Pull up a chair, because these two ratios get quoted together so often that people assume they are the same thing. They are not. One tells you how long your money is stuck. The other tells you whether the whole trip was worth making. You need both, and you need to know how each can be made to look better than it is.
Start with CAC itself
Customer acquisition cost is everything you spent to win new customers in a period, divided by the number of new customers won. “Everything” is the word people shy away from. It should include salaries and commissions of the sales team, SDRs, the presales engineers who run demos, marketing programmes and the marketing team, sales tools, and a fair share of management time. If your finance team shows you a CAC that only counts ad spend, that is a campaign cost, not a customer acquisition cost.
Also decide whether you are counting new customers only. Money spent on upselling existing accounts belongs to a different calculation. Mixing the two makes CAC look lower than it really is for new logos.
Payback: how long is the money stuck?
Take an illustrative company. In one quarter it spends ₹12 crore on sales and marketing and signs new customers worth ₹8 crore of annual recurring revenue (ARR). Its gross margin is 75%. Each month, that ₹8 crore of ARR brings in about ₹66.7 lakh of revenue, of which 75% is gross profit, roughly ₹50 lakh a month. Divide the ₹12 crore you spent by ₹50 lakh and you get 24 months.
The formula, in words: payback in months equals acquisition spend divided by new ARR times gross margin, all divided by twelve. We use gross margin, not revenue, because the money that pays back your spend is what is left after hosting, support and delivery costs. Using revenue instead of margin would have said 18 months here, a flattering error of six months.
Is 24 months bad? It depends on what the customer does next. If customers stay for six years, a two-year payback is fine, provided you can fund the wait. If your customers are small businesses who leave after eighteen months, you never get your money back. Payback does not know this. That is why it is a cash question, not a verdict.
LTV: the guess about the future
Lifetime value tries to answer the question payback ignores. A simple version: annual gross profit per customer, divided by annual churn rate. Suppose a customer pays ₹20,000 a month, gross margin is 70%, so gross profit is ₹1,68,000 a year. If 15% of customers leave each year, the average customer stays about 6.7 years (one divided by 0.15), and LTV is about ₹11.2 lakh. With a CAC of ₹4 lakh, LTV:CAC is 2.8.
Now look at how fragile that is. Change churn from 15% to 30% and LTV halves to ₹5.6 lakh, ratio 1.4. Change it to 10% and LTV is ₹16.8 lakh, ratio 4.2. A five-point error in a number you can only estimate from a short history swings the answer more than any improvement your sales team could make in a year. Be suspicious when a young company with eighteen months of data reports a confident LTV. It is extrapolating.
The commonly quoted target of 3:1 is a rule of thumb that circulates in SaaS circles. It is not a law, and it assumes things about margins and churn that may not hold for your business. Use it as a conversation starter.
Ways each one gets flattered
- Leaving out costs from CAC, such as presales, partner commissions or founder-led selling time.
- Blending free-trial self-serve customers with enterprise deals. The average hides two very different businesses; compute each segment separately.
- Using revenue where margin belongs.
- Putting expansion revenue into LTV while keeping expansion cost out of CAC.
- Matching this quarter’s spend with this quarter’s ARR even though deals closed now were paid for by effort in earlier quarters. With long enterprise cycles, a lagged comparison is fairer.
Which one to trust when
If you are cash-constrained or growing fast on borrowed money, watch payback. It uses only what has happened and a margin you know. If you are deciding whether a segment is worth entering at all, LTV:CAC is the more honest framing, as long as you show the range of churn assumptions rather than one number. A table of three churn scenarios is more useful than a single ratio with a decimal point.
One more habit worth having: when someone quotes either ratio, ask what is in the numerator and what time period the denominator covers. Half of all disagreements about “efficiency” turn out to be disagreements about definitions.
Figures and names in the examples are illustrative.
Leave a Reply