AI Overviews
AI Overviews are AI-generated answers that Google shows above the organic results for some searches.
Google writes each one with a custom version of its Gemini model, summarising several web pages into a short answer and linking to the pages it drew from.
The part most SEO guides skip: those links rarely get clicked. Pew Research Center tracked the real browsing of 900 US adults in March 2025 and found users clicked a link inside an AI summary in about 1% of visits. A citation in an AI Overview is a brand impression, not a traffic source, and B2B teams should value it that way.
How Google picks AI Overview sources
When a search triggers an AI Overview, Google retrieves pages from its index, has Gemini write a summary, and lists the sources inside or beside the answer. The sources are not simply the top organic results. According to Ahrefs’ March 2026 study of 863,000 keywords, only 38% of pages cited in AI Overviews rank in the top 10 for the same query, down from 76% in July 2025.
Google’s own documentation gives the reason: AI Overviews and AI Mode may use a “query fan-out” technique, running several related searches across subtopics as they build the answer. A page ranking 15th for the head term can win the citation by answering a narrow follow-up question more clearly than anything on page 1.
Rollout and coverage
| Date | Milestone |
|---|---|
| May 2024 | Launched in the US |
| October 2024 | Rolled out to 100+ countries |
| May 2025 | 200+ countries and territories, 40+ languages, with Arabic added |
Longer searches trigger them more often. In Pew’s data, just 8% of one- or two-word searches produced an AI summary. That matters in B2B, where buyers type long, specific questions. “SOC 2 requirements for a 50-person SaaS company” is far more likely to show an AI Overview than “SOC 2.”
Traffic impact on B2B sites
Clicks drop across the whole results page. In Pew’s study, users clicked a traditional result in 8% of visits with an AI summary, against 15% without one. Ahrefs’ February 2026 update, using December 2025 data across 300,000 keywords, found the top-ranking page’s click-through rate was 58% lower when an AI Overview appeared.
Google tells a different story. At I/O 2025, Google said AI Overviews had raised usage of Google by more than 10% for the queries that show them, in the US and India. Both can be true. People search more and click less per search.
The loss lands on educational content: glossary entries, how-to guides and top-of-funnel posts. Pages that serve buying decisions lose less, since a buyer comparing vendors still needs your pricing, specifications and proof.
Getting cited
-
Put a direct, self-contained answer in the first 40 to 60 words of the page and of each section. AI Overviews quote short statements that make sense out of context.
-
Give each section one question, with a heading phrased the way buyers phrase the follow-up.
-
Keep structured data matched to the visible text. Google’s documentation on AI features says no special schema, AI text files or extra markup is needed to appear in them. Schema helps only by describing the page accurately.
-
Check your snippet settings. The same documentation lists nosnippet, data-nosnippet, max-snippet and noindex as controls that limit what it can show. I found this on my own site in October 2026: The SEO Framework plugin had set max-snippet:50 site-wide, capping Google at 50 characters per page. Nothing in the WordPress dashboard flagged it.
-
Replace outdated numbers. A page quoting 2022 data loses the citation to one quoting 2026 data.
Measuring AI Overview visibility
Since June 2026, Search Console’s Generative AI performance report shows how often your pages appeared in AI Overviews and AI Mode. It counts impressions only, with no clicks. Pair it with a rank tracker that records whether an AI Overview appears for each keyword and which domains it cites, then compare click-through rates on keywords with and without one. With citations clicked in about 1% of visits, report them as a visibility metric beside rankings, not as a traffic source.
Common mistakes
-
Stopping informational content because clicks fell. The citation still puts your brand in front of buyers at the research stage.
-
Capping snippets by accident through plugin defaults.
-
Treating a page 1 ranking as proof of AI visibility. With 38% overlap, it isn’t.
-
Judging AI Overview work by sessions. The return shows up in branded searches and shortlists.