A hotel owner in Mogappair told us he runs “about 30%”. Asked how he knew, he said his supplier bills came to roughly ₹4 lakh a month and he did roughly ₹13 lakh in sales. That is a reasonable guess. It is not a food cost percentage, and the gap between the two is where the money hides.
Here is how to work it out properly, what the number should be, and — more usefully — what to actually do once you have it.
The formula, and the part everyone skips
Food cost percentage is the share of your food revenue that goes on the ingredients you sold.
Food cost % = Cost of goods sold ÷ Food sales × 100
The part that gets skipped is that cost of goods sold is not the same as purchases. What you bought in a month and what you cooked and sold in a month are different numbers, because inventory sits in between.
COGS = Opening inventory + Purchases − Closing inventory
Worked through, for one month:
| Line | Amount |
|---|---|
| Opening inventory (1st of the month) | ₹1,80,000 |
| Purchases during the month | ₹4,20,000 |
| Closing inventory (last day) | ₹1,65,000 |
| Cost of goods sold | ₹4,35,000 |
| Food sales | ₹14,50,000 |
| Food cost percentage | 30.0% |
Skip the stock count and you are measuring your purchasing rhythm, not your kitchen. A month where you stocked up before Deepavali looks terrible. The month after looks brilliant. Neither is true.
What the number should be
There is no single right answer — it depends on format, cuisine, and where you compete. The working ranges the trade uses:
| Format | Typical food cost |
|---|---|
| Tea stall, cart, small kiosk | 20–25% |
| QSR and takeaway | 25–30% |
| Casual dining | 28–35% |
| Fine dining | 30–40% |
| Vegetarian mess / meals | 30–38% |
| Desserts, sides, beverages | 20–25% |
Two things matter more than hitting a benchmark.
First, the direction of travel. A restaurant that moved from 34% to 31% over a quarter is in better shape than one that has sat at 30% for two years without knowing why.
Second, prime cost. Food cost alone tells you half the story. Add labour to it and you get prime cost — the two largest controllable costs in the business. Above 65% of revenue, rent and utilities will eat whatever is left. Well-run kitchens keep prime cost in the 55–60% band.
Why 2026 is harder than last year
The input side has moved sharply. India’s Wholesale Price Index rose 9.87% year-on-year in June 2026, up from 9.68% in May, with food articles named by the Ministry of Commerce and Industry among the major drivers. The WPI Food Index specifically ran 6.14% year-on-year in June, against 4.49% a month earlier — the rate of food-input inflation more than doubled in a single month (PIB, provisional WPI estimates for June 2026).
What that means on the floor: a dish costed in January against January rates is not costed against August rates. If your menu prices have not moved and your recipe costs have, your food cost percentage has already drifted up — you just have not measured it yet.
This is the argument for weekly measurement. A monthly average hides a three-week slide inside a single number, and by the time you see it you have served twelve thousand covers at the wrong price.
The lever most kitchens never pull
Once you have the number, the usual advice is to attack it three ways: reprice the menu, tighten portions, cut wastage. All valid, all slow, and all visible to your customer.
There is a fourth that is neither slow nor visible, and most kitchens have not touched it: buy the same volume in fewer, larger drops.
Wholesale pricing is tiered by quantity. Buy 3 kg of toor dhall and you pay one rate; buy 50 kg and you pay another — for identical goods out of the identical sack. A kitchen that sends someone out for 5 kg every second day pays the small-lot rate roughly fifteen times a month. Its neighbour, buying the same 75 kg in two drops, does not.
Here is what that difference is actually worth. These are our published Chennai wholesale rates on 5 August 2026, for a month of dry staples in a mid-sized South Indian kitchen:
| Item | Monthly qty | Small-lot rate | Bulk-tier rate | Small-lot cost | Bulk cost |
|---|---|---|---|---|---|
| Idly rice | 300 kg | ₹45.00 | ₹42.69 (26 kg+) | ₹13,500 | ₹12,807 |
| Boiled rice | 200 kg | ₹68.00 | ₹64.62 (26 kg+) | ₹13,600 | ₹12,924 |
| Toor dhall | 60 kg | ₹135.00 | ₹127.50 (50 kg+) | ₹8,100 | ₹7,650 |
| Orid dhall | 50 kg | ₹140.00 | ₹133.00 (50 kg+) | ₹7,000 | ₹6,650 |
| Gram dhall | 30 kg | ₹90.00 | ₹87.00 (10 kg+) | ₹2,700 | ₹2,610 |
| Moong dhall | 20 kg | ₹110.00 | ₹107.00 (10 kg+) | ₹2,200 | ₹2,140 |
| Sunflower oil, 15 L tin | 8 tins | ₹2,710 | ₹2,700 (6 tins+) | ₹21,680 | ₹21,600 |
| Tamarind | 30 kg | ₹150.00 | ₹140.00 (30 kg+) | ₹4,500 | ₹4,200 |
| Jaggery | 30 kg | ₹72.00 | ₹67.50 (30 kg+) | ₹2,160 | ₹2,025 |
| Salt, 1 kg | 40 packets | ₹11.00 | ₹9.20 (25+) | ₹440 | ₹368 |
| Total | ₹75,880 | ₹72,974 |
The difference is ₹2,906 a month — 3.8% off the staples bill — for buying the same goods in a different rhythm. Over a year, ₹34,872. Nothing changed on the menu, nothing changed on the plate, and no customer noticed.
Nobody negotiates their way to a 4% discount on dhall. You schedule your way there.
The honest catch
Two reasons kitchens don’t do this, and both are real.
Storage. Fifty kilos of dhall needs somewhere dry, sealed, and off the floor to sit. Without it the saving evaporates into weevils. Dry staples — rice, dhall, oil, tamarind, jaggery, salt, packing material — are where this works. It does not apply to vegetables, dairy, or anything with a short life, and chasing a tier on perishables costs more in wastage than it saves on rate.
Cash. Consolidating means paying for three weeks of staples on one day instead of spreading it across fifteen. For a kitchen running tight on working capital that is a genuine constraint, not an excuse — which is exactly why credit terms on day-to-day basics matter more than a headline rate. A supplier who gives you terms lets you buy at the bulk tier without breaking the week’s cash flow. Talk to the branch nearest you about what that looks like at your volume.
What to do this week
Four moves, in order
Sunday night or Monday morning — same time, same person, same sheet. You cannot calculate COGS without it, and the discipline matters more than the precision.
Opening stock, plus purchases, minus closing stock, over food sales. One real number beats a year of estimates.
Not last year’s. Ten dishes usually cover most of your covers. Any dish whose cost has moved more than 10% since you last priced it needs a decision: reprice, re-portion, or accept it as a loss leader on purpose.
List everything shelf-stable you buy more than twice a week. Work out the monthly volume, check where the price tiers break, and move to one or two drops a month. Leave perishables exactly as they are.
Frequently asked questions
Frequently asked questions
What is a good food cost percentage for a restaurant in India?
How do I calculate food cost percentage?
How often should I calculate food cost percentage?
Does buying in bulk actually reduce food cost?
Why has my food cost gone up when nothing in my kitchen changed?
Where can I check wholesale prices before ordering?
Rates in this article are Mikro Platforms’ published Chennai wholesale rates on 5 August 2026 and move daily. Check today’s list before you budget against them, and browse the full catalog for anything not on the daily sheet.