How does GA4 detect referral traffic?

GA4 reads the HTTP referrer header sent by the browser. If it points to a domain that isn't classified as a search engine or social platform, GA4 labels the session Referral with that domain as the source.

Why does Stripe or PayPal show up as a referral source in GA4?

When a customer returns from a payment processor's checkout page, GA4 can see that as a new referral session unless the processor's domain is added to your GA4 referral exclusion list, which resets attribution to the payment domain instead of the original source.

What is the GA4 referral exclusion list?

It's a setting in GA4 Admin (under your data stream's configuration) that tells GA4 to ignore specific domains when deciding whether returning traffic should start a new session, preventing false referrals from payment processors or partner sites.

What causes self-referrals in GA4?

Self-referrals usually happen when a business runs multiple subdomains that aren't configured for cross-domain measurement in the same GA4 property, so traffic moving between them gets misread as an external referral.

What Is Referral Traffic in Google Analytics?

Referral traffic in Google Analytics is visits that arrive by clicking a link on another website — a blog post, a partner site, a forum, a directory listing. GA4 detects it by reading the HTTP referrer header the browser sends along with the request. If that header points to a domain other than your own and isn't classified as search or social, GA4 labels it Referral.

How GA4 detects a referral

When a browser follows a link from one site to another, it typically sends an HTTP referrer header identifying the page the visitor came from. GA4 reads that header, checks it against its list of known search engines and social platforms, and if it doesn't match either, files the session under Referral with the sending domain as the source.

This is the same underlying mechanism used for organic social and organic search — the difference is purely which list the domain matches. A link from a blog isn't on either list, so it defaults to Referral.

Why referral exclusions exist

Referral detection has a blind spot: it can't tell the difference between "a visitor discovered you through another site" and "a visitor left your site briefly and came back." Both look identical to the referrer header — a request arriving with a foreign domain as the referrer.

This matters most during checkout and payment flows. If a customer goes from your site to a payment processor like Stripe, PayPal, or Braintree to complete a purchase, then gets redirected back to your confirmation page, GA4 sees that return trip as a new referral session from stripe.com or paypal.com. That resets session attribution, wrongly crediting the payment processor as the traffic source and erasing whatever channel actually brought the customer to you in the first place. It also inflates referral counts while undercounting the real source.

GA4's Referral exclusion list, set in Admin under Data streams, tells GA4 to ignore specific domains when deciding whether to start a new session. Traffic bouncing back from an excluded domain continues the original session instead of starting a fresh Referral one.

What's excluded by default, and what to add

GA4 automatically excludes your own domain (and subdomains, if configured) from referral exclusions, since internal navigation shouldn't count as a referral. It does not automatically exclude third-party payment processors or partner checkout domains — you have to add those yourself.

Add any third-party domain your checkout, auth, or payment flow redirects through: Stripe, PayPal, Braintree, Shopify's hosted checkout, an SSO provider, or a partner's subdomain used for a hosted form. If a domain is part of your functional flow rather than a genuine external referrer, it belongs on the exclusion list.

Self-referrals from uncovered subdomains

A related issue: if your business spans multiple subdomains (app.example.com, shop.example.com, blog.example.com) and they aren't all configured under the same GA4 property with cross-domain measurement enabled, traffic moving between them can register as a self-referral — your own site referring to itself. Cross-domain measurement settings in the data stream configuration fix this by treating those subdomains as one continuous experience instead of separate referral sources. This is distinct from direct traffic, which reflects missing data rather than mislabeled internal navigation.

Gizmo vs Plausible

FAQ

How does GA4 detect referral traffic?: GA4 reads the HTTP referrer header sent by the browser. If it points to a domain that isn't classified as a search engine or social platform, GA4 labels the session Referral with that domain as the source.
Why does Stripe or PayPal show up as a referral source in GA4?: When a customer returns from a payment processor's checkout page, GA4 can see that as a new referral session unless the processor's domain is added to your GA4 referral exclusion list, which resets attribution to the payment domain instead of the original source.
What is the GA4 referral exclusion list?: It's a setting in GA4 Admin (under your data stream's configuration) that tells GA4 to ignore specific domains when deciding whether returning traffic should start a new session, preventing false referrals from payment processors or partner sites.
What causes self-referrals in GA4?: Self-referrals usually happen when a business runs multiple subdomains that aren't configured for cross-domain measurement in the same GA4 property, so traffic moving between them gets misread as an external referral.

If you're here because GA4 is confusing — that's why we built Gizmo.

A dead-simple, cookieless Google Analytics alternative. Free forever, 10k events/mo.

Try Gizmo free See the GA4 alternative