How SaaS affiliate programs pay: recurring vs one-time commissions, cookie windows, churn, a 12-month cohort table and how to check retention before promoting.
The pitch for SaaS affiliate programs is that you get paid forever. Refer a customer once, collect 20% of their subscription every month while you sleep, stack cohort on cohort until the base is large enough to live on. It is a genuinely good model. It is also the model most often mis-sold, because the arithmetic that makes it work only becomes visible somewhere around month eleven, and almost everything written about it skips the part where customers cancel.
Two affiliates can send identical traffic to two SaaS affiliate programs paying identical percentages and end up with a twofold income difference after two years. Nothing in the rate card explains it. The explanation is churn, and it is a number you have to actively extract from the advertiser because nobody volunteers it.
What follows is the mechanics: how recurring and one-time SaaS commissions differ in practice, why self-serve signup makes attribution more fragile than it looks, a twelve-month cohort table you can run your own numbers through, why annual plans pay less nominally and more actually, and a checklist for assessing a program's retention before you commit traffic to it.
Four structures cover almost everything on the market.
Lifetime recurring. A percentage of the subscription for as long as the customer pays. Typically 20 to 30% of monthly revenue. This is the structure worth building a business around, and it is becoming less common as SaaS companies get better at modelling their own acquisition costs.
Capped recurring. The same percentage, but only for 12 or 24 months. Sometimes described in the terms as a commission period rather than a cap, which is easy to skim past. A 12-month cap turns what looks like a recurring program into a deferred one-time payment.
One-time bounty. A flat payment per paying customer, often 50 to 150% of one month's subscription, sometimes structured as a percentage of first-year value. Fast cash, no tail.
Tiered or hybrid. A bounty on signup plus a smaller recurring percentage, or a recurring rate that steps up once you cross a volume threshold. Volume tiers are common and are worth negotiating even at modest scale.
The choice between these mirrors the broader revenue share versus one-time CPA decision, with one SaaS-specific wrinkle: subscription products have unusually predictable retention curves compared with, say, a gambling deposit, which makes the recurring side easier to forecast honestly.
SaaS buys are researched. Somebody reads a comparison post on a work laptop on Tuesday, watches a demo video on their phone on Thursday, starts a free trial from a different browser the following week, and converts to paid fourteen days after that. Your cookie has to survive all of it.
Three details matter more than the headline cookie length:
Coupon and deal-site interference deserves its own mention. If the advertiser's checkout has a promo code field and the internet has a code for it, a share of your buyers will leave the page to look for one and return through somebody else's link. Some programs handle this with attribution rules that protect content publishers; most do not. Ask.
A free trial inserts a delay between the click and the money that is entirely outside your control. Trial-to-paid conversion is the advertiser's job, and it varies widely — a product with a strong onboarding flow converts a much larger share of trials than one that drops users into an empty dashboard.
This makes trial-to-paid rate a number worth asking about directly, because it multiplies straight into your earnings. Two programs with identical commission rates and identical traffic will pay you very differently if one converts a third of its trials and the other converts a tenth. It is also the reason credit-card-required trials, which feel like a conversion killer at the top of the funnel, often produce more affiliate income than card-free ones.
Monthly churn is the percentage of paying customers who cancel in a given month. It sounds like an advertiser's internal metric. For a recurring affiliate it is the single most important input to your income, and it is the one thing rate cards never mention.
The reason it matters so much is that it sets average customer lifetime, which is simply one divided by the monthly churn rate. At 5% monthly churn, the average customer pays for about 20 months. At 10%, about 10 months. At 3%, about 33 months. Your commission per customer scales directly with that number, so a program paying 20% with 3% churn beats a program paying 30% with 10% churn by a wide margin — the exact opposite of what the two rate cards suggest.
Ask about churn and a well-run SaaS company may answer with net revenue retention instead. The two are not interchangeable for you.
Net revenue retention includes expansion revenue — existing customers adding seats or upgrading tiers. A company can have 8% monthly logo churn and still report net revenue retention above 100% because its surviving customers grow fast. That is a healthy business, but it only helps you if your commission is calculated on the customer's current subscription value rather than the plan they originally signed up on. If your commission is locked to the initial plan, expansion revenue is invisible to you and gross churn is the only number that matters.
Confirm this explicitly. The phrase to look for in the terms is whether commission is paid on the customer's ongoing subscription payments or on the initial plan value. It is a small clause with a large effect on a two-year time horizon.
Model it concretely. You refer 10 new paying customers per month, each on a $100/month plan, with a 20% recurring commission — $20 per customer per month. Monthly churn is 5%.
Active customers at the end of each month equal last month's active base times 0.95, plus 10 new.
| Month | Active customers | Monthly commission | Cumulative |
|---|---|---|---|
| 1 | 10.0 | $200 | $200 |
| 2 | 19.5 | $390 | $590 |
| 3 | 28.5 | $571 | $1,161 |
| 4 | 37.1 | $742 | $1,903 |
| 5 | 45.2 | $905 | $2,808 |
| 6 | 53.0 | $1,060 | $3,868 |
| 7 | 60.3 | $1,207 | $5,075 |
| 8 | 67.3 | $1,346 | $6,421 |
| 9 | 74.0 | $1,479 | $7,900 |
| 10 | 80.3 | $1,605 | $9,505 |
| 11 | 86.2 | $1,725 | $11,230 |
| 12 | 91.9 | $1,839 | $13,069 |
Two things stand out. Month 12 pays more than nine times month 1 on identical traffic — that is the compounding everyone talks about. And the same ten customers a month under a $100 one-time bounty would have paid $1,000 every month, $12,000 across the year. The recurring structure only overtakes it in month 11.
That crossover point is the honest answer to whether recurring is better. For eleven months you earn less. If you are reinvesting commissions into paid traffic, earning less for eleven months means buying less traffic for eleven months, which compounds against you in the opposite direction. Recurring is superior over a two-year horizon and inferior over a six-month one.
The base also converges. At 5% churn and 10 new customers a month, the active base stabilises at 10 divided by 0.05 — 200 customers, $4,000 a month. That is the ceiling for this traffic level. Growing past it requires more new customers, not more time.
Change nothing except churn, from 5% to 10%. The active base at month 12 falls to 71.8 instead of 91.9, and cumulative first-year commission falls to roughly $11,080 against $13,070 — about 15% less. Survivable, and easy to miss.
The steady state is where it hurts. At 10% churn the base converges to 10 divided by 0.10 — 100 customers, $2,000 a month. Half the income, permanently, from a difference that cost you only 15% in year one.
This is why churn deserves more of your attention than commission rate. A five-point difference in commission percentage changes your income by a fixed proportion forever. A five-point difference in monthly churn changes the size of the asset you are building. It also explains why a recurring layer works best when paired with something that pays immediately, which is the reasoning behind running a two-layer high-ticket and recurring stack rather than betting everything on the tail.
Most SaaS products discount annual billing by roughly two months. A $100 monthly plan becomes $1,000 a year. At 20% commission that is $200 paid upfront against $240 spread over twelve monthly payments — if the customer survives all twelve.
At 5% monthly churn they probably do not. The expected value of twelve monthly $20 payments at 5% churn works out to roughly $180, below the $200 the annual plan pays immediately. At 10% churn it drops to around $142. The annual plan wins on both realised value and cash timing.
Annual customers also churn less at renewal than monthly customers do month to month, because the decision is made once a year rather than twelve times, and because the upfront payment creates commitment. If a program lets you influence plan selection — through your comparison content, your bonus structure, or simply which plan your link lands on — steering toward annual is usually the highest-return thing you can do that does not involve more traffic.
The catch is the refund window. Annual plans often carry 30-day money-back guarantees with full commission clawback, so your payment is held longer. Check whether annual commissions are paid on a delay and whether a mid-year cancellation triggers a proportional reversal.
Work through this before spending anything. Most of it is a single conversation with the affiliate manager.
Beyond the manager conversation, look at the product itself. Read recent public reviews for a pattern of cancellations. Check whether the company ships — a stagnant product churns. Look at pricing history, because a company that repeatedly raises prices on existing customers generates churn you will absorb. These signals are imperfect but they are available to you before you spend a dollar, and they correlate with retention better than anything on the rate card.
The same evaluation logic transfers cleanly to adjacent tech verticals — the retention questions you would ask a project management tool are the same ones worth asking an AI tool program or a hosting provider, where recurring structures are also standard.
Modelling lifetime revenue on a capped program. Building a two-year forecast on a 12-month commission window is the most expensive reading error in this category.
Treating month-three revenue as the answer. Recurring programs look mediocre for the better part of a year by design. Judging one at month three guarantees you kill things that were about to work.
Ignoring the plan-value lock. Promoting a product whose customers routinely triple their seat count, on a program that pays you on the original plan, means watching your best referrals grow without you.
Assuming last-click protects you. If your model is educational content that sends buyers off to research further, last-click attribution will hand a meaningful share of your work to whoever they touch last.
Not tracking cohort-level retention on your own side. Your commission statement contains the data to calculate your actual churn per traffic source. Affiliates who track retention as a first-class KPI discover that some sources produce customers who stay twice as long as others, which is a bigger optimisation than anything at the click level.
Recurring income is only as good as the tracking underneath it, because a broken postback on month one costs you every month afterward. We build and maintain that infrastructure for partners at no cost, along with the sites and landing pages, so a comparison or review asset can be live and tracked without you touching a server. Partners keep 50% revenue share, and our direct relationships across tech, high-ticket and our other verticals mean we can push for longer cookie windows, commission that follows plan upgrades, and payout bumps that individual affiliates rarely get offered. Where a program's retention does not stand up, we would rather tell you before you build six months of content around it.
Everything runs through Telegram, with a dedicated manager rather than a ticket queue. Start with the Telegram bot, or join the community chat to see what partners are running in the recurring space right now.
Some do, many do not. Lifetime recurring means you earn for as long as the customer pays, which is the strongest structure available to affiliates. More common is a capped window of 12 or 24 months, after which payments stop. A third pattern pays a large one-time bounty instead. Check which one applies before modelling any income, because the difference compounds enormously.
Thirty days is the common baseline and 60 to 90 days is a meaningful advantage. SaaS buyers frequently research across several sessions and devices, then sign up for a free trial before converting weeks later. The more relevant question is whether attribution is anchored at trial start or at first payment, since a 30-day cookie plus a 14-day trial can expire before money changes hands.
Churn sets your ceiling. At 5 percent monthly churn the average customer pays for about 20 months; at 10 percent, about 10 months. Over a single year the two look similar because new signups mask the losses. At steady state the low-churn program pays roughly double. Churn is the single most important number in any recurring program.
Only if you can wait and only if retention is good. A recurring structure typically takes ten to twelve months to overtake an equivalent one-time payout, and it never overtakes at all if churn is high enough. One-time bounties give faster cash flow to reinvest in traffic. Many affiliates run both structures deliberately for that reason.
Four-figure commissions arrive at random. Subscription commissions arrive every month. Run both and the random layer stops deciding whether you can keep going. Here is the math behind it.
Read postFour sales a month can beat two hundred. Here is what qualifies as high-ticket, how commissions get attributed when a salesperson closes weeks later, and the math that makes it work.
Read postThe fastest-growing tech sub-vertical is also the least stable. How to price churn, spot products with staying power, and build content that outlives the offer.
Read post