# Hospodality — Hospitality business maths, worked through Source: https://www.hospodality.com · hello@hospodality.com ## Business maths (55) ### Costing #### Dish costing: a worked example Dish costing means adding up what every ingredient on the plate costs you, then comparing that with the price you charge after VAT has been taken off. In this example we cost a main course of pan-roasted chicken breast with crushed new potatoes, greens and a jus. The chicken supreme is bought at £10.00 per kilo and each portion weighs 200 grams, so the chicken costs 0.2 x £10.00 = £2.00. The new potatoes cost £1.20 per kilo and each portion uses 180 grams, so 0.18 x £1.20 = £0.216, call it £0.22. Greens cost £4.00 per kilo at 100 grams a portion, so 0.1 x £4.00 = £0.40. The jus is batch-made at £6.00 for 3 litres, which is £0.002 per millilitre; a 60 ml portion costs 60 x £0.002 = £0.12. Butter, oil and seasoning are hard to weigh per plate, so many kitchens add a fixed 'sundries' allowance; in this example we add £0.15. The total is £2.00 + £0.22 + £0.40 + £0.12 + £0.15 = £2.89. Now the selling side. The menu price is £18.00 including VAT at 20%. To find the net price, divide by 1.2: £18.00 / 1.2 = £15.00. The VAT portion, £3.00, is not your money; it goes to HMRC. Food cost percentage is cost divided by net price: £2.89 / £15.00 = 0.1927, so about 19.3%. Gross profit is £15.00 - £2.89 = £12.11, which is a GP of 80.7%. That is a healthy dish. A common mistake is to work the percentage off the VAT-inclusive price: £2.89 / £18.00 = 16.1%, which makes the dish look better than it is. Always strip VAT first. A second mistake is to cost with the 'as purchased' weight when the recipe uses the trimmed weight; if the chicken arrived on the bone, the cost per usable kilo would be higher and should be worked out with a yield test first. Third, prices move. If chicken rises to £11.50 per kilo, the chicken cost becomes 0.2 x £11.50 = £2.30, the total becomes £3.19, and the food cost becomes £3.19 / £15.00 = 21.3%. That two-point shift on one dish might look small, but across a menu selling hundreds of portions a week it adds up quickly. Keep a costing sheet for every dish, date it, and review it whenever a supplier price list changes. Many operators set a target food cost range for each section of the menu rather than one number for everything: starters and desserts often run cheaper, steaks and seafood often run higher, and the blend is what matters for the overall margin. - In this example a £18.00 main is £15.00 net of 20% VAT. - Ingredient cost of £2.89 gives a food cost of 19.3% and GP of 80.7%. - Always divide the menu price by 1.2 before working out percentages. - A sundries allowance covers butter, oil and seasoning. - Re-cost dishes whenever a supplier changes a price. #### Cocktail costing: a worked example Cocktail costing works the same way as dish costing, but the units are millilitres and the bottle is the purchase unit. In this example we cost a classic sour-style drink made with 50 ml of gin, 25 ml of fresh lemon juice, 20 ml of sugar syrup and a dash of aquafaba, served over ice with a lemon twist. The house gin costs £21.00 for a 70 cl bottle, which is 700 ml. The cost per millilitre is £21.00 / 700 = £0.03. Fifty millilitres therefore cost 50 x £0.03 = £1.50. Lemons cost £0.30 each and in this example one lemon yields about 30 ml of juice once squeezed, so each millilitre of juice costs £0.30 / 30 = £0.01, and 25 ml cost £0.25. Sugar syrup is made in-house from sugar and water; a litre costs roughly £0.80 to make, so 20 ml cost 20 x £0.0008 = £0.016, call it £0.02. Aquafaba from a tin of chickpeas is close to free if the chickpeas are used elsewhere; allow £0.03. The lemon twist uses the peel of a lemon that has already been juiced, so we allow £0.02 for garnish. Ice is not free either: a bar with its own ice machine pays for water, electricity and maintenance, and a bar buying bagged ice pays for the bags. In this example allow £0.05. The total cost is £1.50 + £0.25 + £0.02 + £0.03 + £0.02 + £0.05 = £1.87. The drink sells for £11.00 including VAT. Net of VAT: £11.00 / 1.2 = £9.1667, call it £9.17. Liquor cost percentage is £1.87 / £9.17 = 20.4%, so the gross profit is £9.17 - £1.87 = £7.30, or 79.6%. Now consider spillage and over-pouring. If bartenders free-pour and the average pour drifts to 55 ml, the gin cost becomes 55 x £0.03 = £1.65, the total becomes £2.02 and the cost percentage becomes £2.02 / £9.17 = 22.0%. On a bar selling 600 of these a month, that drift costs 600 x £0.15 = £90 a month in gin alone. That is why jiggers and stock counts matter. Premium spirits change the picture. Swap in a gin at £36.00 per 70 cl and the cost per millilitre becomes £36.00 / 700 = £0.0514; 50 ml then costs £2.57, and the drink costs £2.94 in total. To keep the same 79.6% margin you would need a net price of £2.94 / 0.204 = £14.41, or £17.29 with VAT. Most bars do not push the full margin through on premium serves; they accept a lower percentage but a higher cash profit per glass. In this example the premium drink at £13.00 makes £10.83 net and £7.89 cash profit, more than the house serve in pounds even though the percentage is lower. - In this example a 70 cl bottle at £21.00 costs £0.03 per ml. - The house cocktail costs £1.87 and sells at £9.17 net, a 79.6% GP. - A 5 ml over-pour on 600 drinks costs £90 a month here. - Ice and garnish have a real cost and belong on the spec sheet. - Premium serves often earn more cash per glass at a lower percentage. #### Pint GP: a worked example Draught beer is bought by the keg or cask and sold by the pint, so the gross profit calculation starts with how many saleable pints a container really gives you. In this example a pub buys an 11-gallon keg of lager for £150.00 before VAT. An imperial gallon holds 8 pints, so 11 gallons hold 11 x 8 = 88 pints. Not every pint is sold. Line cleaning, fobbing, the first pour after a change and the odd spill all waste beer. In this example we allow 4% wastage: 88 x 0.04 = 3.52 pints lost, leaving 84.48 saleable pints. The cost per saleable pint is £150.00 / 84.48 = £1.776, call it £1.78. The pint sells for £6.00 including VAT. Net: £6.00 / 1.2 = £5.00. Gross profit per pint is £5.00 - £1.78 = £3.22, and the GP percentage is £3.22 / £5.00 = 64.4%. Now look at what wastage does. If line cleaning is sloppy and wastage rises to 8%, saleable pints fall to 88 x 0.92 = 80.96, the cost per pint becomes £150.00 / 80.96 = £1.853, and the GP falls to (£5.00 - £1.853) / £5.00 = 62.9%. That 1.5-point fall looks minor until you multiply it out. If the pub sells 1,200 pints of this lager a week, the difference in cost per pint is about £0.077, and 1,200 x £0.077 = £92.40 a week, or roughly £4,800 a year from one tap. Cask ale has its own wrinkles. A 9-gallon firkin holds 72 pints, but sediment and settling mean the realistic yield is lower; in this example allow 6 pints of loss, leaving 66 saleable. If the firkin costs £110.00, each pint costs £110.00 / 66 = £1.667. Selling at £5.40 with VAT, the net is £4.50 and the GP is (£4.50 - £1.667) / £4.50 = 63.0%. Cask also has a shelf life once tapped, often a few days, so a slow-selling ale can lose far more than the planned 6 pints. If only 50 pints sell before the beer turns, the effective cost per pint becomes £110.00 / 50 = £2.20 and the GP drops to 51.1%. The practical lesson is to match the range to the rate of sale: fewer taps that turn over quickly usually beat a long bar of beers that sit in the line. Many pubs track 'yield' weekly by comparing pints rung through the till with pints the cellar should have delivered, and they investigate any gap above a set tolerance. - An 11-gallon keg holds 88 imperial pints. - In this example a £150 keg with 4% waste costs £1.78 per pint. - A £6.00 pint is £5.00 net of VAT, giving a 64.4% GP here. - Doubling wastage on 1,200 pints a week costs about £92 a week. - Cask that turns before it sells can cut GP sharply. #### Wine markup ladder: a worked example A wine list rarely uses one markup across every bottle. Most venues use a ladder: high multiples on cheap wines and lower multiples on expensive ones, so the guest is not punished for trading up and the venue still earns good cash. In this example the ladder works on the cost price before VAT and produces a menu price including VAT. Step one is the house wine, bought at £6.00 a bottle. Using a multiple of 3.5 on cost, the net price is £6.00 x 3.5 = £21.00, and the menu price with VAT is £21.00 x 1.2 = £25.20, rounded to £25.00. At £25.00 the net is £25.00 / 1.2 = £20.83, so the gross profit is £20.83 - £6.00 = £14.83, a GP of 71.2%. Step two is a mid-range bottle bought at £12.00. The ladder uses a multiple of 3.0: £12.00 x 3 = £36.00 net, £43.20 with VAT, rounded to £43.00. Net at £43.00 is £35.83, profit is £23.83, GP is 66.5%. Step three is a bottle bought at £25.00, with a multiple of 2.5: £62.50 net, £75.00 with VAT. Profit is £62.50 - £25.00 = £37.50 and the GP is 60.0%. Step four is a fine wine bought at £60.00, with a multiple of 2.0: £120.00 net, £144.00 with VAT, rounded to £145.00. Net at £145.00 is £120.83, profit is £60.83, GP is 50.3%. Compare the cash. The house wine makes £14.83 per bottle; the fine wine makes £60.83, more than four times as much, even though its percentage is twenty points lower. This is the core idea: percentages pay for the business on volume lines, and cash margins reward trading up. Some operators add a fixed cash margin instead of a multiple at the top end, for example cost plus £50.00 net, which makes premium bottles look like better value and can move slow stock. Glass pricing sits on top of this. A 750 ml bottle gives four 175 ml glasses with 50 ml left over, or five 150 ml glasses exactly. If the house wine sells at £7.50 for a 175 ml glass, four glasses bring £30.00 with VAT, or £25.00 net, against a bottle cost of £6.00. But an opened bottle that is not finished goes to waste, so many venues assume they will lose some of every bottle sold by the glass. In this example, if one bottle in six is poured away half full, the effective cost per bottle rises by about £0.50. Review the ladder at least once a year, and check that each step on the list leaves a sensible price gap so guests can see the choice clearly. - In this example multiples fall from 3.5x at £6 cost to 2x at £60 cost. - House wine makes £14.83 a bottle; fine wine makes £60.83. - Higher-priced bottles carry lower percentages but more cash. - A 750 ml bottle gives four 175 ml glasses plus 50 ml. - Allow for wastage on wines sold by the glass. #### Yield and waste: a worked example Yield is the share of a raw ingredient that ends up on the plate after trimming, boning, peeling and cooking. Costing from the purchase price without a yield test is one of the most common reasons a menu earns less than it should. In this example a kitchen buys a whole salmon weighing 5.0 kg at £14.00 per kilo, so the fish costs 5.0 x £14.00 = £70.00. Step one is the butchery test. After removing the head, frame, skin, pin bones and belly trim, the kitchen weighs the clean fillet at 2.9 kg. The trim is not worthless: the frame and head make a stock, and the belly becomes a staff meal or a tartare special. In this example we give the by-products a notional value of £5.00, so the cost carried by the fillet is £70.00 - £5.00 = £65.00. Step two is the yield percentage: 2.9 / 5.0 = 58%. Step three is the true cost per usable kilo: £65.00 / 2.9 = £22.41. That is 60% more than the £14.00 printed on the invoice. Step four is the portion. The menu uses a 160 gram portion, and the 2.9 kg of fillet gives 2,900 / 160 = 18.125 portions, so 18 full portions with 20 grams over. The cost per portion is 0.16 x £22.41 = £3.59. Had the chef used the invoice price, they would have costed 0.16 x £14.00 = £2.24, and the dish would have been under-costed by £1.35 a plate. Cooking loss is a second stage. If a dish is costed on cooked weight, such as a braised beef shin served at 180 grams cooked, the kitchen must know how much raw meat makes that portion. In this example the shin loses 35% of its weight in braising, so the cooked yield is 65%. The raw weight needed is 180 / 0.65 = 277 grams. At £12.00 per kilo raw, each portion costs 0.277 x £12.00 = £3.32. Waste outside of preparation also counts: over-production, spoilage, returned plates and dropped food. A simple waste log, filled in with item, weight, reason and an estimated cost, often reveals patterns within a fortnight. In this example a kitchen logging £280 of waste in a week against food purchases of £7,000 is wasting 4% of what it buys. Bringing that down to 2% saves £140 a week, about £7,280 a year, without changing a single menu price. Repeat yield tests when a supplier, a specification or a chef changes, because yields move with knife skills and with the size and grade of what arrives. - In this example a 5 kg salmon yields 2.9 kg of fillet, 58%. - True cost per usable kilo is £22.41 against £14.00 on the invoice. - Raw weight needed equals cooked weight divided by cooked yield. - A waste log turns guesswork into numbers within weeks. - Cutting waste from 4% to 2% of £7,000 saves £140 a week. #### Recipe scaling: a worked example Recipe scaling means changing the quantity a recipe makes while keeping the dish the same. The tool is a conversion factor: the new yield divided by the old yield. Multiply every ingredient by that factor and the cost scales with it. In this example a kitchen has a tomato sauce recipe that makes 4 litres and wants to make 10 litres for a busy weekend. The conversion factor is 10 / 4 = 2.5. The original recipe uses 3.2 kg of tinned tomatoes, 400 grams of onions, 40 grams of garlic, 100 ml of olive oil, 20 grams of salt and 10 grams of sugar. Multiplying each by 2.5 gives 8.0 kg of tomatoes, 1,000 grams of onions, 100 grams of garlic, 250 ml of olive oil, 50 grams of salt and 25 grams of sugar. Now the cost. In this example tinned tomatoes cost £1.50 per kilo, onions £0.90 per kilo, garlic £6.00 per kilo, olive oil £8.00 per litre, salt £0.60 per kilo and sugar £1.00 per kilo. The original 4 litres cost: tomatoes 3.2 x £1.50 = £4.80; onions 0.4 x £0.90 = £0.36; garlic 0.04 x £6.00 = £0.24; oil 0.1 x £8.00 = £0.80; salt 0.02 x £0.60 = £0.012; sugar 0.01 x £1.00 = £0.01. Total £6.222, or about £1.56 per litre. The scaled batch should cost £6.222 x 2.5 = £15.555, still £1.56 per litre. A portion of 120 ml costs 0.12 x £1.556 = £0.19. The arithmetic is simple, but scaling has traps. First, seasoning and aromatics do not always scale linearly; salt, chilli and spices are often added at about 80 to 90% of the scaled figure and then adjusted by taste. Second, evaporation does not scale with volume: a bigger pot has proportionally less surface area for its volume, so a large batch may need less reduction time and can end up thinner. Third, equipment limits matter. If the largest pan holds 8 litres, a 10 litre batch must be split, and the split must use the same factor in each pan. Scaling down has the reverse problems. If a pastry section wants to make a quarter of a recipe that uses 3 eggs, the factor of 0.25 gives 0.75 eggs, which in practice means weighing the egg; a medium egg is roughly 50 grams without the shell, so 0.75 eggs is about 38 grams. Many kitchens write recipes by weight for exactly this reason, including liquids. Finally, record the scaled version once it has been tested so that the next cook does not have to repeat the calculation under pressure. - Conversion factor equals new yield divided by old yield. - In this example the factor is 10 / 4 = 2.5. - The sauce costs about £1.56 per litre at any batch size. - Seasoning is often scaled to 80 to 90% and adjusted by taste. - Writing recipes by weight makes scaling far easier. #### Coffee cost per cup: a worked example Coffee looks like a high-margin product, and it usually is, but only if the cost per cup includes everything that goes into it. In this example we cost a 12 oz flat white and a 12 oz oat latte in a café. Start with the coffee. Beans cost £22.00 per kilo. The recipe is an 18 gram double shot, so the theoretical number of doses per kilo is 1,000 / 18 = 55.6. Grinder purging, dialling-in each morning and the odd binned shot waste coffee; in this example allow 10%, leaving 50 usable doses. Coffee cost per drink is £22.00 / 50 = £0.44. Next, milk. A flat white in this example uses about 200 ml of whole milk once steaming loss and the jug leftovers are included. Whole milk costs £1.10 per litre, so 0.2 x £1.10 = £0.22. Oat milk costs £2.20 per litre, so the same 200 ml costs £0.44. Packaging for takeaway adds a cup, lid and sleeve; in this example £0.18. A sugar sachet, stirrer and napkin add about £0.03. The takeaway flat white costs £0.44 + £0.22 + £0.18 + £0.03 = £0.87. The oat latte costs £0.44 + £0.44 + £0.18 + £0.03 = £1.09. Now the price. The flat white sells for £3.60. Hot drinks are standard-rated for VAT whether drunk in or taken away, so in this example both drinks carry 20% VAT. Net price: £3.60 / 1.2 = £3.00. Gross profit: £3.00 - £0.87 = £2.13, a GP of 71.0%. If the oat latte also sells at £3.60, its GP is (£3.00 - £1.09) / £3.00 = 63.7%. Charging a £0.40 plant milk supplement brings the oat latte to £4.00, a net of £3.33, and a GP of (£3.33 - £1.09) / £3.33 = 67.3%. Some cafés choose not to charge extra for plant milk as a policy, and absorb the cost; that is a valid choice as long as it is a deliberate one. The cost per cup is only part of the story. Labour to make each drink, the lease on the espresso machine, water filtration and servicing all sit in overheads, not in the cup cost. In this example a machine lease of £150 a month spread across 3,000 drinks adds £0.05 a cup. Keep a separate calculation for that when deciding whether to add a second group head or a second grinder. Finally, check weights regularly. If a barista doses 20 grams instead of 18, the coffee cost per drink rises by about 11% and the drink may taste worse too. - In this example an 18 g dose from £22/kg beans with 10% waste costs £0.44. - A takeaway flat white costs £0.87 against a £3.00 net price. - Hot drinks are standard-rated for VAT whether drunk in or taken away. - Oat milk doubles the milk cost here, from £0.22 to £0.44. - Machine lease and servicing belong in overheads, not cup cost. #### Menu-wide food cost: a blended worked example A single dish's food cost tells you how that plate performs. The number the business lives on is the blended food cost across everything sold, weighted by how much of each dish actually leaves the pass. In this example a small bistro sells five dishes in a week. Dish A, a soup, sells 150 portions at £7.20 with VAT, which is £6.00 net, costing £1.20 each. Dish B, a burger, sells 300 portions at £16.20, which is £13.50 net, costing £4.05 each. Dish C, a steak, sells 120 portions at £30.00, which is £25.00 net, costing £10.00 each. Dish D, a risotto, sells 180 portions at £15.60, which is £13.00 net, costing £2.60 each. Dish E, a dessert, sells 250 portions at £8.40, which is £7.00 net, costing £1.40 each. Now calculate the sales and cost for each line. Soup: sales 150 x £6.00 = £900; cost 150 x £1.20 = £180. Burger: sales 300 x £13.50 = £4,050; cost 300 x £4.05 = £1,215. Steak: sales 120 x £25.00 = £3,000; cost 120 x £10.00 = £1,200. Risotto: sales 180 x £13.00 = £2,340; cost 180 x £2.60 = £468. Dessert: sales 250 x £7.00 = £1,750; cost 250 x £1.40 = £350. Total net sales are £900 + £4,050 + £3,000 + £2,340 + £1,750 = £12,040. Total cost is £180 + £1,215 + £1,200 + £468 + £350 = £3,413. Blended food cost is £3,413 / £12,040 = 28.3%. Notice that the individual dish percentages are 20%, 30%, 40%, 20% and 20%. A simple average of those five would be 26%, which is wrong because it ignores volume. The steak, at 40%, drags the blend up because it sells well. If a promotion pushed steak sales up to 200 portions, steak sales would become £5,000 and steak cost £2,000; total sales would be £14,040 and total cost £4,213, and the blend would rise to 30.0%. Is that bad? Not necessarily. Gross profit in pounds would rise from £12,040 - £3,413 = £8,627 to £14,040 - £4,213 = £9,827, an extra £1,200. Percentage went the wrong way while cash went the right way. This is why owners should look at both. The theoretical blend from recipe cards should also be compared with the actual food cost from stock records: opening stock plus purchases minus closing stock, divided by net food sales. In this example, if the stock calculation shows £3,800 used against £12,040 of sales, the actual cost is 31.6%, and the gap of £387 points to waste, over-portioning, unrecorded staff food or theft. - Blended food cost must be weighted by volume sold. - In this example the blend is £3,413 / £12,040 = 28.3%. - Selling more steak raised the percentage but added £1,200 of GP. - Actual cost equals opening stock plus purchases minus closing stock. - A gap between theoretical and actual cost points to losses. #### Afternoon tea costing: a worked example Afternoon tea combines many small items, each cheap on its own, into a premium-priced experience. Costing it needs care because a dozen small components add up, and labour is high relative to the food cost. In this example a hotel tea room prices a classic afternoon tea at £42 per person including VAT. The net price is £42 / 1.2 = £35.00. Step one: savouries. Four finger sandwiches per person: egg mayonnaise £0.30, cucumber £0.20, smoked salmon £0.90 and ham with mustard £0.45, totalling £1.85. A small savoury tart: £0.60. Savouries total £2.45. Step two: scones. Two scones per person at £0.25 each, £0.50; clotted cream at 40 grams, priced at £8.00 per kilo, so 0.04 x £8.00 = £0.32; jam at 30 grams, priced at £5.00 per kilo, £0.15. Scones total £0.97. Step three: pastries. Four small pastries per person: a macaron £0.45, a lemon tart £0.50, a chocolate delice £0.55 and a fruit tartlet £0.60, totalling £2.10. Step four: tea. Loose-leaf tea at £40.00 per kilo, 6 grams per pot, so 0.006 x £40.00 = £0.24, and guests often have a second pot, so £0.48. Milk, lemon and sugar £0.10. Tea totals £0.58. Step five: total food and drink cost. £2.45 + £0.97 + £2.10 + £0.58 = £6.10. Food and drink cost percentage is £6.10 / £35.00 = 17.4%, and GP is £28.90, or 82.6%. Step six: the labour. Afternoon tea is labour-heavy. In this example the pastry and garde-manger work takes about 10 minutes of chef time per person, and service, including tea refills, takes about 12 minutes. At a loaded rate of £16 an hour, 22 minutes cost 22 / 60 x £16 = £5.87. Food, drink and direct labour total £6.10 + £5.87 = £11.97, which is 34.2% of the net price. Step seven: the champagne upgrade. Adding a glass of champagne for £14 extra including VAT adds £11.67 net. A 125 ml glass from a bottle costing £24.00 is one-sixth of the bottle, £4.00, if the bottle is fully used. The upgrade earns £11.67 - £4.00 = £7.67 of extra GP. Step eight: waste. Afternoon tea is often prepared in advance, so unsold pastries and sandwiches are a real risk. Pre-booking and a minimum notice period, such as 24 hours, help the kitchen prepare accurately. In this example, if 10% of pastries prepared each day go unsold, the pastry cost rises effectively from £2.10 to £2.10 / 0.9 = £2.33 per guest. - In this example food and drink cost £6.10 against a £35.00 net price. - The GP on food and drink is 82.6%. - Direct labour of £5.87 per guest takes the prime cost to 34.2%. - A champagne upgrade adds £7.67 of GP per guest here. - Pre-booking cuts waste on prepared items. ### P&L #### A small café P&L: a worked example A profit and loss account, or P&L, shows what a business earned and spent over a period and what was left. In this example we build a monthly P&L for a 30-seat neighbourhood café open seven days, all figures excluding VAT. Sales first. The café serves an average of 220 customers a day with an average spend of £8.50 net. Over a 30-day month that is 220 x 30 = 6,600 customers, and 6,600 x £8.50 = £56,100 of net sales. Cost of sales comes next. In this example food and drink purchases, adjusted for stock, come to 30% of sales: 0.30 x £56,100 = £16,830. Gross profit is £56,100 - £16,830 = £39,270, a gross margin of 70%. Labour is usually the biggest cost after that. The café employs staff whose wages, employer National Insurance, pension contributions and holiday pay total £18,500 for the month, which is £18,500 / £56,100 = 33.0% of sales. Now the overheads. Rent is £4,000 a month. Business rates are £700 a month after any relief. Utilities, meaning gas, electricity and water, are £1,600. Card processing fees at 1.5% of sales come to £842. Repairs and maintenance are £500. Cleaning supplies and laundry are £400. Insurance, licences and subscriptions are £350. Marketing is £300. Accountancy and bookkeeping are £250. Packaging not already counted in cost of sales is £600. Total overheads are £4,000 + £700 + £1,600 + £842 + £500 + £400 + £350 + £300 + £250 + £600 = £9,542. Operating profit, before interest, depreciation and tax, is gross profit minus labour minus overheads: £39,270 - £18,500 - £9,542 = £11,228. As a share of sales that is £11,228 / £56,100 = 20.0%. That is a strong month. Below this line come depreciation of equipment, say £600, and interest on a loan, say £250, leaving a profit before tax of £11,228 - £600 - £250 = £10,378. Corporation tax would then apply to the company's taxable profit, which is calculated under tax rules and is not the same as the accounting profit. The P&L is most useful when it is compared with something: last month, the same month last year, or a budget. If next month's sales fall 15% to £47,685 while rent, rates and most overheads stay the same, gross profit becomes £33,380, and if labour can only be trimmed to £17,000, operating profit falls to £33,380 - £17,000 - £9,415 = £6,965, using card fees of £715 in place of £842. A 15% fall in sales has cut profit by 38%. This is operational gearing: fixed costs magnify every swing in sales, in both directions. - In this example 220 customers a day at £8.50 net gives £56,100 a month. - Gross profit is £39,270 at a 30% cost of sales. - Operating profit of £11,228 is 20% of sales. - A 15% fall in sales cuts profit by about 38% here. - Fixed costs magnify swings in sales: operational gearing. #### A 60-cover restaurant P&L: a worked example In this example we build an annual P&L for a 60-cover restaurant open five days a week for dinner and two days for lunch as well, all figures excluding VAT. Covers first. Dinner averages 75 covers a night, which means some tables turn, across 5 nights a week for 50 weeks: 75 x 5 x 50 = 18,750 covers. Lunch averages 40 covers on 2 days for 50 weeks: 40 x 2 x 50 = 4,000 covers. Total covers are 22,750. The average spend per head at dinner is £52 net, and at lunch £32 net. Dinner sales are 18,750 x £52 = £975,000; lunch sales are 4,000 x £32 = £128,000. Total net sales are £1,103,000. In this example food is 70% of sales and drink 30%, so food sales are £772,100 and drink sales £330,900. Food cost runs at 29% and drink cost at 26%. Food cost is 0.29 x £772,100 = £223,909; drink cost is 0.26 x £330,900 = £86,034. Total cost of sales is £309,943, and gross profit is £1,103,000 - £309,943 = £793,057, a gross margin of 71.9%. Labour, including all employer costs but excluding tips and service charge, which in this example are passed to staff through a separate tronc and are not the restaurant's revenue, comes to £375,000, or 34.0% of sales. Overheads in this example are: rent £90,000; business rates £28,000; utilities £38,000; repairs and maintenance £18,000; card fees at 1.6% £17,648; laundry and linen £12,000; cleaning and chemicals £9,000; marketing and reservations software £15,000; insurance and licences £8,000; accountancy, payroll and legal £12,000; small equipment and crockery replacement £10,000; and sundries £6,000. Total overheads are £263,648. Operating profit, or EBITDA, is £793,057 - £375,000 - £263,648 = £154,409, which is 14.0% of sales. Depreciation on the fit-out, say £45,000 a year, and loan interest of £12,000 bring profit before tax to £97,409, or 8.8% of sales. Now test it. Prime cost, meaning cost of sales plus labour, is £309,943 + £375,000 = £684,943, which is 62.1% of sales. Many operators aim to keep prime cost somewhere around the low 60s or below, though the right figure depends on the format. Suppose the restaurant lost 10 dinner covers a night. That is 10 x 5 x 50 = 2,500 covers at £52, or £130,000 of sales. At a 71.9% gross margin that removes about £93,470 of gross profit. If labour could be cut by only £20,000, EBITDA would fall to roughly £154,409 - £93,470 + £20,000 = £80,939. Ten fewer covers a night roughly halves the operating profit, which is why reservations, repeat guests and table turns get so much attention. - In this example 22,750 covers a year produce £1,103,000 of net sales. - Gross margin is 71.9% with a 70/30 food and drink split. - EBITDA is £154,409, or 14.0% of sales. - Prime cost is 62.1% of sales here. - Losing 10 dinner covers a night roughly halves EBITDA. #### A wet-led pub P&L: a worked example A wet-led pub earns most of its money from drinks rather than food. In this example we build an annual P&L for a leased community pub where drink is 80% of sales and a simple food offer is 20%, all figures excluding VAT. Weekly net sales average £14,000, so annual sales over 52 weeks are £728,000. Drink sales are 0.80 x £728,000 = £582,400 and food sales are £145,600. In this example the blended drink GP is 62%, so drink cost of sales is 0.38 x £582,400 = £221,312. Food GP is 65%, so food cost is 0.35 x £145,600 = £50,960. Total cost of sales is £272,272. Gross profit is £728,000 - £272,272 = £455,728, a gross margin of 62.6%. That margin is lower than a restaurant's because draught beer and spirits carry lower percentages than food. The compensation is that a pint needs far less labour than a plated meal. Labour in this example is £160,000, or 22.0% of sales. Machine income from gaming and amusement machines, after the supplier's share, adds £8,000 a year in this example and sits as other income. Overheads: rent to the landlord £65,000; business rates £16,000; utilities £30,000; repairs and maintenance £14,000; card fees at 1.4% £10,192; entertainment such as quiz hosts and bands £15,600, which is £300 a week; music licensing £2,500; television sports subscription £18,000; insurance £5,000; cleaning and glass washing supplies £6,000; accountancy £5,000; marketing £3,000; and sundries £4,000. Total overheads are £194,292. EBITDA is gross profit plus other income minus labour minus overheads: £455,728 + £8,000 - £160,000 - £194,292 = £109,436, which is 15.0% of sales. Now consider the sports subscription. At £18,000 a year it costs £346 a week. If big matches bring in an extra 400 pints a week at a £3.22 GP each, that is £1,288 a week of extra gross profit, so the subscription pays for itself more than three times over. If instead the extra trade is only 100 pints, the uplift is £322 a week, and the subscription loses money. The same test applies to the quiz host: £300 a week needs about 300 / 3.22 = 93 extra pints, or an equivalent mix of drinks and food, just to break even. Wet-led pubs are sensitive to the price they pay for beer, so a tied lease in which the pub must buy beer from the landlord at a set price can change the drink GP dramatically. In this example a tie that lowers drink GP from 62% to 55% would reduce gross profit by 0.07 x £582,400 = £40,768, cutting EBITDA by more than a third. - In this example £14,000 a week gives £728,000 of annual net sales. - Drink is 80% of sales and the overall gross margin is 62.6%. - Labour is 22% of sales, lower than a restaurant's. - A £300 quiz night needs about 93 extra pints to pay for itself. - A tie that cuts drink GP by 7 points removes £40,768 here. #### A 40-room hotel P&L: a worked example Hotel P&Ls are usually laid out by department: rooms, food and beverage, and other operated departments, each with its own revenue, cost of sales and payroll, followed by undistributed costs such as administration, sales and marketing, property maintenance and utilities. In this example we build an annual P&L for a 40-room independent hotel with a small restaurant and bar, all figures excluding VAT. Rooms first. Available room nights are 40 x 365 = 14,600. Occupancy averages 75%, so occupied room nights are 0.75 x 14,600 = 10,950. The average daily rate, ADR, is £120 net, so room revenue is 10,950 x £120 = £1,314,000. Rooms department costs include housekeeping and front desk payroll of £280,000, laundry £55,000, guest supplies such as toiletries and tea trays £33,000, and online travel agent commission. In this example 40% of room revenue comes through OTAs at 18% commission: 0.40 x £1,314,000 = £525,600, and 0.18 x £525,600 = £94,608. Rooms department costs total £462,608, leaving a departmental profit of £851,392, which is 64.8% of room revenue. Food and beverage revenue is £520,000. Cost of sales at 30% is £156,000, F&B payroll is £210,000 and other F&B costs are £30,000, leaving a departmental profit of £124,000, or 23.8%. Other income from meeting room hire and car parking is £40,000 with £8,000 of costs, leaving £32,000. Total revenue is £1,314,000 + £520,000 + £40,000 = £1,874,000. Total departmental profit is £851,392 + £124,000 + £32,000 = £1,007,392. Undistributed costs: administration and general, including the general manager, accounts and IT, £150,000; sales and marketing, including the website and booking engine, £70,000; property operations and maintenance £75,000; and utilities £95,000. These total £390,000. Gross operating profit, GOP, is £1,007,392 - £390,000 = £617,392, which is 32.9% of total revenue. Fixed charges come next: rent or lease £240,000, business rates £50,000 and insurance £22,000, totalling £312,000. EBITDA is £617,392 - £312,000 = £305,392, or 16.3% of revenue. Notice how much the rooms department carries. It earns 70% of revenue but about 85% of departmental profit. That is why hotels focus so hard on occupancy and rate. A five-point rise in occupancy to 80% adds 730 room nights; at £120 that is £87,600 of revenue, and if the variable cost per occupied room, meaning housekeeping, laundry, supplies and commission, is about £35, the extra profit is 730 x (£120 - £35) = £62,050. - In this example 40 rooms at 75% occupancy give 10,950 room nights. - Room revenue at a £120 ADR is £1,314,000. - GOP is £617,392, or 32.9% of total revenue. - Rooms earn 70% of revenue but about 85% of departmental profit. - Five points of occupancy adds about £62,050 of profit here. #### Prime cost: a worked example Prime cost is the total of cost of sales plus total labour cost. It combines the two biggest controllable costs in most hospitality businesses into one number, so managers can see the trade-offs between them. In this example a brasserie has weekly net sales of £30,000. Food and drink cost of sales is £8,700, which is 29.0% of sales. Labour, including wages, employer National Insurance, employer pension contributions and holiday pay, is £10,200, which is 34.0%. Prime cost is £8,700 + £10,200 = £18,900, and as a share of sales £18,900 / £30,000 = 63.0%. What is left after prime cost, here £30,000 - £18,900 = £11,100 or 37.0%, must pay rent, rates, utilities and every other overhead and still leave a profit. If those other costs total £7,800 a week, the operating profit is £3,300, or 11.0% of sales. Now look at why prime cost is more useful than its parts. Suppose the chef moves to more prepared ingredients, such as ready-peeled vegetables and portioned fish. Food cost rises to 31.0%, which is £9,300, an extra £600. But fewer prep hours are needed, and labour falls by £800 to £9,400, or 31.3%. Prime cost becomes £9,300 + £9,400 = £18,700, or 62.3%, and operating profit improves by £200 a week even though the food cost percentage got worse. Looked at in isolation, the food cost number would have triggered alarm. The reverse can also happen. A manager who cuts a kitchen porter shift to save £120 a week might find that pots pile up, chefs spend time washing, waste rises and food cost goes up by £250. Prime cost reveals that the saving cost the business £130. Many operators set a prime cost target and let managers trade between food and labour inside it. In this example the target is 63%. In a quiet week when sales fall to £24,000, food cost stays at 29.0%, or £6,960, because it is variable, but labour has a fixed core of salaried staff and minimum rota cover. If labour only falls to £9,400, it becomes 39.2% of sales and prime cost hits 68.2%. That is the pattern to watch: in a slow week, labour percentage usually rises because part of it does not flex. Review prime cost weekly, not monthly, because a month is too long to discover that the rota has drifted. Use actual figures from stock counts and payroll rather than theoretical ones, and compare them with the budget for that week's forecast sales. - Prime cost equals cost of sales plus total labour. - In this example prime cost is £18,900 on £30,000 of sales, 63%. - Pre-prepared ingredients raised food cost but lowered prime cost. - In quiet weeks labour percentage rises because part of it is fixed. - Review prime cost weekly with actual figures. #### EBITDA: a worked example EBITDA stands for earnings before interest, tax, depreciation and amortisation. It is a measure of how much profit the trading operation makes before the effects of how it is financed, how its assets are written down and how it is taxed. Buyers, lenders and investors use it because it makes businesses with different funding arrangements easier to compare. In this example a cocktail bar has annual net sales of £900,000. Cost of sales is £198,000, labour is £270,000 and overheads including rent, rates and utilities are £297,000. Operating profit before depreciation is £900,000 - £198,000 - £270,000 - £297,000 = £135,000. That is EBITDA, and the EBITDA margin is £135,000 / £900,000 = 15.0%. Now work down. The bar's fit-out cost £300,000 and is being depreciated over 10 years, so depreciation is £30,000 a year. Amortisation of a lease premium paid to take over the site, £50,000 over 10 years, adds £5,000. EBIT, earnings before interest and tax, is £135,000 - £30,000 - £5,000 = £100,000. The owners borrowed £200,000 at 7% interest, so interest is £14,000 a year. Profit before tax is £100,000 - £14,000 = £86,000. Corporation tax is then charged on taxable profits, which are worked out under HMRC rules; capital allowances often replace accounting depreciation in that calculation, so the tax bill is not simply a percentage of £86,000. Check GOV.UK for current rates and allowances. Why does EBITDA matter? Imagine a second bar with the same trading but no loan and an older fit-out that is fully depreciated. Its profit before tax would be £135,000, while the first bar's is £86,000. EBITDA shows they trade identically; the difference is financing and history. Valuations of hospitality businesses are often quoted as a multiple of EBITDA. In this example, if a buyer offered four times EBITDA, the value of the business would be 4 x £135,000 = £540,000, before deducting any debt the buyer takes on. EBITDA has limits. It ignores the cash needed to replace kit, and in hospitality equipment and décor wear out fast. If the bar must spend £25,000 a year on maintenance capital expenditure to stay competitive, a more honest measure of cash generated is £135,000 - £25,000 = £110,000. Some groups also report EBITDA before or after rent, because leases are accounted for differently under some reporting standards; always check which version you are reading. And EBITDA can be flattered by one-off items or by owners not paying themselves a market salary, so adjust for those before comparing. - EBITDA is earnings before interest, tax, depreciation and amortisation. - In this example EBITDA is £135,000, a 15% margin on £900,000. - Profit before tax is £86,000 after depreciation, amortisation and interest. - A 4x multiple values the business at £540,000 before debt. - Subtract maintenance capex for a truer view of cash. #### Break-even covers: a worked example Break-even is the level of sales at which a business makes neither a profit nor a loss. Expressing it as covers per service turns an abstract number into something a team can aim at every night. In this example a 50-cover restaurant has weekly fixed costs, meaning costs that do not change with the number of guests, of £9,000. That includes rent, rates, insurance, salaried staff, a core rota, utilities standing charges, loan repayments and software. Variable costs are those that rise with every guest: food and drink cost, card fees, linen per cover and any hourly staff added for busy services. In this example the average net spend per cover is £45. Variable costs per cover are food and drink at 30% of spend, which is £13.50, card fees at 1.5%, which is £0.68, and other variable costs of £1.82, which includes laundry and disposables. Total variable cost per cover is £13.50 + £0.68 + £1.82 = £16.00. The contribution per cover is the spend minus the variable cost: £45.00 - £16.00 = £29.00. Each guest contributes £29 towards paying the fixed costs. Break-even covers per week are fixed costs divided by contribution: £9,000 / £29.00 = 310.3, so 311 covers. If the restaurant trades 12 services a week, break-even is 311 / 12 = 25.9, so about 26 covers per service. At 50 seats, that is just over half full. Every cover above break-even adds £29 of profit. If the restaurant does 420 covers in a week, profit is 420 x £29 - £9,000 = £12,180 - £9,000 = £3,180. Now test sensitivity. If the average spend drops to £40, food cost at 30% is £12.00, card fees £0.60, other costs £1.82, total £14.42, and contribution is £25.58. Break-even rises to £9,000 / £25.58 = 351.8, so 352 covers. A £5 drop in spend demands 41 more covers a week. If instead rent rises and fixed costs reach £10,000, break-even at the original £29 contribution becomes £10,000 / £29 = 344.8, so 345 covers. The calculation also helps with opening hours. Suppose opening on Monday adds £600 of fixed cost for staff and energy for two services. It needs 600 / 29 = 20.7, so 21 covers across those services just to pay for itself; if Mondays reliably do 40 covers, they earn 40 x £29 - £600 = £560. Use real figures, review them when costs change, and remember that break-even is a floor, not a target. A business that only ever breaks even has nothing left to reinvest, repay debt or survive a bad month. - Contribution per cover equals spend minus variable cost. - In this example £45 spend minus £16 variable cost leaves £29. - Break-even is £9,000 / £29 = 311 covers a week. - A £5 fall in spend raises break-even by 41 covers here. - Extra opening sessions must cover their own added fixed cost. #### Cash flow vs profit: a worked example Profit is what the P&L says the business earned over a period. Cash flow is the money that actually moved in and out of the bank. A business can be profitable and still run out of cash, and that is one of the most common ways hospitality businesses fail. In this example a new bistro trades for its first quarter, January to March. The P&L for the quarter shows net sales of £150,000, cost of sales of £45,000, labour of £50,000 and overheads of £35,000, giving an operating profit of £20,000. On paper, a good start. Now follow the cash. Guests pay by card, and the card provider settles a day or two later, so nearly all of the takings arrive within the quarter. But VAT is also collected: on £150,000 of net sales, the bistro takes in £30,000 of VAT from guests on top, so gross receipts are £180,000. That VAT is owed to HMRC, less the VAT the bistro can reclaim on its own purchases. Many new owners treat that cash as their own and spend it. On the cost side, suppliers in this example give 30-day terms, so of the £45,000 of food and drink, only £30,000 was paid within the quarter. That helps cash. But the bistro also paid a rent deposit of £15,000 and the quarter's rent of £9,000 in advance on the first day. It bought £40,000 of kitchen equipment, which does not appear as a cost on the P&L except as a small amount of depreciation, but which left the bank in full. And it repaid £6,000 of loan capital, which is not an expense on the P&L at all. Here is the cash summary for the quarter. Cash in: £180,000 from guests including VAT. Cash out: £30,000 to suppliers, plus VAT on those purchases, £6,000; labour £50,000; overheads other than rent £26,000, plus VAT on them, £4,000; rent and deposit £24,000; equipment £40,000, plus VAT £8,000; loan capital £6,000. In this example reclaimable VAT is the £6,000 + £4,000 + £8,000 = £18,000 shown, so the VAT owed after the quarter is £30,000 - £18,000 = £12,000. Total cash out in the quarter is £30,000 + £6,000 + £50,000 + £26,000 + £4,000 + £24,000 + £40,000 + £8,000 + £6,000 = £194,000. Net cash flow is £180,000 - £194,000 = -£14,000, and £12,000 of VAT is still due. So a quarter with £20,000 of profit has actually reduced cash by £14,000 and left a £12,000 bill. The fix is a simple 13-week cash forecast, updated weekly, that sets aside VAT and PAYE as soon as they are collected. - Profit and cash flow are different measures. - In this example £20,000 of profit coincides with £14,000 of cash outflow. - Equipment, deposits and loan capital hit cash but barely touch profit. - VAT collected from guests is owed to HMRC, not spare cash. - A rolling 13-week cash forecast catches shortfalls early. #### Budget vs actual variance: a worked example A budget is a plan for sales and costs. Comparing it with what actually happened, line by line, and explaining the differences is called variance analysis. It turns a monthly P&L from a report into a management tool. In this example a gastropub's budget for October expected net sales of £80,000, cost of sales of £24,000 at 30%, labour of £25,600 at 32%, and overheads of £18,000, giving an operating profit of £12,400. The actual results were sales of £74,000, cost of sales of £23,310, labour of £25,900 and overheads of £18,600, giving an operating profit of £74,000 - £23,310 - £25,900 - £18,600 = £6,190. Profit is £6,210 below budget. Step one: sales variance. Sales were £6,000 below budget, 7.5% down. If the gastropub's budget assumed 1,600 covers at £50 and it actually did 1,540 covers at £48.05, the cover shortfall of 60 accounts for 60 x £50 = £3,000, and the lower average spend accounts for 1,540 x (£50.00 - £48.05) = 1,540 x £1.95 = £3,003. Together that is £6,003, which rounds to the £6,000 shortfall. Half the problem is fewer guests and half is lower spend per guest. Step two: flex the budget. Comparing actual costs with the original budget is unfair because sales were lower. A flexed budget recalculates variable costs at the actual sales level. At £74,000 of sales, budgeted cost of sales at 30% would be £22,200. Actual was £23,310, so cost of sales is £1,110 worse than it should have been at that sales level. Actual GP percentage is (£74,000 - £23,310) / £74,000 = 68.5% against a target of 70%. Step three: labour. If labour has a fixed core of £12,000 and a variable part budgeted at 17% of sales, the flexed labour budget at £74,000 is £12,000 + 0.17 x £74,000 = £12,000 + £12,580 = £24,580. Actual was £25,900, so labour is £1,320 over the flexed figure. The rota did not come down as sales fell. Step four: overheads. These were £600 over, mostly a boiler repair, a one-off. Step five: summarise. Of the £6,210 shortfall, lower sales explain the loss of contribution, food cost explains £1,110, labour £1,320 and overheads £600. Step six: act. Investigate the food cost, perhaps through waste records and portion checks, and review the rota against sales. Agree two or three actions, assign an owner to each, and check them next month. Variance analysis works best when done within a week of month end, while people still remember what happened. - In this example profit was £6,210 below a £12,400 budget. - Fewer covers and lower spend each explained about £3,000 of lost sales. - A flexed budget recalculates variable costs at actual sales. - Food cost was £1,110 and labour £1,320 over the flexed budget. - Review variances within a week of month end. ### Labour #### Rota cost vs sales: a worked example A rota is a plan for who works when. Costing it against forecast sales before the week starts is the single most effective labour control most venues have. In this example a gastropub forecasts net sales for next week of £22,000, with Friday and Saturday carrying 45% of the week. Step one is to cost the rota. Hourly staff in this example are paid between £12.00 and £15.00 an hour; to keep the arithmetic simple we use an average of £13.00. Employer on-costs, meaning employer National Insurance above the threshold, employer pension contributions and the cost of accruing holiday, add roughly 20% in this example, so the loaded hourly cost is £13.00 x 1.2 = £15.60. Check current thresholds and rates on GOV.UK, as they change. The draft rota has 560 hourly staff hours across front and back of house: 560 x £15.60 = £8,736. Salaried staff, a head chef and a general manager, cost £1,900 a week fully loaded. Total labour is £10,636, which is £10,636 / £22,000 = 48.3% of forecast sales. In this example the target is 32%, which allows £22,000 x 0.32 = £7,040. The draft is £3,596 over. Step two is to find the hours. Divide the over-spend by the loaded rate: £3,596 / £15.60 = 230.5 hours. That is a big cut, so look at the shape of the week rather than trimming evenly. Monday to Wednesday forecast £6,600 in total. The rota has 210 hours on those days, costing £3,276, which is 49.6% of those days' sales before salaried staff are even counted. Cutting to one fewer server per shift and closing the kitchen at 9pm instead of 10pm saves 70 hours. Thursday and Sunday are trimmed by 40 hours with staggered starts so staff arrive as covers build. Friday and Saturday stay fully staffed, because under-staffing a peak loses sales. Step three is to check the new total: 560 - 110 = 450 hours, costing 450 x £15.60 = £7,020, plus £1,900 salaried, gives £8,920, or 40.5%. Still above target. That tells management something important: either the forecast is too low for the venue's fixed staffing, or the 32% target is unrealistic for this format. Step four is to measure sales per labour hour: £22,000 / 450 = £48.89 of sales per hourly hour. Tracking that weekly shows whether productivity is improving. After the week, compare actual sales and actual hours from the clocking system with the plan, and note where the forecast was wrong. Over time, the forecast becomes more accurate and the rota tighter. The aim is not the cheapest rota but the one that matches staff to demand, hour by hour. - In this example loaded cost is the hourly rate plus about 20% on-costs. - The draft rota costs 48.3% of £22,000 forecast sales. - Divide the over-spend by the loaded rate to find hours to cut. - Protect peak shifts; trim quiet days and use staggered starts. - Sales per labour hour tracks productivity week by week. #### Labour percentage targets: a worked example Labour percentage is total labour cost divided by net sales. It is the most quoted labour measure in hospitality, but a single target applied to every format can be misleading. In this example we compare three businesses run by the same small group, all figures excluding VAT. The first is a counter-service coffee shop with weekly net sales of £9,000 and labour of £2,880. Its labour percentage is £2,880 / £9,000 = 32.0%. The second is a full-service restaurant with sales of £35,000 and labour of £12,600, giving 36.0%. The third is a wet-led bar with sales of £18,000 and labour of £3,780, giving 21.0%. If the group set one target of 30%, the bar would look excellent and the restaurant poor, yet each may be well run for its type. Table service, cooked-to-order food and longer opening hours all need more people per pound of sales. A better approach is to set targets per format and per sales level. Step one: split labour into fixed and variable. In the restaurant, salaried staff and the minimum crew needed to open cost £6,000 a week whatever happens. The rest, £6,600, flexes with covers. Step two: work out variable labour as a percentage of sales. £6,600 / £35,000 = 18.9%. Step three: build a sliding target. At any sales level S, expected labour is £6,000 + 0.189 x S. At £35,000 that gives £6,000 + £6,615 = £12,615, matching the actual figure. At £25,000 it gives £6,000 + £4,725 = £10,725, which is 42.9% of sales. At £45,000 it gives £6,000 + £8,505 = £14,505, which is 32.2%. So the same well-run restaurant should expect around 43% in a slow week and 32% in a busy one. Judging the manager against a flat 36% would punish them for the weather and reward them for a sunny bank holiday. Step four: watch the trend. If, over a month, actual labour runs consistently above the sliding target at every sales level, there is a real problem, perhaps over-staffing, overtime or poor scheduling. If it runs below, check service quality and staff turnover, because a rota that is too lean can burn people out and cost more in recruitment later. Labour percentage should always be read alongside sales per labour hour and guest feedback, and it must include every employer cost, not only gross pay. Agency staff, salaried managers and the owner's own time if they work a shift all belong in the calculation if the figure is to be compared honestly between sites. - Labour percentage is total labour cost divided by net sales. - In this example the bar runs at 21%, the restaurant at 36%. - Split labour into a fixed core and a variable share of sales. - The same restaurant should expect about 43% at £25,000 and 32% at £45,000. - Read labour percentage with sales per hour and guest feedback. #### Holiday accrual: a worked example Workers in the UK are entitled to paid holiday, and in hospitality, where many staff work irregular hours, calculating it correctly matters both legally and financially. The rules have been revised in recent years, including a method of accruing holiday as a percentage of hours worked for irregular-hours and part-year workers, so always check the current guidance on GOV.UK and take advice if unsure. This example shows the principle of accrual and how it affects costs; it is illustrative, not legal advice. The statutory minimum for a full-time worker is 5.6 weeks a year. For an irregular-hours worker, one accrual method calculates holiday at 12.07% of hours worked in each pay period. The 12.07% comes from 5.6 weeks of holiday divided by the 46.4 weeks a year that remain for working: 5.6 / 46.4 = 0.1207. In this example a bartender on irregular hours works 120 hours in a four-week pay period. Their accrued holiday for that period is 120 x 0.1207 = 14.48 hours. At an hourly rate of £13.00, that accrued holiday is worth 14.48 x £13.00 = £188.24. If they take a week off later, they are paid for it from what they have built up. The business cost is real even though nothing is paid until holiday is taken. Over a year, a worker averaging 30 hours a week for 46.4 working weeks works 1,392 hours, and accrues 1,392 x 0.1207 = 168.0 hours of holiday, worth £2,184 at £13.00. For the business, that is a liability that builds up on the balance sheet. A venue with 25 such staff has 25 x £2,184 = £54,600 of holiday cost a year to fund. If managers cost rotas on gross hourly pay only, they understate true labour by around 12%. Holiday pay also has to reflect normal pay in many cases, which can include regular overtime and some other elements, and the rules on this have been tested in the courts; again, check current guidance. Two practical points. First, build holiday into the loaded hourly cost used for rota planning. In this example £13.00 x 1.1207 = £14.57 before employer National Insurance and pension. Second, encourage staff to take holiday through the year rather than letting it pile up for December, when the business is busiest. A spreadsheet or payroll system showing each person's accrued, taken and remaining hours prevents disputes and lets managers plan cover. When someone leaves, any accrued but untaken holiday is generally paid out in their final pay, so the liability does eventually turn into cash. - The statutory minimum for a full-time worker is 5.6 weeks a year. - 12.07% is 5.6 weeks divided by 46.4 working weeks. - In this example 120 hours accrues 14.48 hours of holiday. - Ignoring holiday understates true labour cost by about 12%. - Check GOV.UK for the current rules on calculation and holiday pay. #### Tronc distribution: a worked example A tronc is a pooled arrangement for distributing tips, gratuities and service charges among staff, run by a person called the troncmaster who is independent of the employer's control over the allocation. In the UK, the law now requires employers to pass on all tips, gratuities and service charges to workers without deductions, to distribute them fairly, and to have a written policy; there is a statutory code of practice on fair and transparent distribution. Properly run tronc schemes can also have particular National Insurance treatment, but the conditions are specific, so check current HMRC guidance and take advice before setting one up. This example shows how a points-based distribution might work; it is illustrative only. In this example a restaurant collects £8,400 in a four-week period from a discretionary service charge and card tips. The troncmaster's scheme allocates points per hour worked by role: servers and bartenders 10 points, chefs 8 points, kitchen porters 7 points, and supervisors 12 points. Step one: total each person's points. Suppose the team works as follows over the period. Servers: 6 people, 600 hours in total, 600 x 10 = 6,000 points. Bartenders: 2 people, 240 hours, 2,400 points. Chefs: 4 people, 640 hours, 640 x 8 = 5,120 points. Kitchen porters: 2 people, 280 hours, 280 x 7 = 1,960 points. Supervisors: 1 person, 160 hours, 160 x 12 = 1,920 points. Total points are 6,000 + 2,400 + 5,120 + 1,960 + 1,920 = 17,400. Step two: work out the value of a point. £8,400 / 17,400 = £0.4828 per point. Step three: pay each person according to their points. A server who worked 100 hours earned 1,000 points, worth 1,000 x £0.4828 = £482.80. A chef who worked 160 hours earned 1,280 points, worth £617.98. A kitchen porter who worked 140 hours earned 980 points, worth £473.14. Check the total: role totals are servers £2,896.80, bartenders £1,158.72, chefs £2,471.94, porters £946.29 and supervisor £926.98, which add to £8,400.73. The extra 73p is rounding; real systems round the point value to more decimal places or handle the remainder explicitly. Tips are taxable income for the worker, and income tax is collected through PAYE. Card processing fees on tips are a point to check, because the rules on deductions are strict. The written policy should explain who is included, how points are set and when payments are made, and staff should be able to see how their share was calculated. - UK law requires tips and service charges to be passed to workers in full. - A written tipping policy and a statutory code of practice apply. - In this example £8,400 across 17,400 points is £0.4828 a point. - Points per hour by role keep distribution transparent. - Check HMRC guidance on tronc tax and National Insurance treatment. #### Overtime: a worked example Overtime is work beyond a person's contracted or normal hours. In UK hospitality there is no general legal requirement to pay a higher rate for overtime, but contracts often do, and average pay across all hours worked must not fall below the National Minimum Wage or National Living Wage. Working time rules, including the 48-hour average week, which workers can opt out of, and rest breaks, also apply; check GOV.UK for the current details. This example shows how overtime affects cost. In this example a sous chef is contracted for 45 hours a week at £16.00 an hour, so normal weekly pay is 45 x £16.00 = £720. The contract pays overtime at time and a quarter, so each extra hour costs £16.00 x 1.25 = £20.00 in gross pay. During a staff shortage the sous chef works 55 hours a week for four weeks. Extra hours per week are 10, and the extra pay is 10 x £20.00 = £200 a week, or £800 over four weeks. Add employer on-costs. In this example employer National Insurance and pension add about 15% on these extra earnings, so the true extra cost is £800 x 1.15 = £920. Holiday accrual may also rise, because regular overtime can count towards holiday pay. Now compare with hiring. A commis chef at £13.00 an hour covering 40 hours would cost 40 x £13.00 = £520 a week in gross pay, or £598 with 15% on-costs. If the commis covers the same 40 hours of work that is currently spread as overtime across four senior chefs, each doing 10 extra hours at £20.00, the overtime bill is 4 x 10 x £20.00 = £800 gross, or £920 loaded, against £598 for the commis. The saving is £322 a week, about £16,700 a year. The commis is less experienced, so not every task transfers, but the principle holds: persistent overtime by senior staff is usually more expensive than an extra junior hire. There are hidden costs too. Long weeks increase fatigue, mistakes, accidents and sickness, and they are a common reason chefs leave. Replacing a sous chef could cost thousands in recruitment fees, lost productivity and agency cover. Track overtime hours by person every week. A sensible rule in this example is that any individual doing more than five hours of overtime for three consecutive weeks triggers a review of the rota. Salaried staff paid a flat salary for long hours need watching too: divide their salary by the actual hours worked to check the effective hourly rate stays above the legal minimum. A head chef paid £700 a week working 60 hours earns £11.67 an hour, which may be close to or below the minimum depending on the current rate. - There is no general UK legal requirement to pay an overtime premium. - Average pay across all hours must not fall below the legal minimum. - In this example four seniors' overtime costs £920 against £598 for a commis. - Persistent overtime often costs more than an extra hire. - Check effective hourly pay for salaried staff working long hours. #### Sickness cover: a worked example When someone is off sick, the business usually pays twice: once for the absent worker, through Statutory Sick Pay or any contractual sick pay, and again for whoever covers the shift. The rules on eligibility and the rate of Statutory Sick Pay have been reformed recently, so check GOV.UK for the current position rather than relying on older figures. This example uses placeholder amounts to show the arithmetic. In this example a hotel kitchen has 12 staff. Over a year, the average absence rate is 4% of scheduled hours. Each person is scheduled for about 1,900 hours, so the team schedules 12 x 1,900 = 22,800 hours, and 4% of that is 912 hours lost to sickness. Step one: the cost of paying the absent staff. Suppose that, in this example, sick pay averages £6.00 an hour across all the absent hours. The cost is 912 x £6.00 = £5,472. Step two: the cost of cover. Some absences are absorbed by the remaining team working harder, but in this example 70% of the lost hours, 638 hours, are covered. Half of those are covered by existing staff on overtime at a loaded £19.00 an hour, and half by agency chefs at £24.00 an hour. Overtime cover costs 319 x £19.00 = £6,061. Agency cover costs 319 x £24.00 = £7,656. Total cover is £13,717. Step three: the total cost of sickness is £5,472 + £13,717 = £19,189 a year, or £1,599 per employee. That figure does not include the less visible costs: slower service, more waste and pressure on the rest of the team. Step four: test an intervention. Suppose the hotel invests £4,000 a year in better rotas with guaranteed days off, a return-to-work conversation after each absence, and free staff meals with healthier options. If absence falls from 4% to 3%, lost hours fall to 684, a reduction of 228 hours. The saving scales roughly in proportion: £19,189 x 228 / 912 = £4,797. That more than pays for the £4,000 investment, before counting any effect on staff turnover. Two policy points help. First, keep a simple absence record, because patterns, such as frequent Mondays, become visible only when written down. Second, build a bank of trained cover: part-time or flexible staff who know the kitchen and can be called on before an agency. In this example, if half of the agency hours could be covered by a casual bank worker at a loaded £16.50, the saving would be about 160 x (£24.00 - £16.50) = £1,200 a year. Small, steady improvements in cover planning add up. - Sickness costs twice: sick pay plus cover. - In this example 4% absence costs £19,189 a year for 12 staff. - Agency cover is the most expensive option per hour. - Cutting absence from 4% to 3% saves about £4,800 here. - Check GOV.UK for current Statutory Sick Pay rules. #### Agency vs staff cost: a worked example Agency workers are useful for peaks, events and emergencies, but the hourly charge looks very different from an employee's hourly pay, so compare them on a like-for-like basis. In this example a conference hotel needs extra banqueting servers for a busy season. Option A is to hire through an agency that charges £19.50 an hour. The agency pays the worker, handles their National Insurance and holiday, and makes its own margin. The hotel pays £19.50 for each hour worked and nothing else. Option B is to employ seasonal staff directly at £12.50 an hour. On top of that, the hotel pays employer costs. In this example employer National Insurance, pension contributions and holiday accrual add 22%, so the loaded cost is £12.50 x 1.22 = £15.25 an hour. Check GOV.UK for current rates, since the on-cost depends on thresholds and earnings. Direct employment also has one-off costs. Recruitment advertising in this example is £300, uniforms are £60 per person, and training and induction take 8 paid hours per person, which is 8 x £15.25 = £122. For a team of 10 that adds £300 + 10 x (£60 + £122) = £2,120. Now compare across the season. The hotel needs 10 servers for an average of 25 hours a week for 12 weeks: 10 x 25 x 12 = 3,000 hours. Agency cost: 3,000 x £19.50 = £58,500. Direct cost: 3,000 x £15.25 = £45,750, plus £2,120 of set-up, which totals £47,870. Direct employment saves £10,630, or about 18%. But consider utilisation. Agency staff are booked only for the hours of each event. Direct staff may need guaranteed minimum hours, and on quiet weeks they may be paid for time with little to do. If direct staff are paid for 3,400 hours to deliver 3,000 hours of useful work, the cost becomes 3,400 x £15.25 + £2,120 = £53,970, and the saving shrinks to £4,530. Quality and consistency also count. Staff who know the building, the menus and the standards usually serve better than a rotating pool. On the other hand, agencies can supply people at a few hours' notice for an unexpected large booking. Many venues use a blended approach: a core of direct seasonal staff for predictable volume, and agency staff for peaks above that level. In this example, using direct staff for the first 2,400 hours and agency for the most variable 600 gives 2,400 x £15.25 + £2,120 + 600 x £19.50 = £36,600 + £2,120 + £11,700 = £50,420. Track agency spend monthly as its own line so it does not hide inside general labour. - In this example agency costs £19.50 an hour against £15.25 loaded direct. - Direct staff carry set-up costs for recruitment, uniform and training. - Across 3,000 hours direct employment saves about 18% here. - Low utilisation of direct staff erodes the saving. - A core of direct staff plus agency for peaks is a common blend. #### Labour cost per cover: a worked example Labour cost per cover divides the labour cost of a service, day or week by the number of guests served. Unlike labour percentage, it does not move when menu prices change, so it shows productivity more directly. In this example a restaurant compares four lunch services and four dinner services in one week. Lunches: 220 covers in total, with 96 hourly staff hours at a loaded cost of £15.50 an hour, which is 96 x £15.50 = £1,488. Labour cost per cover is £1,488 / 220 = £6.76. Average net spend at lunch is £26, so labour is £6.76 / £26 = 26.0% of lunch sales. Dinners: 480 covers, with 180 hourly staff hours costing 180 x £15.50 = £2,790. Labour cost per cover is £2,790 / 480 = £5.81. Average spend at dinner is £48, so labour is 12.1% of dinner sales on hourly staff alone. Salaried staff add to both, but are often reported separately. Step two: what the numbers say. Lunch costs nearly £1 more per guest in labour than dinner and earns little more than half the spend. If the restaurant could serve the same lunch covers with 80 hours instead of 96, labour per cover would fall to 80 x £15.50 / 220 = £1,240 / 220 = £5.64, saving £248 a week, about £12,900 a year. Step three: find where hours go. Lunch has quiet first and last hours. If the restaurant opens at 12 noon with full staff but guests arrive mostly between 12.45 and 2pm, staggering start times and sending one server home at 2.30pm saves hours without hurting service. In this example the reshaped rota cuts 12 hours, bringing labour per cover to 84 x £15.50 / 220 = £1,302 / 220 = £5.92. Step four: compare across sites or periods. If a sister restaurant achieves £5.20 per cover at dinner with similar spend and service style, the gap of £0.61 across 480 covers is worth £293 a week. Look at layout, kitchen equipment, menu complexity and the experience of the team before concluding anyone is overstaffed. Step five: watch the floor. Labour per cover can be pushed too low. If guest scores fall and table turns slow because there are too few people on the floor, sales will drop faster than costs. Set a sensible range for each service type and review it alongside guest feedback. Step six: use it for forecasting. With a reliable labour cost per cover, a manager can build next week's rota from the forecast covers: 250 forecast lunch covers at a target of £5.80 gives a labour budget of 250 x £5.80 = £1,450, or about 94 hours at £15.50. - Labour cost per cover equals labour cost divided by guests served. - In this example lunch costs £6.76 per cover and dinner £5.81. - Cutting 16 lunch hours would save about £12,900 a year here. - It does not move when menu prices change. - Set a range per service type and watch guest feedback. ### Pricing #### Menu engineering matrix: a worked example Menu engineering sorts dishes by two measures: popularity, meaning how many sell, and profitability, meaning the cash gross profit each one earns. The four quadrants are often called stars (popular and profitable), ploughhorses (popular, less profitable), puzzles (profitable, less popular) and dogs (neither). In this example a restaurant sells six mains in a month, all figures net of VAT. Dish 1, fish and chips: 400 sold, price £15.00, cost £4.50, GP £10.50. Dish 2, steak: 150 sold, price £28.00, cost £11.00, GP £17.00. Dish 3, pasta: 350 sold, price £14.00, cost £2.80, GP £11.20. Dish 4, duck: 80 sold, price £24.00, cost £7.00, GP £17.00. Dish 5, curry: 220 sold, price £15.00, cost £3.60, GP £11.40. Dish 6, vegetable tart: 100 sold, price £13.00, cost £3.40, GP £9.60. Step one: total sales are 400 + 150 + 350 + 80 + 220 + 100 = 1,300 dishes. With six dishes, an equal share would be 1,300 / 6 = 216.7. A common rule sets the popularity threshold at 70% of that equal share: 0.7 x 216.7 = 151.7. Dishes selling 152 or more count as popular. Step two: work out the weighted average GP. Total GP is 400 x £10.50 = £4,200; 150 x £17.00 = £2,550; 350 x £11.20 = £3,920; 80 x £17.00 = £1,360; 220 x £11.40 = £2,508; 100 x £9.60 = £960. The sum is £15,498. Average GP per dish is £15,498 / 1,300 = £11.92. Dishes above £11.92 count as high profit. Step three: classify. Fish and chips: popular, GP below average: ploughhorse. Steak: 150 is just under the 152 threshold, GP high: puzzle. Pasta: popular, GP below average: ploughhorse. Duck: unpopular, GP high: puzzle. Curry: popular, GP £11.40 below average: ploughhorse. Tart: unpopular and low GP: dog. Step four: act. There are no stars in this example, which is itself useful information. Ploughhorses can take a modest price rise or a cost reduction; raising fish and chips by £1.00 net adds 400 x £1.00 = £400 a month if volume holds. Puzzles need better positioning: move the duck to a prominent spot, have staff recommend it, or rewrite its description. Dogs should be reworked or removed; if the tart must stay for vegetarian guests, re-cost it or improve it so that it earns its place. Re-run the matrix after each menu change, because the averages move once the mix changes, and remember that a dish's role on the menu, such as a vegetarian or children's option, may justify keeping it. - Menu engineering plots popularity against cash GP per dish. - In this example the popularity threshold is 70% of 216.7, about 152. - Weighted average GP is £15,498 / 1,300 = £11.92. - Ploughhorses take price rises; puzzles need better positioning. - Re-run the matrix after every menu change. #### Price increases: a worked example When costs rise, the question is how much prices need to move to protect profit, and how many sales the business can afford to lose before an increase stops helping. In this example a burger bar sells 2,000 burgers a month at £14.40 including VAT, which is £12.00 net. The cost of each burger has risen from £3.60 to £4.20. Step one: see the damage. At the old cost, GP per burger was £12.00 - £3.60 = £8.40, a GP of 70.0%. At the new cost it is £12.00 - £4.20 = £7.80, a GP of 65.0%. Monthly cash GP falls from 2,000 x £8.40 = £16,800 to 2,000 x £7.80 = £15,600, a loss of £1,200. Step two: price to restore the percentage. To get back to 70% GP, the net price must be £4.20 / 0.30 = £14.00, which is £16.80 including VAT. That is a 16.7% rise, which guests will notice. Step three: price to restore the cash instead. To keep £8.40 of GP per burger, the net price must be £4.20 + £8.40 = £12.60, which is £15.12 with VAT; rounded to £15.00, the net is £12.50 and the GP per burger is £8.30. That is a 4.2% rise. In most cases restoring cash margin is the more realistic goal. Step four: test volume. At £15.00 with £8.30 GP, how many burgers must sell to match the old £16,800? £16,800 / £8.30 = 2,024. To match the reduced £15,600 that the business would earn without a price rise, it needs only £15,600 / £8.30 = 1,880 burgers, so it could lose 120 sales, 6% of volume, and still be better off than not raising the price. Step five: compare the bigger rise. At £16.80 with £9.80 GP, the business needs only £15,600 / £9.80 = 1,592 burgers to match the no-rise figure, so it could lose 20% of volume and still break even against doing nothing. But a 16.7% jump risks more than lost burgers; it can change how guests see the place. Practical tips. Raise prices in smaller, more frequent steps rather than one large jump. Move prices to sensible endings, avoiding odd pence that make the menu feel calculated. Pair the rise with something visible, such as a better bun or a new side. Raise items that are less price-sensitive, like sides, sauces and premium add-ons, by a little more, and protect the headline items guests use to judge value. And track volume and average spend for four to six weeks after the change so you know what really happened. - In this example a £0.60 cost rise cuts monthly GP by £1,200. - Restoring the 70% GP needs a 16.7% price rise. - Restoring cash GP needs only about 4%. - At £15.00 the bar can lose 6% of volume and still be ahead. - Smaller, more frequent rises are usually easier on guests. #### Discounts and happy hours maths: a worked example A discount gives away margin, so it needs extra volume to pay for itself. The extra volume needed is much larger than most people expect. In this example a bar sells a house cocktail at £10.80 including VAT, which is £9.00 net. The cost of the drink is £2.00, so the GP per drink is £7.00, a GP of 77.8%. A happy hour offers the drink at £7.20 including VAT, a third off. The net price falls to £6.00, and the GP per drink falls to £6.00 - £2.00 = £4.00. Step one: work out the volume needed to stand still. If the bar normally sells 40 of these drinks between 5pm and 7pm, the normal GP for that period is 40 x £7.00 = £280. At £4.00 a drink, to make the same £280 it must sell £280 / £4.00 = 70 drinks. That is 75% more volume just to break even on the offer. Step two: the formula. The required volume increase is the old GP divided by the new GP, minus one. £7.00 / £4.00 - 1 = 0.75, or 75%. A 20% discount on the same drink makes the net price £7.20 and the GP £5.20, so the required increase is £7.00 / £5.20 - 1 = 34.6%. A 50% discount makes the net price £4.50 and the GP £2.50, so the volume must rise by £7.00 / £2.50 - 1 = 180%. Step three: count the knock-on benefits honestly. Happy hours often aim to bring people in earlier and keep them for full-price drinks and food afterwards. In this example, suppose the offer brings in 30 additional guests who each go on to spend £15.00 net on full-price items at a 70% GP. That adds 30 x £15.00 x 0.70 = £315 of GP. Now the sums look different. If the bar sells 70 discounted cocktails, it makes £280 on them, the same as before, plus £315 from the later spending, a total gain of £315. If it sells only 55, it makes 55 x £4.00 = £220, which is £60 less than before, but the £315 from later spend still leaves it £255 ahead. Step four: watch for cannibalisation. If regulars who would have come anyway at full price simply switch to the cheaper drink, there is no gain. The true test is to compare the whole evening's GP over several weeks, not just the happy-hour window. Check that any promotion complies with the licensing conditions for the premises and with any local authority policy on irresponsible drinks promotions. Two-for-one offers carry the same maths as a 50% discount, so treat them with particular care. - Required volume rise equals old GP divided by new GP, minus one. - In this example a third off needs 75% more volume to break even. - A 50% discount needs 180% more volume here. - Later full-price spending can make a promotion worthwhile. - Check licensing conditions on drinks promotions. #### Service charge and VAT on it: a worked example A discretionary service charge added to a bill is, under UK rules, generally not subject to VAT if it is genuinely optional and the guest can choose not to pay it. A compulsory service charge, for example one applied automatically to large groups and presented as mandatory, is generally treated as part of the price and is subject to VAT. Tips given freely by guests are outside the scope of VAT. The detail matters, so check current HMRC guidance on GOV.UK. Separately, UK law now requires that all tips, gratuities and service charges are passed to workers in full. This example shows how the VAT treatment affects the numbers. In this example a table of four spends £200 on food and drink including VAT. The VAT within that is £200 / 1.2 x 0.2 = £33.33, and the net food and drink revenue is £166.67. Case one: a discretionary 12.5% service charge is added on the VAT-inclusive total. 0.125 x £200 = £25.00. The bill is £225.00. Because the charge is optional, no VAT is due on it, and the full £25.00 is passed to staff, subject to their income tax under PAYE. The restaurant's VAT liability on this table is £33.33. Case two: the same 12.5% is applied as a compulsory charge for a group. The £25.00 is now treated as consideration for the meal and is VAT-inclusive. The VAT within it is £25.00 / 1.2 x 0.2 = £4.17. The business must account for £33.33 + £4.17 = £37.50 of VAT. The amount available to pass to staff from the compulsory charge is the net amount after VAT, £20.83, because the VAT element is owed to HMRC. Staff receive £4.17 less than under the discretionary charge, for exactly the same bill. Over a year, that adds up. If a restaurant collects £150,000 of compulsory service charges, the VAT within them is £150,000 / 1.2 x 0.2 = £25,000. Making the charge discretionary would, in this example, remove that VAT, though it also gives guests the right to decline it. Case three: the guest pays the bill and leaves £20 in cash on the table. That is a tip. It is outside the scope of VAT and must be passed on to workers under the tipping rules. Practical points. Print clearly on the menu and the bill whether a service charge is discretionary. Train staff to remove it without fuss if a guest asks. Keep records showing how service charges and tips are distributed, in line with the written tipping policy. And take advice before changing the treatment of service charges, because the boundary between optional and compulsory depends on how the charge is presented and applied. - A genuinely discretionary service charge is generally outside VAT. - A compulsory service charge is generally subject to VAT. - In this example a compulsory £25 charge contains £4.17 of VAT. - UK law requires tips and service charges to be passed to workers. - Check current HMRC guidance before changing your approach. #### Deposit policies: a worked example No-shows cost restaurants real money, and deposits or card-backed booking guarantees are the usual defence. The question is how big the problem is and what policy is proportionate. In this example a 70-cover restaurant takes 1,800 reservations a month for an average of 2.5 guests each, so 4,500 booked covers. The no-show rate is 6% of booked covers, which is 0.06 x 4,500 = 270 covers. Some of those seats are refilled by walk-ins or the waiting list; in this example, a third are, leaving 180 covers truly lost. The average net spend is £48 and the contribution per cover, after food, drink and other variable costs, is £31. Lost contribution is 180 x £31 = £5,580 a month, about £67,000 a year. That is often the difference between a profitable year and a poor one. Option one: a £10 per person no-show fee charged to a card held at booking. If the no-show rate falls to 3%, 135 covers are no-shows, and with a third refilled, 90 are lost: 90 x £31 = £2,790 lost, a saving of £2,790. Fees collected on the remaining no-shows: suppose 70% of the 135 can be charged, as some will be disputed or waived. That is 94.5 x £10 = £945, so the total improvement is about £3,735 a month. Option two: a £25 per person prepaid deposit for groups of six or more and for Friday and Saturday evenings. In this example a group of eight books on a Saturday and pays £200 upfront. If they arrive, the deposit is deducted from the bill. If they cancel within the stated notice period, the policy refunds or retains it according to the terms. The VAT position of deposits can be complicated: a deposit that is part-payment for a meal may create a VAT point when received, while a genuine no-show fee may be treated differently. Check current HMRC guidance and take advice when drafting terms. Costs of the policy. Booking software with card guarantees may cost more, perhaps £100 a month more in this example, and card processing fees apply to charges taken. Some guests will book elsewhere. If the policy discourages 2% of bookings, that is 90 covers a month, worth 90 x £31 = £2,790 of contribution, which would wipe out most of the gain from option one. That is why many venues apply guarantees selectively: to large groups, to peak nights and to special dates, rather than to every table. Whatever the policy, state it clearly at booking, in confirmation messages and on the website. Send reminders 24 to 48 hours ahead with an easy way to cancel, because a timely cancellation is worth far more than a fee. - In this example 6% no-shows cost about £5,580 of contribution a month. - Contribution per cover, not spend, measures the real loss. - Halving no-shows with a £10 fee improves results by about £3,735. - Guarantees that deter bookings can cancel out their benefit. - Check HMRC guidance on the VAT treatment of deposits and fees. #### Set menu pricing: a worked example A set menu, such as two or three courses for a fixed price, can fill quiet services, speed up the kitchen and help guests feel in control of the bill. Its price has to be built from the dishes on it and the way guests actually choose. In this example a restaurant designs a set lunch with a choice of two starters, two mains and two desserts, offered at two courses for £24 or three courses for £29, both including VAT. Net prices are £24 / 1.2 = £20.00 and £29 / 1.2 = £24.17. Step one: cost each dish. Starters: soup £1.10, terrine £1.90. Mains: chicken £3.20, fish £4.10. Desserts: tart £1.20, ice cream £0.90. Step two: estimate the mix. From past sales, in this example 60% of guests choose the soup and 40% the terrine; 55% the chicken and 45% the fish; 50% each dessert. Weighted starter cost: 0.6 x £1.10 + 0.4 x £1.90 = £0.66 + £0.76 = £1.42. Weighted main cost: 0.55 x £3.20 + 0.45 x £4.10 = £1.76 + £1.845 = £3.605. Weighted dessert cost: 0.5 x £1.20 + 0.5 x £0.90 = £1.05. Step three: cost each package. Two courses are usually starter and main or main and dessert. In this example 70% of two-course guests choose starter and main, costing £1.42 + £3.605 = £5.025, and 30% choose main and dessert, costing £3.605 + £1.05 = £4.655. The weighted two-course cost is 0.7 x £5.025 + 0.3 x £4.655 = £3.5175 + £1.3965 = £4.914. Three courses cost £1.42 + £3.605 + £1.05 = £6.075. Step four: margins. Two courses: GP is £20.00 - £4.91 = £15.09, a GP of 75.4%. Three courses: £24.17 - £6.08 = £18.09, a GP of 74.8%. Step five: the upsell. The third course costs the guest £5 extra, which is £4.17 net, and costs the kitchen about £1.16 extra on average, so each upgrade adds about £3.01 of gross profit. If 40% of 300 set-lunch guests a week upgrade, that is 120 x £3.01 = £361 a week. Step six: check against the à la carte. If the same dishes à la carte would total £34 for three courses, the set menu offers clear value. But if regular à la carte guests switch to the set menu, the restaurant may lose margin. In this example, if 50 à la carte guests a week, spending an average of £30 net with a GP of £21.50, switch to the three-course set at £18.09 GP, the restaurant loses 50 x £3.41 = £170.50 a week. The set menu works when it brings in new guests or fills quiet times, so limit it to lunch or early evening, and track the number of new versus switching guests. - In this example the set lunch is £24 for two courses and £29 for three. - Weighted costs depend on the mix guests actually choose. - Two courses earn a 75.4% GP and three courses 74.8% here. - Each third-course upgrade adds about £3.01 of GP. - Watch for regulars switching from higher-margin à la carte. #### Delivery platform commission: a worked example Third-party delivery platforms give restaurants access to customers at home, but their commission is often a large share of each order. Pricing delivery correctly means working out what is left after commission, packaging and VAT. Commission levels vary by platform, contract and whether the platform or the restaurant provides the rider, so use the figures from your own agreement. In this example a burger restaurant lists on a platform that charges 30% commission on the VAT-inclusive order value, plus VAT on the commission. Step one: an order at the dine-in prices. A customer orders a burger at £14.40, fries at £4.80 and a soft drink at £3.60, a total of £22.80 including VAT. Hot food delivered is standard-rated, and so is the soft drink, so the VAT is £22.80 / 6 = £3.80, and net revenue is £19.00. Commission at 30% of £22.80 is £6.84. VAT on the commission, at 20%, is £1.37; a VAT-registered restaurant can usually reclaim it, so the real cost is £6.84. Packaging for the order costs £0.90. Food and drink cost is £5.70. What the restaurant keeps after cost of sales, packaging and commission: £19.00 - £5.70 - £0.90 - £6.84 = £5.56. As a share of net revenue, that is 29.3%, compared with a dine-in GP after cost of sales of (£19.00 - £5.70) / £19.00 = 70.0%. Step two: but delivery uses no table and less front-of-house labour. Suppose the extra kitchen labour for the order costs £1.50. The delivery contribution becomes £4.06. A dine-in order of the same food might carry £3.50 of service labour and contribute £19.00 - £5.70 - £3.50 = £9.80 before overheads. Delivery earns well under half as much per order. Step three: set delivery prices. Many restaurants raise delivery menu prices, where the platform allows it. Raising every item by 15%, the order becomes £26.22, the net is £21.85, commission is £7.87, and the contribution is £21.85 - £5.70 - £0.90 - £7.87 - £1.50 = £5.88. Step four: volume and capacity. Delivery can make sense when it fills kitchen capacity in quiet hours, because the fixed costs are already paid. If 150 orders a week at £5.88 add £882 of contribution without needing extra fixed staff, it is a real gain. But if delivery orders compete with dine-in guests at peak, slowing the kitchen, they may cost more than they earn. Step five: consider direct ordering. A restaurant's own ordering site removes most of the commission but needs its own marketing and delivery arrangements. Compare the cost per order on each channel monthly. - In this example 30% commission is £6.84 on a £22.80 order. - The restaurant keeps £5.56 after food, packaging and commission. - Hot food delivered is standard-rated for VAT. - A 15% delivery mark-up raises the contribution to £5.88 here. - Delivery works best when it fills spare kitchen capacity. ### Hotels #### Hotel occupancy: a worked example Occupancy is the share of available rooms that were sold over a period. It is the first number most hoteliers check each morning, but it needs to be calculated consistently to be useful. In this example a 40-room hotel reports on a 30-night month. Step one: available room nights. 40 rooms x 30 nights = 1,200 room nights. If two rooms were out of order for 10 nights each because of a leak and a refurbishment, some hotels deduct them: 2 x 10 = 20, leaving 1,180 available. Others keep the full 1,200 so that rooms taken out of service show up as lost potential. Either approach is fine as long as it is stated and used every month. Step two: rooms sold. The hotel sold 900 room nights to paying guests. It also gave 12 complimentary room nights to a travel writer and a contractor, and 8 room nights were used by a duty manager. Paying occupancy is 900 / 1,200 = 75.0%. Including complimentary rooms, occupancy is 912 / 1,200 = 76.0%. House use is usually excluded from both. Step three: weekday and weekend split. In this example Friday and Saturday nights, 8 of the 30 nights, sold 300 of the 900 room nights. Weekend occupancy is 300 / (40 x 8) = 300 / 320 = 93.8%. Weekday occupancy is 600 / (40 x 22) = 600 / 880 = 68.2%. The headline 75% hides a near-full weekend and a soft midweek, and that shape should drive sales effort: corporate accounts, meetings and midweek packages rather than more weekend marketing. Step four: guest occupancy, or density. If those 900 rooms held 1,530 guests, the average number of guests per occupied room is 1,530 / 900 = 1.7. That matters for breakfast covers, linen and amenity costs, and for upselling dinner. Step five: compare with a target. If the budget for the month was 78%, the shortfall is 3 points, which is 0.03 x 1,200 = 36 room nights. At an average rate of £120 net, that is 36 x £120 = £4,320 of room revenue below budget. High occupancy is not automatically good. A hotel that fills every room by discounting heavily may earn less than one at lower occupancy with a stronger rate, because each occupied room carries cleaning, laundry, amenities and often commission. That is why occupancy is always read alongside the average daily rate and revenue per available room. Track occupancy on the books for future dates too, known as pick-up, so you can see whether a weekend three weeks out is filling faster or slower than the same weekend last year, and adjust prices in time. - Occupancy equals rooms sold divided by rooms available. - In this example 900 of 1,200 room nights sold is 75%. - Weekends ran at 93.8% and weekdays at 68.2% here. - State whether complimentary and out-of-order rooms are included. - Three points below budget cost £4,320 of room revenue in this example. #### ADR: a worked example ADR, the average daily rate, is room revenue divided by the number of rooms sold. It tells you the average price actually achieved per occupied room night, after discounts, packages and promotions. In this example a 40-room hotel sold 900 room nights in a month across four segments, all prices net of VAT. Segment one: direct bookings through the hotel's website, 300 room nights at an average of £135. Revenue is 300 x £135 = £40,500. Segment two: online travel agents, 360 room nights at an average of £125, so £45,000. Segment three: corporate contracted rates, 150 room nights at £105, so £15,750. Segment four: a tour operator allocation, 90 room nights at £80, so £7,200. Total room revenue is £40,500 + £45,000 + £15,750 + £7,200 = £108,450. ADR is £108,450 / 900 = £120.50. Step two: watch what goes into room revenue. Many hotels sell bed and breakfast packages. Suppose the direct rate of £135 includes breakfast for two guests, and the hotel values breakfast at £12 per person net. The room element is £135 - (2 x £12) = £111, and £24 should be recorded as food and beverage revenue. If all 300 direct room nights were sold this way, room revenue falls by 300 x £24 = £7,200 to £101,250, and the true room ADR becomes £101,250 / 900 = £112.50. Reporting the inclusive figure as room revenue would overstate ADR by £8. Step three: ADR is gross of commission. The OTA segment earns £125 a night but pays, in this example, 18% commission, which is £22.50. The net ADR from OTAs is £102.50, lower than the corporate rate of £105. Some hotels track a net ADR after distribution costs alongside the headline figure. Step four: see how the mix moves the average. If the hotel replaces the 90 tour operator nights with 90 more direct nights at £135, revenue rises by 90 x (£135 - £80) = £4,950 to £113,400, and ADR rises to £113,400 / 900 = £126.00. No individual price changed, but the average rose by £5.50 because of the mix. Step five: ADR alone can mislead. A hotel that raises its prices and sells fewer rooms may report a higher ADR and lower total revenue. In this example, raising every rate by 10% while losing 12% of room nights gives revenue of 1.10 x 0.88 = 0.968 of the original, a 3.2% fall, even though ADR is up 10%. That is why ADR is always read with occupancy, and why the combined measure RevPAR exists. - ADR equals room revenue divided by rooms sold. - In this example £108,450 over 900 room nights gives an ADR of £120.50. - Strip breakfast out of package rates before reporting room revenue. - A shift in segment mix can move ADR with no price change. - A higher ADR with fewer rooms sold can mean less revenue. #### RevPAR: a worked example RevPAR, revenue per available room, combines occupancy and rate into a single figure. It can be calculated two ways that should give the same answer: room revenue divided by available room nights, or ADR multiplied by occupancy. In this example a 40-room hotel has 1,200 available room nights in a 30-night month and sells 900 of them for £108,000 of room revenue, net of VAT. Method one: £108,000 / 1,200 = £90.00. Method two: ADR is £108,000 / 900 = £120.00, occupancy is 900 / 1,200 = 75%, and £120.00 x 0.75 = £90.00. Both give £90. Now use RevPAR to compare two pricing strategies for next month. Strategy A keeps the rate at £120 and expects 75% occupancy, giving RevPAR of £90.00. Strategy B cuts the rate to £105 and expects occupancy to rise to 85%. RevPAR is £105 x 0.85 = £89.25. On RevPAR alone the two look almost the same. But each occupied room costs money to service. In this example the variable cost per occupied room, covering housekeeping time, laundry, amenities and booking commission, is £35. Expressed per available room, strategy A costs 0.75 x £35 = £26.25, and strategy B costs 0.85 x £35 = £29.75. Profit per available room is £90.00 - £26.25 = £63.75 for A and £89.25 - £29.75 = £59.50 for B. Over 1,200 available room nights, A earns 1,200 x £63.75 = £76,500 and B earns 1,200 x £59.50 = £71,400. Strategy A is £5,100 better, despite selling 120 fewer room nights. There is a counter-argument. Extra guests in strategy B may spend in the bar and restaurant. If each extra occupied room produces £30 of food and drink contribution, B's 120 extra room nights add 120 x £30 = £3,600, narrowing the gap to £1,500. If they produce £50 each, B wins by £900. That is why some hotels look at total revenue per available room, sometimes called TRevPAR, as well. In this example, if the hotel also took £45,000 of food, beverage and other revenue in the month, TRevPAR is (£108,000 + £45,000) / 1,200 = £127.50. RevPAR is most useful for comparing a hotel with itself over time and with a competitive set of similar hotels nearby. If the market's RevPAR rose 6% and yours rose 2%, you lost share even though you grew. Compare like with like: the same days of the week, the same events calendar, and the same treatment of packages and VAT. - RevPAR equals room revenue divided by available room nights. - It also equals ADR multiplied by occupancy. - In this example £120 at 75% gives £90 RevPAR. - A lower rate at higher occupancy earned £5,100 less here once costs were counted. - TRevPAR adds food, beverage and other revenue. #### GOPPAR: a worked example GOPPAR, gross operating profit per available room, measures how much operating profit a hotel makes for every room it has available, whether sold or not. Where RevPAR looks only at room revenue, GOPPAR captures all departments and all operating costs, so it shows whether revenue growth is actually turning into profit. In this example a 40-room hotel has 40 x 365 = 14,600 available room nights in a year. Its total revenue is £1,874,000 across rooms, food and beverage and other income, all net of VAT. After departmental costs and undistributed operating costs such as administration, sales, maintenance and utilities, the gross operating profit is £617,392. GOPPAR is £617,392 / 14,600 = £42.29. For comparison, room revenue is £1,314,000, so RevPAR is £1,314,000 / 14,600 = £90.00. TRevPAR is £1,874,000 / 14,600 = £128.36. The GOP margin is £617,392 / £1,874,000 = 32.9%. Now compare two years. In year two the hotel pushes occupancy with discounting and big group deals. Room revenue rises to £1,380,000, RevPAR to £94.52, and total revenue to £1,950,000. But commission, labour and laundry rise faster than revenue, and GOP falls to £600,000. GOPPAR becomes £600,000 / 14,600 = £41.10. RevPAR rose by 5.0% while GOPPAR fell by 2.8%. Measured on RevPAR alone, the year looks like a success; GOPPAR shows it was not. Step two: use GOPPAR to compare properties. A group owns this 40-room hotel and a 25-room hotel in another town. The smaller hotel has 25 x 365 = 9,125 available room nights and a GOP of £420,000. Its GOPPAR is £420,000 / 9,125 = £46.03, higher than the larger hotel's £42.29, even though its total profit is lower. Per room, it is the better-run or better-placed business. Step three: test an investment. Refurbishing the 40-room hotel's bathrooms will cost £200,000 and, in this example, is expected to raise GOP by £45,000 a year through higher rates. GOPPAR would rise by £45,000 / 14,600 = £3.08 to £45.37. The simple payback is £200,000 / £45,000 = 4.4 years. GOPPAR comes before rent, rates, insurance, interest and depreciation, so it describes the operation rather than the ownership structure. That makes it popular with owners and management companies negotiating fees, which are often linked to GOP. When comparing GOPPAR across hotels, check that costs are classified in the same way, for example whether utilities and maintenance sit above or below the GOP line. - GOPPAR equals gross operating profit divided by available room nights. - In this example £617,392 / 14,600 gives a GOPPAR of £42.29. - RevPAR can rise while GOPPAR falls. - A smaller hotel can have a higher GOPPAR and lower total profit. - Check cost classification before comparing hotels. #### OTA commission vs direct booking: a worked example Online travel agents bring guests a hotel might never reach alone, but they charge commission on every booking. Direct bookings avoid that commission but carry their own costs. Comparing the two properly means counting everything. In this example a hotel sells a room for £150 a night net of VAT, and the average stay is 2 nights, so each booking is worth £300. Through an OTA charging 18% commission, the cost is 0.18 x £300 = £54. The OTA collects payment in some models and pays the hotel later, or the hotel collects and is invoiced for the commission; either way, the hotel keeps £246. Direct booking costs. The booking engine charges 2% of booking value in this example, which is £6. Card processing at 1.5% is £4.50. The hotel spends on search advertising, metasearch listings and email marketing; if that spend is £4,000 a month and produces 250 direct bookings, the marketing cost is £16 per booking. Total direct cost is £6 + £4.50 + £16 = £26.50, so the hotel keeps £273.50. The direct booking is worth £27.50 more, or 9.2% of the booking value. Step two: scale it. In this example the hotel takes 1,800 bookings a year, 55% via OTAs and 45% direct. OTA bookings: 990 x £54 = £53,460 in commission. Direct bookings: 810 x £26.50 = £21,465 in costs. Total distribution cost: £74,925. If the hotel shifts 10% of bookings, 180 of them, from OTA to direct, the saving is 180 x £27.50 = £4,950 a year. Step three: be realistic about why guests book through OTAs. Some come for the loyalty points, the cancellation flexibility or the comparison. Many guests find a hotel on an OTA and then visit its own website, which is sometimes called the billboard effect. Cutting OTA availability sharply can lose more bookings than it shifts. Step four: incentives for direct booking. Common tactics include a small perk such as late check-out, free breakfast or a drink on arrival, a flexible cancellation policy, and member rates for guests who sign up. Check the terms of each OTA contract, because some include rate parity or related clauses, and the rules on what is allowed vary by market and change over time. In this example, offering a breakfast worth £10 net cost to the hotel of £4 per guest, for two guests over two nights, costs £16, still leaving direct bookings £11.50 ahead. Track the cost of acquisition by channel every month, including marketing spend, and not just the commission rate. - In this example OTA commission at 18% costs £54 on a £300 booking. - Direct costs total £26.50 including marketing per booking. - Each booking moved direct is worth £27.50 more here. - Many guests find a hotel on an OTA, then book direct. - Check OTA contracts for parity and related clauses. #### Overbooking maths: a worked example Hotels overbook because some guests do not turn up and some cancel too late to resell the room. Overbooking by the right amount fills those rooms; overbooking by too much means turning guests away, known as walking them, which is costly and damages reputation. In this example a 40-room hotel is full for a Tuesday night. History shows that 5% of guests on non-guaranteed, flexible bookings for midweek nights do not arrive, and we assume each booking behaves independently. Each extra room sold earns £120 net. Walking a guest costs £250 in this example: a room at a nearby hotel, a taxi, a goodwill gesture and an allowance for lost future business. Question: should the hotel accept a 41st booking? If it does, there are 41 bookings for 40 rooms. The 41st booking earns money if at least one guest does not show. The chance that all 41 arrive is 0.95 to the power 41. Using logarithms, 41 x ln(0.95) = 41 x -0.0513 = -2.103, and e to the power -2.103 is about 0.122. So there is a 12.2% chance everyone turns up and one guest is walked, and an 87.8% chance at least one does not, so the extra booking fills a room that would otherwise be empty. Expected gain from the 41st booking: 0.878 x £120 = £105.36. Expected cost: 0.122 x £250 = £30.50. Net expected value: £105.36 - £30.50 = £74.86. Accept it. Now the 42nd booking. It earns money only if at least two guests among 42 fail to show; otherwise someone is walked. The chance of no no-shows among 42 is 0.95 to the power 42, about 0.116. The chance of exactly one no-show is 42 x 0.05 x 0.95 to the power 41, which is 2.1 x 0.122 = 0.256. So the chance of one or fewer no-shows is 0.116 + 0.256 = 0.372. Expected gain: 0.628 x £120 = £75.36. Expected cost: 0.372 x £250 = £93.00. Net: -£17.64. Decline it. The best level in this example is one room over. Notice how sensitive this is. If walk costs were £150 instead of £250, the 42nd booking would be worth 0.628 x £120 - 0.372 x £150 = £75.36 - £55.80 = £19.56, and two rooms over would make sense. If the no-show rate on a guaranteed, prepaid weekend is close to zero, overbooking makes no sense at all. Real systems use the hotel's own history by day of week, season, booking channel and rate type. Always have a written walking procedure, agree arrangements with nearby hotels in advance, and treat walked guests generously. - In this example 5% of midweek flexible bookings do not arrive. - A 41st booking has an expected value of about £74.86. - A 42nd booking has an expected value of about -£17.64. - Lower walk costs justify more overbooking. - Use your own no-show history by day, season and channel. #### Group rate displacement: a worked example A group booking can look attractive because it fills many rooms at once, but if those rooms would otherwise have sold to individual guests at higher rates, the group may displace more profitable business. Displacement analysis compares the two. In this example a 40-room hotel receives a request from a wedding party for 20 rooms on a Saturday night at £95 per room net of VAT. The party also wants a private dinner for 40 guests. Step one: forecast what would happen without the group. Looking at the booking pace and last year's figures, the revenue manager expects to sell 12 of those 20 rooms to individual guests at an average of £130, leaving 8 empty. Step two: room profit with the group. Variable cost per occupied room is £35 in this example. Group room profit is 20 x (£95 - £35) = 20 x £60 = £1,200. Step three: room profit without the group. Transient room profit is 12 x (£130 - £35) = 12 x £95 = £1,140. On rooms alone the group is only £60 better. Step four: count the extra spend. The private dinner is 40 covers at £45 net, or £1,800 of revenue. At a food and drink contribution of 60% after cost of sales and extra staff, that is £1,080. Transient guests also eat and drink, though less predictably; in this example the 12 displaced rooms would have brought 20 guests spending £25 each in the restaurant and bar, £500 at the same 60%, a contribution of £300. Step five: compare. With the group: £1,200 + £1,080 = £2,280. Without: £1,140 + £300 = £1,440. The group is worth £840 more. Accept it. Now change one assumption. Suppose there is a large event in town that night and the forecast is to sell all 20 rooms to individuals at £170. Transient room profit becomes 20 x (£170 - £35) = £2,700, plus restaurant contribution of, say, £500, total £3,200. The group's £2,280 now falls short by £920. The hotel should either decline the group, offer fewer rooms, or quote a higher rate. To match £3,200, the group room rate would need to rise by £920 / 20 = £46, to £141. Displacement analysis is only as good as the forecast. Check booking pace, events, and last year's same night. Also consider the length of stay: if the group checks in on Saturday but individuals tend to book Friday and Saturday together, taking the group may block two-night stays and displace Friday business too. And remember the group's other value: a well-run wedding can bring future bookings from guests who liked the venue. - Displacement compares group profit with the transient profit it replaces. - In this example the group is worth £840 more on a normal Saturday. - On an event night the same group falls £920 short. - Food and beverage spend can decide the case. - Watch for groups that block multi-night stays. #### Hotel break-even occupancy: a worked example Break-even occupancy is the occupancy a hotel needs to cover all its costs. It is a useful reality check when buying a hotel, setting a budget, or deciding whether to close in a quiet season. In this example a 30-room bed and breakfast hotel has annual fixed costs of £600,000. These include salaried staff and the minimum team needed to open, rent, business rates, insurance, utilities standing charges, maintenance, marketing, software and loan repayments. Variable cost per occupied room night is £30 in this example: housekeeping time beyond the core team, laundry, amenities, breakfast food cost, card fees and commission. The ADR is £110 net of VAT. Step one: contribution per occupied room. £110 - £30 = £80. Step two: break-even room nights. £600,000 / £80 = 7,500 room nights. Step three: available room nights. 30 rooms x 365 = 10,950. Step four: break-even occupancy. 7,500 / 10,950 = 68.5%. If the hotel averages 75% occupancy, it sells 0.75 x 10,950 = 8,213 room nights. The surplus above break-even is 8,213 - 7,500 = 713 room nights, and profit is 713 x £80 = £57,040. Step five: sensitivity. If ADR falls to £100, contribution falls to £70 and break-even room nights rise to £600,000 / £70 = 8,571, which is 78.3% occupancy. A £10 fall in rate pushes break-even almost 10 points higher, above the 75% the hotel actually achieves, and the year turns into a loss of (8,571 - 8,213) x £70 = £25,060. If instead fixed costs rise by £50,000, break-even at the original £80 contribution becomes 650,000 / 80 = 8,125 room nights, or 74.2%. Step six: seasonality. Many UK coastal and rural hotels trade strongly in summer and weakly in winter. Suppose the hotel considers closing for January and February, 59 nights. During those months it would sell, in this example, only 35% of rooms: 0.35 x 30 x 59 = 620 room nights, earning 620 x £80 = £49,600 of contribution. If closing saves £55,000 of costs that are fixed while open but avoidable when shut, such as seasonal staff, heating and some utilities, then closing improves profit by £5,400. If it saves only £40,000, staying open is better by £9,600. The key is separating costs that genuinely stop when the doors close from those that continue regardless, such as rent, rates and insurance. Recalculate break-even every year when the budget is set, and check it monthly against actual results. - Break-even room nights equal fixed costs divided by contribution per room. - In this example £600,000 / £80 = 7,500 room nights, 68.5% occupancy. - A £10 fall in ADR raises break-even to 78.3% here. - Seasonal closure only pays if avoidable costs exceed lost contribution. - Separate avoidable costs from those that continue when closed. ### Funding and growth #### Start-up budget for a bar: a worked example A start-up budget lists every cost of getting a venue open and trading, plus a reserve to survive the first few months. Underestimating it is the most common reason new bars run out of money before they find their feet. In this example a first-time operator takes a lease on an empty unit in a market town to open a 70-capacity cocktail and wine bar. All figures are illustrative and exclude VAT unless stated. Professional fees for a solicitor and a surveyor to review the lease and the building: £12,000. Licensing applications and associated costs, including a premises licence application and plans: £1,500 in this example; fees depend on the local authority and the rateable value. Fit-out, covering the bar counter, flooring, lighting, toilets, ventilation, electrics and decoration: £120,000. Bar equipment such as glasswashers, back-bar fridges, an ice machine, a cellar cooling unit and dispense: £35,000. Furniture: £18,000. Till system and IT: £6,000. Sound system: £8,000. Signage: £5,000. Glassware and smallwares: £6,000. Opening stock of drinks and consumables: £12,000. Rent deposit of three months: £15,000. Rent during a two-month fit-out: £10,000. Pre-opening staff recruitment and training: £9,000. Launch marketing: £5,000. Step one: add them up. £12,000 + £1,500 + £120,000 + £35,000 + £18,000 + £6,000 + £8,000 + £5,000 + £6,000 + £12,000 + £15,000 + £10,000 + £9,000 + £5,000 = £262,500. Step two: add a contingency. Building work almost always finds surprises, such as asbestos, drainage or an inadequate electricity supply. In this example the contingency is 10%: 0.10 x £262,500 = £26,250. Step three: add working capital. New venues rarely trade at their target level from day one. The bar's fixed costs are £15,000 a month, and the plan holds three months in reserve: £45,000. Step four: the total. £262,500 + £26,250 + £45,000 = £333,750. Step five: VAT cash flow. VAT on the fit-out and equipment, at 20% on, say, £200,000 of VAT-able spend, is £40,000. A VAT-registered business can usually reclaim it, but the money has to be found first and the refund arrives later. That must be funded, often with a short-term facility. Step six: the funding plan. In this example the founder puts in £100,000 of savings, a bank loan provides £150,000, and two friends invest £83,750 for a minority shareholding. Every source has conditions: the loan needs repaying monthly from the start, and the investors expect a return. A start-up budget should be built line by line from real quotes, reviewed by someone who has opened a venue before, and kept alongside a first-year cash flow forecast. - In this example the core spend is £262,500 before contingency. - A 10% contingency adds £26,250. - Three months of fixed costs adds £45,000 of working capital. - The total requirement is £333,750. - VAT on fit-out must be funded before it is reclaimed. #### Payback period: a worked example The payback period is how long it takes for the cash generated by an investment to repay the money put in. It is simple, easy to explain to a lender or partner, and a good first test of whether a project is worth doing. In this example a bar costs £333,750 to open, including fit-out, contingency and working capital. Step one: forecast the cash the business generates each year. For simplicity this example uses EBITDA less maintenance spending as a proxy for cash available to repay the investment: year one £60,000 while the bar builds a following, year two £85,000, year three £95,000, and £95,000 a year after that. A fuller analysis would deduct tax and interest too. Step two: add it up year by year. End of year one: £60,000 recovered. End of year two: £60,000 + £85,000 = £145,000. End of year three: £145,000 + £95,000 = £240,000. End of year four: £240,000 + £95,000 = £335,000. The investment of £333,750 is recovered during year four. Step three: find the fraction of year four. At the start of year four, £333,750 - £240,000 = £93,750 remains. Year four brings £95,000, so the fraction is £93,750 / £95,000 = 0.987. Payback is 3.99 years, call it four years. Step four: compare with the simple steady-state calculation. Dividing the investment by the mature annual cash, £333,750 / £95,000 = 3.5 years, understates payback because it ignores the slow first two years. Always model the ramp-up. Step five: allow for the time value of money. A pound in four years is worth less than a pound today. Using a discount rate of 8% in this example, year one's £60,000 is worth £60,000 / 1.08 = £55,556 today; year two's £85,000 is worth £85,000 / 1.08 squared, which is £85,000 / 1.1664 = £72,874; year three's £95,000 is worth £95,000 / 1.2597 = £75,415; year four's is worth £95,000 / 1.3605 = £69,827. Cumulative discounted cash after four years is £55,556 + £72,874 + £75,415 + £69,827 = £273,672, still short of £333,750. Year five's £95,000 is worth £95,000 / 1.4693 = £64,655, bringing the total to £338,327, so discounted payback is just under five years. Step six: judge it against the lease. If the lease runs for ten years, a four-to-five-year payback leaves a reasonable period of profit. If the lease has a break clause at year five or a short term, the risk is much greater. Payback ignores everything after the investment is recovered, so use it as a screen, not the final word. - Payback is the time for cumulative cash to repay the investment. - In this example payback is 3.99 years. - Dividing by mature cash alone gives a misleading 3.5 years. - Discounted at 8%, payback stretches to just under five years. - Compare payback with the lease length and any break clause. #### Loan repayments: a worked example Most business loans are repaid in equal monthly instalments that cover interest and part of the capital. Early payments are mostly interest; later ones are mostly capital. In this example a restaurant borrows £100,000 at a fixed annual interest rate of 8%, repaid monthly over five years. Rates and terms are illustrative; real offers depend on the lender, security and the business's record. Step one: the monthly rate and number of payments. Monthly rate r = 8% / 12 = 0.6667%, or 0.0066667. Number of payments n = 5 x 12 = 60. Step two: the standard repayment formula. Monthly payment = P x r / (1 - (1 + r) to the power -n), where P is the amount borrowed. (1.0066667) to the power 60 is about 1.4898, and its reciprocal is about 0.6712. So the denominator is 1 - 0.6712 = 0.3288. P x r = £100,000 x 0.0066667 = £666.67. Monthly payment = £666.67 / 0.3288 = £2,027.64. Step three: the first month. Interest is £100,000 x 0.0066667 = £666.67. The capital repaid is £2,027.64 - £666.67 = £1,360.97, leaving a balance of £98,639.03. In month two, interest is £98,639.03 x 0.0066667 = £657.59 and capital repaid is £1,370.05. Each month, interest shrinks and capital repaid grows. Step four: totals. Over 60 months the restaurant pays 60 x £2,027.64 = £121,658.40, so the total interest is £21,658.40. Step five: compare a ten-year term. With n = 120, (1.0066667) to the power 120 is about 2.2196, its reciprocal about 0.4505, the denominator 0.5495, and the payment £666.67 / 0.5495 = £1,213.28. Total paid is 120 x £1,213.28 = £145,593.60, and interest is £45,593.60, more than double. The longer term eases monthly cash flow by £814.36 but costs £23,935.20 more in interest. Step six: what it means for the P&L. Only the interest is an expense in the P&L; the capital part reduces the debt on the balance sheet. But both leave the bank account. In the first year of the five-year loan, cash out is 12 x £2,027.64 = £24,331.68, of which roughly £7,400 is interest. A business making £30,000 of operating profit would have very little cash left after the loan, before tax and any drawings. Lenders often test this with a debt service cover ratio: cash available for debt service divided by total loan payments. In this example £30,000 / £24,332 = 1.23. Many lenders want a comfortable margin above 1, so check what your lender requires. - In this example £100,000 at 8% over five years costs £2,027.64 a month. - Total interest over five years is £21,658.40. - A ten-year term cuts the payment to £1,213.28 but doubles interest. - Only interest is a P&L expense; capital still leaves the bank. - Debt service cover ratio here is about 1.23. #### Landlord rent review: a worked example Commercial leases often include rent reviews, commonly every five years, at which the rent is reset. How it is reset depends entirely on the lease wording, so the clause deserves careful reading and advice from a solicitor and a chartered surveyor before signing. This example compares two common types; it is illustrative, not legal advice. In this example a restaurant pays a passing rent of £60,000 a year on a 3,000 square foot unit, which is £20 per square foot. Its annual net sales are £700,000, so rent is 8.6% of sales. The lease has a review at year five. Type one: an open market review. The new rent is set at what a willing tenant would pay a willing landlord for the premises on the lease terms, usually assessed from recent comparable lettings. Suppose comparable restaurant units nearby have let at £22 per square foot. The new rent is 3,000 x £22 = £66,000. Many leases make reviews upward-only: if the market had fallen to £18 per square foot, the rent would stay at £60,000 rather than drop to £54,000. Type two: an index-linked review, tied to an inflation measure, often with a collar and a cap. In this example the lease uses an inflation index with a collar of 1% and a cap of 4% a year, compounded over five years. The collar gives a minimum of 1.01 to the power 5 = 1.0510, so the rent cannot fall below £60,000 x 1.0510 = £63,061. The cap gives a maximum of 1.04 to the power 5 = 1.2167, so the rent cannot exceed £60,000 x 1.2167 = £72,999. If inflation over the five years totals 15%, the rent rises to £60,000 x 1.15 = £69,000, within the band. Step two: the effect on profit. Under the open market review, rent rises £6,000 a year. Under the index-linked review with 15% inflation, it rises £9,000. If sales have also grown 15% to £805,000, the index-linked rent of £69,000 is still 8.6% of sales, and the business is no worse off proportionally. If sales have only grown 5% to £735,000, rent becomes 9.4% of sales, and roughly £5,800 of profit has gone to the landlord, the difference between £69,000 and 8.6% of £735,000, which is £63,210. Step three: negotiate at the outset. Points to consider include the review pattern, whether reviews are upward-only, the assumptions the valuer must use, such as ignoring the tenant's own improvements, and how disputes are settled. Diary each review date well in advance, gather your own comparable evidence and instruct a surveyor early. - Rent reviews reset rent, often every five years. - In this example an open market review raises rent to £66,000. - Upward-only reviews never reduce the rent. - A 1% collar and 4% cap bound the five-year rent between £63,061 and £72,999. - Take advice from a solicitor and chartered surveyor before signing. #### Turnover rent: a worked example A turnover rent links some or all of the rent to the tenant's sales. It is common in shopping centres, transport hubs, food halls and some leisure schemes. It shares risk between tenant and landlord: the tenant pays less in a bad year and more in a good one. In this example a café-bar is offered a choice of two lease structures. Option A is a fixed rent of £50,000 a year. Option B is a base rent of £40,000 plus 8% of annual net turnover above £550,000. Turnover is defined in the lease; in this example it means sales excluding VAT, tips and service charges passed to staff. Step one: calculate option B at three sales levels. At £500,000 of turnover, sales are below the £550,000 threshold, so rent is the base of £40,000. At £700,000, the excess is £150,000, so the top-up is 0.08 x £150,000 = £12,000, and rent is £52,000. At £900,000, the excess is £350,000, the top-up is £28,000, and rent is £68,000. Step two: compare with option A. At £500,000, option B saves £10,000. At £700,000, option B costs £2,000 more. At £900,000, it costs £18,000 more. Step three: find the crossover. The two options cost the same when £40,000 + 0.08 x (T - £550,000) = £50,000. That means 0.08 x (T - £550,000) = £10,000, so T - £550,000 = £125,000, and T = £675,000. Below £675,000 of turnover, option B is cheaper; above it, option A is. Step four: express rent as a share of sales. Under option A, rent is 10.0% of sales at £500,000, 7.1% at £700,000 and 5.6% at £900,000. Under option B, it is 8.0%, 7.4% and 7.6%. Option B keeps the rent burden steadier, which is the point of the structure. Step five: think about the forecast. If the business plan expects £650,000 in year one rising to £800,000 by year three, option B saves £2,000 in year one, when cash is tightest (£48,000 against £50,000), and costs £10,000 more by year three (£60,000 against £50,000). An operator who values protection in the uncertain early years might choose B. One confident of strong trade would choose A. Step six: read the small print. The definition of turnover decides what is counted: delivery platform sales, gift vouchers, deposits, staff meals and refunds can all be treated differently. Leases often require monthly or quarterly turnover statements and give the landlord rights to audit, so keep clean, reconcilable till records. Take professional advice before agreeing any turnover rent. - Turnover rent links rent to the tenant's sales. - In this example the fixed and turnover options cross over at £675,000. - Below the crossover the turnover option is cheaper. - Turnover rents keep rent as a steadier share of sales. - The lease's definition of turnover decides what is counted. #### Second site maths: a worked example Opening a second site is one of the biggest decisions an independent operator makes. The first site's success is the obvious argument for it, but the economics of a two-site business are different from those of one site doubled. In this example a neighbourhood restaurant makes £150,000 of EBITDA a year. The owner is considering a second site that will cost £400,000 to open, including fit-out, equipment, fees, contingency and working capital. Step one: forecast the second site on its own. At maturity, expected in year three, it should make £120,000 of EBITDA, a little less than the first because the rent is higher. Year one is forecast at 60% of maturity, which is £72,000, and year two at 85%, which is £102,000. Step two: add the central costs. With one site, the owner does the buying, rotas, accounts and marketing personally. With two, the business needs help: in this example an operations manager at £45,000 including employer costs, and extra accountancy and software of £10,000. Total new central cost is £55,000 a year. Step three: net gain to the group. Year one: £72,000 - £55,000 = £17,000. Year two: £102,000 - £55,000 = £47,000. Year three onwards: £120,000 - £55,000 = £65,000. Step four: allow for distraction. When the owner spends time opening the second site, the first often slips. In this example the first site's EBITDA falls 10%, or £15,000, in year one and recovers by year two. Year one's net gain falls to £17,000 - £15,000 = £2,000. Step five: payback. Cumulative net gain is £2,000 after year one, £49,000 after year two, £114,000 after year three, and then grows by £65,000 a year: £179,000, £244,000, £309,000, £374,000, and £439,000 after year eight. The £400,000 investment is recovered during year eight: at the end of year seven £26,000 remains, which is £26,000 / £65,000 = 0.4 of year eight, giving 7.4 years. That is much longer than the four to five years many investors look for, and longer than many leases leave comfortable. Step six: what changes the answer. Central costs are lumpy; a third site would use the same operations manager, so the central cost per site falls from £55,000 to about £37,000 over three. A cheaper fit-out, perhaps taking over a site already set up as a restaurant, could halve the investment. Stronger trading at maturity, say £150,000, lifts the annual gain to £95,000. Run the numbers with sober assumptions, and test the plan with the first site's figures falling as well as the second's rising. - In this example a £400,000 second site matures at £120,000 EBITDA. - New central costs of £55,000 cut the group gain to £65,000. - Distraction costs the first site £15,000 in year one. - Payback is about 7.4 years on these assumptions. - Central costs spread better across three or more sites. #### Franchise fees: a worked example A franchise lets an operator trade under an established brand, using its systems, menu, suppliers and marketing, in exchange for fees. The usual structure is an initial fee, a royalty on sales and a marketing levy, though each franchisor sets its own terms. Read the agreement carefully and take independent legal and financial advice. In this example a franchisee is comparing a franchised burger restaurant with an independent concept in the same unit, all figures net of VAT. Franchise terms in this example: an initial fee of £25,000, a royalty of 6% of net sales and a marketing levy of 2% of net sales, both paid monthly. The franchisor's brand is expected to bring higher sales: £800,000 a year against £650,000 for the independent. Step one: the franchise's annual fees. Royalty: 0.06 x £800,000 = £48,000. Marketing levy: 0.02 x £800,000 = £16,000. Total: £64,000 a year. Step two: EBITDA before and after fees. Suppose both concepts run at 15% EBITDA before any franchise fees, which is generous for the independent since it also pays its own marketing. The franchise makes 0.15 x £800,000 = £120,000 before fees and £120,000 - £64,000 = £56,000 after, which is 7.0% of sales. The independent makes 0.15 x £650,000 = £97,500, but spends £15,000 on its own marketing in this example, leaving £82,500, or 12.7%. On these assumptions the independent earns £26,500 more each year. Step three: what sales would the franchise need to match the independent? The franchise keeps 15% - 8% = 7% of sales as EBITDA. To make £82,500, it needs £82,500 / 0.07 = £1,178,571 of sales, 47% more than the forecast. Step four: count the benefits that the fee buys. A proven brand can reduce the risk of failure; the franchisor's buying power may lower food costs by a point or two; training, recipes and operating manuals save time; and lenders may lend more readily against a known concept. If buying power improves the franchise's margin before fees from 15% to 17%, EBITDA after fees becomes 0.09 x £800,000 = £72,000, narrowing the gap to £10,500. Step five: the initial fee and fit-out. The £25,000 initial fee is often accompanied by a required fit-out to brand standards, which may cost more than an independent's design. Over a ten-year agreement, the initial fee is £2,500 a year. Also check the terms on renewal, territory protection, required refurbishments and what happens if you want to sell. - In this example fees are a £25,000 initial fee, 6% royalty and 2% levy. - On £800,000 of sales, annual fees are £64,000. - The franchise makes £56,000 against the independent's £82,500 here. - To match, the franchise would need about £1.18m of sales. - Buying power and lower risk are what the fees pay for. #### Equipment lease vs buy: a worked example Kitchens and bars need expensive equipment, and the choice between buying it outright, borrowing to buy it, or leasing it affects cash flow, cost and flexibility. In this example a restaurant needs a combination oven costing £12,000 net of VAT. Option A: buy with cash. The business pays £12,000 now. It expects to use the oven for four years and then sell it for about £2,000. The net cost over four years is £12,000 - £2,000 = £10,000. Servicing costs £400 a year, so £1,600 over four years, for a total of £11,600. Option B: lease. The supplier offers a four-year lease at £330 a month, including servicing, with the oven returned at the end. Total payments are 48 x £330 = £15,840. The lease costs £15,840 - £11,600 = £4,240 more than buying with cash. Option C: buy with a loan. A four-year loan of £12,000 at 9% has monthly payments of about £298.62, using the standard repayment formula: monthly rate 0.0075, 48 payments, (1.0075) to the power 48 about 1.4314, reciprocal 0.6986, denominator 0.3014, and £90.00 / 0.3014 = £298.62. Total payments are 48 x £298.62 = £14,333.76, so interest is £2,333.76. Add servicing of £1,600 and subtract the £2,000 sale value: total cost £13,933.76. That is £1,906.24 cheaper than leasing and £2,333.76 more expensive than cash. Step two: the value of keeping cash. If the £12,000 kept by leasing would otherwise have to be borrowed on an overdraft at a higher rate, or if it would let the business survive a slow winter, leasing may be worth the premium. A new venue with limited cash should usually protect its working capital before it saves a few thousand pounds over four years. Step three: tax and VAT. VAT is charged on lease payments and on purchases; a VAT-registered business can generally reclaim it either way. The tax treatment differs: purchased equipment usually qualifies for capital allowances, while lease rentals are usually deductible as an expense. The detail affects the after-tax cost, so check current rules on GOV.UK or with an accountant. Step four: the flexibility question. A lease may include servicing and replacement if the oven fails, which matters in a kitchen that cannot work without it. Owning the oven means repairs are your problem but you can sell it or move it to another site. Step five: read the lease terms. Some leases cannot be ended early without paying all remaining rentals, so a business that closes after a year could still owe 36 x £330 = £11,880. - In this example buying for cash costs £11,600 over four years net of resale. - Leasing costs £15,840, £4,240 more than cash. - A 9% loan costs about £298.62 a month, £13,933.76 in total net of resale. - Leasing preserves cash, which matters most for new venues. - Check GOV.UK for capital allowances and lease tax treatment. ### Tax and admin (UK) #### VAT basics for hospitality: a worked example Value Added Tax is charged on most sales in UK hospitality at the standard rate of 20%. A VAT-registered business adds VAT to its prices, collects it from customers, deducts the VAT it has paid on its own business purchases, and pays the difference to HMRC, usually quarterly. Registration becomes compulsory once taxable turnover passes a threshold, and businesses below it can register voluntarily. The threshold, the reduced rates that have applied to hospitality at various times, and the detailed rules change, so always check the current position on GOV.UK. This example shows the mechanics using the 20% standard rate. In this example a bistro trades for a quarter. Its sales, including VAT, are £180,000. Step one: output VAT. Because prices include VAT, the VAT within them is one-sixth of the gross figure: £180,000 / 6 = £30,000. The net sales are £150,000. Another way to see it: £150,000 x 20% = £30,000, and £150,000 + £30,000 = £180,000. Dividing a VAT-inclusive figure by 1.2 gives the net amount, and dividing by 6 gives the VAT. Step two: input VAT. The bistro's purchases include some items that carry VAT and some that do not. In this example: drinks stock of £20,000 net, VAT £4,000; cleaning products, packaging, repairs, utilities and other standard-rated costs of £15,000 net, VAT £3,000; a new dishwasher of £5,000 net, VAT £1,000. Most fresh food ingredients bought from suppliers are zero-rated, so they carry no VAT, and wages carry no VAT. Rent may or may not carry VAT depending on whether the landlord has opted to tax the building. Total input VAT is £4,000 + £3,000 + £1,000 = £8,000. Step three: the return. VAT payable is output VAT minus input VAT: £30,000 - £8,000 = £22,000. That is paid to HMRC after the quarter ends, by the deadline shown on GOV.UK. Step four: the impact on pricing. A guest paying £18.00 for a main course pays £3.00 of VAT. The restaurant keeps £15.00, and that is the figure that must cover food, labour, overheads and profit. Step five: records. Making Tax Digital rules require VAT-registered businesses to keep digital records and submit returns through compatible software. Keep every purchase invoice, because without a valid VAT invoice the input VAT generally cannot be reclaimed. Step six: cash discipline. The £22,000 owed was collected from guests during the quarter. Many owners move a sixth of each day's VAT-inclusive takings, less an estimate of reclaimable VAT, into a separate account so the bill is never a surprise. - The UK standard rate of VAT is 20%. - VAT within a VAT-inclusive price is one-sixth of it. - In this example output VAT of £30,000 less input VAT of £8,000 leaves £22,000 to pay. - Most fresh food ingredients bought from suppliers are zero-rated. - Check GOV.UK for the registration threshold and current rules. #### VAT on eat-in vs takeaway food: a worked example VAT on food in the UK depends on what is sold and how. The general principle is that food supplied in the course of catering, which includes food eaten on the premises, is standard-rated, as is hot takeaway food. Much cold takeaway food is zero-rated, though there are many exceptions, such as confectionery, crisps and most drinks other than items like plain milk. The rules are detailed and have been the subject of many disputes, so check HMRC's guidance on GOV.UK before setting up a pricing system. This example shows why the distinction matters to margins using the 20% standard rate. In this example a café sells a cold sandwich for £6.00 whether it is eaten in or taken away. Step one: eaten in. The sale is standard-rated. The net price is £6.00 / 1.2 = £5.00, and VAT is £1.00. Step two: taken away cold. If the sandwich qualifies as zero-rated cold takeaway food, there is no VAT, and the café keeps the full £6.00. Step three: the gap. On the same £6.00 price, the takeaway sale earns £1.00 more for the café. If the sandwich costs £1.80 to make, the eat-in GP is £5.00 - £1.80 = £3.20, or 64%, and the takeaway GP is £6.00 - £1.80 = £4.20, or 70%. Step four: hot food. A toasted sandwich sold to take away is generally standard-rated because it is hot, so it behaves like the eat-in sale: £5.00 net from a £6.00 price. Step five: the till. The till must record each sale correctly, so staff need to ask, or the system must default sensibly, and the business must be able to show HMRC how it decided. If 60% of 400 sandwiches a week are cold takeaway, the zero-rated sales are 240 x £6.00 = £1,440, with no VAT; the 160 eat-in sales are £960, of which £160 is VAT. If the till wrongly recorded everything as zero-rated, the café would under-declare £160 a week, about £8,320 a year, and could face assessments and penalties. If it wrongly recorded everything as standard-rated, it would over-pay 240 x £1.00 = £240 a week. Step six: pricing choices. Some cafés charge different prices for eat-in and takeaway so the net is the same: £6.00 to eat in and £5.00 to take away cold both give £5.00 net. Others keep one price for simplicity and accept the difference. Drinks, meal deals combining hot and cold items, and items like cakes all have their own treatment, so build a VAT map of the menu with your accountant and review it when the menu changes. - Food eaten on the premises is standard-rated for VAT. - Hot takeaway food is generally standard-rated. - Much cold takeaway food is zero-rated, with many exceptions. - In this example a £6.00 cold takeaway sandwich earns £1.00 more than eat-in. - Check HMRC guidance on GOV.UK before setting till rules. #### VAT flat-rate scheme caution: a worked example The VAT Flat Rate Scheme lets eligible small businesses pay HMRC a fixed percentage of their VAT-inclusive turnover instead of calculating output VAT minus input VAT. It can simplify bookkeeping, but for many hospitality businesses it can cost more than standard VAT accounting, so caution is needed. The flat rate percentage depends on the business sector, and there are special rules, including a higher rate for 'limited cost traders' who spend very little on goods. Eligibility limits and the percentages change, so check GOV.UK for current figures. This example uses a hypothetical flat rate of 12.5% purely to show the arithmetic; it is not a real rate for any sector. In this example a small café has VAT-inclusive takings of £120,000 in a year. Its net sales are £120,000 / 1.2 = £100,000, and standard output VAT is £20,000. Standard method. The café's standard-rated purchases, such as drinks, packaging, cleaning products, energy and equipment, total £35,000 net, carrying £7,000 of VAT. Most of its food ingredients are zero-rated. VAT payable under standard accounting is £20,000 - £7,000 = £13,000. Flat rate method, using the hypothetical 12.5%. VAT payable is 0.125 x £120,000 = £15,000. Under the scheme, the café generally cannot reclaim input VAT on everyday purchases, except on certain single capital purchases above a set value. So in this example the flat rate scheme costs £15,000 - £13,000 = £2,000 more a year. Now change the mix. If the café's purchases were mostly zero-rated food, with only £10,000 net of standard-rated costs and £2,000 of input VAT, standard VAT would be £20,000 - £2,000 = £18,000, and the hypothetical flat rate scheme would save £3,000. The answer depends on how much VAT-bearing spend the business has. Step two: the limited cost trader trap. If a business spends very little on goods, as defined in HMRC's rules, it must use a higher limited cost trader percentage. That can make the scheme expensive for businesses whose main costs are labour and rent. Step three: the capital purchase problem. A business planning a refit with a large amount of VAT-bearing spending would, under standard accounting, reclaim that VAT. Under the flat rate scheme, much of it may not be recoverable. In this example a £30,000 refit carries £6,000 of VAT that might be lost. Before joining or leaving the scheme, ask an accountant to compare both methods using a full year of real purchase and sales data, and repeat the comparison whenever the business changes. - The flat rate scheme charges a fixed percentage of VAT-inclusive turnover. - Input VAT on most purchases generally cannot be reclaimed under it. - In this example a hypothetical 12.5% rate costs £2,000 more than standard VAT. - Limited cost traders must use a higher percentage. - Check GOV.UK for eligibility and current percentages. #### Business rates basics: a worked example Business rates are a tax on non-domestic property, collected by local councils. In England and Wales, the Valuation Office Agency assesses a rateable value for each property, broadly an estimate of the annual rent it could command on certain assumptions at a fixed valuation date. The council multiplies the rateable value by a multiplier, sometimes called the poundage, and then applies any reliefs. Scotland and Northern Ireland have their own systems. Multipliers, reliefs and thresholds change, often at each budget and revaluation, and hospitality has at times received special relief, so always check GOV.UK and your council's website for current figures. This example uses a hypothetical multiplier of 0.5, meaning 50p per pound of rateable value, purely to show the arithmetic. In this example a pub has a rateable value of £40,000. Step one: the gross bill. £40,000 x 0.5 = £20,000 a year. Step two: reliefs. Suppose, hypothetically, a sector relief of 40% applies. The relief is 0.40 x £20,000 = £8,000, and the bill becomes £12,000. If the relief were capped at a maximum cash amount per business, the bill would be the larger of the two outcomes; reliefs often have caps and conditions, so check the scheme rules. Step three: what it means for the P&L. If the pub turns over £600,000 a year net of VAT, a £12,000 rates bill is 2.0% of sales. If the relief ended, the full £20,000 would be 3.3%, removing £8,000 from profit. Step four: how pubs are valued. Many licensed premises are assessed on their fair maintainable trade, a measure of the turnover a reasonably efficient operator would achieve, rather than on the rent alone. That means a pub's rateable value can reflect its trading level, and a busy pub may face a higher value at a revaluation. Step five: challenging a valuation. In England, a ratepayer who thinks the rateable value is wrong can use the VOA's check, challenge and appeal process. In this example, if a successful challenge cut the rateable value from £40,000 to £34,000, the gross bill would fall to £34,000 x 0.5 = £17,000, and with 40% relief to £10,200, saving £1,800 a year. Be cautious with firms offering to cut your rates for a fee: understand exactly what they charge and what they will do before signing. Step six: budget for change. Revaluations happen periodically and can move bills sharply up or down. Transitional arrangements sometimes phase changes in. Build next year's rates into the budget as soon as the council's bill arrives, and set up the instalments so they match cash flow. - Business rates are rateable value times a multiplier, less reliefs. - The Valuation Office Agency sets rateable values in England and Wales. - In this example a hypothetical 0.5 multiplier gives a £20,000 gross bill. - Pubs are often valued on fair maintainable trade. - Check GOV.UK and your council for current multipliers and reliefs. #### PAYE and employer National Insurance: a worked example Pay As You Earn is HMRC's system for collecting income tax and National Insurance from employees' pay. The employer calculates the deductions through payroll software, reports them to HMRC in real time on or before each payday, and pays them over each month or quarter. On top of the employee's deductions, the employer pays its own employer National Insurance on earnings above a threshold. Many businesses can claim the Employment Allowance, which reduces their employer National Insurance bill, subject to conditions. Thresholds, rates and the allowance change regularly, so always check GOV.UK. This example uses hypothetical figures to show the principle: a hypothetical employer threshold of £800 a month and a hypothetical employer rate of 15%. In this example a restaurant employs a chef de partie on £2,400 a month gross. Step one: the employee's side. Payroll software calculates income tax using the employee's tax code, and employee National Insurance using the current employee thresholds and rates. These amounts are deducted from the employee's pay and passed to HMRC. They are not an extra cost to the employer; they come out of the £2,400. Step two: the employer's side. Using the hypothetical figures, employer National Insurance is 15% of earnings above £800: (£2,400 - £800) x 0.15 = £1,600 x 0.15 = £240 a month. That is an additional cost on top of the £2,400. Step three: pension. Under automatic enrolment, eligible workers must be enrolled in a workplace pension, and the employer must contribute at least a minimum percentage of qualifying earnings. In this example, suppose the employer contribution comes to £60 a month. Step four: total employment cost. £2,400 + £240 + £60 = £2,700 a month, which is 12.5% more than gross pay. Holiday accrual for staff paid by the hour would add more. Step five: the whole team. If the restaurant's monthly gross payroll is £40,000 across 18 staff, and on average £10,000 of that falls below each person's employer threshold, employer National Insurance using the hypothetical rate is 0.15 x £30,000 = £4,500 a month. If the business can claim an Employment Allowance, that reduces the annual bill by the allowance amount until it is used up. Step six: admin. Run payroll on time, submit Full Payment Submissions on or before payday, pay HMRC by the due date, give employees payslips, and keep records. Tips and service charges distributed through payroll are taxable income and must go through PAYE. Late filing and late payment can lead to penalties and interest, so set reminders for every deadline. - PAYE collects income tax and National Insurance from employees' pay. - Employer National Insurance is an extra cost on top of gross pay. - In this example hypothetical rates add £240 a month on £2,400 of pay. - Automatic enrolment requires employer pension contributions. - Check GOV.UK for current thresholds, rates and the Employment Allowance. #### Corporation tax concept: a worked example Corporation tax is paid by limited companies on their taxable profits. Sole traders and partnerships pay income tax on their profits instead. Taxable profit is not the same as the profit shown in the accounts: some costs are added back because they are not allowed for tax, and some allowances are deducted. The rates, the thresholds at which they apply, and the allowances all change, so check GOV.UK and use an accountant. This example uses a hypothetical corporation tax rate of 20% purely to show the arithmetic. In this example a restaurant company's accounts show a profit before tax of £86,000 for the year. Step one: add back depreciation. The accounts include £35,000 of depreciation on the fit-out and equipment. Depreciation is not allowed as a deduction for tax; instead, the company claims capital allowances under the tax rules. So add back £35,000: £86,000 + £35,000 = £121,000. Step two: add back disallowed expenses. In this example the company spent £2,000 on entertaining suppliers and a potential investor. Business entertainment of non-employees is generally not deductible for tax, so add it back: £123,000. A £500 fine for a late filing is also added back: £123,500. Step three: deduct capital allowances. During the year the company bought £50,000 of kitchen equipment that, in this example, qualifies for allowances that let it deduct the full cost in the year. Deduct £50,000: £123,500 - £50,000 = £73,500. That is taxable profit. Step four: calculate the tax using the hypothetical rate. 0.20 x £73,500 = £14,700. If the company had not bought the equipment, taxable profit would be £123,500, and tax would be £24,700. Timing of capital spending can affect the tax bill, though spending money just to save tax rarely makes sense. Step five: losses. If the company had made a trading loss, it could generally carry that loss forward to reduce future taxable profits, subject to rules. Many new restaurants make losses in their first year, which can shelter some profit later. Step six: payment and filing. Companies file a company tax return and pay corporation tax by deadlines set out on GOV.UK; larger companies pay in instalments. Step seven: extracting profit. After corporation tax, owners may take dividends, which are taxed in their own hands, or salaries, which are deductible for the company but carry PAYE and National Insurance. The right mix depends on personal circumstances and current rates, so take advice each year. - Corporation tax is charged on a company's taxable profit, not accounting profit. - Depreciation is added back and capital allowances deducted. - In this example accounting profit of £86,000 becomes taxable profit of £73,500. - Business entertainment of non-employees is generally not deductible. - Check GOV.UK for current rates, allowances and deadlines.