CAC Payback Period Explained: Why a 3:1 LTV/CAC Can Still Run Out of Cash

Two questions that sound like one

LTV/CAC and CAC payback are usually presented as two views of the same thing, and they are not. LTV/CAC asks whether a customer, over their whole life, returns more than they cost — it is a question about profitability, and it has no clock in it. Payback asks how long you wait for the money back — it is a question about liquidity, and it is nothing but a clock. A business can pass the first test and fail the second badly enough to die, and that combination is common enough to have a name in the venture world: growing broke.

The distinction becomes concrete the moment you notice that CAC is spent up front, in full, before the customer has paid anything. Every new customer is a small loan you make to yourself, repaid in monthly instalments of gross profit. LTV/CAC tells you the loan is a good one. Payback tells you how much cash is tied up in loans outstanding at any moment — and that figure grows with your growth rate, which is why the companies most likely to be strangled by a long payback are the ones doing best on the metric everybody watches.

The formula, and the term everyone drops

CAC payback period = CAC ÷ (monthly ARPA × gross margin). Every part of that matters, but the gross margin is where the errors live. A customer paying $60 a month does not hand you $60 a month of repayment capacity: hosting, support, payment processing, third-party licences and the people who keep the service running all come out first. At a 75% gross margin, $45 a month is what is actually left to repay the acquisition cost.

Drop the margin and the arithmetic flatters you by a completely predictable amount. A CAC of $1,200 divided by $60 of revenue gives 20 months. Divided by $45 of gross profit it gives 26.7 months. The gap is 6.7 months, and it is exactly 25% of the true figure — one minus the gross margin, every time, for any CAC and any price. That is the useful way to remember it: reporting payback on revenue understates it by your cost of goods. At a 90% margin the distortion is a tolerable 10%; at a 50% margin it doubles the answer, and any business with services, hardware or heavy support in the mix is closer to the second case than the first.

Two further details separate a payback figure that survives scrutiny from one that does not. The CAC in the numerator should be the fully loaded one — all sales and marketing salaries, commissions, tools and agency fees, not just media spend — because a payback computed on media alone is measuring a fraction of the loan. And the ARPA in the denominator should be new-customer ARPA, not blended: what matters is what the cohort you just bought actually pays, and in most companies that differs from the average of a book built over years.

A textbook 3:1 that runs out of cash

Company A charges $60 a month, runs a 75% gross margin and pays $1,200 to acquire a customer. Its customers churn at 1.25% a month, so they last 80 months on average and contribute 45 × 80 = $3,600 of gross profit over that life. LTV/CAC is 3,600 ÷ 1,200 = exactly 3.0 — the ratio every investor deck aims at, hit precisely. Payback is 1,200 ÷ 45 = 26.7 months. Nothing about that business is dishonest or badly run. It is simply lending each customer $1,200 and getting it back over more than two years.

Now let it grow. Start at 100 new customers in month one and add 10% more each month. By month 24 the company has around 8,850 customers, roughly $6.4M of annual recurring revenue, and a monthly gross profit approaching $400,000 — every visible sign of a business working. Its cumulative cash position over those 24 months is minus $7.3M. It has never had a cash-positive month, and the monthly hole is still widening: month 24 alone is negative by about $676,000. Growth is not curing the problem, it is the mechanism of the problem, because each month's larger cohort costs its full CAC immediately while the previous cohorts are still only part-way through repaying theirs.

Company B is identical in every respect except that it acquires a customer for $540 instead of $1,200, giving a payback of exactly 12 months. Same price, same margin, same churn, same growth rate, and by month 24 the same 8,850 customers and the same $6.4M of recurring revenue. Its cumulative cash consumption is $1.5M rather than $7.3M — almost five times less, on financials that look identical from the revenue line. The entire difference is the length of the loan, and it is invisible on any metric that does not have a clock in it.

Where churn comes in, and where the cliff is

The standard payback formula quietly assumes the customer is still there each month, paying. Real cohorts shrink, so the true payback is longer than CAC ÷ monthly gross profit whenever churn is not zero. The corrected version asks how many months of a decaying cohort it takes for cumulative gross profit to reach the CAC. For Company A at 1.25% monthly churn, the nominal 26.7 months becomes 32.2 — five and a half months longer, purely because some of the customers who were supposed to be paying in month 30 had left.

Push the churn up and the effect stops being a correction and becomes a wall. At 2% monthly churn the same customer takes 37.7 months to repay and LTV/CAC falls to 1.88. At 2.5% it takes 43.4 months and the ratio is 1.5. At 3% it takes 52.8 months and the ratio is 1.25. And above 3.75% a month, the customer never repays at all — not late, never — because the total gross profit a customer can produce is monthly contribution ÷ churn, which at 3.75% is 45 ÷ 0.0375 = $1,200, exactly the CAC. Any churn above that and the whole life of the customer is worth less than what you paid for them. That threshold is worth computing for your own business: it is simply monthly gross profit ÷ CAC, and it is the churn rate at which acquisition stops being an investment and becomes a donation.

One identity ties the two metrics together

LTV/CAC = 1 ÷ (monthly churn × payback months). Check it on Company A: 1 ÷ (0.0125 × 26.67) = 3.0, and on Company B: 1 ÷ (0.0125 × 12) = 6.7. The identity holds because LTV is monthly gross profit divided by churn, and payback is CAC divided by that same monthly gross profit, so dividing one by the other cancels the gross profit entirely.

Read it the useful way round and it explains the whole disagreement between the two metrics. A ratio of 3:1 does not describe a payback at all — it describes a product of churn and payback. You can reach 3:1 with a twelve-month payback and mediocre retention, or with a twenty-seven-month payback and excellent retention, and those are opposite businesses to run and to fund. This is why an investor who has seen a few of these will not accept the ratio on its own: it can be hit by being genuinely efficient, or by having customers who happen to stay a very long time while you overpay for each one.

Why the line sits near twelve months

The twelve-month rule of thumb is not arbitrary. A payback inside twelve months means each year's acquisition is repaid within that same year, so growth can in principle be funded from the gross profit the business itself generates: the loans you make close before you need to make the next round. Past twelve months, the repayment of a cohort lands in a later financial year than the spend that created it, and the gap has to be bridged with capital — equity, debt, or cash you already had. That is the practical meaning of a long payback. It does not say the business is bad; it says the business is a financing problem as well as a marketing one.

There are four honest ways to shorten it, and they are not equally available. Cut CAC, by fixing the channels where the fully loaded cost per customer is worst — usually the fastest lever and the one most companies have not measured per channel. Raise gross margin, which shortens payback in exact proportion and is the lever that gets forgotten precisely because it sits with engineering and support rather than with marketing. Raise price, which raises ARPA and gross profit together. Or move customers onto annual billing paid up front, which does not change the economics at all but collects twelve months of cash on day one and can turn a 26.7-month payback into an immediate cash-positive sale. That last one is why annual prepay discounts exist, and why they are worth more than they look.

Frequently asked questions

What is a good CAC payback period?

Under twelve months is the benchmark most software investors work to, because it is the point at which a year of acquisition is repaid inside the same year and growth can be funded from the business rather than from a balance sheet. Twelve to eighteen months is workable if you have capital and retention is genuinely strong. Beyond about two years you are running a financing operation as much as a company, and the risk is not that the customers are unprofitable — Company A above returns three times its CAC — but that you must survive the wait. The right target also depends on who you sell to: a self-serve product billed monthly should be far under twelve months, while enterprise contracts with long sales cycles and annual prepayment routinely sit higher and are financed by the prepayment itself.

Should I use revenue or gross profit in the denominator?

Gross profit, always — revenue overstates how fast the CAC is repaid, by exactly one minus your gross margin. At a 75% margin, a payback of 26.7 months shows up as 20 months if you divide by revenue: 25% too short. The intuition is simple: the money that pays back the acquisition cost is what is left after you have paid to deliver the service, not what the customer transferred. If you have to publish the revenue-based version because a partner or an internal convention demands it, publish both side by side and label them, because the two will differ by a fixed and predictable ratio that anyone can then reconcile.

How does annual billing change the payback period?

It changes the cash timing completely and the underlying economics not at all, which is exactly why it is powerful. A customer worth $45 a month in gross profit who prepays a year hands you twelve months of contribution on day one — $540 — so an acquisition costing $540 is repaid the moment the invoice clears, and the 26.7-month arithmetic never gets a chance to bite. Nothing about LTV, margin or churn moved. Two cautions. Discounting the annual plan reduces the gross profit you collect, so a large discount can give back much of what the timing gained. And prepaid cash is not earned revenue: if a share of those customers cancel and you refund, you have to fund the refund out of money you have already spent on acquiring the next cohort.

Does a long payback period mean my CAC is too high?

Not necessarily — the formula has three inputs and any of them can be the culprit. A payback of 26.7 months comes from a CAC of $1,200 divided by $45 of monthly gross profit, and that $45 is itself $60 of revenue times a 75% margin. The same 26.7 months would arise from a perfectly ordinary CAC paired with a price that is too low, or with a margin dragged down by support costs or third-party fees. Before you cut acquisition spend, decompose the figure: compute payback per channel with fully loaded costs, then check whether your gross margin has drifted over the past year. Cutting a channel that was actually your cheapest, because the overall number looked bad, is a common and expensive mistake.

If my growth stops, does the cash problem go away?

Largely yes, and that is the uncomfortable part. Hold acquisition flat at 100 new customers a month with a 26.7-month payback and the monthly cash flow turns positive in month 27 — exactly when the first cohorts finish repaying — after a cumulative trough of about $1.5M. Growth is what makes a long payback dangerous, because every acceleration adds a full CAC today against gross profit spread over the next two years. This is why a company with strong unit economics and a long payback can look transformed the moment it stops growing, and why boards sometimes ask for exactly that. It is also why the honest framing is not that the business is broken but that its growth rate is a spending decision: with a 26.7-month payback, every extra customer you add this month is cash you will not see again until the month after next year.