Google seems to have stopped appending UTM parameters on Gemini links. The cited URLs in Gemini's AI responses no longer carry them. For marketers who rely on UTM tags to tell where traffic came from, that is a small technical change with a large reporting consequence: clicks from Gemini citations will no longer arrive neatly labeled, and teams will need another way to recognize AI-referred traffic when it reaches their site.
The change is straightforward to describe. Cited URLs in Gemini's AI responses no longer carry UTM parameters that Google had previously appended. A user who clicks a cited link will still reach the cited page, but the link itself will not carry the tracking tags that would have identified the click as coming from Gemini in the usual way.
The report is careful in its wording, saying Google seems to have stopped appending the parameters. That caution is worth preserving. The observation is that the cited URLs no longer carry them, and this article does not invent a date for when the change happened or a reason for it. The visible effect, however, is clear enough to plan around.
AI-referred traffic is still a new and contested category in analytics. Teams want to know how much traffic comes from AI answers, which answers drive it, and whether it behaves differently from search or social traffic. UTM tags offered a simple answer to the first question: if the tag says Gemini, the visit came from Gemini.
Without those tags, the first question becomes harder. A visit from a Gemini citation may still show a referrer, but referrer data from AI apps is often incomplete or inconsistent, and many analytics setups will bucket the visit elsewhere. The result is not that AI traffic disappears, but that it becomes harder to count, and harder to defend in a report to a client or a manager.
That counting problem matters because budgets follow measurement. A channel that cannot be measured is a channel that is hard to justify investing in, even when it is actually driving valuable visits. Losing a clean tag therefore has consequences beyond tidiness: it affects how AI visibility is valued inside an organization.
There is a useful parallel here with ChatGPT links, which open without tracking tags. ChatGPT citations have similarly not carried the kind of tags that would make attribution simple, and marketers working with ChatGPT traffic have already had to find workarounds. Gemini moving in the same direction makes that workaround a general requirement for AI traffic, not a platform-specific quirk.
The parallel also suggests this may be a broader norm for AI answers rather than a one-off decision. AI systems present citations as references for an answer, not as campaign links, and the platforms may prefer clean URLs that point to the source itself rather than to a tracked version of it. Whether that preference is deliberate policy or a technical side effect, the effect on marketers is the same.
Adaptation starts with accepting that AI traffic will be messier to measure than campaign traffic. Teams should look at a combination of signals rather than a single tag: referrer data where it exists, landing page patterns, and the timing of traffic changes after AI visibility changes. No single signal will be perfect, but together they can give a reasonable picture.
Segmentation helps as well. Pages that are frequently cited in AI answers can be watched as a group, so that a rise in their traffic can be compared against pages that are rarely cited. That comparison will not prove causation on its own, but it can support a case that AI visibility is contributing, especially when combined with citation monitoring.
For link builders, the change reinforces a point that has been true for some time: a citation in an AI answer is valuable even when it is hard to measure precisely. The visit still happens, the brand exposure still happens, and the reader still encounters the source. Measurement difficulty does not erase that value; it only makes it harder to prove.
The response should be to keep earning citations while improving measurement, not to stop pursuing citations because measurement got harder. Build the assets that AI systems cite, monitor citations directly, and treat analytics as one imperfect window onto AI traffic rather than the only window.
Source: Search Engine Roundtable, October 30, 2026.
Linkbaits helps you create content that earns citations and backlinks. Start with the free tools and build your first linkable asset this week.
Start free trial