Four people, one kitchen, sixty dabbas a day — a worked example

Month by month

The first month is not a business, it is a rehearsal

The blueprint puts first revenue in week three, and that is roughly where it lands — but the first month is 22 subscribers and about ₹42,000, which after provisions, gas and packaging leaves nothing to divide. That is the expected shape, not a failure. A tiffin service is a renewal business: the money arrives in month three because that is when the first cohort renews and the second cohort signs up on top of it, and the two stack. Almost everything a crew does in month one is unpaid work whose return arrives sixty days later — settling the menu, finding out that Thursday’s dish is the one people talk about, learning that the 12:40 delivery is late and the 12:20 one is not. A crew that expects a salary in month one quits in month two, three weeks before the business would have started paying them. The only real protection against that is saying the number out loud in week one: nobody takes anything home until month three, and here is what each of us is living on until then.

Where the subscribers actually come from

Not from Instagram. In this example every one of the first 22 comes from a single office floor, signed up by the person holding the Subscriptions role standing near a lift on three consecutive Mondays with a sample box and a printed menu. Concentration is the whole trick early: sixty dabbas spread across nine buildings is an impossible delivery route, and sixty dabbas in two buildings is a forty-minute one. The route economics decide the sales strategy, which is the opposite of how most people plan it — they pick a price, then discover the delivery cost, then wonder where the margin went. Working the other way round, the question stops being "how do we get more customers" and becomes "who else works in the buildings we already drive to", which is a much easier question and has better answers. The second route in month five is chosen the same way: a hostel cluster where 30 residents eat within one corridor of each other.

What the split looks like when it is real money

The blueprint suggests 35 / 25 / 25 / 15 — Kitchen Lead, Route Lead, Subscriptions, Operations. On month six’s ₹1,94,000, after about ₹1,20,000 of provisions, gas, packaging and fuel, that divides roughly ₹25,900 / ₹18,500 / ₹18,500 / ₹11,100. The Kitchen Lead is working 42 hours a week for it and the Operations role about 12, which is why the shares are not equal. What makes those numbers hold is that they were agreed in week one, when they were percentages of nothing and nobody had yet decided their own contribution was the important one. Attempt the same conversation in month six, with ₹1,94,000 on the table and a kitchen lead who has been awake at five every morning for half a year, and it is no longer a discussion about fairness — it is a negotiation between people who each have a reason to feel underpaid.

The month that goes backwards

Month four is the honest part of this example. Two office accounts relocate and eleven subscriptions leave with them — about ₹28,000 a month gone through nobody’s fault, discovered on a Tuesday when eleven dabbas came back uncollected. A four-person crew feels that immediately, because there is no reserve and there is no month of runway. What makes it survivable is that the split is a share of revenue rather than a fixed wage: everyone earns less that month, together, and nobody is owed money the business does not have. A crew paying itself fixed salaries would have had one person quietly funding the others and resenting it by week two. The recovery in month five is not clever either — the Subscriptions role spent the bad month doing exactly what worked in month one, in a different building.

What sixty dabbas a day actually costs somebody

The number this example does not put in a table is 5:15 am, six days a week, which is when the Kitchen Lead starts if lunch is going out at noon. Provisions are bought the evening before, the cooking runs four hours, packing takes forty minutes with two people and ninety with one, and the delivery window is unforgiving in a way an office job never is: a dabba that arrives at 1:30 pm is a refund and a cancelled subscription. None of that is a reason not to do it, and all of it is a reason to be honest about who is signing up for what. The hours per role in the blueprint — 42, 25, 20 and 12 — are the real ones rather than the encouraging ones, and the crews that come apart are usually the ones where somebody agreed to 42 hours while privately picturing 20.

What went wrong

• Eleven subscriptions vanished in month four when two office accounts changed buildings. Concentration is what makes the route work and also what makes a single move hurt.

• The first packaging order was the cheap containers. About ₹4,000 of leaked sambar and six angry subscribers later, the crew reordered.

• Nobody logged provisions for the first five weeks, so the month-one margin was a guess. The ledger only tells you anything if the boring entries go in on the day.