real-world blocker, and most issues trace back to it.
- Recipe costs look wrong. Likely causes include missing or mis-scanned invoices, incorrect pack sizes or unit conversion errors. Unit conversion errors are the most common costing mistake because recipe quantities must be converted to the same unit as the purchase price. Corrective action involves checking the invoice line items, confirming pack size and unit and rebuilding the affected recipe lines.
- Invoices are missing. Likely causes include suppliers not sending invoices to the dedicated email address or paper invoices not being photographed. Corrective action involves setting up a dedicated invoice email, briefing suppliers and photographing paper invoices the same day they arrive.
- EPOS items do not map cleanly to dishes. Likely causes include till items that combine multiple dishes or legacy items still on the till. Corrective action involves cleaning up the EPOS item list, mapping what is currently sold and remembering that only items sold since the integration was connected are surfaced.
- Margins swing wildly week to week. Likely causes include a single large or unusual invoice or a promotional price not reflected in the EPOS. Corrective action involves checking the invoice and the promotion before reacting. A sudden margin change may be a data issue rather than a customer behaviour change, so anomalies should be reviewed before any menu action is taken.
- Nobody owns the process. Likely cause is unclear responsibility between kitchen and finance. Corrective action involves naming an owner for invoices, an owner for recipes and an owner for the monthly review. Without named owners, the cadence collapses.
For a broader look at how this process reduces waste as well as cost, see How Menu Engineering Reduces Food Waste in Restaurants.
How To Measure Success
“Done correctly” means every dish has a live cost and margin, every EPOS item is mapped, invoices are captured without manual entry and the matrix is reviewed on schedule.
Directional indicators to track include reduced admin time spent on data entry and invoice reconciliation, fewer invoice discrepancies, faster reaction to supplier price changes, clearer margin tracking by dish and more timely pricing decisions. Jelly users cut food costs by 3% on average in the first three months, and gross margins increase on average by 2 percentage points in the same period. One operator improved gross profit from 65% to 72% within 12 weeks on approximately £500,000 in revenue, and Populu lifted GP from 68% to 72% across 16 locations.
Advanced Tips And Next Steps
Once the core workflow is running, several extensions increase its value.
- Delivery menu: Duplicate existing menu items in Jelly and factor in delivery commission overheads to create a separate, profitable delivery menu. Aggregator platforms take 25 to 35 percent off the top of every order routed through them, so a delivery menu priced without accounting for commission will lose margin invisibly.
- Supplier negotiations: Use Price Alert data as evidence when challenging suppliers or claiming credit notes. The data shows exactly which ingredient rose, by how much and from which supplier, so you avoid negotiating blind.
- Multi-site consistency: Use a single system across sites so margin data is comparable. The same dish can be a Star in one location and a Plow Horse in another because guest mix differs. Multi-site operators should run separate matrices for each unit. A central source of truth removes the need to be physically present on every site.
- Accounting integration: Push digitised invoices into Xero with one click to reduce bookkeeping time. Jelly’s Xero integration delivers a 90% reduction in bookkeeping time.
- Onboarding speed: Jelly onboards and generates initial value in the first week. For operators expanding to multiple sites, the central dashboard means margin visibility does not depend on being in the building.
For a comparison of platforms before committing to a full implementation, see Menu Engineering Software Comparison UK 2026 and Best Menu Engineering Software For UK Restaurants 2026.
Compare Jelly With Other Platforms
Frequently Asked Questions
What Are The Four Categories In Menu Engineering?
The four categories are Stars, Plow Horses, Puzzles and Dogs. Stars combine high popularity with high contribution margin and deserve protection and prominent placement. Plow Horses sell well but carry low contribution margin, so you re-engineer the recipe or nudge the price. Puzzles have low popularity but high contribution margin, so you improve description, placement or staff recommendation. Dogs sit low on both popularity and margin, so you consider removing or replacing them. Every dish on your menu falls into one of these four quadrants based on how it performs against the category averages for sales volume and contribution margin.
What Are The Basic Steps Of Menu Planning With Software?
In the context of menu engineering software, the six basic steps are: connect your EPOS to get item-level sales data, map EPOS items to dishes so every sale is attributed correctly, cost every recipe from real supplier invoices including garnishes, oils and wastage, generate the four-quadrant matrix and classify every dish, act on the output through pricing, description, placement or removal decisions and set a review cadence so the process repeats. This workflow, described in full above, turns those steps into a repeatable operating rhythm rather than a one-off exercise.
Do I Need An EPOS Integration To Use Menu Engineering Software?
An EPOS integration is strongly recommended. Without item-level sales data, popularity figures rely on guesswork rather than actual transactions, which means the matrix classifies dishes on incomplete information. Jelly integrates natively with Square, Lightspeed, EPOS Now and Toast, and setup takes approximately five minutes. Without the integration, you can still cost recipes from invoices and track margins, but the sales mix half of the matrix will not reflect real trading patterns.
Can I Run This Across Multiple Sites?
Multi-site use is one of the primary operational benefits of menu engineering software. Jelly provides a single dashboard where margin data from all locations is visible and comparable. Because the same dish can perform differently at different sites, with a Star in one location and a Plow Horse in another, the matrix should be run separately for each site. The underlying recipe costs, invoice data and supplier price alerts remain managed centrally, so an owner or operations manager does not need to be physically present at every site to maintain visibility over kitchen performance.
Conclusion: Turn The Matrix Into Margin
The workflow runs through six steps: connect your EPOS, map dishes, cost recipes from invoices, read the matrix, act on the output and review on a cadence. Each step depends on the one before it. Accurate margins require accurate invoice data, and the matrix only becomes actionable when the sales data feeding it reflects real trading.
The operational value is clarity, control and repeatability. A menu engineered on live data, where every dish has a known cost, a known margin and a quadrant classification, behaves very differently from one inherited from a previous season’s spreadsheet. The review cadence protects that margin over time. Weekly price alerts catch supplier increases before they compound, monthly matrix reviews catch quadrant shifts before they become structural problems and quarterly menu changes ensure the menu reflects current costs and current guest behaviour.
For a walkthrough of what this looks like in practice before committing to a full setup, see Menu Engineering Software Demo: What to Expect.
