Google has spent the past several months rolling out new reporting for how AI-generated search surfaces treat a site. Generative AI performance reporting was announced on June 3, with an August 31 note confirming worldwide availability. It’s a real, useful addition to Search Console. It’s also worth being precise about what it covers, because the gap between “Google shows you more” and “Google shows you what you actually want to know” is where most of the overreach in this category happens.
Here’s the verified scope. The dedicated generative-AI report covers impressions, broken out by pages, countries, and devices. That’s genuinely useful: a site that previously had no documented way to check whether AI-generated results were surfacing it, on which pages, in which countries, now has one. For a category built on the question of whether a brand is being mentioned, cited, or recommended by AI systems, any additional structured visibility into that question deserves to be taken seriously.
Here’s what it does not cover, and here’s where precision matters most. Query-level intent isn’t in scope. Recommendation outcomes, whether an AI mention actually influenced a decision, aren’t in scope. Conversions aren’t in scope. What’s easy to get wrong is assuming clicks aren’t measured anywhere either. They are, just not inside this specific dedicated report: Google already counts AI Mode and AI Overview activity, including click data, within its overall Search Console performance methodology. The new report is a narrower, specialized breakout sitting alongside that broader measurement, not a replacement for it and not evidence that Google has no click visibility into AI surfaces at all.
This distinction matters because of how reporting tools tend to get used once they exist. A number that’s easy to check becomes a number that gets treated as the whole picture, simply because it’s the number available. If a site sees its AI impressions rise in a given country over a given month, it would be easy to read that as proof of growing AI visibility translating into business results. The dedicated report doesn’t support that reading on its own. It supports a narrower, accurate one: impressions changed, in a specific dimension, over a specific window.
There’s a reasonable, narrower use for this report even within its limits. A sudden, large swing in impressions for a specific page, country, or device type is worth investigating, not because the report tells you why it happened, but because it tells you where to look. That’s a legitimate use: treating the report as a trigger for a question, not an answer to one.
There’s a related practical step worth taking now that this reporting exists, separate from any claim about what the numbers mean on their own. A team that wants to use the dedicated generative-AI report well should start logging its own baseline today, impressions by page, country, and device, recorded on a regular cadence, rather than only checking back after something notable happens. Without a logged baseline, a later spike or drop has no clean before-and-after comparison, and the temptation to read significance into a single snapshot goes up. This isn’t a claim about the report itself. It’s a reminder that a new measurement tool is only as useful as the discipline applied around it, and that discipline starts before anything interesting happens, not after.
It’s also worth being accurate about a second, related update. On September 24, Google added multimodal reporting covering Lens, Circle to Search on Android, image uploads to Google Search, and Chrome’s image-search action. This isn’t a separate, standalone system walled off from the generative-AI report. Google’s own documentation describes multimodal data as appearing within the relevant performance reports, including inside the generative AI features report itself. The useful distinction isn’t “two unrelated reports.” It’s that multimodal search is its own modality and mechanism, image-based rather than text-based, and the data for it shows up where it’s relevant rather than existing as one undifferentiated AI number. Keeping that nuance intact matters more than drawing a harder line between the two updates than Google itself draws.
None of this is a criticism of what Google built. Impressions by page, country, and device is a real, checkable signal that didn’t exist in this documented form before. The responsible way to use it is to treat it exactly as what it is: evidence about exposure, not evidence about outcome, and to hold the multimodal update as a related dimension of the same reporting family rather than a wholly separate story.
It’s also worth remembering what the reporting landscape looked like before this update existed. A site previously had no documented way to check whether AI-generated search surfaces were showing it at all, on which pages, in which countries. That absence made it easy for confident claims in either direction, overstating AI visibility or dismissing it entirely, to go unchecked. Judged against that baseline, even a report with real limits is a meaningful improvement in accountability.
For anyone building or buying tools in the AI visibility category, including us, this is a useful discipline to hold onto as more platforms roll out their own version of “here’s some AI reporting.” The value of a new report is real. The value of getting its actual scope right, including correcting an earlier assumption when it’s wrong, is what makes the rest of the analysis trustworthy.
FAQ
What does Google’s dedicated generative-AI performance report actually measure?
It reports impressions, broken out by pages, countries, and devices, for how AI-generated search surfaces treat a site. It was announced June 3, 2026, with worldwide availability confirmed August 31, 2026.
Does Google measure clicks related to AI search at all?
Yes, separately. Google already counts AI Mode and AI Overview activity, including click data, within its overall Search Console performance methodology. The new dedicated generative-AI report doesn’t include that click breakout itself, but that’s a limit of this specific report, not of Search Console as a whole.
Is the September 24 multimodal update a separate report from the generative-AI one?
Not entirely. Google describes multimodal data, covering Lens, Circle to Search, image uploads, and Chrome image search, as appearing within relevant performance reports, including inside the generative AI features report. It’s a distinct modality and mechanism, not a wholly separate reporting system.
Does more AI visibility reporting mean a brand can now prove AI search is driving business results?
Not from these reports alone. They show impressions and exposure by dimension, which is evidence of reach, not evidence of outcome. Proving a business result requires separate evidence these reports don’t provide.
Does Axis Suite import or integrate this new Google reporting data?
No. This piece is about understanding what Google’s own reports show and don’t show. It isn’t a claim about any product integrating Google’s reporting.
What’s the responsible way to use this new reporting?
Treat it as a starting question rather than a finished answer, and get the scope right before drawing a conclusion. A change in impressions tells you exposure moved in a specific dimension over a specific window. Confirming why it moved, or whether it mattered, requires separate evidence these reports don’t supply.
About Axis Suite
Axis Suite is the independent intelligence layer that tracks what AI says about your brand and the observable evidence behind it, retrieval, narrative, recommendation, and corroboration signals, across engines including ChatGPT, Claude, Gemini, and Perplexity. Built by TrendAxis, Axis Suite follows a brand’s path from Mentioned to Cited to Recommended to Chosen. Axis Suite is measurement infrastructure, not a marketing quick fix, every score is built to be explainable and defensible. Explore the Proof Center.