Blog · AI Overviews · 25 Sep 2026
Blog
AI Overviews

AI Overviews:How Googlepicks sources

Since March 2025 Google has answered many searches in Germany with an AI summary above the results. Here is how that summary is built, what it takes for your page to be cited, and what solid research says about the effect on your clicks.

Get in touch
Enquiry
Nikolai Schöbel und Jeremias Burger, Co-Founder Scalableloops

Let's talk about your project.

First we check whether the project fits your business model. Then you get a proposal with phases and effort.

Get your AI visibility checked or call: +49 151 1576 5566
Blog · AI Overviews

Google AI Overviews are AI-generated summaries that Google shows above the classic search results. In Germany they have been live since 26 March 2025, in German and English, for signed-in users aged 18 and over. According to Google, the sources come from the same ranking and quality systems as regular Search: a page has to be indexed and eligible for a snippet, and there are no extra requirements. The effect on traffic is real. In a study by Pew Research Center, users clicked a result on 8 percent of searches that showed an AI summary, against 15 percent on searches without one. Staying visible therefore takes solid SEO, content with genuine substance, and a way of measuring where your brand actually shows up in the answers.

In brief
  • Live in Germany since 26 Mar 2025, in German and English, for signed-in users aged 18+.
  • Google picks sources through its normal ranking systems; the page must be indexed and snippet-eligible.
  • US studies show clearly fewer clicks; Google disputes their methodology.
  • Measurable through the new AI report in Search Console and a dedicated AI visibility analysis.
Published 25 Sep 2026Nikolai Schöbel and Jeremias Burger9 min read
Nikolai SchöbelJeremias Burger

Nikolai Schöbel and Jeremias Burger

Co-founders of Scalableloops GmbH. Nikolai Schöbel leads online marketing and AI strategy, Jeremias Burger the AI architecture. Both build AI systems and train teams on them in their own agency work.

On this page
  1. What are Google AI Overviews?
  2. When did AI Overviews launch in Germany?
  3. How does Google choose sources for AI Overviews?
  4. How do you get your website into AI Overviews?
  5. Do AI Overviews reduce clicks to websites?
  6. Can you keep your site out of AI Overviews?
  7. How can you track your visibility in AI Overviews?
  8. Frequently asked questions
  9. How to check where you stand in AI Overviews
  10. Where the information on this page comes from
Basics

What are Google AI Overviews?

AI Overviews are answer boxes in Google Search that sum up a question in a few paragraphs and link out to websites. In the German interface the feature is called “Übersicht mit KI”. Google shows it when its systems decide that an AI summary will be especially helpful, for example for questions that would otherwise mean reading several sources. For many other searches you still get the familiar list of results.

Two points from Google’s own help pages matter for your planning. Users cannot switch the overview off; they can only use the “Web” filter to see plain links. And Google itself warns that AI answers can contain mistakes and that information should be checked against several sources. For a business, that means what appears about your offer is not automatically correct, and it is worth checking regularly.

The reach is huge. Alphabet CEO Sundar Pichai said in July 2025 that AI Overviews had more than 2 billion monthly users across over 200 countries and territories and 40 languages. Google does not publish how many of them search in Germany.

2 bn

monthly users of AI Overviews, as reported by Google in July 2025, across more than 200 countries and territories.

Alphabet Q2 2025 earnings; TechCrunch
Germany

When did AI Overviews launch in Germany?

Google launched AI Overviews in Germany on 26 March 2025, together with Austria, Switzerland, Belgium, Ireland, Italy, Poland, Portugal and Spain. In Germany they appear in German and in English, for signed-in users aged 18 and over. Google’s English-language blog dates the same announcement 25 March, a difference caused by the time zone of publication.

The date is a useful anchor for your own analysis: comparing search data before and after late March 2025 shows whether clicks have shifted while impressions stayed flat. Keep seasonality and demand in mind, because not every dip in spring 2025 was caused by the overview.

Source selection

How does Google choose sources for AI Overviews?

Google’s description is more down to earth than many guides suggest. Its AI features in Search are rooted in the core ranking and quality systems. To appear as a supporting link in an AI Overview, a page must be indexed and eligible to show a snippet in regular Search. Beyond that, Google says, there are no additional requirements and no special optimisations needed.

One difference from the classic results page is query fan-out. AI Overviews and AI Mode run several related searches across subtopics and data sources for a single question. That lets them show a wider and more varied set of links than one web search would. In practice, a page can be cited because it answers one sub-question particularly well, even if it does not rank at the very top for the main term.

Pew’s US data shows who actually gets cited: Wikipedia, YouTube and Reddit together accounted for 15 percent of the sources linked in AI summaries. Our own AI visibility analyses show a similar pattern: alongside a company’s website, trade portals, directories and press coverage help decide whether a brand appears in the answer. That is an observation from client projects, not a representative study.

What Google saysWhat it means for your website
AI features are rooted in the normal ranking and quality systemsGood SEO remains the foundation; there is no separate AI index
Requirement: indexed and snippet-eligiblePages with noindex or nosnippet cannot be cited
Query fan-out with several sub-searchesAnswer sub-questions clearly and individually, rather than only the head term
No additional requirementsGoogle does not need a special AI file or extra markup
Structured data must match the visible textOnly mark up content that is actually on the page
What to do

How do you get your website into AI Overviews?

Nobody can guarantee a spot in an AI Overview, not even Google. What you can do is create the conditions under which your page is a candidate. Google published a dedicated guide on this in May 2026, and its core message is that optimising for generative search is still SEO. Google explicitly says you do not need new machine-readable files such as llms.txt, content chopped into small chunks, rewrites aimed at AI, or special schema markup. Structured data remains useful for rich results but is not required for the AI features.

What Google does recommend matches what we check in our GEO audit. The difference is the scope: Google only looks at its own Search, while we also measure ChatGPT, Perplexity and other systems that partly work differently. Read more in our articles on AI SEO and on GEO, AEO and LLMO.

  1. 01

    Technical

    Findable

    Keep pages crawlable and indexable, check robots.txt, no nosnippet on pages you want cited.

  2. 02

    Content

    Real substance

    Your own experience, figures and judgement instead of knowledge found on every other site. Google names this as a strength.

  3. 03

    Structure

    Clear answers

    One question per section, with the answer in the first sentences. That is our way of working rather than a Google rule, but it makes content easier to use for readers and systems alike.

  4. 04

    Data

    Consistent details

    Structured data only where it matches the visible text; keep your Business Profile and Merchant Center current if you sell locally or online.

  5. 05

    Reputation

    Get mentioned

    Trade articles, directories and press where your brand appears on its merits. According to Google, bought or artificial mentions tend to backfire.

Clicks

Do AI Overviews reduce clicks to websites?

The most robust figures come from the US. In March 2025 Pew Research Center analysed the browsing of 900 adults, 68,879 Google searches in total. When an AI summary appeared, users clicked a traditional result on 8 percent of visits, compared with 15 percent without a summary. They clicked a link inside the summary on 1 percent of visits. After a page with a summary, 26 percent ended their browsing session, against 16 percent without one.

Ahrefs reached a similar picture by a different route. For 300,000 keywords it compared click-through rates from Search Console data for December 2023 and December 2025. For keywords with an AI Overview, the top result’s click-through rate was 58 percent lower than expected. An earlier analysis from April 2025 had found 34.5 percent.

Google disputes these studies. It called Pew’s approach a flawed methodology with a query set that is not representative of Search traffic, and it says visitors arriving from results pages with AI Overviews tend to spend more time on the site. We have not found a comparable public study for Germany. Expect fewer clicks on purely informational queries, and check the effect against your own data.

StudyDataFinding
Pew Research Center, July 2025900 US adults, 68,879 searches in March 2025Clicked a result: 8% with AI summary, 15% without
Pew Research Center, July 2025Same dataClicked a link inside the summary: 1% of visits
Ahrefs, February 2026300,000 keywords, Dec 2023 vs Dec 2025Position 1 CTR on AI Overview keywords 58% lower
8% vs 15%

How often users clicked a search result when an AI summary appeared, compared with searches that had none.

Pew Research Center, 22 Jul 2025
Control

Can you keep your site out of AI Overviews?

Yes, but not without side effects. According to Google, the nosnippet robots meta tag stops content from being used as a direct input for AI Overviews and AI Mode. It applies to all search results, though, so your page will also appear without a text snippet in regular Search. max-snippet limits how much text may be used, again across all result types.

The Google-Extended token in robots.txt is often misunderstood. It controls whether content is used to train Gemini models, but Google states it has no effect on inclusion in Google Search. AI Overviews draw on what the regular Googlebot crawls. For most businesses, opting out is the wrong move anyway: if you are missing from the overview, your competitors provide the answer.

Measurement

How can you track your visibility in AI Overviews?

Google Search Console now has a dedicated report for generative AI features. It shows how often links to your site appeared in AI Overviews and AI Mode, broken down by page, country and date. It does not show clicks yet. Google is rolling it out gradually, and sites with too few impressions may not see it. Clicks from AI Overviews continue to be counted in the standard Performance report under the “Web” search type.

What Search Console does not show is which questions trigger no overview at all, which competitors and third-party sources get cited instead, and how your brand is described. That is what our AI visibility analysis covers. We run the questions your customers really ask in Google Search and in other AI systems, record how often an AI Overview appears, which sources are cited and whether your brand is mentioned. Because AI answers vary from run to run, we repeat queries before drawing conclusions. Our article on measuring AI visibility walks through the process.

For one engineering company, this kind of analysis showed that narrow technical terms rarely triggered an overview, while questions about standards and applications did so far more often. That is exactly where improving content pays off. It is a single case from our work, not a general rule.

Frequently asked questions

Frequently asked questions

Is “Übersicht mit KI” the same as AI Overviews?

Yes. “Übersicht mit KI” is simply the German name for AI Overviews in Google Search.

When did AI Overviews become available in Germany?

On 26 March 2025, in German and English, for signed-in users aged 18 and over.

Does my website need an llms.txt file to appear in AI Overviews?

Not for Google. Its guide states that no new machine-readable files, AI text files or special markup are needed. Other AI systems may be a different matter, which we cover in our llms.txt article.

Is structured data required for AI Overviews?

No. Google says structured data is not required for its generative AI features. It remains useful for rich results and must always match the visible content of the page.

Will AI Overviews cost me traffic?

US studies suggest so: Pew measured 8 instead of 15 percent clicks on results, and Ahrefs a 58 percent lower click-through rate for position 1. Google disputes the methodology. Track the trend in your own Search Console data.

Can I see whether my pages appear in AI Overviews?

Yes, through the generative AI performance report in Search Console, which shows impressions from AI Overviews and AI Mode. It does not include clicks yet, and not every site has access.

Next steps

How to check where you stand in AI Overviews

  1. 01

    Check the technical basics

    Make sure key pages are indexed and carry no nosnippet or noindex.

  2. 02

    Open Search Console

    See whether the generative AI report is available for your site, and compare clicks before and after March 2025.

  3. 03

    Collect customer questions

    Write down ten to twenty questions prospects genuinely ask before getting in touch, and search for them on Google.

  4. 04

    Close the gaps

    Answer those questions on your website directly and from your own experience. If you want help, our GEO agency handles the measurement, or we train your team in an AI training.

AI Overviews do not reward tricks. They reward pages that answer a real question better than the rest, and if you know what your customers ask, you already hold the most important piece.

or call: +49 151 1576 5566

Further reading
Sources

Where the information on this page comes from

Projekt-Detail

    Got a project in mind?

    We reply personally. First a use-case check, then an architecture proposal.

    Start your inquiry