Skip to content
accs-net.com

Press Esc to close

Referral

Referral traffic in GA4 is any visit that arrives by clicking a link on another website β€” when no UTM tag, no gclid, and no known search/social signal applies. The source is the referring hostname and the medium is referral. This guide covers how Google Analytics 4 identifies referral sources, where to find them in reports, the source/medium distinction across channels, the self-referral problem and the exclusion list fix, referral spam, top sources to pursue, and conversion tracking β€” plus a FAQ that answers the questions people actually ask about GA4 referral traffic.

What Is Referral Traffic in GA4?

In Google Analytics 4, referral traffic is the channel assigned to a visit when the browser sends a Referer header from an external domain and that domain isn’t recognised as a search engine, a social network, or your own site. Think GitHub README links, blog mentions, partner directories, podcast show notes, and Wikipedia citations β€” any context where one website hands a visitor to another.

Two short examples make the distinction concrete:

  • A reader on a tech blog clicks an inline link to your tutorial. No UTM parameters, no ad. GA4 records session_source = blogdomain.com, session_medium = referral.
  • That same blog tags its links with ?utm_medium=email for newsletter sends. GA4 reads the UTM, ignores the Referer, and assigns the visit to Email instead of Referral.

Referral is the default fallback for any external visit that GA4 can’t bucket more specifically. That makes it both a useful signal of organic mentions and a catch-all that can hide misclassified social, email, or paid traffic when tagging is sloppy.

How GA4 Identifies Referral Traffic (utm_medium=referral)

GA4 runs each incoming visit through a short decision sequence to assign source and medium. The order matters, because earlier rules override later ones:

  1. UTM parameters or gclid. If the URL contains utm_medium, utm_source, or a Google Ads gclid, GA4 uses those values and ignores the Referer header entirely.
  2. Search-engine domain match. If no UTM is present and the Referer matches Google’s known search-engine list (google.*, bing.com, duckduckgo.com, yandex.*, baidu.com, ecosia.org, brave.com, etc.), the visit becomes Organic Search.
  3. Social network match. If the Referer matches the social-source list (facebook, instagram, x.com/twitter, linkedin, reddit, tiktok, pinterest, etc.), the visit is Organic Social.
  4. Referral exclusion list. If the Referer matches a domain in your property’s exclusion list, the Referer is dropped and the visit becomes Direct (or whatever earlier signal exists from a previous touchpoint in the same session).
  5. Otherwise β€” Referral. Any external Referer that survives the previous steps is classified as referral, with session_source = the hostname and session_medium = referral.
GA4 referral traffic decision flow: visitor arrives, GA4 checks gclid/utm parameters first, then the Referer header, then matches against search-engine, social, and exclusion-list domains before falling through to referral classification with source as the referring hostname
How GA4 classifies an incoming visit β€” UTMs and gclid evaluated first, Referer fallback last

One consequence worth internalising: any UTM tag wins over Referer. If a partner adds ?utm_source=newsletter&utm_medium=email to their link, GA4 won’t show that visit as referral even if the click came from their website. The medium is whatever the link tells it to be.

Where to Find Referral Traffic in GA4 Reports

GA4 surfaces referral traffic in three main places. Bookmark the first two β€” they answer 80% of the questions analysts get asked about referrals.

  • Reports β†’ Acquisition β†’ Traffic acquisition. Group by Session default channel group and look for the “Referral” row. Click through to break down by session_source to see which domains drive volume.
  • Reports β†’ Acquisition β†’ User acquisition. Same view, but attributed to first-touch instead of last-touch β€” useful for understanding which referral sites discover new users vs reactivate existing ones.
  • Explore β†’ Free Form. Build a custom report with Session source / medium as the row and metrics like engaged sessions, conversions, and revenue. This is how you isolate a single referrer (e.g. news.ycombinator.com / referral) for deep analysis.

For per-page detail, add a Landing page dimension to any of these views. That tells you which URL each referral domain points at β€” essential when you’re tracking a specific PR placement or partnership link.

Referral vs Direct vs Organic vs Social β€” Source Distinction

Mixing these channels up is the single most common GA4 attribution mistake. Here’s the clean reference:

Channel How GA4 detects it Example utm_source / medium Example visit
Referral External Referer Β· no UTM Β· not search/social/excluded github.com / referral Reader clicks a link in a GitHub README
Organic Search Referer matches search-engine list Β· no paid signal google / organic Click from Google SERP free listing
Direct No Referer Β· no UTM Β· or excluded by you (direct) / (none) Typed URL, bookmark, dark traffic
Paid Search gclid or utm_medium = cpc/ppc google / cpc Click on a Google Ads result
Organic Social Referer matches social list Β· no UTM facebook / referral β†’ reclassified to Social Click from a Facebook post
Email utm_medium=email newsletter / email Click from a tagged newsletter
Unassigned None of the rules matched (not set) / (not set) Tagging gap or stripped Referer
GA4 default channel grouping cheat sheet mapping utm_medium values and detection rules to channels: referral, organic search, paid search, organic social, paid social, email, display, affiliates, direct, with example sources and short codes
utm_medium β†’ GA4 channel reference β€” keep this open when auditing campaigns

The practical lesson: if you tag a link, you take responsibility for its channel. An untagged link from a partner shows up correctly as Referral. The same link with utm_medium=partner ends up in Unassigned because “partner” isn’t in GA4’s default channel-group rules. Either tag with a campaign medium GA4 recognises (referral, email, cpc, social, affiliate), or don’t tag at all.

The Self-Referral Problem and How to Fix It

A self-referral is when your own domain appears as the source of a visit. It happens because the user crossed a session boundary on a path that involves your own site β€” typically:

  • A third-party checkout (Stripe Checkout, PayPal, Shopify Payments) that redirects back to your domain after payment.
  • An auth flow (Google login, Auth0, Okta) that bounces the user through an external host.
  • A subdomain hop without proper cross-domain tracking (e.g. blog.example.com β†’ shop.example.com).
  • A redirect that drops UTMs (HTTP β†’ HTTPS, www β†’ non-www, or marketing redirector services).

The fix has two parts. First, add your own domain β€” and any payment/auth providers that bounce traffic β€” to GA4’s unwanted referrals list. Go to Admin β†’ Data Streams β†’ choose stream β†’ Configure tag settings β†’ Show all β†’ List unwanted referrals. Add example.com, checkout.stripe.com, paypal.com, and any other domains that should not start a new session.

Second, fix the underlying tagging. For multi-domain setups, configure cross-domain tracking under the same screen. For redirects, preserve query parameters end-to-end so UTM tags reach the final landing page intact. The exclusion list hides the symptom; cross-domain tracking and redirect hygiene fix the cause.

Quick check: Open Reports β†’ Realtime, click your own site link from another browser, and verify it shows up correctly. If your own hostname appears in the source dimension, the exclusion list is missing or misconfigured.

Referral Spam β€” What It Is and How to Block It

Referral spam is fake traffic created by bots that send requests with a forged Referer header. They rarely render JavaScript, so GA4 sees most of it as bot traffic and filters it automatically β€” but a small share still slips through, especially on low-traffic properties. Symptoms: a sudden spike of zero-engagement sessions from an unfamiliar domain (often buttons-for-website.com, semalt.com, trafficbot.life, etc.).

Three layers of defence work in combination:

  1. Rely on GA4’s bot filter. It’s enabled by default and updated continuously by Google β€” covers the IAB/ABC spider list and most known bots. No setup needed.
  2. Add spam domains to the unwanted referrals list. Same screen as the self-referral fix. New spam will reappear under different domains, so this is whack-a-mole β€” only worth doing for repeat offenders.
  3. Filter at the report layer. In any explore, exclude sessions where the source matches a regex of known spam patterns. Cleaner than altering raw data and easy to adjust.

Don’t waste time on legacy “ghost referral spam” tactics from Universal Analytics β€” GA4’s measurement protocol requires a valid measurement ID and is much harder to spoof at scale.

Top Referral Traffic Sources to Pursue

Referral isn’t passive. Three categories produce most of the high-quality referral volume on B2B and content sites I’ve worked with:

  • Editorial mentions and PR. Industry publications, niche blogs, podcasts with show-notes pages, “best tools” roundups. Pitch with a clear angle and offer a free resource the writer can link to.
  • Partner placements. Integration directories (Zapier, Slack App Directory, Shopify App Store), partner pages on complementary tools, co-marketing landing pages. These compound β€” each placement is evergreen until the relationship ends.
  • Content syndication and community. GitHub READMEs, Stack Overflow answers, Reddit communities, Hacker News submissions, Wikipedia citations. Lower predictability per item but huge upside on the hits.

Track each placement as its own row in Explore by combining Source with Landing page. For paid or co-marketed placements where you want full attribution control, tag the link with utm_source=partnername&utm_medium=referral&utm_campaign=2026-q2 using a consistent UTM parameters that actually track convention. The medium stays “referral” so GA4 keeps it in the right channel.

How to Track Referral Conversions in GA4

Referral conversions tell you which sites send visitors who actually convert β€” not just who clicks. Two reports surface this directly:

  • Reports β†’ Acquisition β†’ Traffic acquisition. Add a Conversions column for your key event (purchase, signup, lead). Sort by conversions to rank referrers by business value, not just session count.
  • Explore β†’ Free Form. Rows: Session source. Metrics: Sessions, Engaged sessions, Conversions, Total revenue. Filter by Session medium = referral. This is your referral-quality scoreboard.

For attribution context, switch the model in Advertising β†’ Attribution settings between Last click and Data-driven. Last-click overcredits the final referrer; data-driven distributes credit across the path and usually shifts value toward earlier referral discoveries. Compare both before reporting numbers to stakeholders.

One pitfall to watch: GA4 reports by default attribute the entire session to the source that opened it, even if the conversion happened minutes later from a different path. If a user lands via referral, leaves, and returns via Direct to convert, the conversion may credit Direct on a last-click model. Cross-reference with Conversion paths under Advertising for the full picture.

Frequently Asked Questions

What counts as referral traffic in GA4?

Any visit where the browser sends a Referer header from an external domain that isn’t a recognised search engine, isn’t a recognised social network, and isn’t on your unwanted-referrals list β€” and where no UTM tag or gclid overrides the channel. GA4 then sets session_source to the referring host and session_medium to referral.

Is social media referral or social traffic in GA4?

It’s classified as Organic Social, not Referral. GA4 maintains a default list of social-network domains (facebook, x.com, linkedin, instagram, reddit, tiktok, etc.) and reclassifies any matching Referer to Social. If a social domain isn’t on the list, you’ll see it under Referral instead β€” that’s the signal to add it as a custom channel-group rule.

What is unassigned in GA4 traffic?

“(unassigned)” appears when GA4 receives a visit with a source/medium combination that doesn’t match any default channel-group rule β€” for example, utm_medium=partner or utm_medium=newsletter. Either change the medium to a recognised value (referral, email, cpc, social, affiliate) or define a custom channel group under Admin β†’ Data display β†’ Channel groups.

How do you find referral traffic in GA4?

Open Reports β†’ Acquisition β†’ Traffic acquisition. The default view groups by Session default channel group; locate the Referral row. Click the row name to break down by session_source and see individual referring domains. For deeper analysis, build a Free Form exploration with Source/Medium as the row dimension and filter Session medium = referral.

How do I fix self-referrals in GA4?

Add your own domain and any payment/auth providers that bounce traffic (Stripe, PayPal, Auth0, Okta) to the unwanted-referrals list under Admin β†’ Data Streams β†’ Configure tag settings β†’ List unwanted referrals. For multi-domain setups, also configure cross-domain tracking on the same screen. Verify the fix in Realtime by clicking your site from another tab and confirming your domain no longer appears as a source.

What is referral spam in GA4?

Fake traffic from bots that send forged Referer headers to inflate numbers or trick admins into visiting spammy domains. GA4’s built-in bot filter catches most of it automatically. For repeat offenders, add the spam domain to the unwanted-referrals list, or filter at the report level using a regex exclusion on Session source.

How do you get more referral traffic?

Three reliable channels: editorial PR (industry publications, “best of” roundups, podcasts), partner directories (integration marketplaces, complementary-tool partner pages), and community placements (GitHub READMEs, Stack Overflow, Reddit, Hacker News). Track each placement by combining Session source with Landing page in Explore β€” this shows which referrers convert, not just which click.

Tom Martin
Written by

Tom Martin

Web analytics specialist with deep expertise in Google Analytics, Tag Manager, and e-commerce tracking. Helping businesses understand their data without the noise β€” practical guides, honest reviews, and real-world implementation experience.