How long until a customer pays you back?

• ·Adds up everything spent to acquire customers in a month: media, sales and marketing salaries, tools and other costs.
• ·Divides by the customers actually won in that month to give fully loaded CAC.
• ·Works out contribution per customer per month: price less the variable cost of serving them.
• ·Divides CAC by contribution to give payback in months.
• ·Uses your churn rate to estimate how long a customer stays, what they contribute over 24 months, and their lifetime contribution.
• ·Compares payback with average customer life, because a payback longer than the relationship is a loss dressed as growth.

Nothing is sent anywhere or stored. If a month was unusual, use a three-month average for spend and customers won.

The ratio of lifetime value to CAC gets quoted more often, and it's the weaker of the two numbers for an early-stage company. It depends on a churn rate estimated from limited history, and dividing one by a small churn rate produces a very large number very quickly. A company with 15 months of data claiming an eight-year customer lifetime is guessing.

Payback is more honest because it's shorter-dated and closer to cash. It tells you how long your money is tied up, which is the question that actually constrains growth when runway is finite.

Use both, with the lifetime figure capped at a horizon you can defend. The calculator shows contribution over 24 months alongside the uncapped figure for exactly this reason.

An illustrative B2B software company. Last month it spent ₹4 lakh on paid media, ₹6.5 lakh on sales and marketing salaries including commissions, and ₹1.2 lakh on tools and travel. It won 18 new customers. Each pays ₹18,000 a month and costs ₹4,300 a month to serve. Monthly churn is 1.8%. Every figure is invented.

Both columns describe the same month. The first is the one that ends up in a deck, and the first follow-up question from any investor makes it disappear. The second is still a good result: 4.7 months against a customer who stays for years is a business worth putting money into.

The useful part is what the second column lets you decide. At ₹65,000 of fully loaded CAC and ₹13,700 of monthly contribution, the company can see exactly how much slower payback would get if CAC rose to ₹90,000 chasing growth, and whether the cash can carry it.

Software sold to businesses

Sales salaries and commissions, SDR costs, marketing, events, tools, and the founder's selling time at seed. Implementation effort is cost to serve rather than acquisition, but it should be counted somewhere.

Services firms and agencies

Business development salaries, proposal and pitch time from delivery staff, travel, and referral fees. Time spent writing proposals that don't convert is a real acquisition cost and almost never counted.

Restaurants and cloud kitchens

Platform advertising, funded discounts, launch marketing for a new outlet, and loyalty costs. Aggregator commission is variable cost to serve rather than acquisition, and both matter.

Healthcare services

Camps, referral arrangements where permitted, digital marketing and booking platform fees. Patient acquisition is often once and retention is the whole economics, so repeat rate belongs beside CAC.

• ·Media and agency spend: your marketing ledger or the platforms themselves, for the month that produced the customers you're counting.
• ·Sales and marketing salaries: payroll, at fully loaded cost including employer contributions, plus commissions actually earned.
• ·Tools: the CRM, outreach, analytics and data subscriptions that exist to win customers.
• ·New customers: your CRM or billing system, counting customers who actually started paying rather than everyone who signed.
• ·Price: average revenue per customer from billing, net of discounts, not the list price.
• ·Variable cost to serve: hosting and usage attributable to customers, payment gateway fees, support cost divided by customers, delivery or implementation where it recurs.
• ·Churn: customers lost in a month divided by customers at the start of it, averaged over several months so one bad month doesn't set the rate.

If support cost per customer is hard to split out, estimate it honestly rather than leaving it at zero. An estimate that's roughly right beats a precise number that's missing.

1. 01Split by channel before doing anything else. A blended 14 months often hides one channel at 6 and another at 30, and the answer is to move money, not to cut it.
2. 02Look at price. A modest rise on new customers shortens payback immediately, and the calculator shows how much volume a rise can afford to lose.
3. 03Look at cost to serve. Support, onboarding and usage costs often grew without anyone deciding to let them.
4. 04Look at retention. Payback and churn are the same problem seen twice; improving month-three retention improves both.
5. 05Look at who you're selling to. Smaller customers usually pay back slower and churn faster, and the segment mix drifts without anyone choosing it.
6. 06Only then look at cutting spend, which shrinks the business but does not by itself make the unit economics better.

Put payback and fully loaded CAC in the monthly pack, by channel where the data allows, with the definition written down. Track the trend rather than the month. Before increasing marketing spend, check what payback did the last time spend went up, because the honest pattern in most companies is that cost per customer rises with volume.

And use it as a spending rule, not just a report. Many companies set a payback threshold above which a channel doesn't get more money until something changes. That single rule prevents most of the expensive growth mistakes an early-stage company can make.

One caution about benchmarks. Payback figures quoted in blog posts and investor decks are rarely calculated the same way, and almost never fully loaded. Comparing your honest number with someone else's flattering one is a good way to talk yourself into spending that your cash position can't support. Compare your number with your own previous quarters, and with the runway you actually have.