Google AI Overview consulting: be the page AI Overviews cite as a source.

We wrote this guide for companies that want their site shown as a source in Google’s AI Overviews, or want to control how their content is used in these answers. Every mechanism we describe for Google AI Overviews and AI Mode is based on Google Search Central documentation, Search Console and Google Search help pages, and Google’s official blog announcements. We verified product names and controls against the source as of September 2026; because Google updates this area often, we check them again during setup. Our own field observations are marked separately under the heading “Expert note”.

This page covers only the Google-specific mechanism, controls and measurement. The broader framework that also covers ChatGPT, Perplexity and Copilot is on our GEO consulting page.

Google AI Overviews vs AI Mode: what is the difference?

A Google AI Overview is the artificial intelligence (AI) answer that Google Search shows above the results for some queries, giving the gist of a topic with links to supporting web pages;¹¹ AI Mode is a separate, conversational search mode that moves forward through follow-up questions.¹⁴ According to Google, AI Overviews appear only for queries where they add value beyond classic Search, so most of the time they don’t trigger;¹ AI Mode is positioned for questions that need more exploration, reasoning and complex comparison.¹ Because the two features may use different models and techniques, the answers and links they show for the same question can differ too.¹

AI Overviews opened to everyone in the US on 14 May 2024,⁹ and AI Mode began opening in the US without a Labs sign-up on 20 May 2025.¹⁰ In Türkiye, both features began rolling out gradually on 18 February 2026, in the Google app and on the results page, under their Turkish interface names AI Bakışı and AI Modu.¹¹ The current availability lists in Google Search Help also include Türkiye and Turkish for both features.¹³¹⁴ According to Google’s June 2026 announcement (updated 31 August 2026), AI Overviews have passed 2.5 billion monthly active users and AI Mode has passed one billion monthly users.⁸

Searches for “google ai overview”, “AI Overviews” and the Turkish “AI Bakışı” all describe the same feature.¹¹ Some writers now describe these answers as a step toward superintelligence (ASI); in practice they are AI summaries built from pages in Google’s index, and the same rules apply to them as to Search. No special optimization is needed to appear in these features;¹ our work consists of applying the defined requirements and controls in full and measuring the result.

How Google builds the answer: grounding and query fan-out

Google builds the AI answer by grounding it in pages retrieved from its own search index and, when needed, by expanding the question into several related searches (query fan-out).² As Google describes retrieval-augmented generation, the core ranking systems retrieve relevant, fresh pages from the index, the model builds the answer from the information on those pages, and it links to the pages that support it.²

In Google’s own example, the question “how to fix a lawn full of weeds” fans out into sub-queries such as “best weed killers for lawns”, “removing weeds without chemicals” and “how to prevent weeds in a lawn”.² With this technique, AI Mode breaks the question into subtopics and searches each one at the same time;⁷¹⁰¹⁴ Google Türkiye explains the same technique in its Turkish announcement under the name “sorgu yelpazesi”.¹¹ Because AI Overviews and AI Mode can use this technique, they can show a wider and more varied set of links than classic Search.¹

That’s why a page can be retrieved and cited through one of the sub-queries rather than the main query. We build the question map from the sub-questions of a decision process, too: comparison, conditions, cost items, implementation steps.

How to appear in AI Overviews: what a page needs to be a source

To be a supporting link in an AI Overview or in AI Mode, a page must be indexed, be eligible to be shown in Google Search with a snippet, and the site must be set to included in Search Console’s generative AI control. The first two are general technical requirements of Search;¹ Google’s generative AI guide adds the third.² The default value of this control is included.⁵ Meeting the requirements doesn’t guarantee being shown; Google says so clearly.¹

In the eligibility audit, we check the following in order:

1. Crawl permission: we confirm that Googlebot is allowed to crawl in robots.txt and at the CDN or hosting layer; Google counts these two layers together.¹
2. Index status: we use the URL Inspection tool in Search Console to check that target pages are indexed and that the right canonical URL has been chosen.
3. Snippet eligibility: we scan for any leftover nosnippet, or its equivalent max-snippet:0, on the page or in the HTTP header, and make sure the theme and plugins aren’t adding these rules.³
4. Content as text: we check that important information sits on the page as text, not inside images.¹

How do snippet controls affect AI Overviews?

Google’s snippet controls also cover AI Overviews and AI Mode: nosnippet stops content from being used as direct input to these features, and max-snippet limits how much of it can be used as input; the limit doesn’t apply where use is separately permitted through structured data or a license.³ When nosnippet is set, the text snippet in web search, Images and Discover is removed as well.³ Crawling, on the other hand, is governed by robots.txt rules written for Googlebot.¹

data-nosnippet removes a specific part of the page from snippets and is valid only on span, div and section elements; Google asks that this attribute not be added or removed later with JavaScript, but written in when the element is added to the DOM.³ When rules conflict, the more restrictive one applies: if the same page has both max-snippet:50 and nosnippet, nosnippet wins.³ All these rules are read when the page is crawled; rules on a page blocked from crawling by robots.txt aren’t seen and are ignored.³

What does the generative AI control in Search Console do?

The Search generative AI control is a Search Console setting where you choose, at property level, whether your site’s links and content can appear in AI Overviews, AI Mode and generative AI features in Discover.⁵ Google began testing the control with a group of site owners in the United Kingdom in June 2026;⁸ as of 31 August 2026 it opened it to all sites worldwide.⁵ The page under Settings has three options:

• Include: content can appear as links in these features and can be used to inform answers; this is the default.⁵
• Exclude: the site’s links and content don’t appear in these features, and the content isn’t used as input to generate answers; no impressions or traffic come from these features either.⁵
• Inherit from parent property: the child property follows the value of the nearest parent property that was set manually; this is the default for properties that have a parent.⁵

According to Google, this control isn’t used as a ranking or inclusion signal in other parts of Search, and it doesn’t override your preferences in services such as Merchant Center and Google Ads.⁵ An exclusion usually takes effect within 1–2 days of the control being switched on; because of caching, it can take longer for some content.⁵

What kind of content does Google expect for AI Overviews?

Google expects content that readers find original, compelling and useful; according to its guide, producing such content will most likely do more for a site’s presence in generative AI search over the long term than all its other recommendations combined.² The guide describes this under the heading of non-commodity content: commodity content repeats what everyone already knows, while non-commodity content offers an expert or experience-based perspective that goes beyond general knowledge.²

Google also recommends content divided into sections with clear headings, and high-quality images and video where relevant; images and video open up extra opportunities to appear in AI features.²

We build the question map from questions heard by sales and support, Search Console queries and question data in Semrush. We add sub-questions that have no answer on the page and strengthen weak answers with the company’s own experience and data.

What do E-E-A-T and structured data do for AI Overviews?

E-E-A-T (experience, expertise, authoritativeness and trustworthiness) describes the qualities that the core ranking and quality systems, which AI Overviews also rely on, look for in content;²¹² structured data, meanwhile, isn’t required to appear in AI Overviews (there’s no special schema.org markup) and continues to be used for rich results.¹² The most important of these qualities is trust.¹² E-E-A-T isn’t a ranking factor on its own; but Google states that on topics that could significantly affect people’s health, financial stability, safety or the well-being of society (YMYL), its systems give more weight to strong E-E-A-T.¹²

The practical route Google suggests is asking “Who, How and Why” of your content: is it clear who wrote it, is there a byline where one is expected, how was the content produced and why.¹² So that the answers to these questions can be read on the page, we make author or expert details, the method and the sources visible; this page’s source list is an example of that in practice.

Google wants structured data to match the text visible on the page.¹ For sites with product and local business information, keeping Merchant Center and Google Business Profile details up to date is also among its recommendations;¹ AI answers can include product listings and local business information too.²

Which “GEO tactics” does Google consider unnecessary?

According to Google, shortcuts such as setting up AI text files like llms.txt, splitting content into small chunks and collecting artificial “mentions” can be ignored for Google Search; many of these recommendations are ineffective or not supported by how Search actually works.² From Google’s point of view, optimizing for generative AI search is still SEO.²

How do you measure visibility in AI Overviews?

AI visibility on Google is read in two places in Search Console: the generative AI performance report shows impressions in these features separately, while the general performance report counts traffic from these features within the “Web” search type.¹⁶ The generative AI report shows how many times links to your site were displayed in AI Overviews and AI Mode, broken down by page, country, device and date, and doesn’t include Search Labs experiments; for sites excluded from these features or with too few impressions, the report may not appear.⁶

A click on a link in an AI Overview counts as a click; all the links in the overview get the same position value.⁷ In AI Mode, each follow-up question is recorded as a new query.⁷ Google points out that conversions and time on site can also be tracked with tools such as Google Analytics.¹ We filter the pages that get AI impressions as landing pages in GA4 and compare their organic sessions and key events with the previous period; we monitor markention.com the same way, with Search Console, GA4 and Data Studio dashboards.

What should a Google AI Overviews specialist know?

Expertise in visibility on Google’s AI features means reading Google’s own documentation and controls correctly: which page is eligible to be a source, how snippet rules affect the overview, and which report counts the visibility. In the order we look for it in a specialist, the knowledge on this page comes down to this:

Tools used in Google AI Overview work

The primary source for visibility on Google’s AI features is Search Console; only Google can provide impressions in the overview broken down by page, country and device.⁶ Alongside it come tools that audit snippet rules at page level, a keyword-level measurement of overviews, and analytics that show what the visit contributes.

The card on the right comes from the latest keyword tracking measurement: it shows for how many tracked keywords a Google AI Overview appears and for how many the site is a source in it. The gap between the two numbers is the next task list; for every keyword with an overview where we aren’t a source, the sources shown are recorded and the page is reworked around that sub-question. The card on the left splits GA4 referrals by AI assistant.

Alongside the Google-specific measurement, the same question set is run on other answer engines. On each analysis run, the GEO screen compares mention rate, source citations to the site, share of voice by brand and the breakdown by engine with the previous run, so whether a question where we are a source on Google is also cited elsewhere can be read in one place. The general method is described on our GEO consulting page.

The first 30 days: how does Google AI Overview work progress?

The job of the first 30 days is to clean up eligibility, build the question map and lay the groundwork for measurement.

1. Week 1 · Eligibility audit: crawl permission, index status, snippet rules, the generative AI control in Search Console and the Google-Extended preference are gathered into one report.
2. Week 2 · Question map: main questions and sub-questions are worked out for priority pages; the sources AI Overviews show today are recorded.
3. Week 3 · Implementation: technical fixes are made, priority pages are restructured to give the answer first, and author and source details are made visible.
4. Week 4 · First reading: Search Console reports and the question set are read again; next month’s list of pages and questions comes out of it.

Changes to snippet rules and pages take effect on recrawl; Google notes that crawling can take anywhere from a few days to a few months.¹

Google AI Overviews consulting pricing: what drives the cost?

The price of this work depends on the number of priority pages, the breadth of the question map, the scope of technical fixes to snippet rules and crawling, how much content rework is included, and the monthly effort assigned to the project. Instead of a fixed price list, we share estimated ranges; we work out the scope together in the first meeting and send a written proposal.

A one-off audit and monthly monitoring can be discussed as separate items; estimated ranges are in the table below.

Google AI Overviews consulting pricing: estimated ranges

Estimated range, excluding VAT; rows marked “/ month” are monthly, the others are one-off. Scope is confirmed in a written proposal after the first meeting. As of October 2026.

SEO, GEO or AI Overviews work: which do you need?

The three engagements share a foundation, but they cover and measure different things. Where to start depends on where your site stands today:

We can plan two or all three together without repeating the shared work.

How to choose an AI Overviews agency or consultant

When choosing a partner for Google AI Overviews, concrete answers to three questions make the decision easier: how is snippet eligibility verified, how are sub-queries found, and which report measures visibility? We add this checklist:

1. Source-based method: every recommended task has a basis in Google’s own documentation; the “GEO tactics” Google considers unnecessary are not sold as paid work.²
2. Guarantee language: no promise of a place in AI Overviews or a set number of impressions; Google chooses the overview’s sources at query time.
3. Access and ownership: Search Console, GA4 and the site stay in your accounts; the partner is added as an authorised user.
4. Decisions stay with you: usage preferences such as nosnippet, Google-Extended or the generative AI control change only with your approval.
5. What the report contains: generative AI impressions in Search Console, overview status per keyword and the work done that month, reported together.
6. Exit terms: the question map, the measurement history and the list of changes are handed over when the engagement ends.

You are welcome to use this list when you talk to us as well. To see how AI assistants beyond Google describe your site, you can also run our free AI visibility analysis. If you’d like to read your site’s place in AI Overviews today together, pick a meeting day that suits you below.