Busy kitchens that also make money.

A cloud kitchen brand runs four kitchens across two cities and does around 900 orders on a good day. The aggregator dashboards show steady growth. A festive discount campaign pushed orders to a record. The team is proud of the ratings. And every Tuesday, when the weekly payouts arrive, the founder notices that the number is smaller than the orders suggested it should be.

The gap is made of things that each look small. Commission. GST on the commission. The share of the discount the brand agreed to fund. Ads bought inside the platform and deducted from the payout. Penalties for late preparation and refunds for missing items. Packaging that went up in price two months ago. None of it shows in the order count, and most of it doesn't show in a monthly P&L either, because the payout arrives as a single net number.

Food businesses run on margins where a few percentage points decide everything. The finance work that matters is taking each order apart by channel and outlet, and making sure the growth being celebrated is growth that leaves money behind.

An illustrative ₹500 order, before GST, with food cost at 32%. The costs are invented but the shape is typical. Dine-in carries rent and service staff that don't appear per order, so its contribution has more fixed cost to cover.

The aggregator order contributes less than a third of the dine-in order and less than half of a direct delivery order. That doesn't make aggregators a mistake. For many brands they're the only realistic way to reach customers at scale, and a cloud kitchen has no dine-in to compare against. But it does mean that a month where aggregator orders grew 20% and direct orders fell can easily be a worse month, and that every rupee of discount on the platform needs to be judged against ₹97, not ₹500.

Most restaurant services in India are taxed at 5% without input tax credit. The exception is restaurants in specified premises, broadly hotels with a room tariff above ₹7,500 a day, which charge 18% and can claim credit. A standalone restaurant, QSR outlet or cloud kitchen generally can't opt for 18% simply to recover its credits.

So the GST you pay is a cost. Commercial rent, kitchen equipment, packaging, aggregator commissions, marketing agencies and software all carry GST, and none of it comes back. A model or budget that shows these costs before GST understates them by up to 18%.

Orders through food delivery aggregators work differently again. Since 1 January 2022 the aggregator, not the restaurant, collects and pays the 5% GST on restaurant services supplied through its platform, under section 9(5) of the CGST Act. The restaurant doesn't charge GST on those orders. It still pays GST on the commission the platform charges it, and that GST is still a cost.

An illustrative new outlet, with every figure invented. Fit-out and equipment cost ₹42 lakh including GST, plus a refundable deposit. Monthly fixed costs are ₹6.8 lakh: rent with GST ₹2.4 lakh, staff ₹3.1 lakh, utilities ₹0.8 lakh, other ₹0.5 lakh. After food, packaging and channel costs, contribution is 38% of net revenue.

The founder's plan used the best outlet's ₹24 lakh from month four. On the brand's own history, the new outlet loses money for most of its first year, needs perhaps ₹15 to ₹20 lakh to fund those losses on top of the ₹42 lakh fit-out, and even at maturity takes around 18 months of profit to earn back its capital.

That can still be a good investment. It just needs to be funded as one, and two openings in the same quarter need twice the cash cushion, arriving at the same time as the festive season's inventory and discounts. The mistake is opening on the best outlet's numbers and finding out in month five.

Most menus are reviewed by popularity and food cost percentage. Both matter, and neither tells you which dishes make the business money. A biryani with 38% food cost that sells 300 times a week can contribute far more rupees than a salad at 22% that sells 40 times. Rupees of contribution per dish, multiplied by volume, is the ranking that decides where menu attention should go.

Channel changes the ranking too. A dish that travels badly generates refunds and poor ratings on delivery, which costs more than its food cost suggests. A dish that needs expensive packaging can be profitable at the table and marginal in a box. Looking at contribution by dish and by channel together often produces a shorter delivery menu, a few price changes on the highest-volume items, and a quiet retirement for dishes that are loved by the chef and ordered by nobody.

Food businesses have an unusual cash shape. Dine-in and direct orders are paid immediately. Aggregator revenue arrives in weekly settlements. Suppliers of fresh produce often need paying within days, while larger suppliers may give a few weeks. Rent is due monthly, sometimes quarterly in advance, and a new outlet ties up a deposit that won't come back for years.

That usually works well day to day, which is exactly why problems arrive suddenly. A festive season needs inventory and discount funding before the extra revenue comes in. Two outlets opening together need fit-out money and months of losses at once. A weekly cash forecast, even a simple one, turns those from surprises into plans.

Aggregator payouts arrive net of many deductions, and errors do happen. A weekly reconciliation compares the orders in your own records against the platform's statement and checks each deduction line.

• ·Order count and gross value, against your point of sale.
• ·Commission rate actually applied, against your agreement.
• ·Discounts charged to you, against the campaigns you agreed to fund.
• ·Ad spend deducted, against what was approved.
• ·Cancellations, refunds and penalties, each with a reason.
• ·GST on fees, matched to the tax invoices the platform issues.

It's tedious work, and it's exactly the kind of tedious work that finds a few thousand rupees a week. Across several outlets and a year, that's real money.

The work adjusts to how food businesses earn. A unit economics review by outlet, channel and menu category. A payout reconciliation process your team can run weekly. A cash forecast that knows about weekly settlements, rent dates and festive inventory. An expansion model that uses your own opening history rather than the best outlet. A fundraising model investors in consumer food businesses will recognise. GST filings stay with your CA.