Google Gemini removes UTM tracking tags from cited links
Google stopped adding the utm_source=gemini tag to links in its AI responses, making it harder for websites to track traffic from its chatbot.
The short version
- Gemini added the utm_source=gemini tag to links in early September 2026 but removed it around September 29.
- Desktop visits can still be identified via referrers, but mobile app clicks often appear as 'Direct' traffic.
- The tag briefly allowed websites to identify Gemini traffic that otherwise lacked a clear source label.
- Google has not provided an official explanation for the removal of these tracking parameters.
Why it matters
For website owners and marketers, this change makes it more difficult to accurately measure how much traffic is being driven by AI assistants. It creates a visibility gap where clicks from mobile apps may be indistinguishable from direct user navigation.
The sudden disappearance of tags
Google's Gemini assistant recently stopped attaching a specific tracking label to the web addresses it cites in its responses. According to server-log data published by search consultant Natzir Turrado and reported by Search Engine Roundtable, the tag utm_source=gemini vanished around September 29, 2026. This occurred just three weeks after Google began adding the undocumented parameter without any official notice.
The tag functioned as a key-value pair added to the end of a URL. Analytics tools like Google Analytics 4 use these parameters to credit a visit to a specific source. When the tag was active, it allowed website owners to see exactly how many users were arriving from Gemini, even if the browser did not provide other identifying information.
Data from Turrado's logs showed that tagged visits climbed into double figures during the third week of September before peaking at 83 on September 28. The following day, the number of tagged visits collapsed to just 3. While some scattered tagged visits appeared in late October, the overall trend indicates a permanent removal of the feature.
How traffic attribution works
Understanding why this matters requires looking at how browsers and analytics tools communicate. On a desktop computer, a click on a Gemini link usually carries a 'referrer'—a piece of data telling the destination site where the user came from. Because of this, desktop visits to websites can still be identified as coming from gemini.google.com.
However, mobile apps often behave differently. The Gemini iOS app opens links within a WebView, which is a browser component embedded inside the app. This specific setup often fails to send a referrer, meaning that without a UTM tag, the visit is categorized as 'Direct' traffic. Direct traffic is a bucket for visits where the analytics tool has no way of knowing if a user typed the URL themselves or clicked a link from an app.
The utm_source=gemini tag was significant because it bridged this gap. It provided a clear label for clicks that arrived with no referrer at all. By removing the tag, Google has made it much harder for publishers to distinguish between a person manually typing a website address and a user clicking a link inside the Gemini mobile app.
Impact on publishers and marketers
This change comes at a time when Gemini is becoming a major player in the AI landscape. Alphabet reported in late 2026 that the Gemini app surpassed one billion monthly active users. With Gemini's share of global AI website traffic reaching 22% in early 2026, the volume of outbound clicks is substantial enough to matter to many businesses.
For marketers, the loss of the tag creates a measurement hurdle. It is now more difficult to compare organic AI visibility against paid channels or traditional search results. When a click leaves an AI assistant but carries no label, it becomes nearly invisible in standard analytics dashboards, potentially leading to an underestimation of AI's influence on web traffic.
The situation is complicated by the fact that different platforms handle this differently. While Google removed the tag, other platforms have had varying experiences. For example, OpenAI's ChatGPT briefly removed its own UTM tags in October 2026 but later restored them after the incident was identified as a bug. Google, however, has not confirmed if the removal was a deliberate choice or a technical error.
The state of AI measurement
Measuring traffic from AI assistants has been a recurring challenge for the industry. In May 2025, Google acknowledged that some links in its AI search experience were unintentionally stripping referrer data. This led to the creation of a custom 'AI Chatbots' channel in Google Analytics, which uses regular expressions to identify common assistant names in referrers.
Despite these efforts, the data remains fragmented. Research from Similarweb suggests that a majority of visits influenced by AI recommendations still arrive through traditional search rather than dedicated AI channels. This highlights the difficulty of tracking the full journey of a user who interacts with an AI before landing on a website.
Because the Gemini tag was not available to all users, the data collected by researchers represents a limited sample. It is unclear if every Gemini user saw the tag or if it was part of a restricted test. For now, the lack of a consistent, documented way to track these clicks leaves publishers guessing about the true value of AI citations.
What to watch for next
The primary question remains whether Google will ever reintroduce the utm_source=gemini parameter. Since the removal happened without any documentation or official statement, it is impossible to know if the company plans to provide better attribution tools in the future. Some observers hope for a return to tagging to help clarify the impact of AI on the web ecosystem.
Website owners might need to rely on more advanced methods to identify Gemini traffic. This could include analyzing server logs for specific user agents, such as the GeminiiOS string. However, these logs are not always accessible to standard website owners and are not part of basic analytics reports.
As AI assistants continue to grow in popularity, the demand for clear attribution will likely increase. Publishers and brands will need to keep a close watch on how Google updates its AI features and whether it eventually provides a standardized way to measure the traffic it generates.
Sources and links
- ReportingPPC Land: the article we sourced
Liked it? Say why
Add a line for other readers.
No account needed. We never show your email. Comments with links, ads or abuse are held and not shown. Your comment shows after a quick check.
Listen to the AI Wire headlines
Listen to the AI WireListen to our daily AI shows every morning: on your commute, in the shower, wherever you are
Listen to the daily showsFirst reported by PPC Land · Read the original. Written by the SI Agent News Desk from the reporting above.Written with AI from the source above. Every number is checked against the source. Spot a mistake? Corrections.
