Google has started giving some site owners a clearer view of where their pages appear inside generative AI features. For SMEs, the useful question is which pages are already being surfaced, in which markets, and whether those appearances line up with the search demand the business actually cares about.
On 3 June 2026, Google announced new Search Generative AI performance reports in Search Console. The reports cover visibility in generative AI features on Search, including AI Overviews and AI Mode, plus generative AI features in Discover. Google says the data remains part of the overall performance report, but the new view separates generative AI visibility so site owners can inspect it directly.
This matters because many SME SEO reports still treat AI search as a vague outside trend. A separate Search Console view makes it possible to move the discussion from “Are we showing up?” to “Which pages are showing up, where, and how consistently?” That is a better starting point for weekly content decisions.
What the new report gives you
Google says the report shows impressions, pages, countries, devices and dates. Each one answers a different measurement question.
- Impressions: how often URLs from your site appeared in generative AI features in Search and Discover.
- Pages: which URLs appeared within AI features.
- Countries: where that visibility occurred.
- Devices: which devices people used when seeing the website, available for Search results.
- Dates: performance over time, with hourly, daily, weekly and monthly granularity.
The scope is important. This is a visibility report, not a full attribution model. It can tell you where Google surfaced your URLs in generative AI features. It does not, on its own, prove revenue, lead quality or final customer intent.
Measure pages before you chase totals
The first useful view for most SMEs is the pages dimension. Total impressions can make the report feel important, but pages tell you whether the right assets are earning visibility.
Start by grouping visible URLs into useful buckets: service explainers, buying guides, comparison pages, local pages, support content and editorial articles. Then compare those groups with the pages that already drive enquiries or assisted conversions elsewhere. If an informational article appears often but your commercial explainers do not, the next content task may be a clearer answer page, a better comparison section or stronger supporting detail on the page that already matters.
This also helps avoid overreacting to single URLs. One page appearing in AI results is useful evidence. A cluster of related pages appearing across the same topic is stronger evidence that Google understands the site as a source for that subject.
Use countries and devices to check fit
Country data is especially useful for UK SMEs that sell into a specific geography. If a business mainly serves the UK but most generative AI visibility is elsewhere, the report may show awareness rather than demand. That does not make the visibility worthless, but it changes how it should be reported.
Device data should be treated as a practical diagnostic. If mobile visibility is high, the visible pages need to answer quickly, load cleanly and make follow-on actions easy. If desktop visibility dominates, the audience may be comparing suppliers, reading longer material or working from office contexts. The report will not tell you that intent directly, but it gives you a better reason to inspect the page experience by device.
Watch trends, not one-day spikes
Google says date reporting is available at hourly, daily, weekly and monthly levels. That granularity is useful, but the first few weeks of data should be read carefully. Google also says the reports are rolling out to a subset of websites while it tests the feature and gathers feedback. Early data may be uneven simply because the report is still being rolled out.
For weekly reporting, compare like with like. Track the same page groups, countries and devices each week. Note known events such as content updates, product launches, seasonality and broader changes in Google AI features. A small spike on one day is less useful than a page group that keeps appearing for the same topic over several weeks.
Pair this report with normal SEO data
The generative AI report belongs next to standard Search Console analysis: the usual query, page, click and conversion checks. If a page gains generative AI impressions but normal search clicks fall, that needs a different conversation from a page that gains AI impressions while also holding or improving qualified traffic.
For context on why this matters, Kahunam recently covered how SME SEO briefs need to change as Google AI Mode scales. The Search Console report gives marketers a measurement layer for that shift, but it still needs to be interpreted alongside business outcomes.
A simple first reporting workflow
Once the report appears in your property, start with a narrow baseline rather than a large dashboard.
- Export or record the top visible pages in generative AI features.
- Tag each page by business role: informational, comparison, commercial, support or local.
- Filter by the country or countries that match the business market.
- Compare mobile and desktop visibility for the same page groups.
- Review weekly trends before making content decisions.
- Cross-check the same pages in standard Search Console and analytics reports.
This keeps the report tied to decisions a small team can actually make: which pages need clearer answers, which topics deserve supporting content, and which markets are showing early visibility.
What not to over-interpret
Don’t treat an impression as a visit. Don’t treat one appearing URL as proof that a topic is won. Don’t assume the report covers every AI search experience or every user journey. Google’s announcement describes a dedicated visibility view for generative AI features, with rollout limited to a subset of websites at launch.
The practical value is still real. Search Console is becoming more useful for AI search measurement, but the first job is disciplined interpretation. Measure page groups, check market fit, watch trends over time, and connect the findings back to normal SEO and business data before changing the content plan.
Author
