Manu Shukla

Posted on Jul 27 • Originally published at ecorpit.com

See your AI Overviews impressions in Search Console: how to read the new report and grow GEO visibility (2026)

#searchconsole #aioverviews #geo #aeo

See your AI Overviews impressions in Search Console: how to read the new report and grow GEO visibility (2026)

Summary. On June 3, 2026, Google added Search Generative AI performance reports to Google Search Console, the first official way to see how often your pages appear inside AI Overviews and AI Mode. The report is split into two parts, Generative AI in Search (AI Overviews and AI Mode) and Generative AI in Discover, and it shows four things: impressions, pages, countries and devices, with data starting May 18, 2026. It shows no clicks, no click-through rate, no average position and no query data yet, and Google says it will add metrics over time. At launch it reached only a subset of website owners in the United Kingdom, an ordering tied to Competition and Markets Authority requirements on Google, with a wider global rollout expected but undated. Barry Schwartz of Search Engine Land noted the features are "rolling out to subset of website owners in the UK and will expand it to more in the future." Alongside the report, Google shipped a control to keep your content out of AI answers without touching organic rankings. This guide covers what the report measures, what it hides, and how to use it to grow visibility in AI answers.

Generative answers now appear on roughly 50% of searches, up from about 25% when AI Overviews launched in 2024, by Search Engine Journal's reporting, so a large share of every audience now meets your brand inside an AI answer first. For three years the honest answer to "how visible are we inside AI Overviews?" was a guess. The impressions were folded into the main Search performance report and could not be separated. This report is the first time Google exposes AI-feature visibility on its own, which matters because being cited inside an AI answer is now its own form of visibility, distinct from a blue-link ranking. You cannot improve what you cannot see, and until June 2026 this was invisible.

What the new report actually shows

The report lives in Search Console under the performance section and carries the same shape SEO teams already know: a time series with dimensions you can filter. The difference is the scope. It counts only appearances inside generative AI surfaces, and it exposes four dimensions.

Impressions tell you how often a page appeared in an AI Overview or AI Mode answer. The pages dimension shows which URLs are being surfaced. Countries and devices break that down by geography and by phone, tablet or desktop. There is a date range so you can watch the trend. That is the whole surface today, and the data begins on May 18, 2026, with no retroactive backfill before that date, so the earliest trend line any site can see starts there.

The important framing: this is a clearer view of data Google already had, not a new collection. AI Overviews and AI Mode impressions were already inside the main Search performance report and remain there. The new report separates them so you are not reading AI-feature visibility blended with ordinary search impressions.

What the report does not show, and why that matters

The gap is as important as the feature. The report has no click data. There is no click-through rate, no average position and no query-level breakdown for AI features. You can see that a page surfaced in an AI answer, but not whether anyone clicked, where it sat, or what the person searched to trigger it.

Metric	Gen-AI performance report	Main Search performance report
Impressions in AI features	Yes	Yes (blended with all impressions)
Pages	Yes	Yes
Countries	Yes	Yes
Devices	Yes	Yes
Clicks	No	Yes
Click-through rate	No	Yes
Average position	No	Yes
Queries	No	Yes (limited sample)

The missing query data is the sharpest limit. Without it you cannot tie an AI-feature impression to an intent, so you are measuring exposure, not demand. Treat the numbers as a visibility signal, not a traffic or conversion metric. The practical move is to read this report next to the main performance report for the same pages, which still carries clicks and queries, and infer the rest. That is the same discipline behind reading ranking versus AI Overview citation data: the two datasets answer different questions and only make sense together.

Where to find it and who has it

At launch the report is limited to a subset of website owners in the United Kingdom. Google has said it will expand more widely but has not given a date. The UK-first order is not a coincidence: it is tied to Competition and Markets Authority requirements on Google around giving publishers control over how their content is used. If your Search Console property is outside the UK, you will not see the dedicated report yet, but the underlying AI-feature impressions are still counted inside your main Search performance report, so you are not flying blind in the meantime.

That geographic caveat matters for Indian and other non-UK teams reading their dashboards this quarter: the absence of the dedicated report is a rollout gap, not a sign that your pages are missing from AI answers.

The two surfaces: AI Overviews, AI Mode and Discover

The report separates generative AI in Search from generative AI in Discover, because they are different products with different behaviour.

Surface	What it covers	Where it is reported
AI Overviews	The AI-generated answer block above search results	Generative AI in Search
AI Mode	The conversational, full-page AI search experience	Generative AI in Search
Discover AI features	AI-generated content inside the Discover feed	Generative AI in Discover (separate)
Search Labs experiments	Early opt-in experimental features	Excluded entirely

Search Labs experiment data is excluded, so anything still behind an opt-in experiment does not appear. For most businesses the Search section, covering AI Overviews and AI Mode together, is the one to watch first, because that is where the bulk of commercial and informational queries now resolve.

The new control: keeping your content out of AI answers

Google shipped a control alongside the report that lets you remove your content from AI Overviews and AI Mode without changing your organic rankings. This is a real decision, not a formality, and it is separate from the older question of whether your content is used to fine-tune AI models.

Option	What it does	Effect on organic rankings
Default (no action)	Content is eligible for AI Overviews and AI Mode	None
Block AI features	Removes content from AI Overviews and AI Mode	None stated
Opt out of model fine-tuning	Stops content being used to train or tune models	Separate control, separate effect

The trade-off is straightforward once you name it. Blocking AI answers usually removes your AI-feature impressions, which is the thing the new report measures. If you believe AI answers are cannibalising clicks you would otherwise win, blocking is a lever. If you believe citation inside an AI answer builds brand visibility that pays back elsewhere, keep the default. The report finally gives you the data to make that call per site rather than on instinct, and it pairs with the strategic question in is GEO and AEO still SEO.

How to use the report to grow AI visibility

The report is a measurement tool, so the workflow is measure, then act. Start by identifying the pages that already earn AI-feature impressions. Those pages are doing something the AI systems reward, and they are your template.

Read what those pages have in common. In practice the pages that get pulled into AI answers tend to lead with a direct answer near the top, cite specific and dated facts, and name their sources. That is the same content shape that wins featured snippets, which is why an AI Overview content strategy and traditional answer-engine optimisation reinforce each other.

Resist two temptations. The first is to chase schema as a shortcut: Google has said structured data is not required to appear in generative AI features, so adding markup will not force you into AI Overviews, and the same goes for llms.txt files, which Google does not use. The evidence for that is worth reading before you spend engineering time on it, and whether llms.txt helps SEO on Google covers it directly. The second temptation is to read impressions as traffic. Until Google adds click data, an impression is exposure, nothing more. Use the report to find where you are visible, then use the main performance report and your analytics to judge whether that visibility is worth anything. The broader playbook sits in our guide to Google AI search optimisation and the SEO strategy pillar for 2026.

Why AI-feature visibility is worth tracking at all

A fair question is whether AI-feature impressions matter when they may never turn into a click. They do, for two reasons. First, generative answers now sit on close to half of searches, by Search Engine Journal's reporting, so a large share of the audience forms an impression of your brand inside an AI answer before any click decision exists. Being named and cited there is a distinct form of visibility, not a consolation prize for a lost click. Second, the alternative to measuring it is guessing, and guessing is how teams either over-invest in blocking AI answers out of fear or ignore a surface that is quietly shaping demand.

The report reframes the goal. A blue-link ranking and an AI-answer citation are now two different games played on the same query. You can rank first and never be cited, or be cited heavily on a page that ranks modestly. Separating the two lets you see which of your pages the AI systems trust enough to quote, which is the clearest signal available of what good looks like for answer-engine and generative-engine optimisation.

The discipline is to hold the number in proportion. Impressions are exposure, not revenue. Read them as a leading indicator: rising AI-feature impressions on a cluster of pages tells you the AI systems are leaning on your content more, which is the moment to deepen that content and make sure the parts an AI is most likely to quote are accurate, current and clearly yours. Falling impressions on a page that used to be cited is an early warning worth a content review before the traffic impact shows up anywhere else.

India-specific considerations

For Indian businesses the report is not yet live, since the rollout started UK-first, but the underlying dynamics already apply. AI Overviews appear across a large share of Indian-language and English queries, and the same impressions are counted in the main performance report today. Teams should start segmenting their existing performance data by the pages they suspect are being cited, so they have a baseline before the dedicated report arrives.

There is a data-governance note too. When you decide whether to allow or block AI features, and when you handle the analytics that track this, the Digital Personal Data Protection (DPDP) Act 2023 governs any personal data in the pipeline. The blocking control itself is about content, not personal data, but the analytics stack around it should be designed with DPDP obligations in mind, the same way any measurement pipeline is.

FAQ

What is the Search Console generative AI performance report?

It is a dedicated report in Google Search Console, launched on June 3, 2026, that shows how often your pages appear in generative AI features. It splits into Generative AI in Search, covering AI Overviews and AI Mode, and Generative AI in Discover. It reports impressions, pages, countries and devices from May 18, 2026 onward.

Does the report show clicks from AI Overviews?

No. The report shows impressions only. There is no click data, no click-through rate, no average position and no query-level breakdown yet. You can see how often a page surfaces in an AI answer, but not whether anyone clicked or what they searched to trigger it. Google says it will add metrics over time.

Who can see the report right now?

At launch it is limited to a subset of website owners in the United Kingdom, with a wider global rollout expected but not dated. The UK-first order is tied to Competition and Markets Authority requirements on Google. If you are outside the UK, the AI-feature data still sits inside your main Performance report.

Is this new data or a new view of existing data?

It is a clearer view of data Google already collected. Impressions from AI Overviews and AI Mode were already counted inside the main Search performance report, and they remain there. The new report separates them so you can see AI-feature visibility on its own, rather than blended with ordinary search impressions.

Can I remove my content from AI Overviews and AI Mode?

Yes. Google added a control that lets you keep your content out of AI Overviews and AI Mode without changing your organic rankings. It is separate from opting out of having your content used to fine-tune AI models. Blocking AI answers usually removes AI-feature impressions, so weigh visibility against control before using it.

Does structured data help me appear in AI Overviews?

Google has said structured data is not required to appear in its generative AI features, so adding schema is not a shortcut into AI Overviews. Schema still helps eligibility for traditional search-result enhancements. For AI answers, clear, well-sourced, directly-answering content matters more than markup, which is what the new report lets you measure.

How should I use the report to grow AI visibility?

Track which pages earn AI-feature impressions, then study what those pages do well: direct answers near the top, clear sourcing and specific data. Compare AI-feature impressions with organic impressions on the same pages to see where AI answers are replacing clicks, and prioritise content that already earns AI visibility for deeper coverage.

Does the report include Search Labs or Discover data?

Search Labs experiment data is excluded entirely, so early tests do not appear. Discover's AI features are reported separately, under Generative AI in Discover, rather than mixed with Search. The Search section covers AI Overviews and AI Mode together, which are the surfaces most businesses care about tracking first.

How eCorpIT can help

eCorpIT helps businesses win visibility in AI answers, not just in traditional rankings. Our senior teams read the Search Console generative-AI reports, map which pages earn AI Overviews and AI Mode impressions, and rework content so it answers the questions those surfaces pull from. We also advise on when blocking AI answers helps or hurts, based on your goals rather than a default. To build an AI-search visibility programme, see our GEO and AEO optimisation service or talk to our team.