Google AI Overviews Are Killing Traffic — Real Data (2026)

Google AI Overviews now appear on over 85% of informational search queries. For publishers ranking on page 1, CTR data shows organic click-through rates falling 30–60% on queries where an AI Overview is displayed. This is not a penalty, a ranking drop, or a manual action — it is a structural change to how search results deliver value. Understanding what is driving the drop, and what to do about it, is now the most important SEO challenge of 2026.

See our companion study of Google AI Overviews’ 38% average click-drop for the full dataset behind this analysis.

How Google AI Overviews Work

AI Overviews (previously called Search Generative Experience, or SGE) use Google’s Gemini model to synthesise an answer directly within the search results page. The model draws from multiple indexed pages, attributes citations to source URLs, and presents a complete answer above the standard blue-link results.

The user receives the answer without clicking. For many query types — definitions, how-tos, comparisons, “best X for Y” — the Overview resolves the information need entirely. The organic results below receive residual traffic from users who want more depth, want to verify the answer, or whose query was not fully resolved by the Overview.

Which Queries Are Most Affected

Not all query types trigger AI Overviews equally. Informational queries — “what is,” “how does,” “why does,” “best way to,” “difference between” — trigger Overviews most frequently and show the steepest CTR declines. Navigational queries (searches for a specific brand or site) and transactional queries (purchase-intent searches) trigger Overviews far less frequently, preserving more organic click volume.

The implication: publishers whose traffic is heavily informational have experienced the most severe impact. Publishers with strong commercial, transactional, or navigational traffic are relatively better positioned — but the scope of informational query coverage is expanding, not contracting.

How to Read the Impact in Google Search Console

The fingerprint of AI Overview impact in GSC is specific: impressions remain stable or increase (your pages still rank), but CTR drops sharply and total clicks fall. If you see a significant CTR decline on queries where your average position is 1–5, and traffic dropped while rankings held, AI Overviews are almost certainly the cause — not an algorithmic penalty.

Filter your GSC Performance report by “Search type: Web” and sort by CTR descending. Pages with previously strong CTR (above 5% for positions 1–3) that have dropped to 2–3% are your highest-priority pages for optimisation. These are the pages where structural changes will recover the most traffic.

How to Optimise for AI Overview Citation

Being cited in an AI Overview is not random — it follows identifiable patterns. Google’s model favours pages that:

• Directly and concisely answer the query in the first 1–2 sentences of the article
• Use structured headings that map to the sub-questions a user might have
• Include specific data, named sources, or original research — not just synthesis of other articles
• Demonstrate E-E-A-T signals: named authors, verified expertise, institutional credibility
• Load fast (LCP under 2.5s) with no significant layout shift on the content area
• Use appropriate schema markup (FAQPage, HowTo, Article) to signal content type

Citation in an AI Overview functions as a brand signal even for users who don’t click through. It drives branded search volume and reinforces topical authority over time — a compounding effect that partially offsets the immediate CTR loss. Publishers consistently cited across a topic area build authority that is difficult for competitors to replicate quickly.

Generative Engine Optimisation (GEO): The Discipline for AI Search

Traditional SEO optimises for ranking position. GEO — Generative Engine Optimisation — optimises for citation and inclusion in AI-generated answers. The disciplines overlap, but they diverge at a key point: GEO prioritises completeness and directness over keyword density and link equity. A page that comprehensively answers a cluster of related queries from a single well-structured URL will consistently outperform a page optimised for a single exact-match keyword.

For a detailed breakdown of GEO tactics, see our Generative Engine Optimisation guide. The core strategic shift: write for the question your reader is actually asking, not the keyword you want to rank for.

What to Do Right Now: Six Priority Actions

The publishers retaining traffic through the AI Overview shift share a common approach: they produce the clearest, most complete, most credible answer available for their target queries. These six steps reduce directly to that principle.

1. Audit top pages in GSC: Identify pages where CTR dropped but average position held. These are your AI Overview impact pages.
2. Rewrite introductory paragraphs: The first 50–100 words should contain a direct, complete answer to the primary query — before any context-setting or preamble.
3. Add structured sub-headings: Each H2/H3 should map to a related question users ask, making it easy for the Overview model to identify and cite your specific sections.
4. Strengthen E-E-A-T signals: Named author, publication date, specific data points with sources, external citations to authoritative references.
5. Implement or audit schema: FAQPage and Article schema increase citation eligibility. Every content post should have Article schema at minimum.
6. Protect transactional pages: Pages with commercial or conversion intent should remain focused on that goal. Avoid adding informational content that dilutes intent signals and draws AI Overview attention to pages you need to convert, not inform.

Frequently Asked Questions

The practical response to AI Overviews is Generative Engine Optimization (GEO) — structuring content to be cited inside AI-generated answers rather than displaced by them. For the broader context of how this transition is reshaping search, see the search engine's endgame.

The same citation-eligibility principles apply beyond Google: see how to appear in ChatGPT Search answers.

Sources: Google Search Documentation | Moz SEO Guide