Knowing your food cost percentage is only the start. Smart food cost management begins with understanding why that number differs from what your recipes say it should be.

Theoretical food cost tells you what your ingredients should have cost if everything went to plan, based on your recipes and sales. Actual food cost tells you what they really cost, based on stock, purchases and consumption. The gap between the two, often called actual vs theoretical (AvT) food cost, is your food cost variance. It usually comes from portioning, waste, purchasing, stock counts or out-of-date recipe data.

Head chefs, finance teams and multi-site operators all need to understand this gap to protect margin. This guide covers the formulas, a worked example, the common causes and a four-step way to close it.

What is theoretical food cost?

Theoretical food cost is the expected cost of the ingredients used to make the dishes you sold in a period. Multiply each dish’s standard recipe cost by the number of portions sold, then add the results together.

Theoretical food cost = Recipe cost per portion × Portions sold

For example, if a dish has a recipe cost of £4 and you sell 100 portions, its theoretical cost is £400.

The figure is only as reliable as its inputs: accurate recipes, portion sizes, current ingredient prices and sales data from your EPOS. If any of these is wrong, the theoretical figure is wrong too.

What is actual food cost?

Actual food cost is the value of stock you really consumed in a period, including everything lost to waste, over-portioning and other operational issues.

Actual food cost = Opening stock + Purchases − Closing stock

Actual food cost % = (Actual food cost ÷ Food sales) × 100

For example, a restaurant opens the month with £5,000 of stock, buys £8,000 and closes with £4,000. Its actual food cost is £9,000. On food sales of £30,000, that is an actual food cost of 30%.

How do you calculate theoretical vs actual food cost variance?

Food cost variance is the difference between actual and theoretical food cost. It shows whether you used more or less stock than your recipes and sales say you should have.

Food cost variance = Actual food cost − Theoretical food cost

Illustrative example: continuing the figures above, the restaurant’s portions sold multiplied by its recipe costs give a theoretical food cost of £8,100, or 27% of food sales. Actual food cost was £9,000, or 30%. The variance is £900, or three percentage points of sales, and that £900 needs explaining.

A positive variance means actual food cost is higher than theoretical. A negative variance means it is lower, which usually points to a counting or recipe data problem rather than an unusually efficient kitchen.

To check the recipe cost side for your own dishes, run them through our free food cost calculator.

What creates the variance between theoretical and actual food cost?

A variance does not automatically mean something went wrong. It is a signal to look at the processes and records behind it. The usual causes are these.

Portioning. When portions are bigger than the recipe specifies, more stock is used for the same number of sales.

Waste. Over-preparation, spoilage and trimmings increase actual consumption without changing the theoretical figure. According to WRAP, food waste costs the UK hospitality and food service sector £3.2 billion a year, an average of £10,000 per outlet.

Staff meals and complimentary dishes. Food eaten by staff or given away without being recorded in the EPOS creates usage with no matching sale.

Stocktake errors. Miscounts, missed storage areas and inconsistent units change the closing stock value, and with it the actual food cost.

Supplier price changes. When ingredient prices change but recipe costs do not, the theoretical figure drifts away from reality. Our guide on how to keep recipe costs accurate when supplier prices change covers this in detail.

Data errors. Incorrect stock movements or transfers, recipe errors and EPOS items mapped to the wrong recipe.

What is the fastest way to close the variance gap?

Closing the gap needs consistent processes across the kitchen, purchasing and finance. We recommend working through it in a fixed order we call the Variance Trail, because data errors can look exactly like kitchen losses. Check the records first, then the people.
Recipe check: is the theoretical figure right?

Update ingredient quantities, yields, portion sizes and costs whenever a dish or a supplier price changes. Standardised recipes make your theoretical food cost a dependable benchmark.

Delivery check: did you receive what you paid for?

Make sure the quantity and price delivered match the order and the invoice, and update ingredient costs when a supplier’s price changes. Supplier price tracking makes those changes visible when the stock arrives.

Count check: is the closing stock value reliable?

Count consistently, using the same units and the same route through your storage areas every time. Investigate unusual movements rather than adjusting the numbers to fit.

Plate check: what is happening on the pass?

Once the records are sound, look at portioning, recorded waste, staff meals and complimentary dishes. Record what is wasted, why, and what is consumed outside normal sales.

Review variance by ingredient, dish and period, and focus on patterns that repeat rather than one-off percentages. If you run more than one site, standardise recipes, count procedures and reporting periods across all of them so the comparison is fair. For how variance fits into the wider control loop, see our complete food cost management system.

How does restaurant software help close the variance gap?

Spreadsheets make it hard to connect recipe costs, sales, purchases and stock movements. Inventory and recipe management software brings those records together, so teams can see where differences arise.

With StockTake Online’s recipe management software, operators keep ingredient quantities, recipe costs and yields in one place. EPOS integrations with Toast, Lightspeed, Aloha EPOS, TISSL, Grafterr and HubRise bring sales in, so theoretical consumption is calculated from what was actually sold. Connected stock control and purchasing track deliveries, stock levels and usage, and reporting compares expected and actual consumption so unusual variances surface earlier.

For multi-site groups, consistent data makes it easier to compare locations and see where processes need attention. The aim is not simply a lower food cost percentage. It is understanding what drives the number, so kitchen, finance and operations teams can act on it.

Where should you start?

Calculate your actual food cost for the last period and compare it with your theoretical figure. Check the recipe costs of your highest-selling dishes in the food cost calculator first, then work through the Variance Trail for the largest gaps.

If you want to see theoretical vs actual reporting running on your own recipes, sales and stock, Book a Demo with the StockTake Online team.

Key takeaways

• Theoretical food cost is recipe cost multiplied by portions sold; actual food cost is opening stock plus purchases minus closing stock.
• Variance is actual minus theoretical. A negative variance usually signals a data problem, not an unusually efficient kitchen.
• Common causes are portioning, waste, unrecorded staff meals and comps, stocktake errors, supplier price changes and data errors.
• Work the Variance Trail in order: recipe, delivery, count, plate.
• Check variance weekly for high-value and high-waste lines, and standardise processes across sites.

Frequently asked questions

What is the difference between theoretical and actual food cost?

Theoretical food cost is calculated from recipe costs and sales: what your ingredients should have cost. Actual food cost is calculated from stock: opening stock plus purchases minus closing stock. The difference between them is your food cost variance.

What is actual vs theoretical (AvT) food cost?

AvT is another name for the same comparison: actual food cost set against theoretical food cost for the same period. The AvT variance shows how much stock was used that your sales and recipes cannot explain.

What is a good food cost variance for a restaurant?

There is no single right figure. It depends on your menu, operation, stocktake quality and business type. The most useful variance is one that is stable and can be explained, so track it over time rather than chasing a benchmark.

Why is actual food cost higher than theoretical food cost?

Usually because of over-portioning, food waste, unrecorded staff meals or complimentary dishes, stocktake errors, supplier price changes that have not reached recipe costs, or errors in recipe and sales data.

How often should a restaurant check food cost variance?

At least monthly alongside the stocktake, and weekly for high-value or high-waste items such as proteins, so problems are caught while they are still small.

How UK Pubs and Bars Control Pour Cost in 2026/27

UK pubs and bars control pour cost by shortening the time between a cost rise and the glass. Heading into 2027, that matters more than the formula itself. Supplier prices change on an invoice, often weeks before anyone updates a serve cost or a menu price, and every pint poured in that gap sells at the old margin.

This guide sets out a weekly control routine for licensees, bar managers and owners of one to three sites. If you need the calculation first, read how to calculate beverage cost percentage, then come back here for the control side.

What does controlling pour cost mean in 2026/27?

Pour cost is the cost of a drink divided by its selling price net of VAT, shown as a percentage. Beverage gross profit is the other side of the same number: a 33% pour cost is a 67% GP. Controlling pour cost means keeping every serve inside the target band you set for it, as costs move underneath you.

Costs are moving again. According to HM Revenue & Customs (28 October 2026), all alcohol duty rates, including the reduced draught relief rates, are uprated in line with RPI by 4.55% from 1 February 2027. At the new draught beer rate of £20.34 per litre of alcohol, duty on a pint of 4% ABV draught beer works out at roughly 46p, so the uprating itself adds around 2p a pint (our arithmetic, for illustration).

The duty line is rarely what hurts. HMRC’s own impact note assumes producers will raise wholesale prices to reflect the change, and those increases land on delivery notes on the supplier’s timetable, not yours. The real exposure is the lag. Supplier price lists may move around the duty date, so put a full price review in the diary for the first full week of February 2027.

Why does pour cost drift when nobody is over-pouring?

Most drift has nothing to do with heavy-handed staff. It comes from three quieter sources.

Price lag. A keg, case or bottle goes up, the invoice is paid, and the serve cost on the spreadsheet stays where it was. The menu price follows weeks or months later, if at all.

Spec drift. A cocktail gains a premium mixer, a house wine moves to a larger glass, a garnish changes. Nobody recosts the serve, so the menu price is anchored to a recipe that no longer exists.

Yield loss. Line cleaning, fobbing, ullage, breakages and unrecorded staff drinks or comps mean fewer saleable measures come out of each keg or bottle than the costing assumes.

A monthly stocktake shows the combined result of all three, but only after the money has gone. Control means catching each one while it is still a decision.

What is the Invoice-to-Glass Loop?

The Invoice-to-Glass Loop is the four-stage routine we recommend for keeping pour cost on target as prices change. Each stage asks one question and needs one piece of evidence. Run it weekly, on the same day, by the same person.

Land: did a cost change arrive this week?

Check every delivery against the last price paid for that line. Any change, up or down, is flagged on the day it lands, with the supplier, the old price and the new price. This is where most pubs lose the thread, because the invoice goes to the accounts pile rather than to the bar manager.

Cost: has every serve using that line been recosted?

One price change can touch a dozen serves: the pint, the half, the double, three cocktails and a pitcher. Update the cost of every serve that uses the line, not just the obvious one, and use realistic yields rather than the theoretical maximum.

Price: does the serve still sit inside its target band?

Compare each recosted serve with the target band you set for its category. Draught, spirits, wine, cocktails and soft drinks usually carry different targets, and the right numbers depend on your venue, your trade mix and your local market. Where a serve has slipped out of band, choose a response from the table below and record the decision with a date.

Pour: is the glass matching the spec?

A correct costing only protects margin if the glass matches it. Each week, count your fastest-moving and highest-value lines, reconcile usage against EPOS sales, and price any gap in pounds. If theoretical and actual pour cost diverge, the problem is in the pour, the till or the cellar, not the price.

How much does a slow price review cost a pub?

Illustrative example: the figures below are invented for clarity. They are not taken from any customer, benchmark or survey.

A pub sells a lager at £5.80 a pint, about £4.83 net of VAT at the standard 20% rate. It pays £135 for an 11-gallon keg (88 pints) and, after line cleaning and fobbing, gets around 95% yield, or about 84 saleable pints. That puts the cost of a sold pint at about £1.62 and pour cost at 33.4%.

The supplier then raises the keg to £143. Cost per sold pint rises by roughly 10p to £1.71, and pour cost moves to 35.4%. On a line selling 400 pints a week, that is about £38 a week of margin gone. If the menu price is reviewed quarterly, around 12 weeks can pass before anyone acts: roughly £460 on one line, before counting the other serves that use the same keg.

To return that pint to a 33.4% pour cost, the selling price would need to move to about £6.15. Whether that is the right call depends on your trade, which is why the decision belongs in the Price stage rather than in a formula.

“The keg went up in March and the pint price moved in June. That gap was the quarter’s GP.”

Illustrative quote, written to reflect a common operator experience. Not a customer testimonial.

Run your own top three lines through the free beverage cost calculator to see where they sit today.

Should you absorb, re-price, re-spec or switch supplier?

When a serve drifts out of band, there are four realistic responses. Most pubs use all four across a menu. The mistake is defaulting to one.

Whichever you choose, write it down. A decision with a date can be reviewed. A price nobody remembers setting cannot.

What does a weekly pour cost routine look like?

This is the Invoice-to-Glass Loop as a weekly checklist. For a single site it is usually one focused session. Groups run it per site and compare.

1. Pick a fixed day and owner, ideally after the week’s main deliveries.
2. List every line whose delivered price changed since last week, with old and new prices.
3. Recost every serve that uses those lines, including halves, doubles, cocktails and pitchers.
4. Sort serves by gap to target and review the five furthest out of band.
5. Choose absorb, re-price, re-spec or switch for each, and record the decision and date.
6. Count your top movers and high-value lines, reconcile usage against EPOS sales, and price the gap in pounds.
7. Assign one fix for the biggest gap and check it again next week.

Where does software fit in pour cost control?

The loop works on paper. It breaks down when the Land and Cost stages depend on someone retyping invoices into a spreadsheet at the end of a long week.

Supplier price tracking records each delivered price against the last one, so a change is visible when the stock arrives. Serve costs can then update from the same data, and EPOS integrations with Toast, Lightspeed, Aloha EPOS, TISSL, Grafterr and HubRise reconcile stock against what was actually sold. Counts run on a phone or tablet, with no extra hardware.

For pubs and bars, bar inventory management software brings the four stages into one view by line, category and site. It does not choose your prices. It shortens the gap between a cost change and the moment you can decide what to do about it.

Where should you start this week?

Start with one number. Take your three best-selling draught lines and your house spirit, run each through the beverage cost calculator, and compare the result with your target. If any are out of band, work through the Invoice-to-Glass Loop for those lines first.

If you are reading this before 1 February 2027, run the loop on your draught and spirit lines now, then again once February delivery prices land. If you want to see the loop running on live supplier prices and EPOS sales across your bar, Book a Demo with the StockTake Online team.

Key takeaways

• Pour cost control in 2026/27 is mainly about speed: the gap between a cost rise and a menu decision.
• Alcohol duty rises 4.55% from 1 February 2027, but the bigger exposure is wholesale increases landing unnoticed on invoices.
• Drift comes from price lag, spec drift and yield loss, not only from over-pouring.
• The Invoice-to-Glass Loop runs four stages weekly: Land, Cost, Price, Pour.
• Every out-of-band serve gets one recorded response: absorb, re-price, re-spec or switch.

Frequently asked questions

What is the difference between pour cost and beverage GP?

Pour cost is a drink’s cost as a percentage of its selling price net of VAT, and beverage GP is the remainder. A serve with a 30% pour cost has a 70% GP. Both should be calculated on net prices, never on the price including VAT.

How often should a pub review drink prices?

Review serve costs weekly and change menu prices when a serve stays outside its target band, rather than waiting for a fixed quarterly review. Weekly checks catch supplier increases while they are still a few pence per serve.

How will the February 2027 alcohol duty rise affect pour cost?

Only slightly on its own. Duty rises 4.55% from 1 February 2027, which on a pint of 4% draught beer is about 2p. The larger effect usually comes from wholesale price increases that follow, so track delivered prices from February rather than the duty rate alone.

Should a pub pass every supplier price rise on to customers?

Not automatically. A small or temporary rise on a price-sensitive line may be worth absorbing, while a permanent rise on a line with price headroom usually justifies a menu change. The point is to decide deliberately and record the decision.

Can software control pour cost without a full weekly stocktake?

Software can flag price changes and recost serves automatically, but actual pour cost still needs counts. Most bars count their top movers and high-value lines weekly and run a full count less often, using EPOS data to target the lines that matter.

Pub stocktaking works best as a short weekly loop. Count your wet stock at the same point each week, record what is physically there, compare it with what deliveries and till sales say should be there, then investigate the gaps that keep repeating. That turns a count into a margin check.

A weekly food cost check is a fixed 30-minute review that shows what changed in your kitchen costs this week, while there is still time to act. You look at six things in order: supplier prices, recipe costs, stock, food cost percentage, GP% and one issue to fix.

Q4 stock planning means ordering for October to December from counted stock, confirmed bookings and supplier lead times, not habit. Cover Christmas by raising stock only on proven lines. Then give every seasonal item an exit plan before you order it. That is how you avoid a storeroom of dead stock in January.

A stock transfer between sites only works when one record moves the stock out of the sending site and into the receiving site at the same time, at an agreed cost, with both managers signing it off. Spreadsheets, WhatsApp messages and paper notes rarely manage that. They capture one end, late, at a guessed value.

The restaurant purchase order process is the controlled path every ingredient order follows, from a request based on stock levels to a checked delivery and a matched invoice. Run it as six gates and each one blocks a specific leak: over-ordering, unapproved spend, price drift or short delivery.

The restaurant procurement process links stock levels, par levels, supplier management and purchasing, so every order reflects what the kitchen actually needs. Done well, it takes you from a stock count to a checked delivery without guesswork, and protects food cost along the way.

Your food cost report says one thing. Your stock count says another. For an Operations Director running more than one site, that gap is not a rounding error. It is where margin quietly disappears.