Published on: June 13, 2026 · 8 min read
The first time I presented a CAC payback number to a leadership team, I said "six months" with total confidence. It was wrong in three different ways at once — and not one of them was an arithmetic error. The math was clean. The number was still a fiction. This is about the three traps hiding inside that one tidy figure, and what I report instead.
CAC payback is supposed to answer a simple question: after we spend money to acquire a customer, how long until that customer pays us back? Shorter is better — it means capital recycles faster and you can grow without constantly raising. It's one of the few unit-economics numbers a CFO and a growth lead will both nod at.
The problem is that the textbook formula — CAC ÷ monthly revenue per customer — quietly smuggles in three assumptions that are almost never true.
Blended CAC is total sales and marketing spend divided by all new customers — including the ones who arrived organically, through word of mouth, SEO, or your existing brand. Those customers cost you (almost) nothing to acquire, and they drag the average down.
That's fine if you only ever want to describe the past. It's dangerous the moment you use payback to decide how much more to spend. The next marginal customer doesn't come from your organic base — they come from paid channels, at the paid price. Planning growth on blended CAC is like budgeting your next road trip using your lifetime average cost per kilometre, most of which you walked.
Illustrative: a business where ~50% of new customers arrive organically. Same spend, very different decision-making number.
The textbook formula uses revenue. But a customer paying you ₹1,000 a month doesn't pay back ₹1,000 of acquisition cost a month — they pay back whatever's left after the cost of serving them. Payment processing, cloud, support, the discount you gave them, the rewards you funded. In fintech and commerce especially, gross margin can be a thin slice of top-line.
Payback has to be computed on gross margin per customer, not revenue. Skip this and you'll declare payback "achieved" while the customer is still deep underwater on a contribution basis.
Even with paid CAC and margin sorted, dividing by an average monthly margin assumes every customer pays at a steady, identical rate forever. They don't. Real cohorts have a retention curve: a chunk churns in the first few weeks, the survivors settle, and the curve flattens. Early churners may never come close to paying back what you spent to acquire them — and they're invisible in a single blended average.
The honest version isn't a single number at all. It's a cumulative margin curve per cohort: spend on day zero, then track how much contribution margin that exact group has returned at month 1, 3, 6, 12. Payback is the month the curve crosses your CAC line. Sometimes it never does.
Same business, same spend, three ways of looking at it. Watch the "payback" move:
| Method | CAC | Monthly value | Payback |
|---|---|---|---|
| Textbook (blended, revenue) | ₹1,400 | ₹1,000 rev | 1.4 mo |
| Paid CAC, still revenue | ₹2,800 | ₹1,000 rev | 2.8 mo |
| Paid CAC, gross margin | ₹2,800 | ₹350 margin | 8.0 mo |
| + cohort-weighted (early churn) | ₹2,800 | ~₹260 eff. | ~11 mo |
Assumes 35% gross margin and meaningful first-90-day churn. The "real" payback is nearly 8× the headline number — driven entirely by which assumptions you make explicit.
The number didn't get worse. The number was always this. The reporting just got honest.
When I present acquisition economics now, the headline isn't a single payback month. It's three things:
None of this requires fancier tooling. It requires refusing to collapse a curve into a point until you've looked at the curve.
I'm honest about the limits. Cohort curves are backward-looking — a cohort acquired in a downturn or under a new pricing model may behave nothing like last year's. Attribution noise means "paid CAC by channel" is an estimate, not a measurement, especially once you account for the organic lift that paid spend itself creates. And payback says nothing about magnitude — a customer can pay back fast and still be worth very little. That last gap is exactly why payback needs a companion metric, which is its own can of worms.