
Affiliate platforms: questions to ask before you migrate (SaaS)
A practical checklist for migrating affiliate platforms in SaaS: what to export, how to compare ledgers, how to handle refunds and clawbacks, and the safest cutover plan.
Read articleA practical guide for SaaS founders who want clean payouts and fewer disputes

This is for SaaS founders using Paddle who want an affiliate program that is easy to run, easy to audit, and does not turn refunds into a mess. You will get a clear decision tree, a setup checklist, and copy/paste policy text.
Paddle affiliate tracking means you can attribute a purchase to a partner and then calculate what you owe them. In practice, you are choosing how the affiliate gets credit (link, coupon, or both), how you store attribution, and how you handle refunds and chargebacks so payouts stay fair.
If you get these rules wrong, the program rarely explodes on day one. It becomes a slow leak. You spend more time double checking payouts than recruiting affiliates.
Below are four common approaches. None is perfect. The goal is to pick one that matches your stage.
Who it fits: you mainly work with creators who share a discount code.
What breaks first: coupon leakage and lost credit when customers buy without the code.
Main risk: you end up paying for deals you did not want to discount.
Who it fits: you want clean attribution without discounting.
What breaks first: cross device and cookie resets reduce credit.
Main risk: you under credit partners and the program feels unfair.
Who it fits: most early programs.
What breaks first: you need a clear rule for conflicts (link vs coupon).
Main risk: confusion if you do not document the rule.
Who it fits: you are under 20 active affiliates and want full control.
What breaks first: refunds complicate history and reporting becomes slow.
Main risk: you stay manual too long and stop growing.
This keeps the program simple, reduces coupon leakage, and still lets you work with partners who require a code.
Primary attribution: We attribute a conversion to the last affiliate link click within 30 days.
Coupon fallback: If a coupon code is used, we attribute the conversion to the coupon owner unless a different affiliate link click happened within the last 24 hours.
Refunds and chargebacks: If a purchase is refunded or charged back, the related commission is canceled. If we already paid the commission, we may deduct it from a future payout.
Payout threshold: We pay commissions once an affiliate's net payable balance reaches $50.
Net payable balance: Net payable balance is calculated after refunds, chargebacks, and adjustments.
Be specific. Ambiguity causes payout disputes.
Start with link attribution. Add coupon fallback only if you need it.
Yes. The question is which attribution method you choose and how you handle refunds.
Use coupons only when you need them. They help creators, but they increase leakage.
For most small programs, $25 to $100 is common. $50 is a solid default.
Cancel the commission on refund. If you already paid it, claw it back from the next payout.
Not at the start. Manual payouts keep you in control while you validate the program.
Paddle affiliate tracking works best when you keep it simple. Pick one attribution method, write the rules down, and build a payout process you can audit.
Next step today: write your link vs coupon rule, set a payout threshold, and run a 2 week pilot with 5 to 10 affiliates.
If you want a simple affiliate program that stays auditable as you grow, TinyAffiliate focuses on tracking and payout ops without forcing automatic payouts.
I share practical notes on affiliate programs for SaaS.No spam. No hype.
Unsubscribe anytime. No spam.
Ready to launch?
If Rewardful feels like overkill, start simple: signup page + links + Stripe-attributed revenue.

A practical checklist for migrating affiliate platforms in SaaS: what to export, how to compare ledgers, how to handle refunds and clawbacks, and the safest cutover plan.
Read articleA founder-friendly guide to affiliate tracking for SaaS subscriptions: which event earns commission, how to handle trials and plan changes, how recurring commissions work, and the tests that catch broken attribution.
Read article