You do not need to tear up your supplier list to cut your restaurant food cost. You need a different set of daily habits around how you buy, store, prep and serve.
The band most Indian kitchens work to is 28% to 35% of revenue. Food is usually the largest cost on the P&L that an owner can move week to week.
Restaurants we work with tend to take 3 to 5 points off that percentage without switching a single supplier.
Key Takeaways
- Work out three numbers first: actual, ideal and the percentage.
- Most of the loss happens before service, in over-ordering and storage.
- Compare rates daily on the expensive and seasonal items; consolidate dry staples with one supplier.
- Count stock daily, rotate it FIFO or FEFO, write recipes in grams.
- Sort the menu by margin and by sales, then stop serving what comes back.
What are the 6 Proven Ways to Reduce Restaurant Food Cost?
Six ways, listed here in the order they pay back fastest. Each one is worked through further down the post.
1. Compare Vendor Rates Before the Morning Order
Check the same item across two or three sources before you buy. No supplier is cheapest on everything, and on a single 10 kg line the gap can reach ₹250. Do this daily on the expensive and the seasonal: paneer and other dairy, oil, proteins, vegetables.
2. Consolidate Your Staples with One Vendor
Volume is the only thing that gives you a say with a supplier, and splitting it across four of them leaves you none. Rice, dry spices and pulses barely move week to week, so there is nothing to lose by putting them all in one place.
3. Count Stock Daily Instead of Weekly
Count what is on the shelf each morning and you order against a number instead of a feeling. Rotate what you find, FIFO or FEFO, so it stays usable. Counting weekly means you can find a problem up to six days after it started.
4. Write Recipes with Gram Weights
A dish with no written card costs whatever the cook on shift decides that day. Gram weights are what make a food cost number repeatable.
5. Rework the Dishes That Sell Well but Earn Little
Plot every dish by how often it sells and what margin it carries. The popular, thin-margin ones are where the menu-level money sits.
6. Stop Serving What Comes Back Uneaten
You buy the papad, the raita and the extra rice whether or not they come back untouched. Make them request-only and trim the portion.
How Do You Calculate Restaurant Food Cost?
Three numbers, and each answers a different question. Two of them need a stock count; the third comes off your recipe cards and your sales report. Work them out before you change anything, or you will not know whether a change helped.
Use the same period every time. Food cost percentage is one of the weekly kitchen metrics, not a monthly one, because a month is too long to wait before correcting anything. Before the formulas, here is what actually gets counted.
What Counts as Opening Stock?
Everything edible on the premises when the period starts, valued at what you paid for it. Walk the store room, the walk-in, the dry shelf and the bar, and count the part-used containers too.
Five lines from one Monday morning count, to show the shape of it:
| Item | Counted | Rate | Value |
|---|---|---|---|
| Rice | 18 kg | ₹62 per kg | ₹1,116 |
| Paneer | 9 kg | ₹320 per kg | ₹2,880 |
| Refined oil | 12 L | ₹145 per L | ₹1,740 |
| Onion | 40 kg | ₹34 per kg | ₹1,360 |
| Part-used containers | mixed | at cost | ₹2,140 |
| These five lines | ₹9,236 |
A real sheet runs to forty or eighty lines depending on the menu, and the total of all of them is your opening stock.
Leave out packaging, cleaning supplies, disposables and gas. Those are real costs, but they are not food cost, and mixing them in makes the percentage meaningless.
What Counts as Purchases?
Every food delivery that arrived during the period, at invoice value, including the cash purchases somebody made on the way to work. Those small market runs are the ones most often missed, and they are the reason a count comes out wrong.
Count a delivery on the day it arrived, not the day you paid for it.
What Counts as Closing Stock?
The same walk as opening stock, on the last day of the period, valued the same way. Whatever you decide to include, include it on both counts, because the calculation below only works if the two ends match.
How Do You Work Out Actual Food Cost and the Percentage?
Actual food cost is what the kitchen really consumed in a period:
opening stock + purchases – closing stock
Food cost percentage turns that into a share of what you sold:
(actual food cost / revenue) x 100
Revenue here means net sales, before the GST you collected. Run it on the gross figure and the answer comes out flattering by more than a point.
What Is Ideal Food Cost?
Ideal food cost is what the recipes say you should have consumed. Multiply each dish’s recipe cost by the number of that dish sold, then add them up. It is the number that tells you whether the actual figure is good or bad.
What Do the Three Numbers Look Like on One Week?
The gap between ideal and actual is the part you can act on. This is the week ending Sunday 13 September 2026 in a busy 80 seat restaurant in Kothrud, Pune (an example).
At 33.1% this kitchen sits inside the normal band, so a monthly report would show nothing wrong. The ideal figure is what exposes the problem, because the recipes say 30%.
That leaves ₹24,000 unexplained in a single week. Check the recipe costings first. Where prices have risen since you wrote them, the ideal is too low and the gap widens on its own. Once the costings are current, the six ways above are where to look.
Where Can Restaurants Usually Save on Food Cost?
Before the food reaches a plate, mostly. Over-ordering and bad storage typically account for more of the loss than anything that happens in service, and the rate you paid is usually the smallest part of it.
That order tells you where the money goes, not where to begin. Rate comparison is the smallest slice on the chart and still the first way on the list, because you can start it on Monday morning without changing anything else.
Why Does Restaurant Food Cost Creep Up When Menu Prices Have Not Changed?
Because your costs move and your menu does not. Ingredient rates change every season. Recipe cards get written once and rarely revisited, so the same dish quietly costs more to make while the price on the menu stays put.
Tomato rates in India have run from around ₹20 per kg to north of ₹100 within a single season. Down To Earth put the last spike down to adverse weather and to growers shifting acreage to crops that pay better. Cooking oil, dairy and pulses move too, though less violently.
Here is how it goes in a bad season. A restaurant owner in Bopal, Ahmedabad costs a paneer tikka at 31% food cost in January 2026, when paneer is at ₹320 per kg. Paneer is roughly half that plate’s cost.
By late March paneer climbs to ₹380 per kg, and the menu price stays at ₹299. That dish now runs at about 34%, and nobody recalculated, because the recipe card still shows the January number.
Multiply that drift across the twenty-five or thirty dishes on the menu that use paneer, and the monthly report makes sense. Owners tell us the same story most seasons.
Raising menu prices every few weeks is not practical. Controlling cost on the buying and kitchen side is.
How Much Can You Save by Comparing Vendor Rates?
Four to five percent of the grocery bill, and you do not have to drop a single supplier to get it. At 4% on a kitchen spending ₹2,80,000 a month on groceries, that is ₹11,200 a month, or ₹1,34,400 a year.
Plenty of kitchens call the same supplier every morning, and that call has been going to the same person for years. The rates get accepted without a second check, because the relationship is comfortable and switching feels like a hassle.
The point is simply to look before you order.
This is the comparison the Purchase Manager add-on inside Petpooja puts on one screen, for a Tuesday morning order in a Kharadi, Pune kitchen.
| Item, standard 10 unit pack | Your regular vendor | Hyperpure | DMart | Cheapest |
|---|---|---|---|---|
| Paneer | ₹3,550 | ₹3,420 | ₹3,300 | DMart |
| Refined oil | ₹1,480 | ₹1,510 | ₹1,440 | DMart |
| Onion | ₹340 | ₹390 | ₹355 | Your vendor |
Rates are an example. Compare your own.
Take all three from your regular vendor and it is ₹5,370. Take the cheapest of each and it is ₹5,080. On three lines alone the gap is ₹290, and a real order runs to forty of them.
Your regular vendor still wins on onions, which is exactly why this is a comparison and not a switch. Purchase Manager flags the cheapest option per item across Hyperpure, DMart and your own vendors, and syncs whatever you buy straight into inventory.
Doing that by hand, line by line, down a forty-line order list is why the habit usually lasts about a week.
Petpooja is India’s biggest and most price effective restaurant POS, behind the success of 1,00,000+ outlets. It has run restaurant billing for 14+ years, and the outlets on it push 60 lakh bills through a day at 0% processing errors.
How Do You Get a Better Rate from a Vendor You Already Use?
Three things make vendors in India willing to renegotiate, and a threat to walk is not one of them.
Volume consolidation. This is way two on the list. A restaurant in Whitefield, Bengaluru splitting its rice, spice and pulse orders across four suppliers has no bargaining room with any of them.
Put the dry staples with one vendor and keep the others as backup. The one holding the volume will quote you better, while the expensive and seasonal items still go to whoever is cheapest that morning.
Faster payment cycles. The standard in restaurant supply is 15 to 30 day credit. Vendors paid within 7 days will often drop the rate, because getting paid sooner eases their own cash flow. If your cash flow can absorb the faster outflow, this is one of the simplest levers available.
Commitment windows. Telling a supplier you will order from them for the next six months gives them revenue visibility. In return, many will lock rates even when seasonal prices move. This works best on dry staples like rice, spices and pulses, where the vendor’s own cost does not swing the way fresh produce does.
How Does Daily Inventory Tracking Lower Restaurant Food Cost?
It replaces guesswork with a number. The Centre for Responsible Business study on HoReCa food waste cites a World Resources Institute case. One operator there reported 5:1 returns from cutting overproduction, out-of-date stock and customer leftovers.
Overproduction starts with guesswork. The manager eyeballs the paneer tray, decides it looks low, and orders 5 kg more than necessary. That stock then sits, edges past its usable window, and goes in the bin by Friday.
Switching from weekly counts to a 15-minute daily check before the morning order changes the maths. Your POS counts every kitchen order ticket (KOT) from yesterday, and the recipes turn those plates into the weight consumed.
The physical count then shows what is actually left. Set a par level for each item, the minimum you want on the shelf, then order enough to bring the count back up to par. It beats a number someone felt was about right.
Our restaurant inventory management guide walks through the count itself, step by step.
Over a month, the same counts give you your inventory turnover ratio, which shows how many times the stock is used up and bought again. A falling ratio means cash is sitting on the shelf as stock instead of moving through the kitchen.
Paper works until the count stops fitting on one sheet, or a second outlet arrives. After that the count, the recipes and the purchase records need to sit in one place, which is what Petpooja’s restaurant inventory management software is for.
Which Stock Rotation Method Should You Use: FIFO, LIFO or FEFO?
FIFO for most of your store room, FEFO for anything with a printed date. LIFO is not permitted for inventory valuation in India, and using the newest stock first is how perishables end up in the bin.
| Method | What it means | Where it belongs |
|---|---|---|
| FIFO, first in first out | The oldest delivery gets used first | Dry goods, packaged items, most of the store room |
| FEFO, first expired first out | The earliest expiry gets used first, whatever the delivery date | Dairy, meat, fish, anything date-stamped |
| LIFO, last in first out | The newest stock gets used first | Not a fit for food stock |
FEFO matters more than people expect, because a delivery date and an expiry date are not the same thing. A carton of cream delivered on Thursday can expire before one delivered on Tuesday, and a kitchen running strict FIFO will reach past the one that needs using.
Either way the mechanics are the same. Label every container with its delivery date and, where there is one, its expiry. Keep the oldest at the front of the shelf, and have the prep team always take from the front.
Where Does FIFO Break Down?
Under pressure. In the Navratri rush or the December wedding season, new bags get dumped on top of older stock in the walk-in, and Tuesday’s paneer spoils behind Thursday’s delivery.
The fix is not training alone. It is shelf organisation: a dedicated zone for each delivery day, with the oldest zone nearest the door and the newest furthest in.
Do Standardised Recipes Really Make That Much Difference?
Yes, and the reason is not what most people think. It is not about making every dish taste identical. It is about making a given dish cost the same to produce, whoever is cooking it.
Without a written recipe card listing gram-level weights, portion sizes wander. The head chef’s dal makhani uses 30g of butter per portion. The night shift cook, who joined last month, uses roughly 45g, because nobody told him otherwise.
The kitchen sends out about 80 plates of it a day. With no card on the wall, a good share of them drift towards the heavier end.
Cost a heavy plate against the card and the gap shows up line by line.
₹8 more butter, ₹4 more cream, ₹12 a plate. If 40 of the 80 daily plates go out that way, it is ₹480 a day, or ₹14,400 a month on two ingredients in one dish.
The rates have not moved in that costing, so the whole ₹12 is portioning. Had the rates moved too, the card would have shown you that instead. Without a card you are just staring at a monthly total that looks too high.
Which Dishes Do the Most Damage to Your Restaurant Food Cost?
The ones that sell well and earn little. Menu engineering calls them plough horses, and finding yours costs nothing but an afternoon with your sales report.
Every dish lands in one of four boxes, set by its sales volume and its margin.
| Category | Sells often? | Good margin? | Action |
|---|---|---|---|
| Stars | Yes | Yes | Keep these front and centre on the menu |
| Plough horses | Yes | No | Rework the recipe or nudge the price up |
| Puzzles | No | Yes | Better placement, staff upselling |
| Dogs | No | No | Remove or replace |
Sorting 30 dishes by hand is slow. There are also AI menu engineering tools that build the grid off your sales data.
Either way, plough horses are the box to work on first. These are the dishes your regulars order every visit, but the margin on each plate is thin. A Hyderabadi biryani selling 40 portions a day at 42% food cost is the usual culprit.
Three changes together can pull that dish back towards the 30% to 35% zone. Adjust the protein-to-rice ratio, swap an expensive garnish, and trim the portion weight by 8% to 10%. No one of them does it alone. Done well, the flavour holds and most regulars will not spot the change.
What About Plate Waste? Is That Really Big Enough to Address?
Plate waste is a smaller slice than ordering or storage, and it matters most where the format gives food away by default. A thali restaurant in Maninagar, Ahmedabad serves papad, raita, pickle and salad with every meal.
The waste piles up when a third of customers leave the raita untouched. That is wasted curd, wasted labour to prep it, and wasted cold storage for the morning batch. Our guide to restaurant waste management covers the bin side of the same problem.
Two adjustments work. First, switch default accompaniments to request-only for items with high return rates. If raita comes back uneaten on a third of covers, stop including it by default. On a fixed thali that is a positioning decision as much as a cost one, so test it on one service first.
The curd in that katori costs about ₹12. You pay it on every cover whether they eat it or not, and one bowl in three goes in the bin, so ₹4 a cover is thrown away. At 150 covers a day that is ₹600 daily, or ₹18,000 a month.
Second, audit rice portions. Steamed rice served alongside a curry is among the most over-served items in Indian restaurants.
If half-finished plates keep coming back, the default serving is too generous. Cut it by 15% and offer a free refill on request. In practice, few diners ask.
Conclusion
Bringing restaurant food cost down does not mean burning supplier relationships. It means working out all three numbers first, then building the six ways above into the week.
Shop the dairy, oil and produce before the morning order, and put the dry staples with one vendor. Count stock daily instead of weekly, and write recipes with gram weights. Work the menu grid, then stop serving the accompaniments and the rice that come back uneaten.
The tightest-run kitchens we work with are not the ones with the cheapest suppliers. They are the ones where the owner knows by Tuesday what the current week’s food cost percentage is tracking at.
Our restaurant P&L template puts that weekly number against rent and labour. The food cost calculator does those three sums for you.
Frequently Asked Questions
The commonly cited range is 28% to 35% of revenue. Quick service and cloud kitchens sit closer to 28%, and fine dining is the exception that runs above the band, because of premium ingredients and the trim loss on them. For everyone else, anything consistently above 35% points to a procurement, portioning or spoilage problem.
Yes. Every method here works on the cost side. Fixing procurement and portioning, based on what restaurants tell us, tends to be worth 2 to 3 percentage points on its own. The rest comes from re-costing recipes and working the menu.
Weekly. Monthly is too late. If tomato prices spiked on the 8th and you find out on the 30th, you have already absorbed three weeks of higher cost. Your POS sales report gives you the sales half of the sum.
Actual is what the store room says you consumed. Ideal is what the recipes say you should have consumed for the dishes you sold. Actual is almost always higher. Before blaming waste, over-portioning or stock going missing, check that the recipe costings are current. A stale ideal widens the gap on its own. Our guide to kitchen inventory cost covers the stock side of the calculation.
One full count period, usually a week. Daily counting and rate comparison show up in the very next order cycle. Recipe cards and menu changes take a month or two, because you are waiting for the dishes to sell.
No, and plenty of single-outlet kitchens run this on a notebook. Software earns its place once you pass roughly fifty ingredients or a second outlet, when the count, the recipes and the purchase records stop fitting in one person’s head.
FIFO uses the oldest delivery first. FEFO uses the earliest expiry date first, whichever day it arrived. They usually point at the same container, but not always, which is why anything with a printed date should be rotated by that date. Run FIFO on dry goods and FEFO on anything date-stamped.
Track what comes back to the kitchen for one full week. Write down which dishes return with food left, and how much. If the same item shows up repeatedly, the portion is too large. On a side like rice, trim the default by 15% and offer a refill. On a main dish, 8% to 10% is about the most you can take before regulars notice.
