Most affiliate marketing advice is written by people selling affiliate marketing courses. This guide is written by a team that runs an affiliate program and pays it out every month, so the numbers here are the ones we actually see.
The short version: the type of commission matters more than the percentage. A 50% one-time payout is usually worth less than a 30% recurring one, and almost nobody does that arithmetic before signing up.
One-time vs recurring commission: the arithmetic
A one-time commission pays you once, when the person you referred makes their first payment. A recurring commission pays you every month that person stays subscribed.
Take a $49/month plan and one referral who stays 12 months:
| Program type | Rate | You earn |
|---|---|---|
| One-time | 50% of first month | $24.50 |
| One-time | 100% of first month | $49 |
| Recurring | 30% for 12 months | $176.40 |
The recurring program pays 3.6× more than the "generous" 100% one-time deal. The reason is simple: you are paid on the customer's lifetime, not on their signup.
This is why the headline percentage is a bad way to compare programs. Ask two questions instead:
- For how long does the commission run? One month, 12 months, or for the life of the account?
- What is the average subscription length? A program paying 40% recurring on a product people quit after two months is worth less than 20% on one they keep for two years.
Why programs cap recurring at 12 months
You will see a lot of programs offering "30% recurring for 12 months" rather than lifetime. That cap is not a trick — it is what makes the program survive.
A platform pays for hosting, payment processing, and support every month a customer stays. Lifetime affiliate commissions on a low-margin subscription eventually cost more than the customer brings in, and programs in that position quietly shut down or cut rates retroactively. A capped window is a program you can still count on in year three.
What you actually need to start
You do not need a large audience. You need a specific audience with a specific problem.
The affiliates who earn the most from ecommerce programs in our data are not the ones with the biggest followings. They are:
- People who teach a skill that produces products. Illustrators, letterers, and designers whose students all need somewhere to sell. The recommendation is a natural next step, not an ad.
- People who run a community. A Discord for sneaker customisers, a WhatsApp group for local artisans, a subreddit for a craft. Trust is already there.
- Freelancers who build stores for clients. Every client project is a referral, and the client stays subscribed for years because their business runs on it.
- People who document their own attempt. "I opened a store and here is month one" outperforms "top 10 platforms" because it is verifiable.
What none of them do is publish a listicle comparing 15 platforms they have never used. Those pages exist in the thousands, rank nowhere, and convert at close to zero.
How to pick a program worth your time
Run any program through these five checks before you write a word about it.
1. Cookie window. How long after someone clicks your link do you still get credit? Thirty days is standard. Seven days is a red flag — most people do not buy on first visit.
2. Attribution on self-serve signups. Some programs only credit you if the customer never touches another marketing channel. Read the terms for "last click" versus "first click" and for exclusions.
3. Payout threshold and method. A $100 minimum payout on a program paying $15/month means you wait seven months for your first cheque. Check the method too — if they only pay to a bank account in one country, and you are not in that country, the commission is theoretical.
4. What happens on refunds and chargebacks. Honest programs reverse the commission when a customer refunds. Programs that do not mention it either eat the loss (unsustainable) or will surprise you later.
5. Whether you can actually recommend it. This is the one people skip. If you would not tell a friend to use the product, promoting it costs you the trust that makes every future recommendation work. That trust is the whole asset.
Cartva's program: 30% recurring for 12 months
We pay 30% of every payment for the first 12 months of each account you refer, with refunds and chargebacks reversed rather than absorbed silently. Sign up, get your link, and track clicks and conversions in your own dashboard.
Promoting an ecommerce platform without an audience
The mistake is starting with the product. Start with the problem the product removes.
Someone searching "how do I sell my designs online" does not want a platform comparison. They want to know whether this is possible for them at all. Content that answers the real question converts; content that answers the product question does not.
Four formats that work, in rough order of effort:
Show your own store. Screenshots of a real store you built, with real products and real numbers, however small. This is the single most persuasive asset you can make, and you can make it in an afternoon. Our guide on how to start print on demand walks the process you would be documenting.
Answer the objection nobody else answers. For first-time sellers outside the US, the blocker is rarely the store builder — it is that they cannot form a company or open a payment account. Content that addresses that specific wall gets read to the end, because almost nothing else addresses it.
Compare honestly, including the losses. A comparison that says "use the other one if you need X" is trusted and gets linked to. One that says every feature is better is trusted by nobody. Our Shopify alternatives comparison is written that way on purpose.
Answer questions where they are already asked. Reddit, Quora, Facebook groups, Discord servers. Not with a link dump — with a real answer, and the link only when it genuinely belongs. One good answer in a thread that ranks in Google will send traffic for years.
What realistic earnings look like
Round numbers, on a program paying 30% recurring for 12 months of a $49/month plan (about $14.70 per referral per month):
| Active referrals | Monthly | Over 12 months |
|---|---|---|
| 5 | $73 | $882 |
| 20 | $294 | $3,528 |
| 50 | $735 | $8,820 |
| 200 | $2,940 | $35,280 |
Two things to hold onto. First, this compounds while referrals stay: month six pays you for month six's signups and everyone still active from months one to five. Second, it decays if they churn — which is why promoting a product people abandon is a bad trade even at a high rate.
Nobody reaches 200 referrals in a month. People reach it over two or three years by publishing consistently on one narrow topic.
Mistakes that cost the most
Promoting more than one competing product. Your audience reads it as "paid to say this," and both recommendations lose their weight.
Not disclosing the relationship. In the US the FTC requires it, and beyond compliance, an undisclosed affiliate link discovered later destroys more trust than the commission ever earned. One plain sentence is enough.
Chasing head keywords. "Affiliate marketing" and "best ecommerce platform" are dominated by sites with thousands of backlinks and full-time SEO teams. You will not rank, and the months spent trying are gone. Long, specific questions — "can I sell print on demand without a company", "how do I get paid selling online from Haiti" — have less competition, and the person typing them is far closer to acting.
Quitting at month three. Affiliate income from content is close to zero for months, then compounds. Almost everyone quits in the flat part.
Frequently asked questions
What is a recurring commission in affiliate marketing?
A recurring commission pays you every month that the customer you referred keeps paying, instead of once at signup. On a $49/month plan at 30% recurring for 12 months, one referral is worth about $176 rather than the $24.50 a 50% one-time deal would pay.
Is affiliate marketing still worth it in 2026?
For thin review sites competing on head keywords, no — those pages have been squeezed by both algorithm updates and AI answers. For someone with a specific audience recommending a product they actually use, yes. The shift is away from traffic volume and toward trust with a narrow group.
How much can you realistically earn from an ecommerce affiliate program?
On a 30% recurring program for a $49/month plan, each active referral pays about $14.70 a month. Twenty active referrals is roughly $294 a month; fifty is about $735. Most people who reach those numbers take one to three years of consistent publishing, not weeks.
Do you need a website to do affiliate marketing?
No. A YouTube channel, a newsletter, a Discord community, or answering questions in forums all work. A site helps because search traffic compounds while social posts decay, but it is not a requirement to be accepted into most programs.
How long do affiliate cookies last?
Thirty days is the common standard, meaning you are credited if the person signs up within 30 days of clicking your link. Shorter windows of seven days or less favour the merchant, since most people do not subscribe on their first visit. Always check this before promoting a program.
What happens to my commission if the customer asks for a refund?
In a well-run program the commission is reversed, because the merchant was never paid either. Programs that stay silent on refunds and chargebacks are either absorbing a cost they cannot sustain long-term or will claw it back in a way you did not expect. Cartva reverses the commission and shows it in your dashboard.
Can I join an affiliate program without a registered company?
For most programs, yes — you sign up as an individual and are paid as one. What varies is the payout method: some programs only send money to a bank account in a specific country, which makes the commission unusable if you are elsewhere. Check the payout options before you invest months in promoting.
Getting started
Pick one program you would genuinely recommend. Pick one narrow audience you already belong to. Publish one honest thing a week about the problem they have, not the product you are paid for.
That is the entire method. It is slow, it works, and it keeps working after you stop pushing.
If you want to see how a recurring program is structured in practice, our own terms are on the Cartva affiliate program page — 30% for 12 months, with a dashboard that shows clicks and conversions as they happen.