Why can an assistant referral appear without a referrer?

The Citead team, updated

In short. An assistant referral can lack a referrer because of browser policy, while surviving URL campaign parameters can still supply an analytics source.

Definition: An assistant referral can arrive without a browser referrer when the linking context suppresses the Referer HTTP header, which carries information about the referring page. Analytics can still identify a traffic source from campaign parameters in the destination URL.

Key facts

What can remove or shorten the referrer?

A referrer policy controls the information sent in the Referer header when a browser follows a link. The W3C editor's draft describes policies delivered through an HTTP response header, an HTML meta element or a link's referrerpolicy attribute. A link with the noreferrer relation can also suppress the header.

The specification gives this response header as an example:

Referrer-Policy: no-referrer

On a referring document, this policy suppresses referrer information for requests made from that document's context. A destination website's policy governs requests initiated from its own context; investigating an incoming click therefore starts with the linking context.

The policy strict-origin-when-cross-origin retains less information for a link between different origins. In the specification's example, a link from https://example.com/page.html to https://not.example.com/ sends https://example.com/ as the referrer. The path disappears, while the referring site remains visible. A link from that same HTTPS page to the example's HTTP destination sends no referrer.

How can analytics show a source without a referrer?

UTM parameters are campaign labels added to a destination URL. Google's manual tagging documentation explains that their values populate traffic-source dimensions. Consequently, a source label can come from the URL even when the browser sends no referrer. That label records the supplied parameter value.

Missing attribution has other causes. Google's direct traffic guidance says redirects can strip UTM parameters and URL shorteners can remove referral details. It also lists typed addresses, links from offline documents and interference from ad blockers among causes of (direct) / (none) traffic. A direct traffic row alone cannot identify which assistant, if any, sent the visit.

How should a missing source be checked?

A useful comparison records the referring page's policy and the final destination URL separately. For a controlled link test, the question is whether referrer suppression occurs, whether campaign parameters survive the redirect path, or whether both signals are absent. Google's direct traffic guidance specifically recommends checking that tracking parameters remain intact when visitors arrive.

For session attribution, Google's Traffic acquisition documentation provides the report under Reports > Acquisition > Traffic acquisition. Its Session source / medium dimension describes acquisition for a new session. The report covers new and returning users, while User acquisition focuses on new users. Choosing the report's session dimension keeps the investigation aligned with the visit being examined.

FAQ

Does a missing referrer always erase the source?

No. Google's traffic-source documentation describes manual tagging through destination URL parameters. You can check whether utm_source and utm_medium reach the landing page even when referrer information is absent.

Does an origin-only referrer identify the conversation page?

Under the W3C policy example, the cross-origin referrer becomes https://example.com/ instead of the full referring page URL. You retain the site origin without the page path.

Sources

  1. W3C, "Referrer Policy", editor's draft, 20 March 2026
  2. Google, "[GA4] Understand (direct) / (none) traffic", accessed 17 October 2026
  3. Google, "Traffic-source dimensions, manual tagging, and auto-tagging", accessed 17 October 2026
  4. Google, "[GA4] Traffic acquisition report", accessed 17 October 2026

Check how a page reads to answer engines. Open the GEO score analyzer

Test a page and see what to fix first

The analysis usually takes less than 10 seconds and needs no account.

400,000+ pages scored in the public benchmark