D2C unit economics: the four numbers that decide if ads can work

Most conversations about whether advertising “works” for a brand are really conversations about arithmetic that was never done. A founder says their agency delivered 4x and it did not feel profitable. Both statements are usually true, and the gap between them lives in four numbers.

They depend on each other in a fixed order. You cannot judge a ROAS figure without a break-even, you cannot get a break-even without contribution margin, and you cannot decide how fast to scale without payback. Work them in that order and most arguments about ad performance simply dissolve.

Start with contribution margin, not gross margin

Gross margin stops at cost of goods, and that is why it flatters everything downstream. Contribution margin deducts every variable cost of fulfilling an order, which is the only version that tells you what is genuinely available to pay for a customer.

Take a ₹2,000 kurta. Cost of goods is ₹800. Shipping is ₹120. The payment gateway takes about 2%, so ₹40. In apparel a returns allowance is not optional — at a 10% return rate that is another ₹200 against every order you take.

That leaves ₹840, or 42%. A founder working from a “60% margin” belief is planning against a number that is eighteen points too generous, and every target built on it will be unreachable.

Two details are worth getting right. Cash-on-delivery, still common in Indian D2C, carries both a handling fee and a materially higher return rate, so a catalogue with heavy COD share needs its own margin figure. And if your products differ widely, one blended margin across the catalogue will hide the products that are quietly losing money.

Break-even ROAS falls straight out of it

Once you have contribution margin, break-even ROAS is one divided by it. Nothing more.

At 42% margin you break even at about 2.4x. At 40% it is 2.5x. At 25% it is 4x. At 20% it is 5x — which means a brand with thin margins running at 4.5x ROAS is losing money on every order while its dashboard looks respectable.

This is the single most useful number to have written down, because it converts every future ROAS conversation from an opinion into a comparison. A 5.5x month is excellent on 42% margin and merely adequate on 20%. Nobody can tell you whether your ROAS is good without knowing this number, and that includes us.

It is also why comparing your ROAS to another brand’s is meaningless. Their break-even is not yours. When you see an agency advertising the returns they generate, the figure is uninterpretable without the margin structure underneath it.

CAC is only meaningful next to what a customer is worth

Customer acquisition cost is total acquisition spend divided by new customers — and the word new is doing real work there. Dividing by all orders counts repeat buyers you already paid for, which can make CAC look roughly half of what it is.

Include everything spent to win the customer: media, agency fees, creative production, and the cost of any acquisition discount. If it was spent to get them, it belongs in the number.

A brand spending ₹2,40,000 on ads and ₹60,000 on fees to acquire 250 first-time customers has a CAC of ₹1,200. Whether that is good depends entirely on the contribution margin on their first order. At a ₹2,400 average order and 45% margin, the customer contributes ₹1,080 — so the first order does not quite cover acquisition. At a ₹4,000 average order, it contributes ₹1,800 and clears it comfortably.

Keep paid CAC and blended CAC separate. Blended divides all acquisition cost by all new customers including organic and referral, and it is the honest number for planning because it is what your bank account experiences. Paid CAC is the one to optimise inside the ad account. Reporting one while thinking about the other is a common and expensive confusion.

Payback decides how fast you can actually go

The first three numbers tell you whether acquisition is profitable. Payback tells you whether you can afford to do it at volume.

If the first order covers CAC, every repeat purchase is upside and you can scale as hard as the account allows. If payback lands within about three months, you can scale steadily. Beyond six months, growth has to be funded from working capital, and that should be a deliberate decision rather than something a founder discovers a quarter later.

This is why two brands with identical lifetime-value-to-CAC ratios can have completely different ceilings. A 3:1 ratio that repays on order one compounds. The same 3:1 repaying at month fourteen is a financing problem wearing a growth costume.

Payback also degrades as you scale, and usually before anything else does. Early spend reaches your most responsive audience cheaply; as budget rises, CAC rises with it. Watching payback monthly is what tells you when you have reached the edge of profitable scale.

Where these numbers get quietly flattered

Four failures account for most of the cases we see.

Using gross margin instead of contribution margin, which sets an unreachable break-even and makes a genuinely healthy account look like a failure.

Building lifetime value from revenue instead of gross profit, which roughly doubles the figure and makes it incomparable to CAC.

Modelling a lifetime a brand has not lived. A two-year-old brand does not know its three-year customer value, and a projection stretched that far is usually a way of justifying a CAC it cannot currently afford. Read the 90-day and 180-day values you actually have.

And reading platform-reported ROAS as though it were revenue in the bank. Meta and Google both count conversions they influenced, and their totals routinely exceed what the business took. Blended ROAS or marketing efficiency ratio will reconcile with the bank; per-platform figures will not.

The order to work in

Calculate contribution margin honestly, including returns. Derive break-even ROAS from it. Measure CAC on new customers with all acquisition costs included. Then check payback before deciding how hard to push.

Do that and you will know within an afternoon whether your current account is profitable, how much headroom it has, and what the constraint actually is. In our experience the constraint is rarely the ad account — it is usually margin or average order value, and no amount of campaign restructuring fixes either.