Google AI Overviews: Measuring Their Impact on Organic Traffic

Google AI Overviews rolled out in France on July 22, 2026, and if you consult on analytics, your inbox probably lit up the same week. The question is always the same: “my organic traffic is dropping, is it the AI Overviews?”. The good news is that you can now start to measure the impact of Google AI Overviews on organic traffic. The bad news is that neither Search Console nor GA4 hands it to you cleanly, and between the two sits a blind spot nobody explains. That is what this guide is for: what to look at, where, and what to add to your client reporting this week.

One framing point up front, because it changes everything: this is not a France-only problem. France was simply the last major market to switch on, held back by the neighboring-rights standoff with publishers. But AI Overviews already trigger on roughly 48 % of queries worldwide and cut organic CTR by 30 to 50 % on positions 1 to 3. The measurement framework I lay out here works for any market. France is the trigger, not the scope.

What AI Overviews are, in two minutes

An AI Overview is that AI-generated answer block Google shows at the very top of the results, above the classic blue links. The user reads the summary and often leaves without clicking a single source. That is what people call “zero-click” search, and it is exactly where your traffic leaks out.

The numbers to keep in mind when framing this for a client: according to BrightEdge, AI Overviews now appear on close to 48 % of queries. On impact, a study across 68,000 real queries measured an average CTR drop of 46.7 % when an AI Overview is present. On informational intents, positions 1 to 3, the observed range sits between 30 and 50 % fewer clicks. In other words, you can hold your number one ranking and still watch your traffic collapse. That split between rank and clicks is what throws everyone off.

The practical consequence for you: tracking rankings is no longer enough. You have to measure AI Overview presence and the CTR gap it creates. The rest of this guide breaks down what each tool gives you, and above all what it hides.

What Google Search Console gives you

This is the real novelty of the year. On June 3, 2026, Google added a dedicated report to Search Console, the “Search Generative AI Performance” report, which finally isolates how your pages perform inside generative surfaces (AI Overviews and AI Mode). Before it, those impressions were buried in the classic performance report, impossible to separate.

What the report gives you concretely is visibility into your impressions inside AI Overviews, broken down by URL, country and device, with granularity ranging from hourly to monthly. So you can see which pages Google deems worthy of citing in its AI answers, and track how that exposure moves over time. For a consultant, it is the first objective building block of an “AI Overviews impact” case.

The problem, and it is a big one, is what is still missing. Here is the honest state of play as I write this.

Translation: Google tells you that you appear, but not how many clicks you get from it, nor on which queries. That is the central blind spot. You know the exposure, never the yield. We will see below how to partly work around this by crossing sources, but keep in mind that the GSC report is a starting point, not a complete answer.

What GA4 gives you, and what it misses

Let us move to the other side of the click. Once someone clicks inside an AI Overview and lands on your site, where does that traffic show up in GA4? This is where the most expensive trap of the whole topic hides.

A click from an AI Overview stays counted in the “Organic Search” channel. Google Search is the source, the session looks like any organic visit, and nothing natively tells it apart from a click on a classic blue link. Do not confuse this with chatbot traffic (ChatGPT, Gemini, Claude as apps), which lands in a separate channel often labeled “AI Assistant” or captured through dedicated referrers. AI Overview traffic is plain Google organic. If you were hunting for an “AI Overviews” line in GA4, you are wasting your time: it does not exist.

What that means for measurement: GA4 will never tell you on its own what share of your organic traffic comes from AI Overviews. What it does give you is post-click behavior. Engagement rate, conversions, pages per session for your organic traffic overall. By crossing how those metrics move against Search Console’s AI impressions, you rebuild an estimate: if your AI impressions climb while your organic clicks stall or fall at equal position, you have your evidence of AI Overview erosion. It is not a direct measurement, it is a triangulation, and for now it is the best tool we have.

If you have not yet connected Search Console to GA4, that is the prerequisite. The method and the SEO analyses you can pull from it are detailed in my guide on the SEO analyses to cross in GA4 and Search Console. Without that bridge between the two, the triangulation below is impossible.

The AI Overviews measurement dashboard

Enough theory, let us get concrete. Here are the KPIs I add to a client report to make AI Overview impact legible, and the source of each. The point is not to pile up metrics, but to tell a story: exposure rising, yield falling.

The most telling KPI for a client is the third crossed with the fourth: showing that average position is stable while CTR collapses. That is the cleanest demonstration that the problem is not your SEO, but the results interface that shifted under your feet. The AI impressions ratio acts as an exposure gauge: the higher it climbs, the more your topic is being absorbed by generative answers, and the longer the pressure on CTR will last.

A field tip: freeze a reference snapshot now, before the effects fully settle on your market. One month of “before” CTR data is worth gold to objectify the “after”. Without that baseline, you are left with hand-waving estimates, which do not survive three minutes in front of a skeptical client.

What it changes in your analytics stack

Now to instrument all this so you are not starting from scratch every quarter. Three concrete workstreams, in priority order.

First, create a dedicated tracking view or segment in Search Console around the AI report, and export its data regularly (ideally to Looker Studio or BigQuery if volume warrants), because GSC’s retention window is short and you will want long history. This is exactly the kind of data you should pull out of the interface before it disappears.

Second, set alerts in GA4 on your organic traffic: an abnormal drop in “Organic Search” sessions while total traffic holds steady is your AI Overviews alarm. Coupled with the GSC report, that alert turns a hunch into a quantified case. This ties into the broader question of the conversion drop in GA4 attribution for 2026: AI Overviews are one more erosion factor to isolate when a client sees their numbers fall with no obvious cause.

Third, fold this measurement into a broader brand-visibility framework. AI Overview traffic and chatbot traffic are two distinct leaks you must never conflate: one is amputated Google organic, the other a channel of its own. The method to capture and label the latter is detailed in my guide on how to track ChatGPT, Gemini and Claude traffic in GA4, and the logic of measuring presence across AI surfaces in the article on your brand’s visibility in AI answers. The three together finally form a complete view, rather than three silos that contradict each other.

My recommendation for the coming week, if you consult: open each client’s Search Console AI report, export a month of baseline CTR history, and set a GA4 alert on organic. You will not have a perfect measurement, nobody does yet, but you will have enough to answer “is it the AI Overviews?” with numbers rather than a shrug. And these days, that is already a real edge.