AI Overviews in Search Console: why growth teams need to recalibrate KPIs

Google confirms that impressions and clicks from AI Overviews and AI Mode are included in the performance report, and ignoring this leads to attributing click drops to the wrong channel.

Google Search Central's official documentation on AI features and your website, updated on December 31, 2025, makes one thing clear for anyone who lives off metrics: the clicks and impressions coming from AI Overviews and AI Mode are already inside the Search Console performance report, mixed in with the organic traffic you've always looked at. The problem isn't technical, it's one of interpretation. Many growth teams in Brazil keep reading the "Clicks" chart as if the SERP were still the 2022 one, when the results page was a list of ten blue links.

What Google actually says about measurement

The central point of the page is this: sites that appear in AI features are counted in the overall search traffic of Search Console, under the "Web" search type in the Performance report. In other words, there is no (yet) isolated, highlighted line saying "this came from AI Overview." Everything goes into the same bucket of impressions and clicks from organic search.

This changes how you read the number. When a query triggers an AI Overview, user behavior on the SERP is different: part of the answer is already in the generated block, and the click to the site only happens if the user wants to dig deeper. Google even states that these clicks tend to be higher quality, in the sense that the user usually spends more time on the site:

"We've seen that when people click from search results pages with AI Overviews, these clicks are higher quality (meaning, users are more likely to spend more time on the site)."

-- Google Search Central

Keep that sentence in mind, because it's the technical argument for not evaluating the new search experience by click volume alone.

Why the campaign KPI attributes the drop to the wrong channel

The typical failure scenario is this: an AI Overview starts appearing for a set of informational queries that used to bring in a lot of clicks. The CTR of those queries falls, because the user resolves the question right in the block. On the team's consolidated dashboard, this turns into a red line reading "organic traffic dropping."

Without separating what is AI Overview from the rest, the analyst tends to look for the culprit in the wrong place: blaming the latest content update, blaming a competitor, blaming a supposed penalty. The real cause is structural, it's the SERP that changed format.

Reading it correctly requires cross-referencing two sources, as the documentation itself suggests by pointing to the guides on analyze traffic changes and combine Search Console and Analytics data:

• Impressions and clicks in Search Console, filtered by query and by the Web type.
• Post-click behavior (time on page, conversion) in Google Analytics.

If impressions stay flat or rise, clicks fall, but time per session and the conversion rate of those who clicked go up, you haven't lost demand: you've changed the profile of who's arriving. That's a data point that should be in the growth report instead of a lone red arrow.

A recalibration roadmap that makes sense

The path I'd follow to avoid misreading next month's data:

1. Freeze a baseline per query. Export clicks, impressions and CTR for your top informational queries before touching anything. Without a before, there's no measurable after.
2. Separate informational intent from transactional intent. AI Overviews show up more for complex topics or questions; more elaborate comparisons tend to trigger AI Mode, not AI Overviews. Mixing the two categories in the same average hides the effect of each one.
3. Add time on page and conversion to the SEO dashboard. If the click is worth more, the KPI can't be just clicks. Bring the Analytics quality metric into the same dashboard.
4. Rewrite the campaign goal. Instead of "grow organic clicks by X%," it's worth aiming for "maintain/grow organic conversions" and "appear as a supporting link in AI Overviews for target queries."

A technical detail from the source helps here: both AI Overviews and AI Mode use the query fan-out technique, firing off several related searches into subtopics to build the answer. In practice, this means one of your pages could be pulled in as a supporting link for a variation of the query you weren't even closely monitoring. Expanding the list of queries you track becomes part of the job.

What does NOT change (and this is where Google cuts the hype)

The part that disarms magical promises from agencies: there's no special optimization for AI Overviews. The documentation is explicit that there's no additional technical requirement, no "AI text" file, no new schema.org markup, and no specific markup to appear in these features.

To be eligible as a supporting link, a page only needs the usual:

• be indexed and able to display a snippet in search;
• have crawling allowed in robots.txt and in the CDN/hosting infrastructure;
• have important content in text format, with structured data that matches the visible text;
• offer a good page experience and be discoverable through internal links.

In other words, the fundamentals of SEO remain fully valid. Anyone selling "optimization for SGE" as a secret technique is selling something the source itself says doesn't exist.

When to limit your presence in AI features

There's the other side: what if you don't want your content feeding these blocks? Google doesn't offer a standalone opt-out button for AI Overviews within search, because, in the documentation's words, AI is "integral to how Search functions." The control is the same as for the traditional snippet:

| Goal | Control | |---|---| | Limit what appears in the preview/snippet | nosnippet, data-nosnippet | | Limit the size of the displayed excerpt | max-snippet | | Remove the page from search | noindex | | Block training/grounding in other Google systems | Google-Extended |

The trade-off is harsh: using nosnippet or noindex to escape AI Overview also drops your eligibility for the regular snippet and, ultimately, for indexing itself. You don't just cut AI out, you cut organic clicks along with it. That's why, in most cases, opting out isn't worth it, and the effort is better invested in measuring correctly what AI is doing with your traffic, not running from it.

If you've applied the preview controls and the content still shows up, the source recommends using the URL Inspection tool to check the HTML the Googlebot received and allowing time for recrawling, which can take from days to months. It's the same logic as always: nothing happens before Google crawls again.

The takeaway for publishers in Brazil

The practical change isn't in code, it's in how you read data. Verifying the site in Search Console and looking at the Performance report already shows the effect, but only those who set a baseline, cross-referenced it with Analytics and adjusted the campaign goal will be able to say whether that falling chart is a problem or just the SERP behaving in a new way. Everything else will keep pointing at the wrong channel.

Source 1: Google Search Central, AI Overviews and your website (https://developers.google.com/search/docs/appearance/ai-overviews)