TL;DR
- Google Search Central added a page on 12 November 2026 on identifying Google Search visits in analytics and server logs by the HTTP Referer header.
- Google Search uses the 'origin' referrer policy, so the Referer holds only scheme and host such as https://www.google.com/, never the path or the query.
- Google Images sends the same Referer values and Discover uses those of the surface it was viewed on, so the Referer cannot split them from web search.
- Google lists four mismatch causes: privacy and consent, on-site redirects that drop Referer or gclid, click-versus-page-load timing, and paid traffic in analytics totals.
- A Search Console logging error lowered generative AI report impressions from 9 to 11 November 2026; Google says it affects data logging only.
Google Search Central published a new page on 12 November 2026, "Identify Google Search traffic in your web analytics or server logs", that explains which HTTP Referer values Google Search sends, why that header is sometimes blank, and four reasons your analytics numbers will not match Search Console. The short answer it gives: Google Search uses the "origin" referrer policy, so the Referer carries only the scheme and host (for example https://www.google.com/), never the path or the search query, and analytics tools count visits differently from the way Search Console counts clicks. Separately, Google's Search Console data anomalies page notes a logging error that lowered impressions in the generative AI performance report from 9 to 11 November 2026.
What Google published on 12 November 2026
The new document sits in the Monitoring and debugging section of Google Search Central documentation: Identify Google Search traffic in your web analytics or server logs. Its footer shows a last-updated date of 12 November 2026, and the Search Central changelog entry reads "Added a new page about identifying traffic from Google Search using HTTP Referer headers", as quoted by Search Engine Roundtable on 12 November. PPC Land reported the changelog describes the purpose as documenting how site owners can identify and track traffic from Google Search in their web analytics systems.
Its sections cover how Google Search sets the Referer header, why the Referer is sometimes blank, how to match the Referer in analytics or logs, and why your data might not match Search Console. None of it introduces a new feature. It is documentation of how existing behaviour works, which is exactly what makes it useful in a reporting dispute.
How Google Search sets the Referer header
According to the Search Central page, Google Search uses the "origin" referrer policy. That means the Referer header your server receives contains only the scheme and host of the referring page. The URL path and any query parameters are left out.
Two consequences follow directly from the page:
- The Referer does not include the search query the user typed. Google points to the Search Console Performance report for query data.
- Because only the host is sent, you identify Google Search traffic by matching the host value, not a path or parameter.
The Referer values to match
The page lists the values each Google surface sends. The table below reproduces them.
| Where the click came from | Referer value Google lists |
|---|---|
| Google Search on the web | https://www.google.com/ or a local Google domain, such as https://www.google.co.uk/ |
| Google app on Android | android-app://com.google.android.googlequicksearchbox/https/www.google.com when the link opens in a browser tab, or https://www.google.com/ or a local Google domain |
| Google app on iOS | https://www.google.com/ or a local Google domain |
| Google Images and Discover | The same values as above: Images uses the same origins, and Discover uses the values of the surface where the user saw it |
The last row matters more than it looks. Google says Images sends the same Referer values, so its traffic is included when you match these patterns, and Discover traffic uses the values of whichever surface the user viewed Discover on. In other words, by Referer alone you cannot separate web search, Images and Discover.
Google also says that to count visits from Google Search, you should only look at requests for the landing pages themselves. Google's reason is that the browser also sends a Referer header when it fetches resources such as images, CSS and JavaScript files, but for those requests the value is your own page's URL.
Why the Referer is sometimes blank
The Search Central page lists four situations in which a visit from Google arrives with no Referer at all:
- Private browsing and anti-fingerprinting protections.
- Opening a result in a new tab, in some browsers.
- Moving from a native mobile search widget to an external browser.
- Privacy extensions or strict browser settings.
Google adds that JavaScript-based tracking might not log every visit, because privacy settings, consent management and ad blockers can stop the script from running. For server logs, the page says to match the Referer field directly. It does not rank logs above JavaScript analytics for completeness, but the practical implication, which is analysis, is that server logs avoid the consent and ad-blocker losses while still suffering from blank Referers.
Four reasons your data won't match Search Console
Google lists four reasons, and each one is a different mechanism.
1. Privacy and client configuration
Browser settings, extensions, ad blockers or consent choices can block analytics scripts, strip query parameters or suppress the Referer header.
2. Redirects on your own site
Google names three redirect problems. Client-side redirects, such as JavaScript or meta refresh, may replace the Referer with your own page's URL, so the visit looks internal. Redirects that do not preserve the query string drop parameters such as gclid, which is commonly used for ad tagging. And a redirect that sends a stricter Referrer-Policy header can remove the Referer entirely.
3. Different measurement methods
Search Console records a click when a user selects a link. Analytics tools record a visit only after the destination page loads and runs the tracking code.
4. Organic versus paid
Search Console reports organic clicks only and excludes Google Ads. Depending on configuration, your analytics system may include paid traffic in its Google totals.
What the page does not cover
The document is about classic Google Search referrers. It does not mention AI Mode, AI Overviews or Google News, and it says nothing about Gemini. It also does not give any expected size for the gap between analytics and Search Console. Google offers reasons, not a tolerance band, so there is no official "normal" difference to quote.
Whether a click from a link inside an AI Mode answer sends the same Referer as a classic result is not stated on this page. Treat any claim either way as unconfirmed until Google documents it.
The 9 to 11 November Search Console data note
In the same week, Google added an entry to the Search Console data anomalies page. It says a logging error caused a decrease in impressions in the generative AI performance report in Search, for data from 9 to 11 November 2026, and that "this issue affects data logging only". Search Engine Roundtable reported it on 12 November.
The timing overlaps the November 2026 core update, which Relevant Audience reported on 11 November 2026 began on 10 November and may take up to two weeks. A drop in generative AI impressions on 9 to 11 November should therefore be read as a reporting error first, not as a core update loss. Relevant Audience covered that report's arrival on 12 August 2026, when it became visible on every property. It shows impressions, pages, countries, devices and dates for generative AI features, with no clicks or queries.
A practical reconciliation checklist
Based on the four reasons Google lists, here is how to work through a gap between your analytics and Search Console. The steps are drawn from the page; the order is a suggestion.
- Exclude paid traffic from your analytics Google total before comparing it with Search Console.
- Check that your analytics matches all the Referer values Google lists, including local Google domains and the Android app value starting android-app://com.google.android.googlequicksearchbox.
- Count only landing-page requests when working from server logs.
- Test your redirects. Look for JavaScript or meta refresh redirects, redirects that drop the query string, and redirects that set a stricter Referrer-Policy.
- Remember that Images and Discover visits look like web search visits by Referer.
Relevant Audience covered related reporting gaps earlier this autumn, including how Direct, Unassigned and Cross-network traffic work in GA4 on 4 November 2026 and why Google Analytics and Google Ads numbers differ on 1 November 2026. On 15 October 2026 it reported claims, still unconfirmed, that ChatGPT clicks were arriving without UTM tags or a referrer and landing in Direct.
If your GA4 setup needs to be checked against these Referer rules, or redirects on your site may be stripping attribution, a technical SEO audit is the right place to test it.
What this means for Thai marketers
For Thai sites, Google's rule about local Google domains applies: the page says the Referer can be https://www.google.com/ or a local Google domain, and google.co.th is Thailand's local Google domain. A filter that matches only www.google.com may undercount. The page does not mention Thailand specifically. The other point worth carrying into client reporting is timing: Thai teams reporting November numbers will be dealing with the core update that began on 10 November and the generative AI report logging error from 9 to 11 November at the same time, and both should be labelled in any month-on-month comparison. For help building organic reports that hold up to this kind of scrutiny, see the Relevant Audience SEO Thailand service.
Frequently asked questions
Why don't my GA4 organic numbers match Search Console clicks?
Because the two tools measure different things at different moments, according to Google's 12 November 2026 page. Search Console counts a click when the user selects the link, while analytics counts a visit only after the page loads and the tag fires; privacy settings, ad blockers, consent choices, redirects and paid traffic in analytics totals widen the gap further.
Can I see the search query in the Referer header?
No. Google Search uses the "origin" referrer policy, which sends only the scheme and host, so the query is never included. Google points to the Search Console Performance report for query data.
Can I separate Google Images and Discover traffic using the Referer?
No. Google says Images sends the same Referer values as Search, and Discover uses the values of the surface where it was viewed, so the Referer alone cannot tell them apart.
Does the page cover AI Mode or AI Overviews traffic?
No. The Search Central page does not mention AI Mode, AI Overviews, Google News or Gemini, so how those clicks are referred is not stated there.
Was the drop in generative AI impressions on 9 to 11 November a ranking change?
No, Google says it was a logging error. The Search Console data anomalies page says the error decreased impressions in the generative AI performance report for 9 to 11 November 2026 and "affects data logging only".
Google has now written down, in one place, the reasons analytics and Search Console disagree. That will not close the gap, but it gives everyone in a reporting conversation the same reference. If you want your organic reporting checked against it, the Relevant Audience team can help.
