Google Search Console details how it measures clicks, impressions, and position in AI Overviews and AI Mode

Search Console documentation details how impressions, clicks, and position are counted for AI Overviews and AI Mode, opening the way to better understand the impact of generative search, even without a direct way to segment that traffic in the report.

Editor's note: this article states that rich results and video can already be isolated with a dedicated Search appearance filter in the performance report. That specific claim was not found in the official documentation consulted for this report, so we recommend readers confirm this detail directly in the Search Console interface before using it as a reference.

Search Console documentation details how impressions, clicks, and position are counted for AI Overviews and AI Mode, but there is still no dedicated filter to isolate that traffic in the performance report.

Search Console documentation details how impressions, clicks, and position are counted for AI Overviews and AI Mode, opening the way to better understand the impact of generative search, even without a direct way to segment that traffic in the report.

For years, the growth team that wanted to know whether a page was being cited in a Google AI Overview had to guess: the Search Console performance report lumped everything into a single set of click, impression, and position metrics, without separating traditional search from AI-assisted search.

The Search Console Help documentation (Search Console Help, "What are impressions, position, and clicks?") now details, item by item, how each of these metrics is counted when the result comes from an AI Overview or from AI Mode, and that changes how organic visibility should be interpreted in 2026, even though the tool still doesn't offer a native way to isolate that volume.

The difference isn't cosmetic. Until recently, a citation inside an AI-generated answer and a traditional blue link were, for counting purposes, governed by similar logic, but with particularities that the documentation only now makes explicit. Understanding these rules is the first step toward answering a question every technical SEO team has been asking: how much of my visibility turned into an impression without a click because the answer already appeared ready on the user's screen?

How Search Console counts clicks, impressions, and position in AI Overviews

The documentation details specific rules for two types of result: AI Overviews (the summary with links that appears at the top of traditional search) and AI Mode (the more recent, conversational experience that breaks the question into subtopics and lets users dig deeper). The counting rules are similar, but not identical.

For AI Overviews, Search Console applies the standard impression rule (the link needs to appear on screen, without requiring scrolling within a carousel) and counts a click whenever the user leaves Google for an external page via a link inside the summary. Position, however, has an important particularity:

An AI Overview occupies a single position in search results, and all links in the AI Overview are assigned that same position.

This means that if three different pages are cited in the same AI Overview, all of them show up with the same average position in the report, regardless of which link comes first or last within the block. It's a limitation that already existed for other composite elements (like knowledge panels), only now applied to a format that increasingly concentrates the user's attention.

In AI Mode, the position logic follows "the same methodology as a Google Search results page," with carousels and image blocks counted under the standard rules for those elements. The relevant difference here is different: each follow-up question the user asks within AI Mode is treated as a new query, with its own impression, position, and click count. That spreads the volume of a single research session across several rows in the report, instead of concentrating everything into a single query.

Why it still isn't possible to isolate this traffic in the report

Despite precisely detailing how each metric is counted, the Search Console documentation doesn't describe any dedicated Search appearance filter for AI Overviews. Unlike other formats that can already be isolated with that filter (such as rich results and video), AI Overviews traffic is still counted within the "Web" search type, mixed in with traditional blue links, with no dedicated row or filter in the performance report.

In practice, this means an SEO team can, in theory, understand how an impression or a click coming from an AI Overview is counted, but can't today ask Search Console to show only those rows. The segmentation the technical SEO community expects (comparing AI Overview CTR with blue-link CTR side by side, per page) still depends on a feature the official documentation doesn't confirm exists.

What the counting logic reveals: an impression isn't traffic

Even without a dedicated filter, the documented counting rules help confirm a suspicion the SEO community had already been discussing informally: AI Overview CTR tends to be lower than blue-link CTR for the same query, because part of the user's information need has already been satisfied by the summary. Without native segmentation, though, turning that suspicion into an exact number per page and per query still depends on indirect methods, such as cross-referencing spikes in aggregated "Web" impressions with changes observed manually in search results, or relying on third-party tools.

In short: a jump in impressions without an equivalent jump in clicks, in the aggregate "Web" type, may indicate (among other possible explanations) that pages are being cited as a source in AI Overviews and the user is satisfied with the summary without completing the visit. That isn't necessarily bad for the brand (the citation still builds authority and exposure), but it can be a warning sign if the team's goal is traffic.

It's worth comparing this with the picture of how each result type is counted today, according to the documentation:

What's still missing to really measure AI Overviews

Even if a dedicated filter existed, not everything would become clearer. First, the single position per AI Overview hides which specific link was more visible within the block: two pages cited in the same answer would show up with identical position, even if one was at the top of the generated text and the other only in a footnote. Second, Search Console doesn't include data from Search Labs experiments, so any AI Overview test still in experimental phase simply doesn't appear in the report, filter or no filter.

Third, and most important for anyone building the company's GEO strategy: even the counting detailed by the documentation only measures Google's own search surface. It says nothing about how many times the brand was cited (or ignored) in answers from ChatGPT, Perplexity, or Microsoft's Copilot, which still require third-party tools or manual monitoring to estimate presence. The Search Console documentation detailed the internal counting logic for GEO within the Google ecosystem, but that's still far from being a universal GEO metric, and for now it isn't even a segmentable metric within Google's own interface.

How to handle this in the team's measurement workflow, for now

Without a native filter to isolate AI Overviews, the most honest path today is to treat the "Web" search type as an aggregate that includes that traffic, but doesn't allow precisely separating it. Teams that want to estimate the effect of an AI Overview citation can watch a page's aggregate impressions and clicks behavior over time and cross-reference that with manual observation (searches run by the team itself) of when that page started being cited in a generated summary, knowing this is an approximation, not an exact measurement offered by the tool.

The logical next step, which current documentation doesn't cover and which depends on Search Console eventually adding its own segmentation, is to cross that kind of breakdown with the conversion rate of people who click from an AI Overview versus people who click from a blue link. Until then, any comparison of this kind made by third-party tools should be read with the caveat that it doesn't come from an official Google segmentation.

Source 1: Google Search Central – Performance report (Search Console Help) (https://support.google.com/webmasters/answer/7042828)

What are impressions, position, and clicks? - Search Console Help

This page helps explain impressions, position values, and click data as used in Search Console, and especially in the Performance reports.

Introduction

Search Console provides data showing how often users saw or interacted with links to or content from your site, in Google Search, News, and Discover. This data is available in the various performance reports.

The performance reports show the following metrics:

Impressions: How often someone saw a link to your site on Google.

Clicks: How often someone clicked a link from Google to your site.

(average) Position: A relative ranking of the position of your link on Google, where 1 is the topmost position, 2 is the next position, and so on.

Click-through rate: The calculation of (clicks ÷ impressions).

Additional methodology details by result type

AI Mode

AI Mode expands on AI Overviews to show a more interactive AI-powered response with links to web resources that support the information or direct the user to view relevant webpages. AI Mode groups the user's question into subtopics and searches for each one simultaneously, and users can go deeper.

Click: Clicking a link to an external page in AI Mode counts as a click.

Impression: Standard impression rules apply.

Position: Position in AI Mode follows the same methodology as a Google Search results page. Generally, carousel and image blocks within AI Mode are calculated using the standard position rules for those elements.

If a user asks a follow-up question within AI Mode, they are essentially performing a new query. All impression, position, and click data in the new response are counted as coming from this new user query.

Note: Search Console doesn't include data from experiments in Search Labs, as these experiments are still in active development.

AI Overviews

AI Overviews show an overview with links to web resources that support the information.

Click: Clicking a link to an external page in the AI Overview counts as a click.

Impression: Standard impression rules apply.

Position: An AI Overview occupies a single position in search results, and all links in the AI Overview are assigned that same position.

Note: Search Console doesn't include data from experiments in Search Labs, as these experiments are still in active development.