Two programs, same product at $49 a month. One pays 25% for twelve months. The other pays 15% for the life of the customer, which sounds like the cheaper deal because the rate is lower.
At 5% monthly churn the first costs $112.61 per referral and the second costs $147.00. The one that looks cheaper is 30% more expensive, and the rate is the reason nobody notices.
TL;DR: A recurring commission is not rate times price times window, because some of those customers cancel first. At 5% monthly churn a 12-month window delivers 9.19 paid months, so budgeting on 12 overstates the cost by a third. The figure to decide on is commission as a share of customer lifetime value, which should usually land between 10% and 30%. The affiliate commission calculator runs the arithmetic including the payout ramp.
Why does the window overstate what you pay?
Because a commission window is a maximum, not a schedule.
Picture a hundred customers referred in January, on a product with 5% monthly churn. In February, 95 of them are still paying. In March, about 90. By the twelfth month roughly 54 remain. The commission is paid on each month each customer actually pays, so what you owe is the sum of those survivors, not a hundred customers times twelve months.
That sum has a name, expected paying months, and for a window of n months at churn c it is the sum of (1 - c) to the power of each month. At 5% over twelve months it comes to 9.19.
| Monthly churn | Paid months in a 12-month window | Commission at 25% of $49 | Customer LTV | Share of LTV |
|---|---|---|---|---|
| 2% | 10.76 | $131.86 | $2,450.00 | 5.4% |
| 3% | 10.21 | $125.01 | $1,633.33 | 7.7% |
| 5% | 9.19 | $112.61 | $980.00 | 11.5% |
| 8% | 7.90 | $96.83 | $612.50 | 15.8% |
| 10% | 7.18 | $87.90 | $490.00 | 17.9% |
Two things in that table are worth sitting with.
Higher churn makes the commission cheaper and the program worse. The payout falls from $131.86 to $87.90 as churn rises, because the customers leave before you have to pay for them. But lifetime value falls faster, so the share of the customer you are handing over climbs from 5.4% to 17.9%. The cheap-looking payout is the expensive program.
The absolute number moves much less than people expect. Between 2% and 10% churn, a fivefold difference in retention, the commission only moves by about a third. The window caps it. It is the denominator that moves.
What should the commission actually be measured against?
Customer lifetime value, which turns an unanswerable question into an arithmetic one.
"Is 25% too much" has no answer. "Are we willing to give away 11.5% of each customer to acquire them" does, and it is directly comparable to what the same customer costs through paid advertising. Most programs should land between 10% and 30% of LTV. Above that, the program is competing with your margin rather than with your other channels.
Running the same product through different windows shows how much of that is decided by the window rather than the rate:
| Commission window | Paid months | Commission per referral | Share of LTV |
|---|---|---|---|
| First payment only | 1.00 | $12.25 | 1.3% |
| 6 months | 5.30 | $64.90 | 6.6% |
| 12 months | 9.19 | $112.61 | 11.5% |
| 24 months | 14.16 | $173.46 | 17.7% |
| Lifetime | 20.00 | $245.00 | 25.0% |
Same rate, same price, same churn. The window moves the cost by a factor of twenty, and it is the term most programs copy from a competitor without checking.
Is a lifetime commission a bad idea?
It is expensive in a way that arrives years after the decision, and the arithmetic is counterintuitive enough to be worth showing.
Compare 25% for twelve months against 15% for life, which is the trade most programs think they are making when they offer "a lower rate but forever":
| Monthly churn | 12 months at 25% | Lifetime at 15% | Which costs more |
|---|---|---|---|
| 2% | $131.86 | $367.50 | Lifetime, by 179% |
| 3% | $125.01 | $245.00 | Lifetime, by 96% |
| 5% | $112.61 | $147.00 | Lifetime, by 31% |
| 8% | $96.83 | $91.88 | The 12-month deal, barely |
The crossover is around 8% monthly churn. Below it, the lower lifetime rate is the more expensive offer, and the better your retention gets the worse that deal becomes. A lifetime commission is a bet against your own product improving.
There is a second problem that does not show in a table. A lifetime commission has no end, so the cost is not forecastable: you are committing a share of revenue from customers who have not been referred yet, for a period you cannot bound. A long fixed window gives affiliates most of the upside at a number you can put in a budget.
When does the real monthly cost arrive?
Later than the program launches, which is why the first few months feel affordable.
With ten new referrals a month on a twelve-month window at 5% churn, every month adds a cohort while the earlier cohorts are still being paid for. The bill climbs until the oldest cohort reaches the end of its window:
| Month | Monthly payout |
|---|---|
| 1 | $122.50 |
| 2 | $238.88 |
| 3 | $349.43 |
| 6 | $649.02 |
| 9 | $905.89 |
| 12 | $1,126.12 |
Month one costs an eighth of the steady state. A program judged on its first quarter looks cheap, and the number that matters is the one at the end of the window, which for a lifetime commission never arrives at all.
This is also the honest way to present a program internally. "Affiliates cost us $350 last month" is true in month three and misleading, and the correction arrives about the time someone has planned around it.
What is EPC and why do affiliates ask for it?
Earnings per click: the total commission earned divided by the clicks sent. It is the number an affiliate uses to decide whether promoting you is worth their traffic, because it is the only figure that compares two offers with different rates, prices and conversion rates in one number.
If yours is low, there are exactly two causes and they need different fixes. Either the page is not converting, in which case the commission rate is irrelevant and the landing page is the problem, or it converts fine and the commission is too small to be worth their audience. Raising the rate to fix a conversion problem is the expensive mistake, because it raises the cost of every sale you were already making.
Worth knowing in the other direction: EPC is a property of the traffic as much as the offer, so an affiliate's EPC on your program and yours on another are not comparable. Use it to compare offers within one affiliate's traffic, which is exactly how they use it.
What does the arithmetic leave out?
Three costs that sit on top of the commission and are often larger than people expect for a small program.
Platform fees. An affiliate platform charges either a monthly fee or a percentage of commissions paid, sometimes both. On a small program the fixed fee can exceed the commissions.
Payout fees. Paying fifty partners in twelve countries is not free, and the per-payment cost matters when the average payout is small, which is the argument for a payout minimum in the affiliate agreement.
Time. Recruiting partners, reviewing applications, answering questions about numbers, and policing the prohibited conduct clause. This is the one nobody budgets and the one that decides whether a small program is worth running at all.
Add those before comparing the channel against paid acquisition, or you are comparing a fully loaded cost against a commission-only one.



