What Is Referral Traffic? Definition, GA4 Rules and Examples

Referral traffic is the visits that reach your site when someone clicks a link on another website, such as a blog post, a news article, a review site, a forum thread or a partner page. Your analytics tool recognizes it from the referrer, the address of the previous page that the browser passes along, and files the visit under the linking domain. Visits from search engines, ads and typed URLs are not referral traffic.

This guide covers how referral tracking works, exactly what Google Analytics 4 counts as Referral (it is narrower than most definitions), where to find the report, why referrers go missing, and how to judge each referring site by the conversions and value it sends, not just the clicks.

What Is Referral Traffic?

Think of referral traffic as word of mouth with a paper trail. Another site decided to link to you, one of its readers clicked, and your analytics recorded who sent them. The site that sent the visitor is the referral source (or referrer). The page they landed on is the landing page.

Three things make a visit a referral in practice:

• The visitor clicked a link on a page outside your own domain.
• The browser sent a referrer that names that outside site, or the link carried campaign tags that say referral.
• The source is not one your analytics tool already sorts into another channel, such as a search engine, a known social network or an AI assistant.

That third point is where most confusion comes from, and it is covered in detail in the GA4 section below.

Examples of Referral Traffic

Social networks and AI assistants also send visitors through links, and some tools report them as referrals. GA4 does not: it gives them their own channels, as explained next.

How Referral Traffic Is Tracked

The HTTP referrer

When a visitor clicks a link, the browser requests your page and usually includes a Referer header (the misspelling is part of the standard) describing where the click came from. Your analytics script reads the same information from the page and records the referring domain as the session’s source, with medium set to referral.

Why you usually see the domain, not the exact page

Modern browsers limit what the referrer reveals. When a site sets no policy, browsers apply strict-origin-when-cross-origin, which sends only the origin (for example https://example.com/) to another site, and sends nothing at all when the click goes from an HTTPS page to an HTTP page, according to MDN’s Referrer-Policy reference. A site can also set no-referrer and send nothing. So “which article on that blog linked to me?” often has no answer in your analytics. You get the domain, and you find the exact page by searching that site or checking a backlink tool.

UTM parameters

When you control the link (a partner page, a guest post byline, a sponsored newsletter), add UTM tags so the visit is labeled no matter what the browser sends:

https://yoursite.com/pricing?utm_source=partnerblog&utm_medium=referral&utm_campaign=integration-page

Tags override the referrer in most tools, so use them consistently. Pick a medium GA4 recognizes: referral for referrals, affiliate for affiliates, email for newsletters. An unrecognized medium such as partner lands in Unassigned.

What Counts as Referral Traffic in GA4

In Google Analytics 4, Referral is one channel in the default channel group, and its rule is short: the medium is referral, app or link. But GA4 checks other rules too, and a link from a known social network, search engine, video site, shopping site or AI assistant is classified by those lists instead. Google publishes the rules in its default channel group documentation.

So if your social posts get clicks, they show under Organic Social, not Referral. And if you are tracking AI assistants, GA4 now reports the assistants on its list in their own channel. Our guide to AI conversion tracking covers what that channel can and cannot see.

Referral vs Organic, Direct and Social Traffic

How to Find Referral Traffic in GA4

1. Go to Reports > Acquisition > Traffic acquisition. This report counts sessions. User acquisition counts only how new users first arrived, so use Traffic acquisition for ongoing referral performance.
2. In the table, the default dimension is Session default channel group. Find the Referral row for total sessions, engagement and key events from referrals.
3. Switch the dimension to Session source / medium and type “referral” in the search box. Each row is now one referring domain.
4. Add Landing page + query string as a secondary dimension to see which of your pages each site sends people to.
5. Pick the key event you care about in the key events column (a lead form, a purchase) and sort by it, not by sessions.

Why Referral Traffic Matters

• It shows which relationships work. Partnerships, guest posts, PR mentions and directory listings either send people or they do not. The referral report tells you which.
• Visitors arrive with context. Someone who clicked a recommendation on a site they trust often knows what you do before they land. Whether that becomes higher conversion is something to check per source, not assume.
• It spreads your risk. A site that relies only on Google is exposed to every ranking change. Referral sources are other doors into the same business.
• The links can help search. Links from relevant sites help search engines discover and evaluate your pages. The visits are a separate benefit; see the section on where sources disagree.

Measure Referral Traffic by Conversions and Value, Not Clicks

A referral report sorted by sessions rewards the sites that send the most people. The useful question is which sites send people who become leads or customers. Give each conversion a value, then compare referrers by value per session:

Illustrative example. A B2B software company assigns $150 to each demo request (see how to calculate conversion value to set your own). Over 30 days, three referrers send:

The blog roundup sends 69% of the referral sessions (1,200 of 1,750) but only 22% of the value ($900 of $4,050). The partner page sends under 9% of sessions and a third of the value. If you had one hour to spend, improving the partner page listing or adding a second partner is worth far more than chasing another roundup. Next, look at the landing pages: if the blog sends readers to an article with no clear next step, a better call to action on that page may lift its numbers.

The same method works for organic search and AI assistants, which is the subject of our pillar guide on SEO conversion tracking.

Why Referrers Go Missing

Plenty of real referrals never show up as Referral. They fall into Direct instead:

• Apps. Clicks from email clients, chat apps (WhatsApp, Slack, Messenger) and many mobile apps usually carry no referrer. This is often called dark social.
• Documents and offline. PDFs, slide decks and QR codes have no previous web page to report.
• Referrer policies. HTTPS to HTTP clicks send nothing under the browser default, and sites that set no-referrer send nothing anywhere.
• Redirect chains. Some link shorteners and redirects lose the original source along the way.

You cannot fix other sites’ settings, but you can tag every link you control with UTMs and serve your own site over HTTPS.

When Referral Numbers Look Wrong: A Debugging Checklist

A sudden referral spike, or a referrer that seems to drive most of your sales, is usually a tracking issue. Check these in order:

For the payment-gateway case, GA4 lets you exclude up to 50 domains per web data stream under Admin > Data collection and modification > Data streams > Web > Configure tag settings > List unwanted referrals, which tells Analytics to ignore that referrer, according to Google’s help page on unwanted referrals. It applies to new data only, so expect old reports to keep the bad credit.

Where the Top Results Disagree

• Is social traffic referral traffic? Some guides say yes, others say it depends on the platform. Both are true for different tools. In GA4 specifically, clicks from sites on Google’s social list go to Organic Social, not Referral. Check how your own tool groups them before comparing numbers.
• Is there a “good” referral share? One popular page calls 7% of traffic the line between good and bad, with no source. We found no published benchmark to back a single number, and the right share depends on your business model. A B2B tool listed in many directories and a local plumber will look nothing alike. Track your own trend and value per session instead.
• Do referral visits improve SEO? Several guides blur the link and the visit. Search engines look at links, not at how many people clicked them, and links marked nofollow, sponsored or ugc may pass little or no ranking value. Measure referral traffic on its own merits.

How to Get More Referral Traffic

• Start with sources that already convert. Use the value table above. More of what works beats new experiments.
• Complete your listings. Review sites, directories, marketplaces and integration pages where buyers compare options.
• Write for other sites. Guest posts, expert quotes and podcast appearances put a link in front of a new audience.
• Publish things worth linking to. Original data, calculators, templates and clear reference guides earn links without outreach.
• Help in communities. Answer questions on forums and communities where your buyers ask them. Link only when it answers the question.
• Fix the landing pages. Check which of your pages referrers point to and make sure each has a clear next step.

FAQ