Written by: JJ Tan, Founder, Jelly | Last updated: 1 July 2026
Key Takeaways
- Gross profit equals total revenue minus COGS. UK restaurants should target 55–65 % GP margins and keep prime costs at or below 65 % of revenue.
- Weekly GP analysis using five repeatable stages, invoice capture, theoretical dish costing, stock counts, variance tracking and Flash Reports, typically recovers 2–3 percentage points of margin and removes 10–20 hours of manual work.
- Accurate inputs are essential. Supplier invoices, standardised recipes with yields, real-time POS data and weekly stock counts must sit with clearly defined owners to prevent data gaps.
- Automated price alerts, variance tracking and same-week decision-making tackle common pitfalls such as supplier price creep, portion drift, unrecorded wastage and delayed monthly reports.
- Operators using Jelly achieve faster, more accurate GP analysis without spreadsheets. See how it works for your operation.
Core Inputs You Need Before You Start
Accurate GP analysis requires four inputs to be in place before any calculation is run. The table below shows who owns each input, where it comes from and how often it must be updated. Every input needs a named owner and a clear update frequency so gaps do not appear in your data.
| Input | Source | Owner | Frequency |
|---|---|---|---|
| Supplier invoices (line-item level) | Email or photo capture | Owner / Finance | Every delivery |
| Standardised recipes with yield % | Recipe book / costing tool | Head Chef | On menu change |
| POS sales data (item level) | POS integration | Owner / Finance | Real-time |
| Opening and closing stock counts | Physical count | Head Chef | Weekly |
Ownership sits at the heart of reliable GP numbers. The owner or finance manager controls invoice capture, Flash Report review and supplier payment. The head chef owns recipe accuracy, portion standards and weekly stock counts. When these roles blur, data gaps appear and GP figures lose credibility.
Why Weekly GP Analysis Protects Your Margin
UK restaurant operators targeting sustainable profitability should aim for a gross profit margin of 55–65 % on food and beverage revenue, with prime cost, COGS plus labour, held at or below 65 % of revenue. Operators who rely on monthly accountant reports to monitor these figures work with data that is four weeks old, long after a supplier price increase has already eroded the week’s margin.
A weekly GP analysis routine removes that lag. Operators who follow it consistently recover 2–3 percentage points of gross margin, eliminate 10–20 hours of manual spreadsheet work each week and gain the ability to react to price changes in the same week they occur.
Step-by-Step Process for Weekly GP Analysis
Stage 1: Capture and Digitise Invoices
Objective: Create a single, accurate record of every ingredient cost at line-item level.
Action: Photograph or email every delivery invoice into a central system the same day it arrives.
Success criteria: No invoices processed from memory or estimated figures, every SKU, quantity and unit price recorded.
Manual keying of invoices into spreadsheets is the single largest source of GP error. A missed line item or a transposed price silently distorts every downstream calculation. Automated invoice scanning removes this risk and creates a timestamped audit trail for supplier queries.
Stage 2: Calculate Theoretical GP Per Dish
Objective: Establish what each dish should cost based on current ingredient prices and standard recipe yields.
Action: Link each recipe ingredient to its live invoice price, then apply yield and wastage percentages.
Formula: Theoretical dish cost = Σ (ingredient quantity used × current unit price) ÷ yield %
The table below shows how three common menu items translate recipe costs into GP percentages. Small differences in recipe cost create significant GP variation when selling prices sit in a similar range.
| Dish | Recipe cost (£) | Selling price (£) | Theoretical GP % |
|---|---|---|---|
| Pan-roasted salmon | 4.80 | 18.00 | 73.3 % |
| Beef burger | 3.20 | 14.00 | 77.1 % |
| Mushroom risotto | 2.10 | 13.50 | 84.4 % |
Success criteria: Every dish on the menu has a theoretical GP % that updates automatically when a new invoice is scanned.
Stage 3: Run Weekly Inventory for Actual COGS
Objective: Measure what was actually consumed, including wastage, staff meals and unrecorded usage.
Action: Count all stock at the same time each week, with Sunday close as a common choice. Calculate actual COGS using the formula below.
Formula: Actual COGS = Opening stock + Purchases − Closing stock
Success criteria: Stock count completed in under 90 minutes, with the COGS figure available by Monday morning. With both theoretical and actual COGS now calculated, the next stage compares the two to reveal where margin is leaking.
Stage 4: Compare Theoretical vs Actual Variance
Objective: Identify where actual COGS exceeds theoretical COGS and quantify the gap.
Action: Subtract theoretical COGS from actual COGS for the week, then express the result as a percentage of revenue.
| Metric | This week (£) | % of revenue | Status |
|---|---|---|---|
| Revenue | 18,500 | 100 % | – |
| Theoretical COGS | 6,290 | 34.0 % | Target |
| Actual COGS | 6,845 | 37.0 % | ⚠ +3 % |
| Weekly GP variance | 555 | 3.0 % | Investigate |
A variance above 2 % of revenue warrants immediate investigation. Common causes include portion drift, unrecorded wastage, supplier price creep or theft.
Stage 5: Build a Weekly Flash Report and Act on Red Flags
Objective: Produce a single decision-ready document that the owner and head chef review together each Monday.
Action: Compile GP %, actual versus theoretical variance, top five cost movers and any Price Alert flags from the week.
Weekly Flash Report checklist:
- ☐ GP % for the week vs prior week and target
- ☐ Actual vs theoretical COGS variance (£ and %)
- ☐ Top three ingredient price increases flagged
- ☐ Any dish now below minimum GP threshold
- ☐ Credit notes raised or pending from suppliers
- ☐ Actions assigned with named owner and deadline
Success criteria: Report reviewed and actions assigned before the first service of the week.
Common GP Problems and How to Fix Them
Supplier price creep. Suppliers adjust line-item prices incrementally, often by pennies, across dozens of SKUs. Without automated price alerts, these increases accumulate unnoticed for weeks before appearing in a monthly P&L. A Price Alert system that flags every price movement on the day the invoice is processed fixes this.
Portion drift. A recipe specifies 180 g of protein, yet the kitchen plates 210 g. Over 200 covers a week, that 30 g drift costs an additional £180–£300 depending on the protein. Weekly variance tracking surfaces this pattern within one cycle.
Missing wastage recording. Wastage that is not logged inflates apparent COGS and makes the theoretical versus actual variance look worse than the underlying operation warrants. Assign one person per shift to record spoilage against the relevant ingredient code.
Multi-site inconsistency. When each site uses different spreadsheet templates or counts stock on different days, group-level GP comparisons become meaningless. A single platform with standardised inputs across all locations resolves this quickly.
Delayed accountant reports. Monthly management accounts act as a compliance tool, not an operational one. By the time a report confirms that GP dropped in March, it is already April. Weekly Flash Reports replace this lag with same-week visibility.
Delivery menu under-costing. Delivery platforms such as Deliveroo and UberEats charge average commissions of 30 %, which must be factored into dish-level GP calculations for delivery SKUs. A dish priced for 70 % GP in the dining room generates closer to 40 % GP on a delivery platform if the menu price is not adjusted.
How to Measure Success
| Metric | Baseline | Target (12 weeks) |
|---|---|---|
| Weekly GP % | Current average | +2–3 percentage points |
| Theoretical vs actual variance | > 3 % | ≤ 1.5 % |
| Admin hours per week | Current baseline | < 2 hrs |
| Credit notes recovered | £0 tracked | Quantified monthly |
| Time-to-decision on price changes | 4 weeks (monthly report) | Same week |
One operator using Jelly improved gross profit from 65 % to 72 % within 12 weeks on approximately £500,000 in revenue. Stuart Noble, Head Chef at Cairn Lodge Hotel, cut food costs by 5 % within a month after introducing automated invoice scanning and live dish costing.
Advanced GP Tactics and Automation
Menu engineering using the Stars and Plowhorses framework. Once POS sales data links to dish-level GP, every item on the menu can be plotted on a two-axis grid, popularity, covers sold, against GP contribution, pounds per cover. Stars are high-popularity, high-GP dishes, so protect their recipe integrity and feature them prominently. Plowhorses are high-popularity but low-GP dishes, so re-engineer the recipe, adjust the portion or raise the price. Dogs, low popularity and low GP, are candidates for removal.
Delivery menu costing. Build a separate delivery menu with prices that absorb the platform commission discussed earlier while maintaining a minimum GP target. Jelly’s Cookbook allows operators to duplicate existing menu items and apply a commission overhead, which produces a separate GP calculation for each channel without repeating recipe-build work.
Supplier negotiation using Price Alert data. When a supplier increases the price of a key ingredient, a Price Alert generates a timestamped record of the change, the previous price and the cost impact per dish. This data converts a subjective negotiation into a factual one. Amber restaurant in East London saves £3,000–£4,000 per month through credits, better buying and tighter menu controls driven by exactly this approach.
Jelly as the automation layer. Each of the five stages above can run manually, yet the combined admin burden reaches 10–20 hours per week. Jelly automates invoice capture, live dish costing, theoretical versus actual variance and Flash Report generation. POS integrations with Square, Lightspeed, EPOS Now and Toast deliver item-level sales data in real time, so the Sales Mix report stays current. The result is the same analytical rigour with a fraction of the admin, at a flat rate of £129 per location per month.
Ready to tighten GP control without living in spreadsheets? Explore Jelly in a live walkthrough.
Frequently Asked Questions
How often should a restaurant analyse its gross profit?
Weekly analysis is the minimum effective frequency for operational GP control. A weekly Flash Report gives operators enough data to spot trends, such as a rising COGS variance or a dish that has slipped below its GP target, and act before the problem compounds. Monthly reviews, typically driven by accountant reports, arrive too late to influence the decisions that caused the variance. Daily GP visibility, which Jelly’s Flash Report supports, is the gold standard for operators with higher revenue or tighter margins.
Who should own the GP analysis process, the owner or the head chef?
Both roles carry distinct responsibilities. The owner or finance manager owns the invoice capture process, the Flash Report review and any supplier payment or credit-note decisions. The head chef owns recipe accuracy, portion standards and the weekly stock count. GP analysis breaks down when either party assumes the other is handling their half. A shared platform where both roles have direct visibility, rather than a spreadsheet passed between them, removes the most common source of friction and delay.
How does GP analysis work across multiple sites?
Multi-site GP analysis requires standardised inputs, including the same recipe templates, the same stock-count schedule and the same invoice-capture process at every location. Without standardisation, site-to-site comparisons are distorted by inconsistent methodology rather than genuine operational differences. Jelly supports multi-site operators by centralising all invoice, recipe and sales data in one platform, which allows group-level GP reporting alongside site-level drill-down. Operators approaching two to five sites find this particularly valuable as they lose the ability to be physically present at every location.
Does Jelly integrate with existing POS and accounting systems?
Jelly integrates natively with Square, Lightspeed, EPOS Now and Toast via real-time API, which delivers item-level sales data the moment a transaction completes. Connecting any supported POS takes approximately five minutes. On the accounting side, Jelly pushes digitised invoices directly into Xero with a single click, with Sage integration in development. These integrations mean that GP calculations draw on live cost and sales data at the same time, rather than requiring manual exports and reconciliation between systems.
See how Jelly fits your existing setup, and connect with our team to map your integration.