Written by: JJ Tan, Founder, Jelly

Key Takeaways

  • Multi-site food cost control depends on clean data and a fixed weekly rhythm. A weekly control loop runs on an automated invoice-to-recipe-to-POS-to-GP data chain.
  • The four levers of control are centralised procurement, standardised recipes, actual vs theoretical (AvT) variance tracking, and site accountability with per-site scorecards.
  • A consistent weekly loop (Monday stock count, midweek invoice capture, Thursday price-drift check, Friday variance review) keeps every site accountable instead of relying on group averages.
  • AvT variance under 2 percentage points is the benchmark for tight control. Anything above 3 points is escalated at site level before it compounds.
  • Jelly automates the entire data chain, onboards in the first week, and delivers real-time GP visibility at a flat £129/month per location for UK restaurant, pub and hotel groups.

See Jelly In Action For Your Group — Book A Demo

What Multi-Site Food Cost Control Actually Means

Multi-site food cost control means measuring and managing food spend across every location through four levers: centralised procurement, standardised recipes, actual vs theoretical (AvT) variance tracking, and site accountability. The core principle is simple. Every site’s actual food cost is measured against what its recipes and sales say it should be. Any gap is investigated at site level before it turns into a group-level problem.

The Four Levers Of Multi-Site Food Cost Control

Centralised Procurement Across Every Site

One supplier price list across all sites is the foundation of comparable food cost reporting. UKHospitality’s procurement guidance states that in multi-site businesses, risk and margin leakage often stem from fragmented buying, inconsistent supplier choices and varying commercial terms, and that without a consolidated view, pricing pressure and working capital drag can accumulate unnoticed.

Consider a four-site group that allows one site to order through an unconnected supplier portal. Those purchases never reach the group figure. The site’s food cost stays invisible to the ops director until month-end, when the damage has already landed. Strong control means every site buys from the same contracted catalogue at the same price, with every invoice flowing into one system.

Before you move on, name one owner for supplier relationships so accountability is clear.

Standardised Recipes Across Locations

Every site needs the same spec, the same yield and the same wastage assumption. A 15g over-portion of protein across 400 covers a week is not a rounding error. At scale, a single extra 15g per portion across 400 wraps a week equals 6kg of ingredient, or roughly £81 a week on one menu item alone. Multiply that across a five-item menu and three sites and the number becomes material within a month.

Strong control means one recipe version, costed from the latest matched invoice, applied group-wide. For the mechanics of building and maintaining those recipes across sites, see How To Manage Food Costing Across Multiple Restaurant Sites.

Sushi Revolution uses Jelly to set separate target gross profits on dine-in and delivery menus, accounting for 30% delivery commissions, and has seen actual gross profits run 2–3% higher on average as a result. That kind of concept-specific targeting only works when recipes are standardised and costed from live invoice data.

Actual Vs Theoretical (AvT) Variance Tracking

AvT variance is the gap between what recipes and sales say food should have cost and what it actually did cost. A variance under 2 percentage points is the benchmark for tight food cost control in a multi-site operation; a gap of 3–4 points indicates an identifiable issue; a persistent gap of 4–5 points or more is a strong signal of a systemic problem such as stale recipe costs, a high-spoilage site, or consistent over-portioning.

Strong control means variance under 2 percentage points at every site, reviewed weekly. Any site above 3 points becomes the primary investigation target for the following week.

Site Accountability With Scorecards

A scorecard per site, reviewed on a fixed cadence, converts data into action. Use food cost variance by site as the primary accountability metric. A consistent blended gross profit figure can mask individual sites running 6–8 percentage points below that average for months at a time. A blended group figure can hide a struggling site, so site-level thresholds are what drive decisions.

Before you build the scorecard, name one owner for each lever. Without a named owner, the weekly review has no one to escalate to when a site breaches its threshold.

How To Run A Weekly Food Cost Review Across Sites

This weekly loop is the operating spine for a 2–10 site group. The single-site version appears in How To Lower Food Cost Percentage: A Weekly Control Loop. For a group, the same pattern applies, but every location follows the same timings with clear owners.

  1. Monday — Stock Count. The site manager or kitchen lead counts the highest-value lines. Every site counts on the same day and at the same point in the service cycle, ideally after close and before the first delivery of the new period. If one site counts on Sunday evening and another on Wednesday morning, they compute food cost over different windows and create margin differences that operations cannot explain.
  2. Midweek — Invoice Capture. All supplier invoices are emailed or photographed into one system. Month-end entry distorts the picture. A Thursday delivery keyed in on Monday makes one week look brilliant and the next look like a crisis, even though nothing changed in the kitchen.
  3. Thursday — Price-Drift Check. Ops or finance flags supplier price increases by ingredient and supplier. The NIQ and Prestige Purchasing Foodservice Price Index for July 2026 recorded ongoing volatility across UK hospitality food and drink prices, with drought conditions threatening domestic vegetable yields and global grain markets reacting to geopolitical stress. Weekly price-drift checks become a core control, not an optional extra.
  4. Friday — Variance Review. The ops director compares each site’s AvT and GP against target. Any site above 3 percentage points becomes the investigation target for the following week. This is a site-by-site comparison, with the group average used only as context.
  5. Monthly — Group Scorecard. The owner and finance director review the weighted group figure alongside every site’s individual performance. The group average provides the backdrop. The site figures drive the actions.

The loop fails if invoices arrive late or inconsistently. Put these five entries in the group calendar with a named owner against each before moving on.

The Data Chain: Invoice To Recipe To POS To GP

AvT accuracy depends on invoice line-item capture and POS-to-recipe mapping. Late or manual invoice entry breaks variance tracking before it starts. Reliable gross-profit reporting in a multi-site restaurant group depends on a connected data chain of three links: complete invoice ingestion from every supplier, correctly mapped recipes tied to those invoices, and a live POS connection feeding sales actuals — breaking any one link still produces a GP number, just not a reliable one.

The automated chain works as follows. A supplier invoice is captured by email or photo and every line item, including quantity, SKU, price and tax, is digitised. Recipe costs update automatically. POS sales are mapped to dishes. GP is recalculated in real time. The five-minute POS setup connects Square, Lightspeed, EPOS Now or Toast, and a one-click invoice push sends data to Xero (Sage coming soon).

Jelly is the system that makes this chain automatic for UK restaurant, pub and hotel groups. For a full comparison of multi-site food cost control software options, see Best Restaurant Cost Control Systems: UK Multi-Site 2026.

Restaurant365 is the most-cited food cost control vendor in Parse.gl’s AI mention-rate analysis at 35%, ahead of MarginEdge (21%), Apicbase (10%), Crunchtime (7%) and Supy (6%). All of these describe the four levers above. The documented trade-offs for a 2–10 site UK group are these. Jelly onboards and generates initial value in the first week once suppliers send invoices to a dedicated email address. Apicbase and Supy have longer setup timelines. Jelly charges a flat £129/month per location with no per-user or per-feature charges, so costs stay predictable as the group grows. It is built for growing 2–5 site kitchens in restaurants, pubs and boutique hotels rather than enterprise chains, which keeps the interface clean enough for the least tech-savvy chef to use without training overhead.

Check how many of your suppliers currently reach one system. Any gap of 10% or more signals a broken chain.

See Your Group’s GP In Real Time — Book A Demo

How To Calculate Food Cost Across Multiple Sites

Food cost % = cost of goods sold ÷ net food sales × 100. Net sales must exclude VAT. Under HMRC Notice 709/1, restaurant and café meals and qualifying hot takeaway food are standard-rated for VAT, so a VAT-inclusive price is divided by 1.2 at the standard 20% rate to get the net figure. A £14 dish is £11.67 net. Using the gross price flatters food cost percentage by several percentage points and makes every site’s figures incomparable.

AvT variance for actual vs theoretical food cost multi-site = actual food cost % − theoretical food cost %. Actual food cost is calculated as: (opening stock + purchases + transfers in − transfers out − closing stock) ÷ VAT-exclusive net food sales × 100. Theoretical food cost is: Σ (recipe cost × units sold) ÷ total net food sales × 100.

Inter-site transfers must be valued and recorded once, or variance is distorted. If stock moved between sites lands in an operating expense account at the receiving site and a revenue account at the supplying site, it creates a cost and revenue that never existed externally and both site margins are wrong. Portion consistency is a separate variable. Any week with a high proportion of agency cover warrants a closer look at the portioning line of the variance report.

The UKHospitality × Christie & Co Benchmarking Report 2022 found food gross profit averaged 67% across its survey — a food cost of 33% of net sales — with every one of its eight market segments falling between 65.9% and 69% food GP. Published UK guidance commonly puts food cost at roughly 28–35% for full-service restaurants and around 25–33% for pub kitchens, with QSR concepts cited in the region of 20–32% depending on the source; these are trade or consultant ranges rather than an official national benchmark. For pub-specific benchmarks, see Multi-Site Pub Cost Control: A Practical Playbook.

Confirm every site’s sales denominator is VAT-exclusive before comparing anything across the group.

What Is A Good Food Cost Percentage For A Multi-Site Restaurant Group

There is no single target. The right approach is concept-specific bands, with each site compared against its own history and its best-performing sister site rather than an industry table. A site that has run at 30% for two years is a different situation from one that has climbed from 30% to 34% over six months. The trend matters as much as the absolute figure.

Management thresholds for a UK group of 2–10 sites:

  • ≤2 percentage points AvT variance: well managed
  • Above 2 points: investigate
  • Above 3 points: escalate to ops director
  • Persistent 4–5 points: systemic problem requiring structural review

Write your thresholds down and publish them to site managers so accountability is explicit rather than assumed. Thresholds only work if the data behind them arrives in time, and month-end accountant reports are too late.

How To Stop Supplier Price Drift Across Sites

Supplier price drift needs catching weekly, not at month-end. If a supplier raises the price of a key ingredient by 8% and the invoice is approved without updating the item’s purchase price and every linked recipe, an operator running food cost at 28% can discover at month-end that actual cost is 33% with no clear explanation. By then, four or five weeks of margin have already been lost.

The NIQ and Prestige Purchasing Foodservice Price Index for August 2026 recorded a 0.4% month-on-month rise in UK hospitality food and drink prices, with the Meat and Poultry, Milk, Cheese and Eggs, and Oils and Fats categories all posting increases. This kind of category-level volatility is exactly what a Thursday price-alert review is designed to catch before it compounds.

Jelly’s Price Alert feature flags every invoice line where a price has changed, showing the amount and the supplier. That gives the ops director or head chef the evidence to call a supplier, negotiate better rates and claim credit notes before the next Friday variance review reveals the damage. Stuart Noble, Head Chef at Cairn Lodge Hotel, cut food costs by 5% in a month using Jelly’s price-alert capability. Amber restaurant in East London saves £3,000–£4,000 per month through a combination of invoice automation, price-change alerts and real-time recipe costing.

Diarise a Thursday supplier challenge with the price-alert report open.

30-Day Rollout For A 2–10 Site Group

A sequenced rollout gives accurate data quickly without overwhelming sites. Going live on all fronts at once often produces noise instead of control.

  1. Week 1: Connect suppliers to a dedicated invoice email address and connect the POS. Every supported POS, including Square, Lightspeed, EPOS Now and Toast, takes approximately five minutes to connect. From day one, price alerts and spending insights are live.
  2. Week 2: Build recipes from scanned ingredients and map POS items to dishes. Jelly populates ingredients automatically from scanned invoices, so chefs click rather than type. What previously took 28 minutes per menu item now takes approximately 3 minutes.
  3. Week 3: Set site scorecards and AvT targets. Publish the management thresholds to every site manager. Confirm that every site’s sales denominator is VAT-exclusive.
  4. Week 4: Run the first Friday variance review and the first Thursday supplier challenge. Any site above 3 percentage points becomes the investigation target for week five.

The first three actions are simple: set up the invoice inbox, connect the POS, and name the Friday reviewer. Book the week-four review before week one starts. If it is not in the calendar, it will not happen.

Plan Your 30-Day Rollout With Jelly — Book A Demo

Frequently Asked Questions

What Is A Good Food Cost Percentage For A Multi-Site Group?

There is no universal figure. As noted above, published UK guidance puts food cost at roughly 28–35% for full-service restaurants and 25–33% for pub kitchens as a percentage of net food sales, with QSR concepts often somewhat lower. The more useful benchmark is concept-specific. Compare each site against its own history and against the best-performing comparable site in the group. A site that has held 31% for 18 months is performing well. A site that has drifted from 31% to 35% over the same period has a problem that needs investigating regardless of where the group average sits.

How Is AvT Variance Calculated?

Actual food cost is calculated from physical stock movement: opening stock plus purchases plus transfers in, minus transfers out and closing stock, divided by VAT-exclusive net food sales for the same period, expressed as a percentage. Theoretical food cost is calculated bottom-up from recipes: the recipe cost of each menu item multiplied by the number of units sold, summed across the menu, divided by net food sales. AvT variance is actual food cost percentage minus theoretical food cost percentage, expressed in percentage points. A positive result means the site used or lost more food than its recipes predict. For the calculation to be valid, both sides must use the same period, the same recipe version, the same supplier-price basis, and inter-site transfers must be recorded on both sides.

Who Should Own The Weekly Review?

The ops director owns the Friday variance review and the Thursday supplier challenge. The site manager or kitchen lead owns the Monday stock count. Finance owns the monthly group scorecard. Splitting ownership by cadence and by decision type, operational versus commercial, prevents the review from becoming one person’s weekend task and ensures accountability is distributed across the management structure.

How Quickly Can A Group Be Live On Jelly?

As covered above, Jelly onboards and generates initial value in the first week once suppliers send invoices to a dedicated email address. Price alerts and spending insights are available from day one. Recipes can be built in week two from ingredients already populated by scanned invoices. The first Friday variance review is achievable by week four. The flat £129/month per location covers the full platform from day one.

Conclusion

The weekly control loop only works when the data underneath it updates automatically. Manual invoice entry, spreadsheet recipe costing and month-end accountant reports all break the chain before the loop can run. Jelly is the best multi-site food cost control software for UK restaurant, pub and hotel groups that need to control food costs across sites because it automates invoice line-item capture and keeps dish and menu GP live in real time. As covered above, it onboards in the first week at a flat £129/month per location. For a full ranking of systems in this category, see Best Restaurant Cost Control Systems: UK Multi-Site 2026.

Control Multi-Site Food Costs With Jelly — Book A Demo