Marketers face a growing hole in their analytics because traffic from Google's AI Overviews is not reported separately. Observed behaviour shows these clicks landing in Google Analytics 4 and Search Console under the 'google / organic' source, indistinguishable from a traditional blue link. This makes direct measurement impossible.
The solution, however, is not to give up but to get smarter. Marketers must adopt a proxy method, correlating changes in organic traffic for AI-prone queries with their measured AI visibility to estimate its true impact.
The scale of this blind spot is significant. A Honeyb capture on 2 October 2026 found that Google showed an AI Overview on 93% of commercial searches (Honeyb capture), meaning your analytics are blind to the first interaction most of your customers have with your brand on Google.
Other studies confirm the trend. Semrush reported in October 2026 that AIOs were present for over 80% of major brand queries it tracks (Google AI Overviews now appear for most major brands), meaning even the largest, most-searched-for brands are not immune to this measurement black hole.
AI Overview Traffic: The Unseen Majority
93%
Of commercial searches show an AI Overview
Source: Honeyb capture, 2 October 2026
70%
Chance an AIO changes between views
Source: Ahrefs
71%
Of AIO sources are outside the top 10
Source: Honeyb capture, 2 October 2026
Why Your Analytics Tools Are Flying Blind
The problem is one of attribution by design. Observed behaviour in analytics platforms indicates that clicks within AI Overviews are attributed as standard organic traffic, as Google provides no separate source designation. In GA4, this traffic is bundled into the 'Organic Search' channel grouping, and in Search Console, it is just another click under the 'Web' search type. There is no filter or report to isolate it.
Any guide suggesting a simple fix with custom channel groupings is missing the point: without distinct referral data from Google, there is nothing to group.
This leaves marketing teams in a difficult position. A significant and growing portion of user journeys now happen inside an AI answer, but the primary tool for measuring search performance offers no way to quantify it. Relying on last-click attribution from GA4 alone means you are likely misinterpreting a large and growing share of your acquisition funnel.
How to Estimate AI Traffic With the Proxy Method
If you cannot measure direct traffic, you must measure influence. The proxy method estimates the impact of AI Overviews by correlating two different datasets: the traffic you can see in Google Search Console, and the AI visibility you must measure with a third-party tool. It is an estimation, but it is far better than guessing.
The process involves three steps. First, identify a basket of commercially important keywords for which Google consistently shows an AI Overview. Second, use an AI visibility tool like Ahrefs, Semrush, or our own platform, Honeyb, to track how frequently your domain is cited in the AI Overviews for those keywords over several weeks. This is your AI visibility score.
Third, in Google Search Console, filter your performance report to show only the organic traffic from that same basket of keywords over the same period. By plotting your AI visibility trend against your organic traffic trend, you can spot correlations. If your visibility rises and traffic for those terms follows, you have a strong signal that your AIO presence is driving clicks. Conversely, a drop in visibility followed by a traffic dip for those keywords suggests you have lost ground to a competitor in the answer box.
To build your keyword set in Google Search Console, use the query filter to isolate terms relevant to your products or services. You can use regular expressions to match groups of related keywords. Pay close attention to pages with high impression counts but declining click-through rates, as these are often prime candidates for queries now dominated by an AI Overview. This gives you a stable cohort of terms to monitor.
Why Spot-Checking Is Dangerously Misleading
This systematic approach is critical because one-off checks are dangerously misleading. An AI Overview has a 70% chance of changing between observations for the same query, according to research from Ahrefs. The answer you see today may not be what your customer sees tomorrow. A single screenshot of your brand in an AI Overview is not a reliable performance indicator; only a trended, multi-query visibility metric can smooth out this volatility and reveal the true picture of your performance.
Want to get recommended by AI?
Check your AI search visibility, then let the Honeyb agent write, fix, and earn what gets you recommended. Free to start.
Once you have your keyword list and date range, export the traffic data from GSC and the visibility data from your chosen AI tracking tool. The goal is to align them by date in a single spreadsheet. With two columns of data, one for AI visibility percentage and one for organic clicks, you can create a simple line chart to visualise the correlation. This chart becomes the central piece of evidence for your influence report.
The table below compares the proxy method to the default, but flawed, approaches.
| Method | What It Measures | Pros | Cons |
|---|---|---|---|
| GA4 / GSC Default | 'google / organic' traffic | Simple, uses existing tools | No separation for AI Overview clicks |
| Manual Spot-Checking | A single AIO answer at one moment | Quick, requires no extra tools | Highly unreliable due to answer volatility |
| The Proxy Method | Correlation of visibility & traffic | Measures influence, not just clicks | Requires third-party tools, is an estimate |
Is This Different for ChatGPT or Perplexity?
While the challenge is universal, the mechanics differ for other AI engines. Traffic from ChatGPT, Perplexity, or Claude should, in theory, appear in your analytics with its own referral source. In practice, it is often messy. In October 2026, for instance, many analysts noted a sharp drop in ChatGPT referral traffic, which was later attributed to a change in how the platform handled outbound links, stripping them of tracking parameters and causing the traffic to be misclassified as 'Direct' by analytics platforms, as reported by Search Engine Journal.
For marketers, this means even direct referral data from AI tools cannot be trusted at face value without also monitoring visibility at the source.
This stands in contrast to platforms like OpenAI, which are actively working to improve attribution for advertisers. In October 2026, OpenAI announced expanded measurement capabilities for its ad platform, including conversion data integrations with attribution partners (OpenAI announcement). While this applies to paid ads within ChatGPT, it signals that some players in the AI space recognise the need for clearer data, a stance Google has yet to adopt for organic AI Overviews.
Each engine presents its own attribution quirks. The core lesson is the same: you cannot rely solely on referral traffic data to understand your brand's performance in AI. Measuring visibility at the source, through consistent tracking of the answers themselves, is the only resilient strategy. For a deeper dive on setting up tracking for these other sources, see our guide to AI traffic analytics in GA4.
Reporting on AI: From Clicks to Influence
Without a clean 'AI Overviews' row in your traffic reports, you must change the conversation with your team and leadership. Stop promising direct traffic attribution from AI search. Instead, report on AI visibility as a key performance indicator, just as you would for brand awareness or share of voice. Present the correlation data from the proxy method as evidence of influence.
Your new dashboard might feature a line chart showing AI visibility percentage week-over-week, next to a bar chart of organic clicks for the same keyword set. Annotate key events, like a content update or a competitor gaining a citation, to explain shifts in the trend. As Ahrefs noted in an October 2026 analysis, much of AI's influence happens 'out of sight' of traditional analytics, making visibility the primary metric to track (How to Measure AI Search Visibility When Attribution Falls Short).
Frame it as a measure of how successfully the brand is shaping the answers customers receive. This shifts the focus from chasing individual clicks to the more strategic goal of becoming the default recommendation in your category, a core principle of a solid AEO (answer engine optimisation, or getting the AI to cite you) strategy. It is a more honest and durable way to measure what is fast becoming a dominant force in how people find information and make buying decisions.
The inability to track AI Overview clicks directly is not a temporary glitch; it is a feature of the new landscape. By adopting a proxy method and shifting your reporting to focus on visibility and influence, you can build a far more accurate and strategic picture of your performance in AI search.
The quickest way to find out where you stand is to look: run a free AI visibility check and see which answers name you today.
Sources cited in this report
