Food cost percentage = ingredient cost ÷ selling price × 100. For Indian restaurants the healthy band is 28–32% overall, QSR and street formats run 25–30%, fine dining tolerates 30–35%. If your number sits above the band (or worse, you don't know your number), this guide is the fix: the formula, the benchmarks, and the seven moves that reliably cut 3–5 points.
The formula, and the version most restaurants get wrong
The per-dish version is simple: cost every ingredient in the recipe at purchase price, sum per portion, divide by menu price. A plate costing ₹62 sold at ₹220 runs 28.2%.
The version that matters monthly is the actual food cost:
Actual food cost % = (opening stock + purchases − closing stock) ÷ food sales × 100
The difference between the two numbers is the most important metric in your restaurant. Theoretical (recipe-based) food cost is what the month should have cost; actual is what it did. The gap, variance, is wastage, over-portioning, unbilled dishes and theft, priced in rupees. Restaurants that track only the theoretical number feel profitable while leaking; restaurants that track both know exactly where the month went.
Run your per-dish numbers now:
Food-cost % = ingredient cost ÷ selling price × 100. Indian ideal band: 28–32% overall (street/QSR runs lower, fine dining higher). CountStand computes this live per dish from recipes and flags variance against actual stock.
What's the ideal food cost percentage by format?
| Format | Healthy band | Why it differs |
|---|---|---|
| Street / chaat / snacks | 20–28% | High-margin items, low protein share |
| QSR / fast food | 25–30% | Combos and beverages subsidise proteins |
| Café / bakery | 25–32% | Coffee margins offset dairy and imports |
| Casual dining | 28–32% | The classic full-menu band |
| Fine dining | 30–35% | Premium proteins; ambience carries price |
| Cloud kitchen | 25–32% | Must absorb packaging + commissions below the line |
| Bar-led venues | 18–25% on beverage | Liquor margins reshape everything |
Averages hide leaks: a menu can average 30% while its top seller runs 41% because paneer prices moved and nobody re-costed the dish. Track per-dish, not just overall, it changes which problems you see.
Why is your food cost higher than it should be?
Seven causes cover nearly every case; each has a specific fix.
- No standard recipes. If portioning lives in the cook's wrist, every plate is a different cost. Fix: gram-level recipe cards with yields; the kitchen follows weights, not vibes.
- Yield blindness. You buy whole chicken at ₹180/kg but serve deboned at an effective ₹300+/kg after cleaning loss. Costing at purchase price understates everything. Fix: cost at post-prep yield.
- Supplier price creep. Tomatoes doubled, the butter brand changed, nobody re-costed the gravy base. Fix: monthly re-costing from actual purchase prices, automatic if purchases flow through your inventory system.
- Wastage without a log. Spoilage, over-prep, plate returns, invisible unless written down. Fix: a two-tap wastage log with reasons, reviewed weekly. What gets logged gets managed.
- Over-portioning at rush. The extra ladle at speed is 3–8% of cost in busy QSRs. Fix: portion tools (ladles, scoops, weighed proteins) sized to the recipe card.
- Theft. Cancelled-KOT scams, stock walking out the back, free plates for friends. Industry rules of thumb put un-instrumented leakage at 2–5% of revenue. Fix: variance analysis, theoretical vs actual per ingredient, per week, plus KOT audit trails. Suspicion is not a system; measurement is.
- Mispriced menu. Costs moved, prices didn't; or the delivery menu ignores the commission. Fix: quarterly repricing from live costs, the menu price calculator does the arithmetic, including the aggregator gross-up.
Weigh and re-cost your five best-selling dishes this week, at post-prep yields and current purchase prices. In most restaurants at least one flagship dish is quietly running 8–10 points above target, and it's usually the one that "everyone knows" is profitable.
How do you cut food cost without customers noticing?
The moves that work don't touch the guest experience: re-engineer the two or three high-cost dishes (a garnish that costs ₹14 and adds nothing; a protein gram-count 15% above the card), shift the menu's centre of gravity toward high-margin stars in placement and server suggestions, negotiate or dual-source the five ingredients that dominate spend, prep to par levels driven by actual forecast rather than fear of running out, and convert trim into specials (bones to stock, ends to staff meals, priced, not vanished).
What doesn't work: across-the-board portion cuts (guests notice immediately and the reviews say so), quality downgrades on signature dishes, and one-time audits without weekly measurement, the leak returns in six weeks when attention moves on.
The uncomfortable summary
Food cost isn't controlled in the kitchen; it's controlled in the measurement loop: recipes → live costing → variance → weekly review. Owners who run that loop hold 28–32% without heroics. Owners who don't, donate 2–5% of revenue to entropy, every month, invisibly. The loop is exactly what CountStand automates: recipe-linked deduction at sale time, daily variance flags with names and amounts, and the morning WhatsApp brief that tells you which dish moved. The calculator above finds today's number; the system keeps it honest forever.
What is a good food cost percentage for a restaurant in India?
28–32% overall for most full-menu restaurants; 25–30% for QSR; 30–35% for fine dining. More important than the level: knowing your per-dish numbers and your theoretical-vs-actual gap weekly.
How do I calculate food cost percentage monthly?
Actual food cost % = (opening stock + purchases − closing stock) ÷ food sales × 100. Compare it with your recipe-based theoretical percentage, the gap is your controllable leak (wastage, over-portioning, theft), and its size decides what to fix first.
Is 40% food cost too high?
For almost every Indian format, yes, 8–12 points above the healthy band means margin is leaking through recipes, yields, prices or theft. The exceptions are deliberate loss-leaders and premium-protein fine dining; even there, 40% overall (versus on select dishes) is a red flag.
What is food cost variance?
The difference between what sales SAY you should have used (theoretical, from recipes) and what stock counts SHOW you used (actual). Variance per ingredient per week, priced in rupees, is the single most actionable number in restaurant operations, it names the leak.