Restaurant food cost in Pakistan: working out what a plate actually costs

What is food cost percentage?

Food cost percentage is what the ingredients of a dish cost you, divided by what you sell that dish for, multiplied by 100. If a chicken karahi costs Rs 420 in chicken, ghee, tomatoes, onions and spice, and you sell it for Rs 1,200, the food cost is 420 ÷ 1,200 × 100 = 35%. The remaining 65% — Rs 780 — is your gross profit on that plate, and it has to cover the rent, the gas and the wages before any of it is yours. It is a per-dish number, not a shop-wide one, and that distinction is the whole point.

You can do this on paper tonight. Write down every ingredient in one plate and how much of each, price each at what you last paid per kilo, and add them up. Do it for your five best sellers and you'll learn more in an hour than a month of sales totals told you.

What is a good food cost percentage?

The band most kitchens work to is 28–35% — that is the range NexusTill's own costing uses as healthy. Below it you may be under-portioning or over-pricing; well above it the dish is eating the margin meant to cover rent and wages. But it's a starting point, not a verdict: a tandoor roti and a mutton dish will never sit at the same percentage. What matters more than hitting a number is knowing which number each dish is at, and whether it moved.

Why does a single average food cost hide the problem?

Because an average is the one number that can look healthy while the business bleeds. If your kitchen averages 32%, that figure is made of dishes at 22% and dishes at 48%, and the good ones are quietly paying for the bad ones. Worse, the 48% dish is often a popular one — so the more you sell, the more you lose per plate. A restaurant-wide food cost tells you the kitchen is roughly fine; it never tells you which dish to fix.

Where does the money actually go in a Pakistani kitchen?

In four ordinary places, none of them dramatic: portioning drift, waste and spoilage, staff meals, and theft — plus price rises the menu never caught up with, covered below. Most kitchens have all of them running at once, and three are nobody's fault.

The heavy hand

The biggest one and the hardest to see. Your recipe says 40g of ghee; your cook, fifteen years in and cooking by feel, uses 55g because that's how his hand moves. That's not laziness — it's an experienced cook making the dish taste right. Fifteen grams does not sound like anything. Multiply it by sixty karahis and it is nearly a kilo of ghee a day that was never in the costing — and the same drift lives in the meat portion.

Waste, spoilage and the trim

Vegetables that turned before they were used. A degh of daal that didn't sell. Chicken trimmed harder than the recipe assumed, so a 250g portion started as 320g of purchased weight. None of it appears in your recipe; all of it was paid for.

Staff meals

The kitchen eats, and it should. But if eight people eat twice a day from the stock you sell from, that's a real ingredient cost with no sale attached. It doesn't have to be reduced — it has to be counted, so you know whether your 38% is a portioning problem or the staff lunch you never priced.

Theft

It exists, it's usually small and steady rather than dramatic, and it's the last thing to look for. Most owners suspect theft when the real answer is a heavy hand and a six-month-old menu price. Rule out the others before you accuse anybody.

Why does food cost drift after you've set the menu?

Because most restaurants cost a dish exactly once — the week they set the menu — and never again, while chicken, ghee, onions and tomatoes carry on moving underneath them. A dish costed at 30% in January can be at 45% by June, and nothing in your day tells you: the dish still sells, the kitchen still looks busy. You just make half as much on every plate, and the first sign is a month where the sales were fine and the bank balance wasn't.

How do I find which dish is losing money?

Cost every dish, sort by food-cost percentage with the highest at the top, and work on the worst five. The fix is nearly always one of three things, and you'll know which within a minute: the main ingredient's price moved, the portion in the kitchen isn't the portion in the recipe, or the menu price was set by looking at the restaurant down the road instead of your own costs.

1. 1Write a real recipe for each dish — the quantities the kitchen actually uses, not the ones you'd like it to. A recipe that lies to you is worse than no recipe.
2. 2Price each ingredient at what you paid most recently, then work out the plate cost and the percentage against your current menu price.
3. 3Sort by percentage, highest first, and ignore the average completely.
4. 4For each of the worst five, pick one: raise the price, change the portion, or accept it deliberately because it brings people in. Re-run the list whenever buying prices move — in Pakistan, monthly.

What does a nightly closing count tell you that a recipe can't?

A recipe tells you what should have been used. A closing count tells you what actually went. The gap between those two numbers is the part of your food cost no recipe can explain — the heavy hand, the waste, the staff meals, the theft — and you cannot see it any other way. If the recipes say 18kg of chicken and the shelf says 22kg left the fridge, that 4kg is the real conversation, and you can only have it if somebody counts.

It doesn't have to be everything — count the expensive, fast-moving things at the same time every night. Some nights the gap is noise; a gap always in the same direction is not.

Who does not need any of this?

A small dhaba with six dishes and one cook who buys fresh every morning genuinely does not need a costing system. If you buy today what you'll cook today and hold the whole menu in your head, you already have a tight feedback loop — this morning's market spend against tonight's drawer. It starts to matter when the menu outgrows your head, when somebody other than you is portioning, or when a second kitchen means you can't be in both.

What can software actually do about food cost — and what can't it?

Software can measure it. It cannot stop a heavy hand. A system can hold your recipes, re-price them as purchase costs change, deduct the chicken and ghee the moment a dish is sold, and price the gap at the nightly count. What closes that gap is a conversation with the cook and a scale in the kitchen. The software makes the problem visible and puts a rupee figure on it; the fixing is management.

Concretely, in NexusTill: you build a recipe once and it gives you the plate's unit cost, selling price, gross profit and food-cost percentage — the same cost ÷ price × 100, except it re-works itself as ingredient costs move. Selling a dish pulls its raw materials off the shelf, and wastage is recorded with a reason. The nightly count prices the leak in rupees: where a cook-to-order sale took out more than was booked in, it values the shortfall at cost and shows the gap against what was counted.

Restaurant mode is on Growth at Rs 6,999 a month — recipes, food-cost %, cook-to-order stock, raw materials, wastage and the nightly leak count, plus tables, the kitchen display and modifiers. It is not on Starter. Kitchen inventory per branch and stock transfers sit on Scale at Rs 11,999.

Before paying for any of it, do the paper version. Cost your five best sellers by hand this week. If the numbers land where you expected, you have a well-run kitchen. If two are above 40%, you've found the money — and the only question left is whether you want to redo that arithmetic every month by hand. The trial is 7 days, needs no card, and doesn't convert into a paid plan on its own.

Your sales total tells you the restaurant was busy. Only the food cost tells you whether being busy was worth it.

Written by The NexusTill team.