← Blog

Getting cited in AI Overviews: what you can measure

When Google answers a search with an AI summary, a few paragraphs of text appear at the top of the page with source links beside them. In Türkiye the feature is called AI Bakışı. According to a Webrazzi report dated 18 February 2026, Google started gradually rolling out AI Overviews and AI Mode to users in Türkiye that day.

The marketing team's question is simple: is one of those source links us, or a competitor? This post covers what Google's own documentation says, how to set up measurement and where the measurement can mislead you.

What Google says

Google Search Central's "AI features and your website" page keeps it short. It says there are no additional requirements or special optimizations needed to appear in AI Overviews or AI Mode. To be shown as a supporting link, a page has to be indexed and eligible to be shown in Google Search with a snippet. There are no further technical requirements.

The same page says you don't need new machine-readable files, AI-specific text files or special schema.org markup. The recommendations are familiar: allow crawling in robots.txt and in your CDN settings, use internal links, keep important content in text form, and make sure structured data matches the visible text on the page.

So by Google's description, the road to AI Overviews runs through regular search results. A page that is not indexed, or has its snippet turned off, cannot be a supporting link. That is worth remembering before you open a separate budget line called "content for AI".

Search Console won't show it separately

The first place people look is Search Console. According to the same Google document, traffic from AI Overviews and AI Mode is counted in overall search traffic and appears in the Performance report under the Web search type. You can't use this report to tell which click came from a link in the summary and which came from a classic result.

On the same page Google also says clicks from results pages with AI Overviews are higher quality and that these users tend to spend more time on the site. It shares no data to back that up. Don't treat that sentence as an input for decisions until you see it in your own analytics.

The setup: a keyword list and the results page

Drawing conclusions from a single search misleads you, because the summary doesn't appear on every search and it varies by location and language. A sturdier approach is to read the results pages for the keywords you already track, regularly and with your target market's settings, and record these on every read:

  • Did an AI Overview appear on the results page.
  • If so, which domains the source links point to.
  • Whether your domain is among those sources, and which competitors are there.
  • Your organic position for the same keyword.

The last item matters. If your organic position is good but you're not in the summary, the first place to look is how the page answers the question. If a competitor is in the summary and also ahead of you organically, the job is classic SEO. Without seeing both lists side by side, it's hard to decide which page deserves the effort.

Reading the results

The summary can appear for a keyword one day and not the next. So don't report a single day's read as "we're in the summary now"; talk in terms of a few weeks of data. A ratio per keyword, such as "cited in how many of the last four reads", moves far less than a single-day yes or no.

Also don't record a day when the read failed as "no summary". If the results page could not be read, there is no information for that day. We covered this distinction for another surface: Tracking a competitor's social media

Opting out is also possible

Some brands don't want their content used in the summary. Google points to its preview controls for this: nosnippet, data-nosnippet, max-snippet and noindex. These also affect how the page appears in regular search results. A page that turns off its snippet to stay out of the summary loses its description in the classic result too, so weigh that cost before deciding.

To limit training and grounding use in Google's systems outside Search, the document points to a separate token, Google-Extended. Visibility in Search is governed by Googlebot's robots.txt rules.

Reading the competitor side

The interesting data is on the competitor side. If the summary has been citing the same two competitors for a keyword for weeks, open those pages and look: do they answer the question in the first paragraph, do they use a table, when was the page last updated? The point isn't to copy them. You see which content format works for that question and rebuild your own page accordingly.

Note the week a competitor drops out of the summary as well. Did they change the page, remove it, or did another source take its place? Without knowing that, your reading of your own gains is incomplete.

With Adversee

When Adversee's SEO module reads the results page for tracked Google keywords, it also records SERP features and the AI Overview. The keyword table shows your position next to the positions of the competitors you track, and the weekly AI mention data shows Google AI Overview mentions and source domains for Türkiye. A keyword that could not be read shows as not read; it is never counted as zero.

Start tracking competitor moves automatically.

Getting cited in AI Overviews: what you can measure · Adversee