Written by: JJ Tan, Founder, Jelly | Last updated: 23 June 2026
Key Takeaways
- Volatile supplier prices and delayed margin data quietly erode gross profit, so UK restaurants benefit from a structured seven-step pricing process.
- Accurate contribution margin calculation depends on itemised invoices, POS sales data and a recipe library with yield percentages to prevent drift.
- The four-quadrant menu engineering matrix groups dishes as Stars, Plowhorses, Puzzles or Dogs so operators can prioritise repricing, promotion or removal.
- Pricing psychology, small price tests and a monthly review rhythm typically deliver GP gains of 2–3 percentage points within 90 days.
- Jelly automates invoice scanning, live dish costing and matrix classification, and you can see it in action with a quick demo to understand its impact on menu profitability.
Before You Reprice Your Menu
Menu pricing work only delivers reliable results when three inputs are in place. You need access to itemised supplier invoices (digital or photographed), a point-of-sale system that records item-level sales, and a recipe library that maps ingredients to dishes with accurate yield and wastage percentages. Without all three, contribution margin figures drift and the process breaks down.
Assign a single owner, typically the Head Chef or Operations Manager, to keep the system moving. This person updates recipe costs when invoices change and runs the monthly review. This single point of accountability is critical because shared ownership without a named lead is the most common reason optimisation stalls after the first cycle.
The Contribution Margin Formula
Contribution margin (CM) shows how much revenue a single dish contributes toward fixed costs and profit after variable costs are deducted.
Contribution Margin = Selling Price − Variable Cost per Dish
Variable cost includes food cost and any variable labour directly linked to preparation. For most UK operators, food cost is the dominant variable and the easiest lever to adjust.
Consider a pan-seared salmon dish that sells for £18.00. Ingredient cost is £5.40, which is a 30% food cost percentage, so the contribution margin is £12.60. If a supplier raises salmon prices by 15% in Q1 2026, a realistic scenario given ongoing North Atlantic supply pressures, the ingredient cost rises to £6.21 and CM falls to £11.79 without a menu price change. That £0.81 reduction per cover, multiplied across 200 covers per week, removes about £842 per month in margin from a single line item.
Gross profit percentage and contribution margin are related but distinct. GP% shows the proportion of revenue retained, while CM shows the absolute pounds contributed per dish. Both matter, with GP% used for benchmarking and CM used to decide which dishes to reprice first.
Once you have accurate CM figures for every dish, you can classify them by both profitability and popularity. This is where the menu engineering matrix becomes your main decision-making framework.
The Four-Quadrant Menu Engineering Matrix
Menu engineering, developed by Kasavana and Smith, classifies every dish on two axes: popularity, measured as number of covers sold relative to the menu average, and profitability, measured as contribution margin relative to the menu average. The result is four clear categories.
| Category | Popularity | Profitability | Action |
|---|---|---|---|
| Stars | High | High | Protect and promote, and avoid unnecessary price increases |
| Plowhorses | High | Low | Reduce portion cost or move price gradually |
| Puzzles | Low | High | Improve visibility through menu placement and description |
| Dogs | Low | Low | Remove or replace, as they consume kitchen labour for minimal return |
Calculate the menu average CM and the menu average cover count. Any dish above both averages is a Star. Above average covers but below average CM is a Plowhorse. Below average covers but above average CM is a Puzzle. Below both is a Dog. Run this classification monthly, not annually, because a dish can move quadrant within weeks when supplier prices shift.
7-Step Menu Optimisation Checklist
Step 1 — Calculate Contribution Margin per Dish
Objective: Establish a current, accurate CM for every menu item. Inputs: Itemised invoices and recipe cards with yield percentages. Action: Pull the most recent invoice price for each ingredient, apply the yield factor, add the ingredient costs to get dish cost, then subtract this from the selling price. Good looks like: Every dish has a CM figure updated within the last invoice cycle.
Step 2 — Classify Every Menu Item
Objective: Populate the four-quadrant matrix. Inputs: CM per dish and POS item-level sales data for the last 30 days. Action: Calculate menu average CM and menu average covers, then assign each dish to Star, Plowhorse, Puzzle or Dog. Good looks like: A complete matrix with no unclassified items.
Step 3 — Apply Pricing Psychology
Objective: Increase perceived value and reduce price sensitivity. Inputs: Current menu layout and price points. Action: Remove currency symbols (£) from printed and digital menus to reduce price salience. Once price awareness is lower, use anchor pricing by placing a high-CM Puzzle dish next to a premium item so the Puzzle feels reasonable by comparison. Finally, round prices to .00 or .50 rather than .99, which signals quality in full-service environments and supports the premium anchors.
Good looks like: Menus reviewed by someone unfamiliar with the operation who cannot immediately spot the highest-margin items from price formatting alone.
Step 4 — Adjust Portion or Ingredient Specifications
Objective: Improve CM on Plowhorses without a visible price increase. Inputs: Recipe cards and supplier price history. Action: Reduce protein portion by 10–15g where presentation still works, swap a secondary ingredient for a lower-cost equivalent of similar quality, or renegotiate pack size with the supplier to reduce unit cost. Good looks like: Plowhorse CM moves above menu average without a spike in customer complaints.
Step 5 — Create a Separate Delivery Menu
Objective: Protect GP on third-party delivery channels where commissions run at 25–35%. Inputs: Dine-in menu and delivery commission rate. Action: Duplicate dine-in items, add commission overhead to each dish’s variable cost, then reprice to maintain target CM. Remove low-CM, high-complexity dishes that do not travel well. Sushi Revolution uses separate target gross profits for dine-in and delivery menus, accounting for 30% delivery commissions, achieving actual gross profits 2–3% higher on average. Good looks like: Delivery menu CM matches or exceeds dine-in CM on a per-cover basis.
Step 6 — Test Price Changes on Low-Risk Items
Objective: Confirm customer price elasticity before broad repricing. Inputs: Matrix classification and POS sales velocity. Action: Increase the price of two or three Puzzles by £0.50–£1.00, then monitor cover count over four weeks. If volume holds within 5%, extend the increase to similar items. Good looks like: A documented price-test log with before and after CM and cover data.
Step 7 — Set a Monthly Review Cadence
Objective: Turn optimisation into a repeatable habit rather than a reactive fix. Inputs: Updated invoices, POS sales mix and the previous month’s matrix. Action: Schedule a 60-minute monthly session to refresh CM figures, reclassify the matrix and act on any items that have shifted quadrant. Good looks like: A standing calendar invite with a named owner and a completed matrix from the prior month as the baseline.
Common Menu Pricing Mistakes
Spreadsheet drift. Manually entered recipe costs fall out of sync with actual invoice prices within days of a supplier change. A dish costed in January using January prices shows an incorrect CM by March without a manual update, and those manual updates are often missed.
Missed price alerts. Supplier price increases usually appear inside invoice line items with no clear notification. Without a system that flags changes at the SKU level, operators only notice margin erosion weeks after it starts.
Inconsistent unit conversions. Mixing kilograms, litres and portions in the same recipe card without a consistent conversion layer produces CM figures that are arithmetically correct but operationally wrong. A 5kg bag of flour costed per gram must be converted to the exact gram weight used per dish, including wastage, before the CM calculation is valid.
How to Measure Menu Optimisation Success
Four directional KPIs show whether the process is working. Track GP margin lift, aiming for the 2–3 point improvement outlined earlier. Measure hours saved on costing per week, with a target reduction from 10–20 hours to under 2. Count supplier credits claimed per month, which reflect how well price alerts are used. Monitor time from supplier price change to menu price update, with a target of under 48 hours.
Track these monthly alongside the matrix. If GP margin is not moving after 60 days, the most likely causes are stale recipe costs, unclassified new menu items or price changes that have not flowed through to selling prices.
Advanced Tips for Faster Menu Decisions
The seven-step process above works manually, but automation compresses the review cycle from monthly to daily. When invoice scanning, price alerts and POS-linked sales mix reports run automatically, you see issues before they become problems. When every invoice is scanned at line-item level the moment it arrives, whether by email or photograph, ingredient costs update in real time and dish CM figures recalculate automatically. A price alert that fires the same week a supplier increases a SKU price gives operators the data to negotiate a credit note or switch supplier before the margin impact compounds.
Jelly automates this entire flow. Invoice scanning feeds live dish costing, live costing feeds the sales mix report, and the report surfaces the matrix classification without manual input. The onboarding timeline mentioned earlier breaks down to five minutes for POS connection, with the remaining time spent on invoice scanning setup and recipe library migration. Pricing is a flat £129 per site per month with no per-user charges. Operators using Jelly report an average GP improvement of 2 percentage points within the first three months, and one operator moved from 65% to 72% gross profit within 12 weeks on approximately £500,000 in revenue.
Recap and Your Next Step
Restaurant menu pricing for stronger profit margins follows a clear sequence. You establish accurate contribution margins, classify every dish using the four-quadrant matrix, apply pricing psychology to reduce price sensitivity, adjust portion specifications on Plowhorses, build a separate delivery menu, test price changes in small steps and review results monthly. This process is repeatable and measurable. The main constraint is data freshness, because stale invoice costs and delayed sales mix reports stop most operators from sustaining gains beyond the first cycle.
Jelly removes that constraint by automating the invoice-to-margin pipeline, so the matrix stays current and decisions happen in hours rather than weeks.
Frequently Asked Questions
What is a good contribution margin for a restaurant dish in the UK?
There is no single universal figure, because contribution margin varies by cuisine type, service format and price point. The more useful benchmark is your own menu average, because any dish above that average CM contributes more than its share to fixed costs and profit. As a starting point, most full-service UK restaurants target a food cost percentage of 28–35%, which implies a CM of 65–72p for every £1 of revenue before labour and overheads. The goal of menu engineering is to shift the sales mix toward higher-CM dishes, raising the blended average across all covers.
How often should a UK restaurant update its menu prices?
A monthly review cadence works as a minimum for operators with multiple suppliers and a menu of more than 20 items. In periods of active supplier inflation, which has characterised the UK food supply chain through 2024–2026, a two-week review cycle suits high-cost protein and dairy lines better. The practical constraint is data freshness. If invoice costs are updated manually, monthly is realistic. If invoice scanning is automated, weekly or even daily visibility is achievable without extra admin time.
What is the difference between food cost percentage and contribution margin?
Food cost percentage expresses ingredient cost as a proportion of selling price. Contribution margin expresses the absolute pounds remaining after ingredient cost is deducted from selling price. A dish with a low food cost percentage but a low selling price can have a lower CM than a dish with a higher food cost percentage but a higher selling price. Both metrics matter, with food cost percentage used for benchmarking against industry norms and for supplier negotiation, and contribution margin used to decide which dishes to promote, reprice or remove from the menu.
How do delivery commissions affect menu pricing strategy?
Third-party delivery platforms typically charge 25–35% commission on the gross order value. This commission is a variable cost that must sit inside the dish’s cost base when you calculate CM for the delivery channel. A dish priced at £14.00 with a £4.00 ingredient cost has a dine-in CM of £10.00. On a platform charging 30% commission, the effective revenue is £9.80, which reduces CM to £5.80, a 42% reduction. Operators should build a separate delivery menu with prices adjusted to restore target CM, and remove dishes that cannot reach acceptable margins even at higher delivery price points.
Can menu engineering work for a pub or boutique hotel as well as a restaurant?
The four-quadrant matrix applies to any operation with a defined menu and item-level sales data. For pubs, the analysis usually separates food and beverage menus and may weight popularity differently because drinks often drive covers. For boutique hotels, the matrix is most useful for the restaurant and bar menus, while room service menus often need a separate analysis because labour cost per cover is materially higher. The contribution margin formula and classification logic stay identical across all formats, while the inputs and target CM thresholds change by operation type.