Start with one customer, because the whole sector is just this arithmetic times a growth rate.
How long does a SaaS customer take to pay back?
Take a synthetic Indian SaaS company (call it anything). It spends ₹60,000 of sales and marketing to win one customer. That's its CAC, and under every GAAP it's fully expensed this year, because there's no "customer asset" on the balance sheet. The customer pays ₹48,000 a year. Serving them costs 20% of that in cloud and support, so they contribute ₹38,400 of gross profit a year. Here's the cumulative cash from this one relationship:
| Day 0 | End yr 1 | End yr 2 | End yr 3 | |
|---|---|---|---|---|
| Cumulative gross profit | 0 | 38,400 | 76,800 | 1,15,200 |
| Less CAC | −60,000 | −60,000 | −60,000 | −60,000 |
| Cumulative position (₹) | −60,000 | −21,600 | +16,800 | +55,200 |
Payback comes in 18.75 months (60,000 ÷ 3,200 of monthly gross profit). Every month the customer stays after that is nearly pure margin. You'd take that trade every time. An 18-to-19-month payback on an asset that can live eight or ten years is a return most industrial projects would kill for.
Advertisement
Why does a growing SaaS company report a loss?
Now watch what the P&L does when the company grows its acquisition 50% a year. Assume customers arrive on day 1 of each year and there's zero churn for now. The KPI panel relaxes that. Fixed R&D + G&A runs ₹30/35/40 lakh:
| ₹ lakh | Year 1 | Year 2 | Year 3 | Year 3 (freeze growth) |
|---|---|---|---|---|
| New customers | 100 | 150 | 225 | 0 |
| Install base | 100 | 250 | 475 | 250 |
| Revenue (₹0.48 L each) | 48.0 | 120.0 | 228.0 | 120.0 |
| Gross profit (80%) | 38.4 | 96.0 | 182.4 | 96.0 |
| S&M (₹0.60 L per new customer) | 60.0 | 90.0 | 135.0 | 10.0 |
| R&D + G&A | 30.0 | 35.0 | 40.0 | 40.0 |
| EBIT | −51.6 | −29.0 | +7.4 | +46.0 |
| EBIT margin | −108% | −24% | +3% | +38% |
Read the last two columns together. The same company, in the same year, is either a 3%-margin business or a 38%-margin business, depending on one management choice: how hard to press the acquisition accelerator. The install base is a profitable annuity. The reported loss is the cost of enlarging it, all expensed at once. Nothing in SaaS analysis matters more than this idea, and it cuts both ways:
- Cohorts refute the bull error, where slogans can't. "Losses are just investment" is true only if each cohort really pays back. That means CAC payback is short, retention holds, and the margin structure at maturity is real. A company whose cohorts never cover their CAC is also "investing in growth". It's just destroying value at scale.
- Cohorts refute the bear error too. Dismiss every loss-making SaaS on P/E grounds and you miss that the loss is discretionary. Your job is to recompute the freeze-growth column: gross profit of the existing base, minus retention-level S&M, minus R&D/G&A. That number, sometimes called steady-state or pro-forma profitability, is the earnings engine the P&L hides.
How do prepaid contracts and commissions affect the numbers?
Two mechanical notes before the panel. First, when customers prepay annually, cash timing is friendlier than the P&L. Cash arrives up front, revenue is recognized monthly over the year, and the gap is held on the balance sheet as deferred revenue, the liability you met in revenue recognition. A growing SaaS with annual prepay runs on negative working capital. Its customers finance the growth, and CFO runs ahead of earnings. So watch the sign. Deferred revenue building is health. Deferred revenue shrinking while management talks up ARR is a leak.
Second, under ASC 606/IFRS 15 (and Ind AS 115), sales commissions, which are a chunk of CAC, are in fact capitalized as "costs to obtain a contract" (US GAAP codifies this in ASC 340-40) and amortized over the expected benefit period. That means reported S&M slightly understates cash CAC in fast growth. It rarely changes the verdict. Still, when you compute CAC payback from the P&L, check the contract-cost note. If precision matters, rebuild CAC from cash spend.