Google has finished rolling out a generative AI performance report to every Search Console property. Since 31 August 2026, site owners can see how often their pages appeared inside AI Overviews, AI Mode, and the AI features in Discover. For anyone measuring AI search, this is the first official view of that exposure inside a tool most SEO teams already use.
The report has a clear limit. It shows impressions, meaning how often a page appeared. It does not show how many clicks those appearances produced, and it does not show which queries triggered them. That leaves a gap between being seen and being chosen, and much of the work of generative engine optimization sits in that gap.
What the Report Gives You
The report answers one question: is my content appearing in AI-generated results at all? For a business that has never checked, that is a useful starting point. If key pages appear in AI Overviews and AI Mode, the content is being retrieved and considered. If they do not appear, the cause is usually upstream, in crawlability, page structure, or how clearly the answer is stated.
That is valuable diagnostic information. It tells you whether the problem is visibility or conversion. A page that never appears cannot be cited. A page that appears often but earns no visits may have a different issue, such as an answer that satisfies the question so completely that the user does not need to click.
What the Report Does Not Tell You
Without click data, you cannot say how much traffic an AI appearance is worth. Without query data, you cannot tell which questions triggered it. Both gaps matter. A page may appear for a high-intent question and generate no measurable click, while another appears for a minor query and sends a steady stream of visitors. The report shows neither difference.
This means it should be read alongside your existing analytics, not in place of them. Compare the pages that appear in the new report with the pages that already earn clicks from search. Where they overlap, you have a clearer picture. Where they diverge, you have a question to investigate.
Three Steps for Reading It Well
Start with the pages that matter most
Pick the five to ten pages that drive your most valuable search traffic or represent your core offering. Check whether they appear in AI Overviews and AI Mode in the report. Those pages are where visibility loss would cost the most.
Check the structure of pages that do not appear
If an important page is missing from AI features, review how it answers its main question. Is the direct answer near the top? Are the headings phrased as the questions people ask? Is the claim backed by a specific, attributable source? These features make content easier for AI systems to extract.
Confirm that crawlers can read the page
Make sure your robots.txt file does not block the crawlers that feed AI systems. A rule written for one crawler can unintentionally exclude others. This is a simple check, and it rules out a common cause of missing visibility.
Measure Over Time, Not in One Snapshot
AI search results change often, so a single check tells you little. Record which pages appear each month, note changes after you update content, and watch for patterns across page types. Over several months, the trend is more useful than any single report.
Pair this with a clear goal. If the aim is brand presence in AI answers, impressions are a reasonable measure. If the aim is leads or sales, you will need your own analytics to connect appearances to outcomes, because the report alone does not do that.
Keep the Limits in Mind
Google’s new report is a meaningful step toward measuring AI search, but it is incomplete. It gives you exposure without outcomes. Use it to decide where to look, then use your analytics and direct testing to decide what matters.
If you want to know where your site appears in AI answers, including ChatGPT, Perplexity, and Claude as well as Google, dpanell’s free AI Search Visibility Audit is a practical place to start. It shows the gaps that are worth fixing first.
