Why is my actual food cost higher than recipe cost?

The gap (typically 4-10 points) comes from portion drift (chefs over-serve), prep and line waste, unrecorded comps and staff meals, inventory shrinkage/theft, and aggregator refunds that write off prepared food. Recipe cost assumes perfect portions and zero waste; kitchens don't work that way.

What is a normal gap between recipe and actual food cost?

2-4 points is well-controlled. 5-7 points needs investigation. Above 8 points is a structural leak; usually portion drift plus unrecorded comps. On ₹20L monthly revenue at 32% target, each extra point is ₹20,000/month.

How do I find where the leak is?

Run a 30-day theoretical vs actual audit: (opening inventory + purchases − closing inventory) = actual usage. Compare to recipe cost × covers sold = theoretical usage. Reconcile the delta by category: weigh 10 plates per top dish for portion drift, audit waste logs, review comp/discount reports, and check aggregator refund SKUs.

Do aggregator refunds affect food cost percentage?

Yes, indirectly. A refunded order still consumed ingredients; the revenue is reversed but the food was made and often discarded. High-refund outlets (biryani, delivery-only gravies) can add 1-2.5 points to effective food cost if not tracked separately.

How fast can I close an 8-point gap?

Portion SOPs + scales on top 10 dishes recover 3-5 points in 2-3 weeks. Waste logging and FIFO rotation add 1-2 points over 30 days. Comps policy and staff meal tracking add 1-2 points. Theft and persistent over-portioning need management attention beyond tools.

Plate-cost leak: why recipe says 32% but P&L says 40%

The recipe file says butter chicken costs ₹72 on a ₹225 menu price; 32% food cost. The month-end P&L shows 40%. That 8-point gap on ₹22L monthly revenue is ₹1.76L walking out of the kitchen without a line item. It's not one problem; it's five leaks that compound. Here is what each one costs, how to spot it, and the order to fix them.

Recipe cost vs P&L food cost

Recipe (theoretical) food cost = sum of recipe costs × covers sold, at standard portions and current ingredient prices. Actual (P&L) food cost = (opening inventory + purchases − closing inventory) ÷ revenue. The delta is your leak. For how to build accurate recipe costs, see recipe costing for Indian kitchens. For the shrinkage audit framework, see inventory shrinkage in Indian kitchens.

Metric	Example (₹22L monthly revenue)	Target
Recipe food cost %	32%	Your menu engineering target
Actual P&L food cost %	40%	What purchases ÷ sales show
Leak (delta)	8 points = ₹1,76,000/mo	Under 4 points
Revenue at risk if unfixed	₹21L/year	—

The five leaks

Leak	Typical share of gap	₹ impact (8-pt gap)	How to spot it
Portion drift	35-45%	₹62,000-79,000	Weigh 10 plates vs recipe; chefs over-serve 10-18%
Prep + line waste	20-30%	₹35,000-53,000	No waste log; trim and burn consistently high
Unrecorded comps + staff meals	15-25%	₹26,000-44,000	POS comps low but protein purchases high
Theft + receiving gaps	5-15%	₹9,000-26,000	SKU-level cycle count variance
Aggregator refunds	10-20%	₹18,000-35,000	Refund rate >3% on specific dishes

Leak 1; portion drift

The most common leak. Without scales and visual portion guides, line cooks naturally serve 10-18% above recipe standard; extra paneer, larger chicken pieces, heavier gravy ladles. On paneer butter masala at ₹85 recipe cost, 15% drift adds ₹13 per plate. Across 400 covers/week on top sellers, that's ₹2.1L/month.

Fix; digital scales at every station for top 10 dishes. Laminated portion photos. Weekly random plate weigh (10 per dish).
Recovery; 3-5 points of food cost in 2-3 weeks for most outlets.

Leak 2; waste

Prep waste (trim, overproduction), line waste (remakes, dropped plates), and expiry waste (FIFO failures) rarely get logged. Indian kitchens with gravy-led menus over-produce bases; 15-20% of prepared gravy can be discarded if sales miss forecast. A 10kg butter chicken base batch costs ₹1,800-2,400; wasting 2kg daily is ₹12,000-18,000/month invisible.

Fix; daily waste log (3 categories: prep, line, expiry). Batch sizes tied to forecast, not habit. FIFO labels on all prep.
Recovery; 1-2 points over 30 days.

Leak 3; unrecorded comps and staff meals

Owner comps for regulars, manager comps for complaints, staff meals, and 'testing' portions don't always hit the POS. POS comp report shows 1-2%; actual is often 3-5% of covers when you audit protein purchase vs theoretical usage. Staff meals alone on a 16-person team (₹60-80/meal × 2 meals × 26 days) can be ₹50,000-65,000/month unrecorded.

Fix; all comps require manager PIN + reason code. Staff meals logged and capped. Owner comps budgeted monthly.
Recovery; 1-2 points.

Leak 4; theft and receiving gaps

Outright theft is 5-15% of total shrinkage; smaller than portion drift but visible on high-value SKUs (paneer, ghee, chicken, alcohol). Receiving gaps (short deliveries not caught at gate) add to the same bucket. Weekly cycle counts on top 15 SKUs surface this within 30 days.

Fix; weigh at receiving. Lock high-value storage. Weekly spot counts on paneer, chicken, ghee.
Recovery; 0.5-1.5 points depending on severity.

Leak 5; aggregator refunds

A refunded biryani or leaky gravy container still consumed full ingredients. Revenue is reversed; food cost isn't. Outlets with 4-6% refund rates on delivery-heavy menus add 1-2.5 points to effective food cost. Track refund rate by SKU weekly; anything above 3% needs packaging or recipe fix, not more discounting.

Fix; separate packaging for high-refund dishes. Portion consistency (refunds spike when portion looks small). Temperature checks before dispatch.
Recovery; 0.5-1.5 points on delivery-heavy menus.

The 30-day audit

Day 1-2; physical inventory count on all major SKUs. Record opening values.
Day 3-30; log all purchases. Track covers sold by dish from POS.
Day 31; closing count. Actual usage = opening + purchases − closing.
Day 32; theoretical usage = Σ(recipe cost × dish covers). Delta = leak in ₹.
Day 33-35; attribute delta: weigh plates (portion), review waste log (waste), comp report (comps), SKU variance (theft), refund report (aggregator).

Don't re-cost recipes until you've closed the leak. Lowering paneer grams on paper while chefs still over-portion just gives you a false sense of control. Fix execution first; then refresh recipe costs quarterly.

What 'good' looks like

Control level	Recipe vs actual gap	Monthly leak (₹22L revenue)
Excellent	2-3 points	₹44,000-66,000
Acceptable	4-5 points	₹88,000-1,10,000
Investigate	6-8 points	₹1,32,000-1,76,000
Code red	9+ points	₹1,98,000+

Closing an 8-point gap to 4 points on ₹22L revenue recovers ₹88,000/month; more than most marketing campaigns return. Start with portion drift on your top 5 sellers; it's the highest-ROI fix in the kitchen.

Re-cost your menu with true plate economics →