Category: SEO | Reading time: 12 minutes | Last updated: April 2026
In late 2023, Google started testing a feature that would eventually reshape the entire search landscape. By 2025, AI Overviews (Google’s AI-generated answer boxes) were appearing on a substantial share of search queries. If you publish content to the web, the impact is already visible: traffic dropping, click-through rates collapsing, and the rules of SEO shifting under your feet. This is not a passing trend. AI Overviews are now a permanent part of Google Search, and they are fundamentally changing how content gets discovered, shared, and monetized.
But here is what most articles get wrong: AI Overviews are not destroying SEO, they are redirecting the traffic game. Yes, organic CTR for the top-ranking page drops sharply on queries with AI Overviews. The Pew Research panel study published in July 2025 documented users clicking through only about 8 percent of the time when AI summaries appeared, versus around 15 percent without them. Ahrefs’ analyses across hundreds of thousands of keywords show similar relative declines, and Seer Interactive has tracked even larger CTR drops on certain query categories. Yet pages cited as a source inside the AI Overview gain a direct advantage: more clicks, higher-intent traffic, and stronger conversion signals than non-cited pages. The new reality: you do not get traffic from ranking first anymore, you get traffic from being the most useful source the AI thinks to pull from.
What AI Overviews are: from featured snippets to AI synthesis
AI Overviews are Google’s AI-generated summaries that answer search queries directly. They appear at the top of search results on mobile and desktop, synthesizing information from multiple web pages into one cohesive answer. Unlike a featured snippet, which pulls and formats text from a single source, an AI Overview is a synthesized response that can combine facts, context, and citations from many sources. For users this is fast and convenient. For publishers it is the existential challenge of the mid-2020s.
The prevalence has accelerated dramatically. Pew Research and other panel-based measurements consistently document that AI Overviews now appear on a substantial share of US informational queries, and the share continues to grow. The feature spans nearly every search vertical: news queries, how-to questions, product research, local searches, even commercial-intent queries. If your site depends on organic traffic, AI Overviews now affect a large portion of your potential visitor base, regardless of vertical.
The CTR shift
The data is consistent across independent studies. When an AI Overview appears for a query, the click-through rate on the number-one organic result drops substantially. Ahrefs’ analyses of large keyword datasets, Seer Interactive’s CTR studies, and Pew Research’s panel-based behavior measurements all converge on the same conclusion: AI Overviews redirect a meaningful share of clicks away from the traditional organic results. Paid search has been hit at least as hard as organic on the queries where AI Overviews dominate.
The reversal is in the citation. Pages cited inside the AI Overview maintain or grow their traffic, while pages that ranked but were not cited lose ground. The prize is no longer the position-one ranking; it is the citation inside the box. The competitive question shifts from “can we rank first” to “can we be the most useful source the AI cites for this query”.
Zero-click search at high rates
One more metric to absorb. On AI Overview queries, the median zero-click rate is now substantially higher than on traditional queries. Most users get their answer from the synthesized summary and never click any result. Press Gazette’s analyses of Google traffic to news publishers documented sizeable year-over-year declines in publisher traffic in 2025, with US publishers among the hardest hit. This is the environment publishers are operating in: most queries no longer route traffic to any source, regardless of ranking position.
Which pages get cited
Here is where the SEO playbook gets rewritten. Ahrefs’ analyses of millions of AI Overview citations show that a substantial share of cited pages do not rank in the top ten for the corresponding query. Many AI Overview sources do not appear in the first-page results at all. This upends the traditional logic. Ranking first does not guarantee citation, and being ranked outside the top ten does not eliminate citation chances. The AI is not following rank-order logic; it is selecting sources based on relevance, specificity, trustworthiness, and accessibility, regardless of where they rank in the standard organic results.
The consequence is that you can rank well and still get nothing in the AI surface, and you can rank modestly and still capture meaningful AI-driven traffic if your content is genuinely the most useful answer for the query.
How to get cited: five strategies
Getting cited in AI Overviews is not a mystery, but it requires a fundamental shift in approach. Google has not published official guidelines, but research from the SEO community analyzing large citation samples has revealed consistent patterns. Five strategies actually work.
Put your best answer in the first third of the page
Independent analyses converge on the finding that a substantial majority of AI Overview citations come from content in the top portion of a page. Burying the answer in section five, below case studies and tangential explanation, makes the AI less likely to pull from it. The answer needs to be immediately accessible. The pattern that works: a concise direct answer in the first 100-150 words of the article (typically following an H2 that mirrors the query), then the deeper exploration below.
Use schema markup for machine readability
Schema markup does not guarantee citation, but it raises the odds. The right schema tells Google’s AI what the content is: a how-to, an FAQ, a research article, a product review. Use Article schema for editorial pieces, FAQPage for Q&A content, HowTo for procedural guides, Table-aware structures for data comparisons. Schema does not make you rank higher; it makes the AI’s text-extraction step more reliable on your content.
Build topical authority and E-E-A-T signals
AI Overviews rely on Google’s E-E-A-T framework: Experience, Expertise, Authoritativeness, Trustworthiness. The AI does not cite content from sites it does not trust, regardless of how well-written. Build authority within your topic cluster through consistent, interconnected content. A coherent body of expertise on one subject (a clean cluster of cornerstone plus spokes, properly interlinked) gets recognized as a reliable expert source rather than a one-off page. Our topic clusters guide covers the cluster construction in detail.
Earn editorial backlinks from high-authority sites
Backlinks remain critical for AI Overview citations. Pages cited in AI Overviews show meaningfully stronger link profiles than non-cited pages. Backlinks signal trustworthiness to both Google’s ranking systems and its AI models. The implication: the link-building work that supports traditional rankings also supports AI citation, and there is no shortcut around it.
Create unique data, original research, rare expertise
The AI favors sources with primary information. If you synthesize or repeat what others have said, citation is harder because you are competing in a crowded space. Original research, surveys, proprietary data, and analyses no one else has performed make you a uniquely citable source. The Pew, Backlinko, Ahrefs, Web Almanac, and First Page Sage style of original-data publication is exactly the format AI Overviews repeatedly cite, because the AI cannot synthesize new primary data; it can only quote it.
Structuring content for AI extractability
Beyond the strategies, the mechanics of HTML structure matter. AI Overviews parse content’s structure: headers, lists, tables, paragraph flow. Google’s AI is not just reading the text, it is reading the DOM. The patterns that maximize extractability:
Consistent heading hierarchy: H1 for title, H2 for main sections, H3 for subsections, no skipped levels.
Bullet lists for comparisons and benefit content. The AI extracts list items more accurately than running prose.
Data tables with clear headers for comparative or numerical content.
Short paragraphs (80-125 words) instead of dense blocks.
Introductory sentences at the start of sections that directly answer the question being searched, before the supporting context.
Is SEO ranking dead?
No. But ranking is no longer the primary goal. Ranking still matters because it builds authority signals the AI values, queries without AI Overviews still exist (a meaningful share), ranking well increases the odds of being cited, and some industries and query types show minimal AI Overview penetration. The “rank first, get traffic” equation is broken on the queries where AI Overviews dominate. You can rank first and get little traffic from AI queries. You can rank seventh and dominate AI citations. The optimization target shifts from rank position to citation suitability.
The strategic pivot
The shift from ranking to citation is not optional, it is the only path forward. The mental reframe: instead of “can I rank first for this keyword”, ask “can I become the most useful source the AI will cite for this query”. These are different problems with different solutions.
The concrete changes:
Prioritize content that answers questions directly and early. Do not write for position one; write to be cited.
Invest in primary research, original data, or proprietary methodology that no competitor has. This is the differentiator.
Build topical authority through clusters of interconnected, complementary content. The AI values expertise depth.
Focus backlinks on authority and relevance, not volume. One backlink from a trusted source is worth ten from mediocre sites.
Structure content for machine readability. Use schema, tables, lists, hierarchical headings.
Accept that overall organic click volume may decline, but AI citations can offset that with better conversion and higher-intent traffic. Users clicking from AI citations tend to be more qualified, because they have already seen the synthesized answer and clicked specifically for additional detail or verification.
Conclusion: adapt or decline
AI Overviews are not a passing feature, they are the new search landscape. By 2026, they appear on a substantial share of queries, and penetration continues to grow. The traditional ranking-first-gets-clicks logic no longer works on the affected queries. The same data points to a clear path forward: create authoritative, well-structured, original content with deep topical authority and you will be cited. Being cited is now the practical equivalent of ranking well in the older model. Publishers who adapted their strategy in 2025 are already harvesting traffic gains from AI citations. Publishers waiting for “things to go back to normal” are losing market share daily. The question is simple: are you going to be among the cited sources, or among the two-thirds losing traffic to an AI landscape you did not plan for. The answer starts with one decision: restructure your content strategy for AI citation, starting today.
LaFactory builds AI-citation-first content strategies that capture visibility in both traditional rankings and AI surfaces. Contact us to scope a content audit and roadmap that prepares your site for the AI search era.