Google AI Overviews now appear in roughly 48% of all Google searches, up 58% year on year as of February 2026. For marketers and SEO leads, that figure is not an abstraction. It means the most visible position on the results page is increasingly occupied by a generated summary, and the organic listings that used to capture first-click attention have been pushed further down the screen.
The instinct many teams have is to look for a new optimisation playbook, a set of AI-specific signals to target. The evidence does not support that. Google’s own documentation confirms that the same core ranking and quality systems powering normal Search also determine which pages are cited in AI Overviews. The real question is not what new signals exist, but why some pages get cited and others do not, even when both rank well.
This guide explains how AI Overviews choose sources, what structural and content changes improve citation chances, why ranking first helps but does not guarantee inclusion, and how to measure performance in a way that supports actual business decisions.
TL;DR: Google AI Overviews use the same core Search signals as normal rankings, so classic SEO remains foundational. Getting cited also requires passage-level answer clarity, broad topic coverage, and citation-friendly formatting. Ranking #1 improves the odds but does not guarantee inclusion. Measurement should track visibility and commercial outcomes, not clicks alone, because AI Overviews consistently reduce click-through rates even for cited pages.
What Are Google AI Overviews?
Google AI Overviews are AI-generated summaries that appear at the top of the Search results page for eligible queries. They synthesise information from multiple sources and display citations inline, allowing users to read a direct answer without scrolling to organic listings.
Three things define how they work in practice:
-
Position: AI Overviews sit above the traditional #1 organic result, making them the functional top of the SERP for any query where they appear.
-
Scope: They are most common on informational and research-intent queries, though Google continues to expand the query types they cover.
-
Source attribution: Each summary cites the pages it drew from, which is where the citation opportunity sits for publishers.
The commercial concern is not the technology itself but the traffic shift it creates. Pew Research found that click-through rates drop from 15% to 8% when an AI summary is present on the results page. That is a 47% relative decline in clicks, on any query where an AI Overview appears.
| Metric | Without AI Overview | With AI Overview |
|---|---|---|
| Average CTR (Pew Research, 2025) | 15% | 8% |
| Position-1 CTR (Ahrefs, Dec 2025) | 0.073 | 0.016 |
| Organic CTR on informational queries (Seer Interactive, Sep 2025) | 1.76% | 0.61% |
For pages that used to rely on informational traffic to feed commercial funnels, this is a material shift. Understanding the source-selection process is the first step to addressing it. Kobestarr Digital’s AEO guide covers the underlying mechanics in depth.
How Does Google Choose AI Overview Sources?
Google has been explicit on this point. According to Google Search Central, “the same core ranking and quality systems that power normal Google Search also determine which pages are cited and surfaced in AI Overviews.” There is no separate AI Overviews algorithm to reverse-engineer.
That said, the selection process has two distinct stages, and conflating them is where most optimisation strategies go wrong.
Stage 1: The Candidate Set
The candidate set is the pool of pages Google considers eligible for an AI Overview on a given query. Entry requirements mirror standard organic eligibility:
-
The page must be crawlable and indexed.
-
It must not block snippet generation via
nosnippetormax-snippet:0directives. -
It must meet Google’s core quality and relevance thresholds for the query.
-
It must satisfy the language understanding systems Google uses, including BERT, MUM, and Neural Matching.
Ranking well for the target query is the most reliable way to enter the candidate set, but it is not the only route. Pages outside the top 10 can still be cited if they provide a better-matched passage for a specific sub-question the Overview is synthesising.
Stage 2: Citation Selection
Once a candidate set is assembled, Google selects which pages to cite based on how well individual passages serve the summary being generated. This is where the overlap between organic rankings and AI Overview citations has been declining.
| Factor | Candidate Set Entry | Citation Selection |
|---|---|---|
| Crawlability and indexability | Required | Assumed |
| Organic ranking position | Strong predictor | Helpful, not decisive |
| Passage-level answer clarity | Moderate influence | High influence |
| Topical breadth and depth | Moderate influence | High influence |
| Structured formatting (lists, tables, FAQs) | Low influence | High influence |
According to Search Engine Journal, covering related subtopics now carries more weight in citation selection than holding one strong ranking for a single keyword. The practical implication is that a page ranking 8th with well-structured, passage-ready answers can outperform a page ranking 2nd with dense, unstructured prose.
If this shift is affecting existing traffic, Kobestarr Digital’s analysis of is your traffic dropping because of AI Overviews explains how to diagnose the pattern in Search Console.
How Do You Optimise for AI Overviews?
Optimisation for AI Overviews is not a separate discipline. It is a refinement of existing SEO practice, applied with a clearer understanding of what citation selection rewards. The following framework covers the four layers that determine whether a page moves from the candidate set into active citation.
1. Secure Technical and Organic Eligibility First
No amount of passage-level work matters if the page cannot be crawled, indexed, and ranked. Before any structural changes:
-
Confirm the page is indexed and not blocking snippets.
-
Check internal linking so the page carries sufficient authority signals.
-
Verify that the page satisfies the search intent for the target query, not just the keyword.
-
Ensure the content is fresh enough for time-sensitive queries, since Google’s systems weight recency for rapidly evolving topics.
2. Structure Passages for Direct Answers
The single most actionable change for most pages is restructuring how answers are written at the passage level. AI Overviews extract passages, not pages. A passage that cannot be lifted cleanly and read in isolation will not be cited, regardless of how well the page ranks.
The passage pattern that works:
-
Open each major section with a 40-60 word direct answer to the question the heading poses.
-
Follow with supporting evidence, data, or context that expands on the answer.
-
Use question-shaped headings (H2 and H3) so Google’s systems can match the passage to user queries.
-
Avoid pronoun-heavy references that require surrounding context to make sense (“this approach,” “as mentioned above”).
3. Broaden Topical Coverage
Citation selection rewards pages that cover a topic cluster comprehensively, not pages that optimise for a single keyword. Search Engine Journal’s analysis found that covering related subtopics and formats carries more weight in AI Overview citation than holding one strong ranking for a narrow term.
In practice, this means:
-
Map the full question cluster around the target topic, including adjacent questions users ask on forums and in Reddit threads.
-
Add FAQ sections that address those adjacent questions directly on the page.
-
Use tables and comparison lists for queries where users are weighing options.
-
Include schema markup (FAQPage, HowTo, Article) to make structured content machine-readable.
4. Use Evidence and Attributed Sources
AI Overviews favour pages that cite original evidence, named studies, or authoritative sources. Vague claims (“many experts agree”) are harder to lift than attributed facts (“according to [named source], X% of queries now trigger AI Overviews”). Writing that mirrors real user language, including the concerns visible in forum threads, tends to match Google’s passage-retrieval patterns more reliably.
Pre-publication citation checklist:
-
Every major claim links to a primary or authoritative source
-
Statistics include the source name and year
-
Question-shaped headings match real user queries
-
Each section opens with a self-contained 40-60 word answer
-
FAQPage or HowTo schema is implemented where relevant
-
No snippet-blocking directives are present
For teams working across multiple AI search channels, the same structural principles apply to other platforms. Kobestarr Digital’s guide on how to rank in ChatGPT covers the retrieval mechanics specific to that environment.
Tools that track AI Overview citation rates can help prioritise which pages to restructure first. Searchable monitors AI visibility across Google AI Overviews, ChatGPT, and Perplexity, giving page-level citation data that is otherwise difficult to extract from Search Console alone.
Does Ranking #1 Get You Into AI Overviews?
Ranking first remains the strongest single predictor of being cited in an AI Overview, but it is not a guarantee. The data makes the distinction clear.
Research published in 2026 shows that the #1 organic result is cited in AI Overviews approximately 43% of the time. By position 20, that citation rate falls to around 7%. High rankings improve the odds considerably, but the majority of #1 results are still not cited for the queries where an AI Overview appears.
The overlap between top-10 organic rankings and AI Overview citations has also been declining. Studies from Search Engine Journal put current overlap at 17-38%, down from approximately 54% in late 2025. This means that for a growing share of AI Overview citations, the cited page is not in the top 10 organic results for that query.
| Belief | Reality |
|---|---|
| Ranking #1 guarantees AI Overview inclusion | #1 results are cited only ~43% of the time |
| Only top-10 pages can be cited | Pages outside the top 10 are cited in 62-83% of cases |
| Better rankings always mean more clicks | Position-1 CTR dropped 58% when AI Overviews are present (Ahrefs, Dec 2025) |
| A separate AI Overview strategy is needed | Google confirms the same core Search systems apply |
The most accurate mental model is this: organic rankings buy eligibility and improve citation probability, while citation-ready passage structure wins the actual inclusion. Chasing rank alone, without addressing passage clarity and topical depth, leaves a significant portion of the citation opportunity on the table.
How Do You Measure AI Overview Performance?
Measurement is where most teams fall short. Tracking clicks alone will show a decline almost regardless of whether the AI Overview strategy is working, because AI Overviews reduce CTR structurally. The right framework tracks three layers.
Layer 1: Visibility
Google Search Console now includes dedicated reporting for AI features, logging impressions and clicks from AI-generated results separately. Filter by “AI Overviews” in the Search Type dropdown to see which queries are triggering AI Overview appearances and whether the site’s pages are being cited.
Key signals to track:
-
AI Overview impressions by query
-
Click rate on AI Overview citations versus standard organic listings
-
Query cohorts where AI Overviews appear but the site is not cited (the citation gap)
Layer 2: Citation Proxies
Search Console does not yet show citation rates directly. Proxy signals include:
-
Branded search volume trends (cited pages often drive brand queries downstream)
-
Direct and assisted conversion rates on pages known to be cited
-
Changes in query impressions after structural content updates
Layer 3: Commercial Outcomes
Lower CTR does not automatically mean lower value. The relevant question is whether cited pages outperform uncited pages on the same query type, and whether the traffic that does arrive converts at a higher rate because it is more informed.
| Measurement Layer | What to Track | Tool |
|---|---|---|
| Visibility | AI Overview impressions, query triggers | Google Search Console |
| Citation proxies | Branded uplift, assisted conversions | GA4, Search Console |
| Commercial outcomes | Conversion rate on AI-cited queries | GA4 with query annotations |
| Competitive citation gaps | Queries where competitors are cited and you are not | Searchable |
Teams working at scale across AI search channels will find it worth separating AI Overview performance from standard organic reporting entirely. Kobestarr Digital’s AEO agency service builds these measurement frameworks as part of a structured visibility programme.
Key Takeaways
-
AI Overviews use traditional Search signals for eligibility. Google confirms no separate ranking formula exists. Crawlability, indexability, and core quality signals still determine whether a page enters the candidate set.
-
Citation wins require more than a strong ranking. Passage-level answer clarity, topical breadth, and structured formatting (lists, tables, FAQs, schema) are the primary factors in citation selection.
-
Ranking #1 improves citation probability but does not guarantee it. The #1 result is cited approximately 43% of the time. Pages outside the top 10 account for 62-83% of AI Overview citations.
-
AI Overviews reduce clicks structurally. Tracking clicks alone will show a decline regardless of strategy. Measurement should cover visibility, citation proxies, and commercial outcomes separately.
-
The candidate-set vs citation-selection distinction matters. Optimising for one without the other leaves significant citation opportunity unaddressed.
-
Measurement is now available directly in Search Console. Google’s AI features reporting allows query-level tracking of AI Overview impressions and citations for the first time.
Frequently Asked Questions
How do I appear in Google AI Overviews?
Ensure the page is indexed and snippet-eligible, then structure content so individual passages can be extracted and read in isolation. Use question-shaped headings, open each section with a concise 40-60 word direct answer, and implement FAQPage or HowTo schema where relevant. Broad topical coverage and attributed evidence improve citation selection beyond organic rank alone.
Do Google AI Overviews use the top organic results?
Partly. High organic rankings are the strongest predictor of citation, but they are not the only route. Studies show that 62-83% of AI Overview citations come from pages outside the top 10 organic results for the same query. Citation selection rewards passage clarity and topical depth, not rank position alone.
Can you opt out of appearing in Google AI Overviews?
There is no direct opt-out for AI Overviews specifically. Google’s documentation states that pages can limit the content shown by using nosnippet, data-nosnippet, or max-snippet directives, but these controls affect all snippet-based features, not AI Overviews in isolation. Removing indexation entirely would also remove the page from all Search results.
Do AI Overviews reduce clicks to my website?
Yes, consistently. Pew Research found CTR drops from 15% to 8% when an AI summary is present. Seer Interactive recorded a 61% drop in organic CTR on informational queries in September 2025. The relevant question is not whether clicks fall, but whether cited pages outperform uncited ones and whether conversion signals improve despite lower raw traffic.
What content wins citations in AI Overviews?
Content that combines strong organic eligibility with citation-ready structure performs best. Specifically: question-shaped headings, self-contained passage answers of 40-60 words, comprehensive coverage of adjacent subtopics, structured formats (tables, numbered lists, FAQs), attributed statistics, and schema markup. Pages that mirror real user language, including the phrasing found in forums and search threads, tend to match Google’s passage-retrieval patterns more reliably.
About the author: Kobi Omenaka is the founder of Kobestarr Digital and a specialist in Answer Engine Optimization and AI search visibility. Kobi works with in-house marketing teams and business owners to build citation presence across Google AI Overviews, ChatGPT, and Perplexity, using transparent frameworks tied to agreed commercial KPIs.
If AI Overviews are affecting traffic to pages that matter commercially, the starting point is understanding which queries are triggering them and whether the site is being cited or bypassed. Get a free AI visibility audit to see where the gaps are.