KPW Tech Brief
AI Search Reporting Arrives: What Google and Bing Now Show You
Two search engines opened part of the AI answer box to measurement, and most of the coverage got both the date and the metric wrong.
Read aloud by your browser. Nothing is downloaded and nothing is sent anywhere.
Google announced dedicated Search Generative AI performance reports in Search Console on June 3, 2026, and Microsoft began a global preview rollout of AI Visibility Insights in Bing Webmaster Tools, including Citation Share, on June 16, 2026. Much of the secondary coverage dates these to July, or describes Google's version as a toggle inside an existing report. The primary sources say otherwise: two dedicated reports, one for Search and one for Discover, rolling out to a subset of properties.
The more useful correction is what the numbers actually are. Google's AI reports show impressions, broken down by page, country, device and date. They do not show clicks, queries, click-through rate, or average position for AI Overviews and AI Mode. Bing's Citation Share is described by Microsoft as an observational metric, not a ranking system, not traffic share, and not a window into competitor domains.
So the black box is not open. A door was propped ajar, and it is worth knowing exactly how far.
The reports shipped in June, and they are not a toggle
On June 3, 2026, Google published Introducing Search Generative AI performance reports in Search Console on the Search Central Blog. The post describes two dedicated reports, not one filter bolted onto the existing Performance view. One covers Search, reporting impressions from AI Overviews and AI Mode. The other covers generative AI in Discover.
Thirteen days later, on June 16, 2026, Microsoft published New AI Visibility Insights in Bing Webmaster Tools, announcing four capabilities: Intents, Topics, Citation Share, and Compare. The post says they are "now beginning to roll out in preview within Bing Webmaster Tools globally today."
A good deal of the third-party writing on both places them in July 2026, and some dates Citation Share to April. If you are trying to reconcile what a vendor told you against what you can see in your own account, the month matters, because the second half of both announcements is about staging.
Google states the reports are initially rolling out to a subset of Search Console properties while it tests and gathers feedback, with coverage expanding over time. Microsoft's phrasing, "beginning to roll out in preview," carries the same caveat. Global intent, staged access. If you log in and the report is not there, that is not a misconfiguration on your end. It is the rollout.
Two of the largest search companies on earth shipped their first real AI answer visibility reporting inside a two-week window, and a fair share of the trade coverage put both in the wrong month. That tells you something about how much of the commentary is reading the announcement and how much is reading other commentary.
Google shows impressions. It does not show clicks.
This is the part that changes what you can actually do with the data. Google's June 3 announcement describes impressions of your content within AI Overviews and AI Mode, broken down by page, country, device and date. Clicks are not among the listed metrics. Neither are queries, click-through rate, or average position for those AI features.
Marketing posts claiming you can now see how much traffic AI Overviews sends you are describing a report that does not exist yet. What you have is a visibility metric. It answers "is my content appearing," and it declines to answer "is anyone coming."
Google's definition of the impression is worth reading closely, because it is narrower than the casual version. In its documentation for the generative AI reports, an impression occurs when a link to your website is shown to a user inside a generative AI feature such as an AI Overview or AI Mode. It is not counted every time an AI answer renders on a query in your topic area. It is counted when your link is on screen. Commentary that treats any appearance of an AI answer as an impression for everyone in the space is measuring something Google is not reporting.
One further wrinkle landed in mid-August. Google's Latest Google Search Documentation Updates page records that AI Mode now counts toward totals in the main Search Console Performance report, with the AI feature documentation updated to point at the section explaining how AI Mode traffic is included in overall search metrics. So AI Mode is not sitting in a sealed compartment. It contributes to your global numbers while also appearing in the dedicated report.
Practically, that means a year-over-year comparison of your Search Console totals is now comparing two slightly different definitions of the same chart. Worth knowing before you present that chart to anyone.
Bing's Citation Share is a denominator, not a leaderboard
Bing Webmaster Tools already reported total citation counts for AI-generated answers, meaning how often your content was cited. Citation Share is layered on top of that rather than replacing it. Microsoft defines it as the percentage of citations attributed to your site out of all citations shown across all sites for the same grounding query. A grounding query is the query the AI system actually used to retrieve source material, which is not necessarily the sentence the user typed.
Microsoft is unusually direct about what the metric is not. The post calls Citation Share "an observational metric, not a ranking system or a competitive scoreboard," and states plainly that it does not expose competitor domains, does not represent traffic share, and does not assign quality scores. Coverage framing it as a share-of-voice leaderboard is describing a product Microsoft explicitly says it did not build. You get your slice. You do not get the names on the other slices.
The two companion features are classification tools. Intents sorts the grounding queries behind AI answers by user purpose, informational versus transactional being the example Microsoft gives. Topics clusters those queries into thematic groups. Some explainers have described these as ranking factors or traffic segmentation. Microsoft frames them as classification and clustering of the queries used to ground AI answers, which is a considerably more modest claim and a more honest one.
Where this gets useful for an ordinary business is the intent split. If your content is being cited heavily on informational queries and almost never on transactional ones, that is a specific, checkable finding about what role your site plays in AI answers. It is not a ranking. It is a description of the job your pages are currently doing, which is often not the job you thought you hired them for.
All four features remain in preview during the global rollout, so availability varies by account. Two search engines, two different metrics, both staged, neither audited by anyone outside the company reporting them.
It answers whether your content is appearing, and it declines to answer whether anyone is coming.
The AI answer box stopped being unmeasurable in June 2026, and became partially measurable, in impressions on one side and citation percentages on the other. Google reports whether your link appeared. Bing reports what fraction of the citation space you held. Neither reports what you probably want most, which is whether any of it sent you a customer. Track what is there, and be precise about the gap.
I watch search reporting for the same reason I watch payment rails: it is the plumbing that decides whether the work you already did gets seen. A lot of small businesses in Kansas built a website years ago, got it ranking for a handful of local terms, and reasonably assumed the job was done. What changed in June is that there is now a second scoreboard, and for most properties it is still rolling out, so you may not even know whether you are on it.
The honest position is that this data is thin. Impressions with no clicks, and a citation percentage with no competitor names attached. But thin data you can check daily beats confident guessing, and it beats paying someone for a dashboard that infers AI visibility from the outside when Google and Microsoft now report a version of it directly.
What I would do with an afternoon: log into Search Console and Bing Webmaster Tools, verify the property is properly claimed, and see whether the AI reports are present yet. If they are not, that is the rollout, not you. If they are, note the baseline now, because a baseline taken during a preview is more useful than one taken after everyone has adjusted.
You do not need to understand how a grounding query is constructed to prepare for that, but you do need a site that is actually claimed, properly structured, and answering the questions your customers ask in language a machine can quote.
Questions people are asking
When did Google Search Console add AI visibility reporting?
Google announced Search Generative AI performance reports on June 3, 2026, in a Search Central Blog post. There are two dedicated reports, one for Search covering AI Overviews and AI Mode, and one for generative AI in Discover. Several third-party articles date this to July 2026, which does not match Google's own post.
Does Google Search Console show clicks from AI Overviews?
No. Google's June 3, 2026 announcement lists impressions broken down by page, country, device and date. It does not list clicks, queries, click-through rate, or average position for AI Overviews and AI Mode. Any tool or article promising AI Overview click data is describing something Google has not published.
What exactly counts as an impression in Google's AI reports?
Google defines it as an instance in which a link to your website is shown to a user within a generative AI feature, such as an AI Overview or AI Mode. It is tied to your link being displayed, not to the AI answer simply rendering on a related query.
What is Bing's AI Citation Share, and can I see my competitors in it?
Citation Share is the percentage of citations attributed to your site out of all citations shown across all sites for the same grounding query. Microsoft describes it as an observational metric, explicitly not a ranking system, not traffic share, and it does not expose competitor domains. It sits alongside the existing raw citation count rather than replacing it.
Why should a small business owner care about any of this?
Because AI answers are now a distinct place your business either appears or does not, and for the first time you can check which. It costs nothing to claim your property in Search Console and Bing Webmaster Tools and look. Both feature sets are still rolling out in stages, so if the reports are not visible in your account yet, that is expected, not a problem with your site.
Sources & verification
Every factual claim above traces to one of these. Tiers are defined on our methodology page.
KPW Tech Brief is plain-English analysis of the infrastructure changes underneath AI, search, payments and business records. Every claim sourced and dated.
About Kansas Prairie Webworks. A Central Kansas web design, local SEO, and AI automation studio in Salina. We build and run the systems we write about here. More about us.
Disclosure. This article is for general educational and technology-discussion purposes only. It is not investment, legal, tax, or financial advice. Digital assets, tokenized securities, equities, stablecoins, and related products involve risk. Regulatory status described here reflects publicly available information as of August 19, 2026 and may have changed. Consult qualified professionals before making decisions.