Unit economics: does each sale actually make money?

There is a particular kind of trouble that looks exactly like success from outside: revenue climbing month after month while the bank balance drifts downward. Orders are up, the shop is busier, the owner is working harder than ever — and there is less money available at the end of each month than there was at the start. The instinct in that situation is almost always to sell more, because selling more is what fixed every previous problem. If the underlying unit loses money, selling more is the one thing that reliably makes it worse.

Unit economics is the arithmetic that tells you which situation you are actually in. It asks a single question about one sale, in isolation: after everything that particular sale cost you, how much cash did it leave behind? That leftover figure is the contribution margin, and its sign matters more than its size. A positive margin means volume is genuinely worth chasing, because every additional unit contributes something toward your fixed costs. A negative margin means every unit you sell hands away money you will have to find from somewhere else — savings, a loan, or a supplier you pay late.

A variable cost is one that appears because a specific sale happened and would not have appeared otherwise. For a product business that usually means the cost of the goods themselves, packaging and label, the payment gateway's cut, delivery or courier charges, any marketplace or aggregator commission, and per-order handling if you pay someone by the order rather than by the month. Two are routinely forgotten. The first is returns: if a predictable share of orders comes back, the cost of that is a real per-order cost even though it does not appear on every order. The second is the discount you actually gave rather than the price on the tag.

Fixed costs stay out of this particular sum — rent, salaries, your software subscriptions, the electricity bill. They are real and they have to be paid, but they do not change when you sell one more unit, and mixing them in makes the per-unit figure move every time volume moves, which defeats the purpose. Keep them out here and bring them back in at the break-even step, where they belong. The test for any cost is simple: if you sold one fewer unit this month, would this number change?

Take a product listed at ₹900 and actually sold, after a routine discount, at ₹850. Suppose you buy it for ₹500, packaging and label run ₹40, the courier charges ₹80 on average across the destinations you ship to, and the payment gateway takes 2% — that gateway percentage is the one figure here you should replace with your own, because it is printed on your settlement statement and it varies by provider and payment method. Two percent of ₹850 is ₹17. Adding those gives ₹637 of variable cost against ₹850 of revenue, leaving a contribution margin of ₹213, or a little over 25% of the selling price.

Every number in that paragraph is illustrative arithmetic, not a benchmark, and the exercise is only worth anything with your own figures in the slots. What the structure gives you is a place to notice things. The ₹50 discount was 19% of the margin. If a marketplace took 15% commission on the same sale, that is ₹127 gone and the margin drops to ₹86 — the same product, sold through a different channel, is a substantially different business. Run the sum once per channel rather than once for the product.

Services feel harder to reduce to a unit, but the calculation works once you decide what the unit is: one job, one visit, one project, one month of a retainer. Revenue per unit is what the client actually pays for that job. Variable costs are materials consumed, travel to the site, subcontractor or freelancer fees, and the practitioner's time if they are paid per job rather than a monthly salary. If your staff are salaried, their cost is fixed, which makes the per-job contribution look flattering — and that flattery is the trap.

The correction is to treat capacity as the real constraint. A salaried team has a finite number of working hours in a month, so the figure that governs the business is contribution per available hour, not contribution per job. A job paying ₹8,000 with ₹1,000 of materials looks better than one paying ₹5,000 with ₹500 of materials, until you notice the first takes three days and the second takes half a day. Divide by the hours each consumes and the ranking often inverts. This is why service businesses can be fully booked and still not profitable: they have filled their capacity with the wrong jobs.

Once you have a reliable contribution margin per unit, break-even is one division. Add up the fixed costs you must pay in a month — rent, salaries, subscriptions, utilities, loan instalments, and a realistic figure for your own drawings, which owners routinely leave out and then wonder why the business never funds their life. Divide that total by the contribution margin per unit. The answer is the number of units that must sell before the month stops losing money.

With ₹1,80,000 of monthly fixed costs and the ₹213 margin from earlier, break-even is about 845 units, or roughly 28 a day. That number is the point of the whole exercise, because it is checkable against reality in a way a margin percentage is not. Ask whether 28 a day is consistent with the footfall you actually see, the stock you can actually hold, and the hours you can actually work. If break-even sits above the most you have ever sold in your best month, the problem is not effort or marketing — it is the price, the cost base, or the fixed cost base, and no amount of selling will resolve it.

Unit economics is unusually honest as business tools go, because every input is a number from your own records rather than an estimate about the world. That also fixes its limits. It cannot tell you whether demand exists at the price you plugged in; it takes your price as given and works forward from it. It says nothing about what it costs to find the customer — the money spent on advertising, samples and discounts to acquire a buyer sits outside the contribution margin, and a healthy margin can still be swallowed whole by acquisition cost. Keep the two figures adjacent and separate rather than merging them.

It also assumes costs scale in a straight line, which they do not: suppliers give volume discounts, couriers reprice by weight band, and one more order some days needs no extra effort while on others it needs an extra pair of hands. And it is a snapshot. Purchase prices move, gateway rates change, courier charges rise. A margin calculated eighteen months ago and never revisited is a historical document, not a current fact. Recalculate when any input visibly moves, and treat the output as a decision aid rather than a verdict.