How Affiliate Tracking Works
Affiliate tracking records which partner referred each customer so the right partner is credited and paid. It typically uses unique links that set a cookie or pass an identifier, referral or coupon codes entered at checkout, and server-side events that tie the referral to a customer account. Durable tracking binds the partner to the account and follows later payments.
Step 1: the tracking link
Each partner receives a link containing an identifier, usually a parameter added to your URL. When a visitor clicks it, your site or tracking software reads the identifier and stores it. The link may also pass through a redirect on a tracking domain that records the click before sending the visitor on. The click alone earns nothing; it simply records that this visitor came from this partner.
Step 2: storing the referral
Traditionally, the identifier is stored in a browser cookie with an expiry, often called the cookie window or attribution window. If the visitor converts before the cookie expires, the partner is credited. First-party cookies, set by your own domain, are more reliable than third-party cookies set by a tracking network, which many browsers now restrict. Local storage and URL parameters carried through the signup flow are other options.
Step 3: recording the conversion
When the visitor signs up or buys, the stored identifier is attached to the conversion. In client-side tracking, a script or pixel on the confirmation page reads the cookie and reports the conversion. In server-side tracking, your backend sends the conversion event, with the partner identifier, directly to the tracking system. Server-side tracking is more reliable because it does not depend on scripts running in the browser or surviving ad blockers.
Referral and coupon codes
Codes give partners a way to be credited without any cookie. The buyer enters a partner's code at signup or checkout, often in exchange for a discount. Codes work well for podcasts, video, events, and word of mouth, where clicking a link is awkward. Their weakness is leakage: codes shared on coupon sites can be used by buyers who never encountered the partner, which is why some programs restrict discount codes or monitor where they appear.
Account-level binding
For subscriptions and B2B products, the most durable approach is to bind the partner to the customer account when it is created. The referral identifier, from a link or code, is saved on the account, and every later payment, renewal, upgrade, or expansion is credited to that partner by looking up the account. This removes dependence on cookies after signup, ends disputes about attribution windows for renewals, and makes recurring commission accurate.
Where tracking fails
Common failure points include cookies blocked or cleared, buyers switching devices between clicking and buying, redirects stripped by browsers or email clients, checkout flows on a different domain that lose the parameter, and payments processed outside the tracked flow, such as invoices for enterprise deals. Each failure means a partner is not credited, which damages trust. Testing the full path, from link to payment, for every product and plan catches most of these issues.
Connecting tracking to billing
Tracking is most accurate when conversion and payment events come from the billing system itself. Sending payment events from a billing provider, with the customer account and amount, lets the tracking system credit the partner for real revenue and adjust for refunds automatically. Idempotency keys on each event prevent retries or replays from counting the same payment twice.
Privacy and consent
Tracking must comply with privacy and consent rules in the regions you operate in. That can mean asking for consent before setting non-essential cookies, describing referral tracking in your privacy policy, and minimising the personal data shared with partners. First-party, server-side approaches that store only the referral identifier on the account tend to need less personal data than third-party tracking networks.
Tracking for subscriptions and renewals
Subscription businesses need tracking that lasts as long as the customer. If a program pays recurring commission or ranks partners on lifetime revenue, every renewal, upgrade, and expansion payment must be attributed to the original partner. That is only practical when the partner is stored on the customer account and payment events reference the account. Refunds and downgrades should flow through the same path so that commissions and partner metrics adjust automatically rather than by manual correction.
Testing your tracking
Before launch and after any change to signup, checkout, or billing, run test referrals end to end: click a partner link, sign up, pay, upgrade, refund, and confirm each event appears against the right partner. Test on different browsers and devices, with ad blockers enabled, and through every payment path, including invoices and sales-assisted deals. Give partners a way to report missing credit, and investigate reports promptly.
Frequently asked questions
- How long do affiliate cookies last?
- It varies by program. Cookie windows can range from a single day to several months, and the program terms should state the length. Longer windows favour partners whose audiences take time to decide. Account-level binding removes the window problem for repeat payments once the customer has signed up.
- What is server-side affiliate tracking?
- Server-side tracking sends conversion or payment events from your backend directly to the tracking system, instead of relying on a script running in the buyer's browser. It is more resistant to ad blockers and cookie restrictions and can include accurate revenue data from your billing system.
- How do I track affiliate sales without cookies?
- Use referral or coupon codes entered at checkout, carry the referral identifier through the signup flow and save it on the customer account, and send conversion events from your server or billing system. Binding the partner to the account at signup lets every later payment be credited without any cookie.