Google made AI visibility measurable: Search Console reports and the opt out control reach all sites
Per Google's official help documentation, as of August 31, 2026 the generative AI performance data in Search Console has been rolled out to all websites worldwide. The same day, the on off control for Search generative AI also went global.
The rollout that started in June is complete
Google's Search Console help documentation states that as of August 31, 2026 these insights have been rolled out to all websites worldwide.
That means the gradual rollout which began in June 2026 is now complete.
In its August 31 story Search Engine Land reported that both the reports and the control were rolling out globally.
Impressions yes, clicks no: the limits of the report and the control
The report covers two surfaces: AI Overviews and AI Mode. Google Discover has its own separate generative AI performance report, and the Search report does not include Discover.
The nature of the data matters. The report provides impression data only, not click data. An impression counts how many times links to your site were shown to a user inside a generative AI feature on Google Search.
The data can be broken down across four dimensions: pages, countries, dates and devices. There is no query dimension. Pages are grouped by the final address after redirects and by canonical URL, and dates are given in Pacific time.
Google keeps another note on the same help page: because the report rolls out over time some properties may not have access yet, and the report can look empty if the site has not received enough impressions in generative AI features. A global rollout does not mean every site sees data straight away.
A second feature went global on the same day. The Search generative AI control under Settings in Search Console lets you choose whether your site is used in these features.
The control covers three areas: AI Overviews, AI Mode and generative AI features in Google Discover.
The control works at Search Console property level. A property inherits the setting from its closest parent that has changed it, and the owner of a child property can follow the parent or make a separate choice for that property and its subpages.
Google states that this control only affects whether your content can appear in certain Search generative AI features, and that it is not used as a ranking or inclusion signal affecting other parts of Search.
Once an opt out goes live, content is excluded within 1 to 2 days. Some content can take longer because of caching and propagation across Google's systems.
The conclusion is the same in both sources: sites that opt out receive neither impressions nor traffic from these features.
You cannot manage the visibility you cannot measure
The assessment in this section is ours. Appearing inside AI answers was long something you could feel but not measure. Before measurement arrived, every claim in this area was a matter of belief.
The data being limited to impressions is not a shortcoming, it is the point. If an AI answer resolves the user's question without a visit to the page, the link between visibility and traffic weakens. The report shows exactly that gap.
This is why judging brand visibility and traffic targets in the same table is now misleading. A page whose impressions rise while clicks stay flat is not failing, it may be doing a different job.
The absence of a query dimension is the report's practical limit. You know which page appeared, not which question it answered. For now that gap has to be filled with judgment on the content side.
The opt out control offers a real choice with a clear price: turn it off and those surfaces send neither impressions nor traffic. Because Google writes that it is not a ranking or inclusion signal for other parts of Search, the decision can be made on strategy rather than fear of a penalty.
The control working at property level, with inheritance from a parent, adds flexibility to that decision. You do not have to switch off the whole site, a separate choice can be made for a child property.
What it means for businesses in Türkiye
The assessment below does not appear in the source, it is our reading. The documentation contains no Türkiye specific information.
The country breakdown in the report is a practical gain. A business able to read impressions coming from Türkiye separately can judge its AI visibility inside its own market.
For local service businesses the first job is to take a baseline. Without recording today's impression level, the effect of any later content change cannot be measured.
If the report looks empty that may not be a fault. Google notes the site may simply not have received enough impressions. For a small business that means the starting point of measurement is zero, and that is data too.
On the content side the conclusion is this: being cited in AI answers is about clear definitions, visible sourcing and pages that answer the question directly. Those were already good content principles, they now have a measurable counterpart.
Few businesses will consider using the opt out. Outside publishers who sell their own content through a subscription, turning visibility off means closing your own door for most SMEs.
The UNALSOFT take
On the web design and content side we have for a while been building pages to be readable not only by a human reader but by answering systems too: clear definitions, an explicit heading hierarchy, verifiable sources and an answer that follows immediately after the question. This update puts a number on that approach for the first time. Its real value for us is this: we no longer have to speculate when we talk about AI visibility, we can open the report and look. Our advice to clients is to open it today and record the current level. Without a starting point for measurement, every improvement after it stays a story.
Are you visible inside AI answers?
Let us measure your site's generative AI visibility and review your content structure accordingly.