How to Build an Affiliate Program for Your Online Course in 2026: The Commission Math That Actually Attracts Partners

Most course creators treat affiliates as free money: sign up a few friends, hand out a coupon code, wait for sales that never come. Then they conclude “affiliates don’t work for courses.” What actually failed was the math. An affiliate program is not a marketing channel you switch on — it is a second pricing model layered on top of your first one, and if the numbers don’t work for the partner, no amount of enthusiasm will move your course. This is the commission math nobody runs before they launch, and the operational reality that decides whether partners promote you once or forever.

Why “just give them 30%” is the wrong starting point

The default advice is to copy whatever percentage you saw on someone else’s affiliate page — usually 30% or 40%. But a percentage is meaningless without knowing your unit economics. A 40% commission on a course with a 20% refund rate and a paid-ads-funded affiliate is a program that loses money on every sale. The right way to think about it is contribution margin: what is left after refunds, payment processing, platform fees, and your own cost to deliver the course, before you decide what you can afford to give away.

Run the numbers on a concrete example. Say your course sells for $300. Payment processing takes roughly 3% ($9). Your platform’s transaction and pricing structure matters here — a subscription LMS with per-transaction fees behaves very differently from a pay-once checkout tool. Assume a blended 5% platform cost ($15). Refunds run 8%, so on average you lose $24 per sale to returns. Your delivery cost — support, community moderation, hosting — is maybe $10 per student. That leaves you around $242 of contribution margin on a $300 sale before any affiliate commission.

Now a 40% commission is $120. That still leaves you $122, which is healthy. But the same 40% on a $97 course with a 15% refund rate leaves you almost nothing. The percentage didn’t change; the math underneath it did. Your commission rate should be derived from margin, not borrowed from a competitor.

The number affiliates actually care about (and it isn’t your rate)

Here is the counterintuitive part. Serious affiliates — the ones with an audience worth having — do not choose programs by commission percentage. They choose by earnings per click (EPC): how much money they make, on average, every time they send one visitor to your page. EPC is your commission multiplied by your conversion rate. A 50% commission that converts at 1% produces a lower EPC than a 30% commission that converts at 4%.

This flips the usual creator anxiety on its head. You do not win affiliates by being the most generous; you win them by having the best-converting offer. A course with a strong webinar or launch sequence that converts cold traffic at 4-6% is far more attractive to a partner than a cheaper course that converts at 1%, even at half the commission. When you pitch affiliates, lead with your conversion data and average order value, not your rate. If you don’t know your conversion rate yet, you are not ready to recruit affiliates — you are ready to fix your funnel.

What a “good” EPC looks like

There is no universal benchmark because it depends entirely on how cold the traffic is, but a rough frame: for warm email traffic to an evergreen course, an EPC above $1 will keep most affiliates interested; above $2-3 and you’ll have partners competing to promote you. Below $0.50 and you are asking people to work for free. Calculate it honestly using your real numbers, because affiliates will calculate it themselves after their first promotion — and if their actual EPC undershoots what you implied, they never come back.

Cookie windows, attribution, and the fights you can avoid

The single most common source of affiliate resentment is attribution. Someone promotes your course, a buyer clicks their link, then buys three weeks later after seeing your own email — who gets paid? If your cookie window is 24 hours, the affiliate gets nothing and feels cheated. If it’s 365 days, you may pay commission on sales your own marketing closed.

The pragmatic middle for courses is a 30 to 60 day cookie with last-click attribution, disclosed plainly up front. Longer windows (90+ days) are a genuine recruiting advantage for high-consideration courses where buyers deliberate, and they cost you less than you fear because most conversions happen within the first week anyway. The real rule is not the specific number — it’s that you write it down, publish it, and never quietly change it after affiliates have started sending traffic. Attribution disputes kill programs faster than low commissions do.

Self-referral and coupon leakage

Two leaks quietly drain course affiliate programs. First, affiliates buying through their own link to pocket the commission — decide whether you allow it (many creators do, treating it as a launch discount) and state it. Second, coupon-code affiliates whose codes end up on deal-aggregator sites, so you pay commission on buyers who were already going to purchase at full price. If you use codes, keep them unique and un-guessable, and monitor where they surface.

Who to recruit — and the order that works

The instinct is to open the program to everyone. That produces a hundred affiliates who each send two clicks and generate support tickets. In practice, roughly 90% of affiliate revenue for a small course comes from a handful of partners. Your job is to find those few, not to maximize signups.

Start with your own successful students. Someone who finished your course and got a result is a more credible promoter than any influencer, and they already understand the value. Reach into your existing email list and offer the program first to graduates. Second, look for creators whose audience overlaps yours but whose product doesn’t compete — the person teaching an adjacent skill is your ideal partner. Recruit these individually, with their EPC estimate in hand, before you ever build a public “become an affiliate” page. A public page is the last step, not the first.

The operational cost creators forget

An affiliate program is not passive. You are now running a small partner business: onboarding, giving partners swipe copy and assets so they don’t misrepresent your course, paying on time, handling refund clawbacks, and answering “why wasn’t I credited” emails. Budget real time for this. The programs that fail aren’t the ones with the wrong commission — they’re the ones where the creator recruited partners, then went silent, leaving affiliates with no assets, no updates, and payouts that arrive late. Partners promote whoever makes them look good and pays reliably. Be that person and a modest program compounds; neglect it and even a generous one dies.

A simple launch sequence that respects the math

Put it together and the path is straightforward. First, calculate your contribution margin and set a commission you can defend. Second, measure your real conversion rate and average order value so you can quote an honest EPC. Third, write your terms — commission, cookie window, attribution, self-referral, payout schedule — and publish them once. Fourth, recruit five to ten aligned partners by hand, arming each with assets and their expected earnings. Fifth, support them relentlessly through their first promotion and pay early. Only then, if it’s working, open a public page. The creators who do this in order build a channel that quietly outperforms paid ads for years. The ones who skip to “just give them 30%” spend six months concluding, wrongly, that affiliates don’t work.

Frequently asked questions

What commission rate should I offer affiliates for an online course?

Derive it from contribution margin, not a fixed percentage. Subtract payment processing, platform fees, average refund losses, and delivery cost from your price, then decide how much of what remains you can share. For many digital courses this lands between 30% and 50%, but the correct number is whatever leaves you a healthy margin after refunds — a rate that loses money at scale is worse than no program.

What’s a good cookie window for a course affiliate program?

Thirty to sixty days with last-click attribution suits most courses. Longer windows of 90 days or more can be a recruiting advantage for high-consideration purchases and cost less than expected, since most conversions happen within the first week. The key is disclosing it clearly and never changing it retroactively.

How many affiliates do I actually need?

Far fewer than you think. For most small courses, roughly 90% of affiliate revenue comes from a handful of committed partners. Recruiting five to ten aligned promoters and supporting them well beats opening the floodgates to hundreds of one-click affiliates who generate support work and little revenue.

Why won’t affiliates promote my course even with a high commission?

Because affiliates choose programs by earnings per click, not commission rate. EPC is your commission times your conversion rate, so a generous rate on a poorly converting offer still pays them little. Fix your conversion rate and average order value first; a strong-converting funnel at 30% beats a weak one at 50% every time.

Should I let affiliates buy through their own link?

That’s your call, but decide and state it explicitly. Many creators allow self-referral and treat it as an informal launch discount; others prohibit it. Ambiguity is the problem, not the policy itself — write it into your terms so no one feels misled after the fact.