Google has quietly begun appending UTM parameters to outgoing links from Gemini, giving site owners and SEOs a reliable way to attribute referrals from the AI assistant for the first time.

The change matters because traffic from AI chatbots like Gemini frequently shows up as "direct" in analytics packages such as GA4. Direct is the fallback classification used when no referrer or UTM data explicitly ties a visit to a source. Until now, a large share of Gemini-driven visits has been invisible to standard attribution.

The move arrives amid growing frustration with Google over collapsing Search referral volumes, driven by AI Search results and Gemini itself answering queries that once sent users to publisher sites.

Google confirms the change

A Reddit user opened a discussion after spotting the new tagging, writing: "Google Gemini add UTM to results! Gemini Start using UTM ! finally , its very new 24hours or so. Like chatgpt UTM."

Google Search Advocate John Mueller replied: "Nice. I see it too. Doesn't Gemini have a referrer too though? I thought you could already track them."

Redditor CrawlWarden explained why the referrer alone was insufficient: "There is a referrer, but it mostly survives on desktop web only, reported pass through from the Gemini mobile app runs around 9 percent and Android assistant invocations strip it entirely, so a large share of that traffic has been landing in Direct instead of GA4's ai assistant medium. A UTM survives the in app webview and shows up in raw server logs too, which is why this matters more than it looks even though the referrer technically already existed."

Mueller responded: "If someone has more information on what happens with the referrer (ideally with something I can reproduce), I'm happy to forward that to the team. It would be great to have these retained (in addition to the UTM-tagging)."

Why marketers should care

The UTM addition lets SEOs and site owners measure exactly how much traffic Gemini sends — no more guessing. Better referral attribution also enables segmentation of AI-referred visitors, so teams can compare referrals across AI referrers and against traditional search within the same analytics framework.

The practical payoff is budget logic. With accurate numbers in hand, marketers can justify resource allocation to AI channels with real data rather than assumptions.

Questions remain

ChatGPT adds UTM parameters only to links grounded in web-sourced answers, not to responses drawn from its training data. Google has not documented its implementation, so the exact circumstances that trigger UTM tagging in Gemini remain unknown. The tags presumably appear when Gemini cites a website directly, but nothing is confirmed until Google documents the behavior.

Until then, marketers tracking AI referrals should audit their GA4 and server logs now that a previously invisible traffic stream has started carrying identifiable parameters.